Public Health Nutrition: 19(12), 2135–2148

doi:10.1017/S1368980015003183

Review Article

Correlates of nutrition label use among college

students and young adults: a review
Mary J Christoph1, Ruopeng An1 and Brenna Ellison2,*
1
Department of Kinesiology and Community Health, University of Illinois at Urbana–
Champaign, Urbana, IL, USA: 2Department of Agricultural and Consumer Economics,
University of Illinois at Urbana–Champaign, 321 Mumford Hall, 1301 W. Gregory Drive,
Urbana, IL 61801, USA

Submitted 13 July 2015: Final revision received 15 September 2015: Accepted 13 October 2015: First published online 9
November 2015

Abstract college students

Objective: Nutrition labels are an essential source for consumers and young
to obtain nutrition-related information on food products and serve adults. Future
as a population-level intervention with unparalleled reach. The research is
present study systematically reviewed existing evidence on the warranted to
correlates of nutrition label use among college students and young assess the role
adults. of label use on
Design: Keyword and reference searches were conducted in improved
PubMed, EBSCO, PsycInfo, Cochrane Library and Web of Science. dietary
Inclusion criteria included: study design (randomized controlled decisions.
trial, cohort study, pre–post study or cross-sectional study); population
(college students and young adults 18–30 years old); main outcome
(nutrition label use); article type (peer-reviewed publication); and
language (English). Setting: College/university.
Subjects: College students and young adults.

Results: Sixteen studies based on data from college surveys in four

countries (USA, UK, Canada, South Korea) were identified from
keyword and reference search. Reported prevalence of nutrition
label use varied substantially across studies; a weighted average
calculation showed 36·5 % of college students and young
adults reported using labels always or often. Females were
more likely to use nutrition labels than males. Nutrition label use
was found to be associated with attitudes towards healthy diet,
beliefs on the importance of nutrition labels in guiding food
selection, self-efficacy, and nutrition knowledge and education.
Conclusions: The impact of nutrition labelling on food purchase
and intake could differ by population subgroups. Nutrition
awareness campaigns and education programmes may be
important mechanisms for promoting nutrition label use among

Improving diet quality is a key health promotion
strategy. Since 1980, a major theme of the
US federal dietary guidelines has been to
increase consumption of nutrient- rich foods
and reduce consumption of energy-dense
foods(1). However, a large majority of the
American population fails to meet these
guidelines, with insufficient consumption of
nutrient-rich foods such as fruit and
vegetables and excessive discretionary calorie
intake(2).
Nutrition labels are an essential source for
consumers to obtain nutrition- and health-
related information on food products and serve
as a population-level intervention with
unparalleled reach(3). A substantial proportion
of US consumers report regular use of nutrition
labels to guide
*Corresponding author: Email brennae@illinois.edu
Authors 2015
Keywords
Nutrition labels
La
bel use
Dining hall
University
students
Young adults
their food selection(4–7). The perception on
the credibility of nutrition labels appears high,
whereas findings on the relationship between
nutrition label use and diet quality remain
largely inconclusive(4–7). Multiple systematic
reviews suggest that nutrition labelling
alone may not effectively reduce calorie
selection or intake in general populations(8,9),
although labelling appears somewhat
effective when paired with interpretational
aides such as statements about daily nutritional
needs(9). The substantial variability in study
results could be partially due to het-
erogeneities in nutrition label use and dietary
habits across population subgroups. Children,
adolescents, obese older adults, individuals
with less education and/or nutrition
© The
 Correlates of college students’ label use 2137
nutrition label use (Nutri- tion Facts labels, labels within dining
knowledge, people with lower disposable income and
halls or nutrition labels
those with limited health awareness are found less likely
to use labels and/or effectively process the nutrition
infor- mation presented(7,10,11).
One population subgroup that has received less atten-
tion in the literature is college-aged students and
young adults. These individuals are often included in
the general adult population studies; however, there
is evidence to suggest this particular subgroup
warrants specific atten- tion. During the college time
period, many young adults are making the transition
from living at home with their family members to
living independently. This transition forces young adults
to start developing their own habits, routines and
preferences (including food and dietary decisions)
(12)
, many of which persist into adulthood.
Unfortunately, two patterns have emerged for this
age group: weight gain(13) and decreased dietary
quality(14). Nelson et al.(12) noted that the transition from
adolescence to adulthood is associated with
decreased fruit and vegetable consumption, increased
fast food and soft drink consumption, and lower levels of
physical activity. More concerning, longitudinal studies
show that poor dietary quality in young adulthood is
associated with long-term risk of cardiovascular(15)
and metabolic disease(16).
Nutrition labels may serve as an important preventive
tool for college students and young adults by
encouraging the formation of habitual behaviours that
could profoundly impact their food preferences and
diet quality later in life(12). To date, much of the
research assessing compre- hension, predictors and the
impact of nutrition label use on food behaviours and
intake focuses on the general adult population.
Limited research in young adults suggests that
individuals in this subgroup may use nutrition labels,
but frequency and predictors of usage are not well
known. While the impact of food environment
interventions has been reviewed in college
students(17), no reviews have focused on predictors or
correlates of label usage in college students and young
adults. Documenting factors that influence nutrition
label use in this subgroup is particularly important for
informing targeted nutrition interventions and improving
the effectiveness of nutrition education pro- grammes
and awareness campaigns. The objective of the present
study was to systematically review existing scien- tific
evidence on the correlates of nutrition label use among
college students and young adults 18–30 years of
age.

Methods

Study selection criteria

Studies that met all of the following criteria were

included in the review: study design was a randomized
controlled trial, cohort study, pre–post study or cross-
sectional study; population was college students and
young adults 18–30 years of age; main outcome was
 Correlates of college students’ label use
in general); article type was a peer-reviewed
publication; and language was English.
Studies were excluded from the review if they
met one or more of the following criteria: case
reports or case–control studies; non-English
publications; non-peer reviewed articles;
experiments that require nutrition label
reading as a prerequisite for study participa-
tion; and studies that assess participants’
preference for alternative label formats,
belief on the accuracy of information
presented on labels, label comprehension
or intent to use some hypothetical rather
than actual labels.
2137
dissimilar nature of study designs and outcome
measures (i.e. alternative definitions and instruments on
nutrition label use). Analysis included a narrative
review of the included studies with general themes
summarized(18), in addition to a weighted average of
nutrition label usage prevalence for the thirteen studies
with similar Likert-scale responses for label usage
frequency (most studies used 3-, 4- or 5-point scales).
Responses were grouped into three categories: (i)
always or often; (ii) sometimes; and
(iii) rarely or never. Overall prevalence for each
category was calculated by dividing the number of
students in each category by the total sample size of all
the included studies.

Search strategy

Keyword search was performed in

PubMed, EBSCO, PsycInfo, Cochrane
Library and Web of Science. The search
algorithm included all possible
combinations of keywords from the
following three groups: (i) ‘nutrition’,
‘calorie’, ‘food’, ‘diet’ or ‘menu’; (ii) ‘label’,
‘labeling’ or ‘labelling’; and (iii) ‘college
student’, ‘university student’, ‘young adult’,
‘university cafeteria’ or ‘college cafeteria’.
Articles with one or more of the following
keywords were excluded: ‘supplement’,
‘pharmacology’, ‘medication’, ‘allergy, ‘mice’
or ‘cigarette’. Titles and abstracts of the
articles identified through keyword search
were screened against the study selection
criteria. Potentially relevant articles were
retrieved for evaluation of the full text.
A cited reference search (forward reference
search) and
a reference list search (backward reference
search) were also conducted based on the
articles identified from the keyword
search. Articles identified through
forward/ backward reference search were
further screened and evaluated using the
same study selection criteria. Reference
searches were repeated on all newly
identified articles until no additional
relevant article was found.

Data extraction and synthesis

A standardized data extraction form was

used to collect the following methodological
and outcome variables from each included
study: author(s), publication year, study
design, setting, sample size, sample
demographics, response and/or completion
rate, participant recruitment criteria,
nutrition label use, correlate(s) of nutrition
label use, main findings and conclusions.
A meta-analysis could not be conducted due to the
 Table 1 Definitions of study quality criteria

Item Criterion of study quality

1 Study design and data collection procedures clearly documented (yes = 1, no = 0)*
2 Sample size (400 or more participants = 1, less than 400 participants = 0)†
3 Response or completion rate reported (yes = 1, no = 0)
4 Survey instrument validated (yes = 1, no = 0)‡
5 Considered demographic correlates of usage (yes = 1, no = 0)§
6 Considered non-demographic correlates of usage (yes = 1, no = 0)||
7 Used regression analysis to examine the relationship between label usage and multiple predictor variables/correlates
simultaneously (yes = 1, no = 0)
8 Non-restricted population (yes = 1, no = 0)¶
*At a minimum, how subjects were recruited and mode of data collection (e.g. in-person, online) should be indicated.
†Using the most conservative estimate of 50 % nutrition label usage prevalence, a power analysis indicates that at the 95 % confidence level, a sample size of
384 people would be needed to detect significant differences in usage.

‡Surveys were considered validated if they were either adapted from previously published surveys or first pilot-tested with the population of interest.

§Demographic correlates included: gender, age, class, education level, race/ethnicity, BMI and marital status.

||Non-demographic correlates included: knowledge, attitudes, beliefs, self-efficacy, behaviours and nutrition education.

¶Indicates the population was not restricted by any factor other than age or student classification.

Study quality assessment

* Since usage prevalence in college students and young adults has not been previously reviewed,
The quality of each study included in the review we used the most conservative estimate of 50 % prevalence for our power analysis calculation to
was assessed by the following eight criteria, adapted gauge necessary sample size to detect significant correlates of usage at the 95 % confidence level.
from the US National Institutes of Health(19)
recommendations and tailored specifically for assessing
the cross-sectional stu- dies included: (i) study design
and data collection proce- dures were clearly
documented (yes = 1, no = 0);
(ii) sample size (400* or more participants = 1, less than
400 participants = 0); (iii) response or completion
rate was reported (yes = 1, no = 0); (iv) survey
instrument was vali- dated (yes = 1, no = 0); (v)
demographic correlates of usage (yes = 1, no = 0)
were considered; (vi) non-demographic correlates of
usage (yes = 1, no = 0) were considered;
(vii) regression analyses were performed to examine the
relationship between label usage and multiple predictor
variables/correlates simultaneously (yes = 1, no = 0);
and
(viii) non-restricted population wherein participants were
not excluded from eligibility for any factor other than
age or student status (yes = 1, no = 0). Full definitions
for all study criteria are available in Table 1. Given
these criteria, total study quality score ranged between
0 and 8. Study quality score helped measure the
strength of study evidence but was not used to
determine the inclusion of studies.

Results

As Fig. 1 shows, among a total of 235 unduplicated

articles identified through keyword and reference
searches, 197 were excluded in title and abstract
screening. The remaining thirty-eight articles were
reviewed in full texts, in which twenty-two studies
were excluded due to the following reasons: age
ineligibility (n 3)(20–22), no assessment of nutrition label
use (n 11)(23–33) and an ineligible study design (n 8),
which included six experiments that required partici-
pants to read a nutrition label(34–39), one semi-structured
interview(40) and one case–control study(41).
The remaining sixteen articles were included
in the review.

Basic characteristics of selected studies

Tables 2 and 3 summarize the sixteen peer-
reviewed journal articles included in the
review. All but two studies were published in
2005 or later. Studies were conducted in four
countries: the USA (n 13)(42–54), the UK (n 1)
(55)
, Canada (n 1)(56) and South Korea (n 1)(57).
Among the US-based studies, two were
conducted in the West (Cali- fornia, Oregon)
(48,54)
, four in the South (Georgia, Missouri,
Louisiana, Texas)(42,49,51,53), four in the
Northeast (Penn- sylvania, New York, Vermont,
Connecticut)(43,46,47,50) and three in the Midwest
(Nebraska, Minnesota, Ohio)(44,45,52). All studies
were surveys conducted in college or
university settings. Ten of these surveys were
administered in person, four online(43,45,54,55),
one via telephone(47) and one did not state
the administration method(51).

Prevalence of nutrition label use

Although most of the studies had slightly different

measures of label use, the majority of studies
shared similar response categories for the
frequency of label usage (3-, 4- or 5-point
Likert scales that were typically anchored by
‘never’ and ‘always’). In order to merge the
different scales, we look at usage frequency
using a 3-point scale (‘always/often’,
‘sometimes’, ‘rarely/never’).† Thirteen out of
the sixteen studies were included in this
analysis; the remaining three were excluded
for only reporting usage as a dichotomous
(yes/no) variable(57), not reporting overall
usage frequency(42) and only reporting usage
as a continuous number from the frequency
scale(52).‡ Table 4 provides the

† It should be noted that some studies used slightly different wording (e.g.
‘none of the time’, ‘every single time’, ‘always or almost always’);
how- ever, all scale points could be classified into one of these three
categories. For exact wording, refer to the ‘Measure of nutrition label use’
column in Table 2.
‡ Cha et al.(42) and Misra(52) did not report the breakdowns of their 5-
point frequency scales. Both authors were contacted by email but were
unable to provide this information by the deadline given.
 Articles identified in keyword and reference search (n 235)
Articles retrieved for full text evaluation (n 38)
Articles included in the review (n 16)
Fig. 1 Study selection flowchart
breakdown for each frequency category across the
thirteen studies. Using this information, we calculated a
weighted average of label usage frequency among
college students and young adults. As shown in Table
4, 36·5 % of college students reported using nutrition
labels always or often. Almost the same percentage
(36·7 %) of students reported using labels sometimes,
whereas 26·8 % reported rarely or never using labels.
Correlates of nutrition label use
Twelve of the thirteen studies that assessed gender
dif- ferences in nutrition label use found being
female was significantly associated with higher label
use(43–50,52–54,56). In the remaining study, Cha et al.(42)
found that gender and food label use were not
correlated; however, this study surveyed only 103
participants, 70 % of whom were female. One study
had female participants only and thus was unable to
test for gender differences in nutrition label use(57),
whereas the other two studies did not report the
presence/absence of gender differences(51,55).
Seven studies reported participants’ BMI or body
weight status based on BMI (i.e. underweight, normal
weight, overweight and obesity), five of which were
self- reported(47,48,50,55,57) and two based on objectively
mea- sured height and weight(42,45). Martinez et al.(50)
found that overweight or obese college students were
significantly more likely to use nutrition labels in making
lower-calorie and healthier food choices in dining
halls. In contrast, Li et al.(48) and Krukowski et al.(47)
found no association between BMI and nutrition label
use among college stu- dents and young adults. While
Cooke and Papadaki(55) did not report on any relationship
between BMI and nutrition label use directly, they
documented that BMI was related to nutrition
knowledge and attitudes towards healthy eating –
two variables that were related to label use.
Specifically, they found that normal-weight college
stu- dents had higher nutrition knowledge than their
under- weight counterparts, whereas overweight college
students had lower attitudes towards healthy eating
than their underweight counterparts. Other studies
found no differ- ence in height or weight(57) or average
BMI(45) between
Articles excluded on basis of title and abstract (n 197)
Articles excluded as did not meet inclusion criteria (n 22)
label users or non-users. Although Rasberry et al.(53)
did not assess BMI, they found frequent label users
were nearly three times more likely than non-users
to select ‘weight control’ as a reason for using
labels.
The findings on nutrition label use in relation to age
and student classification (undergraduate students
including freshmen, sophomores, juniors and seniors, and
graduate students) remain mixed. Age was found to be
positively associated with nutrient label use in two
studies(42,52). In terms of student classification, one study
found that juniors and seniors were more likely to use
nutrition labels com- pared with freshmen and
sophomores(51), whereas Jasti and Kovacs(46) found that
graduate students were less likely to use nutrition labels
than undergraduates, possibly due to the higher
proportion of international students among the grad- uate
student body. A third study by Misra(52) found under-
graduate and graduate students were equally likely to
use labels; this result is somewhat surprising given the
author also found age was positively related to a higher
label reading behaviour score. Others found no
significant changes in nutrition label use by age(45,49) or
student classification(48).
There is limited evidence on nutrition label use in
relation to race/ethnicity or marital status. Two
studies reported that white(45) or non-Hispanic white
students(46) were more likely to use nutrition labels
than all other races/ethnicities, whereas another study
reported no dif- ference in nutrition label use by
race/ethnicity(48). Only one study assessed marital
status but did not find it to be associated with
nutrition label use(48).
A few studies examined nutrition label use in relation
to attitudes, beliefs and self-efficacy. Four
studies(45,49,52,55) found that attitudes towards healthy
eating or preparing healthy meals positively predicted
nutrition label use. Smith et al.(56) found that the only
significant predictor of nutrition label use in both
genders was the belief on the importance of nutrition
labels in guiding food selection, although beliefs in the
truthfulness of labels and diet–disease relationships
also significantly predicted usage in men. Jasti and
Kovacs(46) found that belief on the importance of
eating a low-fat diet predicted label use, while
Rasberry et al.(53) found that health reasons and
looking for specific information related to
usage. Increased
Table 2 Basic characteristics of the studies included in the review Cor
rel
Study design and
Author(s), year setting Recruitment and sampling procedure Sample characteristics Correlate(s) studied Measure of nutrition label use ate
s
Cha et al. (42)
,
Cross-sectional; Self-referral and flyers posted at 103 inactive overweight or obese young Age, gender, ‘How often do you use the food label when of
colleges, education,
2014 colleges and universities, student clinics and via adults; 78·6 % female; mean age BMI, health literacy, making a food selection?’ (5-point Likert coll
universities, university emails to recruit students who of 23·99 (SD 3·25) years; 70 % self-efficacy scale where 1 = ‘never’, 5 = ’always’) eg
Atlanta, Georgia, were physically inactive and African- American; 59·2 %
USA overweight/obese. undergraduates e
Those under physician-supervised diet/
exercise regimens or with diabetes or stu
blood diseases excluded. Out of 234 de
recruits, 106 enrolled in the study; three
participants were removed from data nts
analysis ’
Conklin
Cross-sectional; Housing and Food Services sent 3290 first-year 1317 college freshman who ate Gender, information ‘I use nutrition and ingredient information onlab
in
(43)
et al. , land-grant students an email invitation for the campus dining commons; 49·3% sources, knowledge, food products’ (3-point Likert scale
2005 university, study during the second week of female motivations where 1 = ‘never’, 3 = ‘always’); and
Pennsylvania, USA classes. 1317 students (40 %) ‘Nutrition and ingredient information
responded supplied in the dining commons affects my
choice of food’ (7-point Likert scale where
1 = ‘strongly disagree’, 7 = ‘strongly
agree’)
Cooke and
Papadaki(55), Cross-sectional; UK 134 UK universities from the Universities 524 college students; 75 % female; Attitudes, knowledge ‘How often do you read the nutrition labels
2014 universities and Colleges Admission Service database mean age of 24·9 years; mean on
were invited; thirty-seven sent the survey BMI of 23·1 kg/m2; 77 % food labels before purchasing foods or
web-link to students. Survey closed when undergraduates beverages?’ (4-point Likert scale
a convenience sample of 500 participants where 1 = ‘never or rarely’, 4 = ‘always
was reached or almost always’; students responding
3 or 4 were categorized as frequent
label users)
Driskell
et al.(44), Cross-sectional; Participants entering or leaving dining halls 205 college student dining hall Gender, education, Current use of Nutrition Bytes (dining hall
2008 University of were recruited during a midweek lunch patrons; 44·3 % female; 83 % knowledge labelling system). If user, frequency of
Nebraska, USA period during week 7 of the semester. were 19–21 years of age; 87 % usage was asked (3-point Likert scale
Twenty respondents were excluded from undergraduates where 1 = ‘never’, 3 = ‘nearly always’)
data analysis for not listing gender or
being under age 19 years
Graham and
Laska(45), Cross-sectional; Data collectors gave students on campus an Age, gender, BMI, ‘How often do you read the nutrition labels
2012 community college online survey web-link and code 1201 college students; 598 from race/ ethnicity, on food labels before purchasing foods
and 4-year community college and 603 from dietary behaviours, or beverages?’ (4-point Likert scale
university, public 4-year university; 52·5% nutrition knowledge, where 1 = ‘never or rarely’, 4 = ‘always
Minneapolis/St. female; mean age of 21·5 years; attitude towards or almost always’; students responding
53·4 % non-white preparing healthy 3 or 4 were considered frequent users)
Paul, Minnesota, meals
USA

Jasti and
Cross-sectional; Students from five undergraduate and graduate 222 college students; 56 % Age, gender, race/ Participants were asked if they used food
female; labels
Kovacs (46)
, Queens College, non-nutrition classes were surveyed at mean age of 23·9 (SD 5·6) years; ethnicity, shopping, to help them make better food choices
2010 Queens, New the end of classes in summer and 63 % non-white; 82 % attitudes, knowledge, (3- point Likert scale where 1 =
York, USA autumn. undergraduate nutrition education ’always’, 3 = ‘never’; students
Those under 18 years old were responding 1 or 2 categorized as
excluded. Overall response rate of 83 users)
%
Krukowski
Cross-
sectional; A random sampling telephone survey was used 316 college students; 56 % Gender, BMI, Usage frequency of food labels was
female; food measured
(47)
et al. , Vermont, USA on two populations: (i) community sample mean age of 20·5 (SD 3·5) behaviours on a 3-point Likert scale where 1 =
2006 and (ii) college sample.* Those under 18 years; 22·3 % overweight ’often’, 3 = ‘rarely’
years old were excluded. Overall response (BMI ≥ 25·0 kg/m2)
rate of 60·4 % for college sample
Li et al.(48), Students were recruited in general education 488 college students; 65 % female; How often they read food labels before
2012 Cross-sectional; Age, gender, marital
Oregon State classes and given the link to an online mean age of 19·6 (SD 2·4) years; status, race/ethnicity, buying (5-point Likert scale where 1 =
University, USA survey at the beginning and at the end of 75 % white; 63 % freshmen; mean BMI, class ‘never’, 5 = ‘always’)
academic terms. 582 students completed BMI of 23·8 (SD 4·3) kg/m2
the survey; however, ninety-four 21
observations were removed for
incomplete responses 39
Table 2 Continued 21
Study design and 40
Author(s), setting Recruitment and sampling procedure Sample characteristics Correlate(s) studied Measure of nutrition label use
year

Lim et al.(57),
Cross-sectional; 300 female students were recruited; those with 275 female college students; mean Age, height, weight, class, Participants were asked if they read
age nutrition
2015 women’s university, incomplete responses (n 15) and majoring of 20·6 years; mean height and beliefs labels when selecting or
Seoul, South in food or nutrition (n 10) were excluded, weight of 162·4 cm and 52·6 kg; purchasing processed foods or
Korea for a total of 275 students 34·2% freshmen, 29·4 % snacks; ‘yes’ responses = users;
sophomores, 16·4 % juniors, 20·0 ‘no’ or ‘don’t know’ responses =
% seniors non-users
Marietta Undergraduates taking a basic life science Age, gender, attitude,
Cross-sectional; class were recruited and surveyed in class 208 undergraduates; 67·8 % female; knowledge, nutrition ‘When you purchase a food product for the
(49)
et al. , Southeast Missouri mean age of 21·7 (SD 5·6) years education first time, do you look at the Nutrition
1999 State University, Facts label on the package?’ (5-point
Cape Girardeau, Likert scale where 1 = ‘never’, 5 =
Missouri, USA ‘always’)

Martinez
et al.(50), Cross-sectional; Yale Students were recruited over a 4-month 487 undergraduates; 56 % female; Gender, BMI ‘How often do you look at food labels on
2012 University, period and received $US 2 for a 5 min mean BMI of 22·5 (SD 2·9) packaged foods?’ (5-point Likert scale
Connecticut, USA survey about the dining halls. Non- kg/m2; 17 % overweight or where 1 = ‘never’, 5 = ‘always’)†
undergraduates and those under 18 obese
McLean- years old were excluded
Meyinsse Cross-sectional; Class year ‘How often do you read Nutrition Facts
et al.(51), Louisiana, USA Students were surveyed on campus during 441 college students; 58 % labels?’ (4-point Likert scale where 1 =
2011 spring and autumn semesters female; ‘never’, 4 = ‘often’)

mean age of 20 years; 35 %

Misra (52)
, 2007 Cross-sectional; two Survey was mailed to 1500 students
(1000 undergraduate, 500 graduate)
Midwestern randomly selected from the registrars’
universities, USA student directories. The return rate was
36 % (184 graduate and 353
undergraduate students)
freshmen; 87 % African-
American
537 college students; 67 %
female; mean age of 23 (SD 6·1)
years; freshmen and juniors (66
%); 21 % were overweight and
8 % obese
Age, gender, supplement Four questions based on the Label
use, nutrition education, Reading Survey assessed the general
attitudes, knowledge use of the Nutrition Facts panel – exact
wording not provided (5-point Likert
scale where 1 = ‘never’, 5 = ‘always’)

Rasberry
et al.(53), Cross-sectional; Students were recruited from eighty-eight 1294 college students; 52 % Gender, weight Four items related to usage – exact wording
2007 large university in physical education activity programme female; control, knowledge, for all items not provided (4-point Likert
Texas, USA courses; 1294 out of an estimated 48 % freshmen/sophomores attitudes, beliefs scale where 1 = ‘never’, 4 = ‘always’)
2756 eligible participants (47 %
response rate) completed surveys.
Those under age 18 years were
excluded
Smith et al.(56),
2000 Single-stage cluster Second-year classes offered in the College 553 college students; 46·8 % female; Gender, beliefs ‘How frequently do you use the nutrition
sampling; of Arts and Science (twenty-five classes 80·1 % between the ages of 18 information panel?’ (5-point Likert scale
University of contacted, of which seventeen participated) and 24 years where 1 = ‘none of the time’,5 = ‘every
Saskatchewan, were randomly sampled; the response single time’). Participants were also
Canada rate was 92 % asked about use of nutrient claims using
same 5-point scale
Wie and
Cross-sectional; 4- Contacted faculty randomly, who distributed an 342 college students; 75 % female; 54 % Gender, ‘I consider the calorie content of menu items
nutrition at
Giebler (54)
, year university in online survey. Students (n 819) in the were aged 21–30 years and 31 % education restaurants before making my choices’
2014 Sacramento, rosters were invited to participate via an were aged 20 years or younger; (5- point Likert scale where 1 =
California, USA email including a survey and one follow- 70 % juniors/seniors ‘never’, 5 = ‘always’)
up email reminder sent to faculty. The
response rate was 342 out of 819 (42 %)
*For the purposes of the present review, we focus on the college sample results. However, in some cases, the authors only report results on the combined sample (these instances are identified in the findings
section in Table 3). MJ
†Exact usage questions not provided in the manuscript; the wording of this measure was obtained via personal communication with the corresponding author. Ch
rist
op
h
et
Table 3 Main findings and conclusions of the studies included in the review
Cor
rel
ate
Author(s), year Main finding Conclusion
s
Cha et al.(42)
, 2014 Food label usage was significantly correlated with education, age and eating
behaviour self-efficacy. In a regression, health literacy and self-efficacy explained Strategies to enhance health literacy, self-efficacy and food label of
almost 10 % of the variability in food label use; in the full model, only self-efficacy use should be developed in an effort to improve dietary coll
was significantly correlated with label usage quality
eg
Conklin et al.(43), 2005 21·7 % of students said they use nutrition and ingredient information on food e
products always (56·6 % responded sometimes, 21·6 % responded never).
Almost 30 % of students agreed that nutrition information supplied in the dining One-third of students used labels to help make food choices in stu
commons affects their food choice. Females were more likely to use labels. the dining commons; females were more likely use labels and de
Using dining hall labels was significantly related to a self-reported good report eating in the dining hall because information was
available nts
understanding of nutrition, using nutrition labels on food products and obtaining
information on food choices from weight-loss programmes. Usage was not related ’
to wanting to be healthy now, avoiding poor health later, food allergies or losing lab
weight
Cooke and Papadaki(55), 2014 34·7 % of students said they always or almost always read the nutrition labels on
food before purchasing foods/beverages, 28·8 % said often, 23·3 % said
sometimes and 13·2 % said never or rarely.* Logistic regression results showed
that for every one-point increase in nutrition knowledge and attitudes score, the Nutrition knowledge and attitudes significantly predict usage
odds of nutrition label use significantly increased by 0·03 (3 %) and 0·2 (20 %)
times, respectively. Usage was significantly correlated with knowledge (0·203)
and attitudes (0·137)
Driskell et al.(44), 2008 Nutrition Bytes labels were used by 58·5 % of the 205 subjects. More women
(79·1 %) than men (42·1 %) reported using Nutrition Bytes labels (P = 0·0001).
Education on nutrition labels resulted in label usage; however, nutrition knowledge
was not related to label usage. Reasons for using labels were general knowledge
(51·7 %), concern about overall health (49·2 %), calorie counting (46·7 %) and The majority of patrons used the Nutrition Bytes labelling
concern about a certain nutrient(s) (43·3 %). Common reasons for non-use were system. Females and those who had been educated on
‘it will not change my mind about food items I select’ (64·7 %) and not having nutrition labels were more likely to use labels. More efforts on
enough time (27·1 %) label education and linking nutrition, diet and health may
improve label use, diet quality and health outcomes
Graham and Laska(45), 2012 35 % of students were classified as frequent label readers. Females and white
students were significantly more likely to be frequent label readers; age and BMI
were not related to frequent label reading. Nutrition label readers had greater nutrition
knowledge, better estimates of how many calories and fruit/vegetable servings are
needed to be healthy, and were more likely to perform healthy dietary Frequent use mediated the relationship between the importance
behaviours compared with the infrequent label readers. Regression analysis of preparing healthy meals and dietary quality, showing that it
revealed attitude towards preparing healthy meals was significantly related to may be a tool that people who value nutrition may use to
frequent label reading change behaviour

Jasti and Kovacs(46), 2010 24 % of students reported always using food labels, 61 % reported sometimes and 14 % White students, undergraduates, females, grocery
shoppers and
reported never using labels. Females, non-Hispanic whites and undergraduates those with prior nutrition education or who thought a low-fat
were significantly more likely to use food labels than their male, non-white and diet was important used labels more frequently
graduate counterparts, respectively. Additionally, students with prior nutrition
education and grocery shoppers were more likely to use food labels. Logistic
regression showed good trans-fat knowledge, importance of eating a low-fat diet,
younger age group and positive grocery shopping status were strong predictors
of food label use; prior nutrition education, gender and college level were not
Krukowski et al.(47), 2006 48·4 % of college students said they always used food labels; 22·9 % said sometimes and Women were more likely to use food labels than men.
Eating home-
28·7 % said rarely. Females were significantly more likely than males to use labels eating 50 % or more home-
often (P < 0·001). For the combined (college and community) sample, participants cooked meals were more likely to
21
report looking at food labels often compared with those who had <50 % home- cooked meals and less fast food were related to greater food
cooked meals (P < 0·001), and those who ate fast food more than once (v. 0 or 1 label use while weight status was not
meal) in the past week were more likely to report rarely looking at food labels (P <
0·001). Weight status did not influence food label use
Table 3 Continued
Author(s), year
Li et al.(48), 2012
label usage, in
Lim et al.(57), 2015
Marietta et al.(49), 1999
were more
Martinez et al.(50), 2012
McLean-Meyinsse et al.(51),
2011
Misra(52), 2007
21
42
Main finding Conclusion
On a 5-point scale (where 1 = ‘never’,5 = ‘always’), the average label usage score was 2·9 Females were more likely to use labels than males;
(SD 1·3); 14·6 % of students reported always using labels; 21·7 % reported often, turn, was related to more healthful eating habits
20·5% reported sometimes, 26·8 % reported rarely and 16·4 % reported
never.* Females exhibited a significantly higher usage score than males.
Regression analysis further revealed that only gender was significantly associated
with label usage (females more likely to use than males; P = 0·002); ethnicity,
marital status, class year and BMI were not associated with usage
Label users (37·8 %) and non-users (62·2 %) did not differ in age, class year, Label users had more favourable beliefs about using labels and in
height or weight. Label users had higher positive behavioural beliefs and beliefs in the immediate benefits of reading labels; however, beliefs about
the benefits of label usage than non-users, and lower scores for negative long- term benefits (e.g. disease prevention) did not differ
behavioural beliefs (e.g. ‘using labels is annoying’). Label users also had between users and non-users. Label users had higher
significantly higher normative belief scores, indicating these individuals perceived perceived control and did not perceive as many constraints
more influence from significant others regarding the use of nutrition labels relative to using labels. Nutrition education efforts should focus on
to non-users. Finally, label users reported greater control over constraints such increasing perceived control over label use
as time, knowledge, impulsive eating, food expense and preference for specific
foods. Non-users had significantly lower confidence in understanding information
such as nutrients, % daily value, nutrient content
per serving size, and serving size
20·2 % of students said they always looked at labels when purchasing a food product for Females and those with higher knowledge and attitudes
the first time; 23·6 % said often, 26·4 % said sometimes and 29·8 % said rarely or likely to use labels. More label education efforts are
never. Females looked at labels significantly more often than men and were more recommended to increase label usage and improve food
likely to use labels to fit a food into their daily diet. Regression analysis revealed choice
knowledge, attitude, age, gender and previous experience with nutrition labels
explained 14 % of the variation in label use, with attitude (r2 = 0·08, P < 0·001)
and gender (r2 = 0·04, P < 0·01) being significant independent predictors.
Knowledge (r = 0·87, P < 0·0001) and attitude (r = 0·27, P < 0·0001) were
correlated with label use
19·5 % of students said they always look at food labels on packaged foods; 28·4 Most students read labels at least sometimes, and nearly half
% said often, 26·1 % said sometimes, 20·8 % said rarely and 5·1 % said never.* reported selecting a lower-calorie and/or healthier option
Most students (88 %) reported nutrition information sometimes, often or always because of the labels. Females and overweight and obese
affects their food choices. Women were more likely to look at information online, students were more likely to report that labels influenced them
looked more frequently at posted information in dining halls and reported that to choose lower- calorie/healthier options
dining hall labels influenced them to choose healthier options. Overweight/obese
individuals were more likely to report that labels influenced them to make lower-
calorie and/or healthier food choices in dining halls relative to normal-weight
individuals
31·3 % of students said they read food labels often, 28·6 % sometimes, 16·8 % Upperclassmen used labels more frequently
rarely, 11·8 % never and 11·6 % did not comment. Juniors and seniors were more
likely to use labels sometimes or often than freshmen and sophomores
On a 5-point scale (where 1 = ‘never’, 5 = ‘always’), the average label usage education, knowledge and label
was 3·28 (SD 1·12). Undergraduates had higher nutrition knowledge and more reading behaviour
positive attitudes towards labels than graduate students, but were equally likely to
use labels. Two-thirds used Nutrition Facts when purchasing a food for the first
time, comparing products and purchasing foods with health claims. Regression
analysis showed that being female, older, having a positive attitude towards food
labels, using nutritional supplements and previous nutrition education were related
to a higher label reading behaviour score and accounted for 44 % of the variance
in label reading behaviour. Attitude mediated the relationship between nutrition
Most students used labels, had positive attitudes towards labels and thought they were useful, accurate and older age and being female were significantly and positively
understandable; however, a third could not perform label comparisons. Undergraduates were more associated with label usage
knowledgeable and positive towards labels than graduate students; positive attitudes, prior nutrition education,

MJ
Ch
rist
op
h
et
Table 3 Continued
Author(s), year
Rasberry et al.(53), 2007
Smith et al.(56), 2000
were
Wie and Giebler(54), 2014
*Percentage breakdowns for each frequency category were not provided in the original manuscript. These numbers were provided through personal communication with the corresponding author.
Cor
rel
ate
Main finding Conclusion
s
15·2 % of students said they always look at nutrition facts panels when purchasing of
foods; 27·2 % said often, 43 % said sometimes and 14·6 % said never. Label Females were more likely to use labels than males. Frequent
users exhibited greater knowledge of nutrition, label use and diet–disease label use was related to health reasons, looking for specific coll
relationships and more favourable attitudes towards label use than non-users. nutrition information, weight control, favourable attitudes, and
Females also exhibited significantly higher label usage than males. Logistic knowledge of nutrition and diet–disease relationships. Infrequent eg
regression results revealed that frequent label use was predicted by: health use was related to desiring specific foods, time constraints e
reasons (e.g. diabetes control, balancing diet, being healthier), looking for and not caring about nutrition information stu
specific nutrient information (e.g. calories, fat, carbohydrates), weight control
and knowledge. Three variables predicted infrequent use: buying the foods one de
wanted regardless of nutrition content, time constraints and didn’t care nts
29·5 % of students said they use the nutrition information panel on food labels every single Females use food labels more often than males. Label users ’
lab
time or more than half the time; 12·3 % said about half the time and 56·4 % more likely to believe in the importance of nutrition information
said less than half the time or none of the time. Females were significantly on food labels than non-users
more likely to use labels than males. Both male and female nutrition label users
were more likely to believe in the importance of nutrition information on labels. For
females, usage was related only to believing in the importance of nutrition
information on food labels when making food choices; for males, believing in the
truthfulness of the nutrition information panel and of a relationship between
dietary fibre and cancer were significantly higher in users relative to non-users
14·3 % of students said they always consider the calorie content of menu items Females were more likely to change their perceptions and
at restaurants before making their choices; 26 % said frequently, 25·7 % said behaviour following food label exposure. Nutrition education
sometimes, 20·2 % said rarely and 12·9 % said never. Females were more likely to may improve attitudes towards nutrition labels, although more
report perception and behaviour change following label exposure. Nutrition majors or educational efforts may be needed to turn perceptions into
non-nutrition majors who were enrolled in a nutrition class exhibited more positive behavioural change
perceptions of calorie counts on menus than non-nutrition majors who had not
taken a nutrition course
21
43
 Correlates of college students’ label use 214
self-efficacy was also documented to be positively reported understanding of nutrition were associated
(43)

associated with nutrition label use(42,57). with greater usage. Rasberry et al.(53) documented
Several studies found that nutrition knowledge nutri- tion label use to be related to improved knowledge
and education predicted nutrition label use. Driskell linking diets to certain diseases. Wie and Giebler (54)
et al.(44) found that education on nutrition labels, but not reported college students majoring in nutrition were
nutrition knowledge, was related to higher usage. more likely to use nutrition labels compared with their
Others found that nutrition education(52), nutrition counterparts with other majors.
knowledge(53,55) and self- A few studies assessed nutrition label use in relation
to behaviours; Conklin et al.(43) reported that
students obtaining information on food from
Table 4 Estimated label usage prevalence, by study and in weight-loss pro- grammes were more likely to use
aggregate
labels. Performing healthy dietary behaviours(45),
increased grocery shop- ping(46), eating more meals at
Label usage prevalence home and not eating fast
St T Alwa Som Rare food in the past week(47),
ud ot ys/of etim ly/ne
y al ten es ver and use of nutritional
le supple-
ad
N (n) (n) (n) ments(52) were also
auth positively related to label
or* usage.
Conklin(43) 1317 286 746 285
Cooke(55)† 524 333 122 69
Driskell(44) 205 14 96 95

Graham(45)† 1193 418 442 333 Study quality

Jasti(46) 222 54 136 32 Table 5 reports
Krukowski (47) the316
overall study
153 quality assessment
72results 91 as well as the results
Li(48)† 488 177
for each of the sixteen
100 211
studies included in the
Marietta(49) 208 review. On average,
55 62 studies scored 6·2 out of
Martinez(50)† 486 233 8 points (range: 3–8),
127 126 but the distribution of
qualifica-
Mc 390 tions differed
Le substantially across
an criteria. The large
- 138

126

126
(51)
Meyinsse usage was measured as a
dichotomous yes/no question
rather than usage frequency.
Rasberry(53) 1285 †Percentage breakdowns for each
545 frequency category were not
553 provided in the original manuscript.
187 These numbers were provided
through personal communication
Smith(56) 543 with the corresponding author.
163
68
312

Wie(54) 341
142
116
83

All studies 7518 2747

– 36·5%

*Three of the sixteen studies

could not be included in the
review. Cha et al.(42) and Misra(52)
did not report the breakdowns
of their 5-point frequency scales.
Both authors were contacted by
email but were unable to provide
this information by the deadline
given. Lim et al.(57) was not included
in this calculation because their
 Correlates of college students’ label use 214
majority (94 %) of %) documented the *Graham and Laska(45) do use regression analysis where label usage is
modelled as a function of attitude towards healthy eating, gender,
studies clearly study response or age and race/ ethnicity; however, only the regression coefficient for
documented the com- pletion rate and attitude is reported in the paper.

study design and data twelve (75 %) adopted correlate. Only half of the nutrition knowledge and
collection procedures; a previously vali- studies used regression education. In contrast,
however, sample size dated or pilot-tested analysis to examine the findings on nutrition label
was much more measure on nutrition relationship between use in relation to age,
variable, ranging label use. Almost all label usage and multiple student classification,
from 103 to 1317. studies (94 %) predictor race/ethnicity, marital
Only nine of the reported usage in variables/correlates status and body weight
sixteen studies (56 %) relation to at least one simultaneously. Finally, status were largely
met the sample size demographic correlate most of the studies (88 inconclusive.
quality criterion of and 81 % of studies %) examined free-living Our results are
400 or more reported usage in college students and consistent with studies
participants. relation to at least one that examined
young adults in general
Thirteen studies (81 non-demographic label use in the general
(no popu- lation
population. Guthrie et al.
restrictions). (58)
and Campos et al.(3)
Table 5 Study quality assessment for each study included reported prevalence
in the review and average quality across all studies
estimates of 71 % and 75
Discussion
Documen Rep Cons % in US populations
ted ort Val Con ider (label usage is defined as
ed ida sider ed U N The present study
using information at least
andSdatares ted ed non- se on To systematically reviewed
collectio
a pon or dem dem d - tal sometimes), similar to our
existing evidence on the
m se pil ogra ogra reg res prevalence estimate of
pl or ot- phic phic res tric correlates of nutrition
73·8 % for young adults
e co tes corre corr sio ted label use among college
mpl te lates elat n who use labels at least
students and young
etio d of es of sometimes. Additionally,
n adults 18–30 years of
Cowburn and Stockley(6)
age. A total of sixteen
and Campos et al.(3)
Stud p s r s u u a populati studies based on college
y r i a u s s n on conducted reviews
surveys in four countries
lead o z t r a a a points examining predictors/
auth c e e v g g l (USA, UK, Canada and
correlates of label use in
or e e e e y South Korea) were
d y si the general population and
identified from keyword
u s found many of the same
r and reference search.
relationships we
e Reported prevalence of
s
identified in the present
nutrition label use
review. Similarly to our
varied substantially
1 0 1 results, women were more
Cha(42) (43) across studies; however,
Conklin 1 1 1 likely to use labels than
Cooke(55) 1 1 1 a weighted average
men, as were individuals
Driskell(44) 1 0 1 across all studies
Graham(45)* 1 1 0 with high nutrition
revealed 36·5 % of
Jasti(46) 1 0 1 knowledge or nutrition
college students and
Krukowski(47) 1 0 1 education, positive
1 1 1 1 young 1adults reported 0 1
attitudes 1
towards diet7 and
L 1 0 1 1 using labels
1 always1 or 0 0 5
i health, or who practised
( often (36·7 % said
4 healthy eating habits and
8 sometimes, 26·8 % said
) dietary behaviours. Both
rarely or never). Twelve
reviews also found that
L of the thirteen studies
label use was related to
i that assessed gender
general education level
m differences reported that
and income; however,
(
females were more
5
7 due to our restricted
) likely to use nutrition
Marietta(49) 1 0 0 population, we were
labels. Nutrition label
Martinez(50) 1 1 1 unable to assess the
McLean-Meyinsse(51) 0 1 0 use was also found to be
relationship between these
Misra(52) 1 1 1 associated with attitudes
Rasberry(53) 1 1 1 variables and label usage.
towards a healthy diet,
Smith(56) 1 1 1 Surprisingly, the reviews
belief on the importance
Wie(54) 1 0 1 on the general
1 1 0 1 of nutrition labels in
population did not
guiding food selection,
Average of all studies 0·94 0·56 0·81 report the relationship
self-efficacy, and
0·94 0·81 0·50 0·88 between BMI or weight
Correlates of college students’ label use 214
status and label usage, quality in college rely on self-reported
which may not reflect
which was a common students and young data; label use may have
actual use. Although a
variable of interest for adults(42,45,46,48,55). Four been self-reported, but
majority of US
many of the studies of the five studies found diet quality and food
consumers report
included in our that label use led to choice were often
regular use of nutrition
review. improved dietary directly observed.
facts labels, in-store
It is important to quality(42,45), lower Generally speaking, this
observations suggest
note that in many of consumption of fried body of research
actual use during food
the studies reviewed foods(46), decreased fat provided a less
pur- chase can be
(both in young adult intake and increased optimistic view on the
lower(59). Moreover,
and general adult fibre intake(48). Inter- ability of labels to
whether consumers can
popula- tions) label use estingly, Cooke and impact dietary quality.
understand and use Papadaki(55) found that Systematic reviews in
is typically based on
nutrition facts label is label use was adults suggested that
self-reported data,
contingent upon the negatively related to simply posting calorie
purpose of the task(4– dietary quality when information may not
7)
. Regular label users nutrition knowledge impact calorie
can understand some and attitudes were purchases or
of the terms but may controlled for. Looking consumption(8,60); thus,
be confused by other beyond standard the link between label
types of information. A nutrition labels on use and dietary
majority appears packaged foods, there behaviour remains
capable of retrieving were also several unclear.
basic facts and making studies that had A few limitations are
simple calculations/ examined the link present in the existing
comparisons between between calorie label literature on the
products using use in restaurants and prevalence and
numerical informa- tion food choice. While this correlates of nutrition
on the label, but their portion of the literature label use in college
ability and accuracy considered a different students and young
decline as the meal setting (eating adults. There remains
complexity of the away from home instead no standardized
task increases. of eating at home), it had instrument to assess
Further, while it is the advantage of nutrition label use, and
critical to understand examining the often there is a lack of
which types of relationship between distinction between the
individuals are more label use and food nutrition facts labels on
likely to use nutrition choice in more natural food packages and
labels, the question settings (actual nutrition/calorie labels on
remains whether label restaurants). In contrast food venue (e.g. fast-food
use actually leads to to the majority of studies outlet, full-service
improved dietary included in Campos et restaurant, cafeteria or
behaviour. In their al.’s(3) review, these dining hall) menus and
review of the general studies did not solely
between nutrition label
population, Campos et
al.(3) discussed several formats (e.g. front-of-pack, adults have not been
studies that found an traffic light label, front-of- assessed in any of the
association between pack studies included in the
nutrition label use and + traffic light label)(49,50,56). review, such as health
diet. Some studies While many studies and/or risk behaviours (e.g.
found label users to examined two or more of smoking,* drinking, drugs
have healthier diets the correlates of use), mental and/or
overall while others nutrition label use, very physical health, and
found label users had few studies assessed a neigh- bourhood or campus
lower intake levels of comprehensive list of food environment. In
certain nutrients (e.g. psychosocial factors that addition, only three of the
fat, cholesterol) than enabled within-study included studies were
non-users. In the comparison. Many published outside the
present review, five potentially important USA, and even within the
studies examined the correlates of nutrition included thirteen US-
relationship between label use among college based studies, geographic
label use and dietary students and young regions such as the
 Correlates of college students’ label use 214
Southwest, West and among college students 15, 385–399.
analysis and wrote the first
Northwest were under- and young adults; 8. Kiszko K, Martinez O,
draft; R.A. and B.E. oversaw
represented. Lastly, our however, they should be Abrams C et al. (2014)
the analysis and edited all The influence of
review only observed rigorously evaluated to
drafts. Ethics of human calorie labeling on food
correlates of reported determine whether label orders and
subject participation: Not
use; whether and how use leads to consumption: a review
applicable.
usage relates to dietary improvements in diet of the literature. J Community
intake must be further quality. Health 39, 1248–1269.
assessed, particularly 9. Sinclair SE, Cooper M
comparing the effect of References
Acknowledgements & Mansfield ED (2014)
different labelling The influence of menu
1. US Department of labeling on calories
schemes, comparing and Financial support: This Agriculture & US selected or
combining label inter- research received no specific Department of Health consumed: a sys-
ventions with other and Human tematic review and
grant from any funding agency Services (2010)
nutrition interventions, Dietary Guidelines for
meta-analysis. J Acad
in the public, commercial or not- Nutr Diet 114, 1375–
and surveying large Americans, 2010, 7th
for-profit sectors. Conflict of 1388.
samples over time. ed. Washington, DC:
10. Satia JA, Galanko JA &
interest: none. Authorship: US Government
Neuhouser ML (2005)
Printing Office.
M.J.C. formulated the Food nutrition label
Conclusion 2. Krebs-Smith SM,
use is associated with
research question, Guenther PM,
Subar AF et al. demographic,
The present study performed the behavioral, and
(2010) Americans do
reviewed correlates of not meet federal psychosocial factors
* ‘Cigarette’ was used as an exclusion and dietary intake
nutrition label use term since initial searches included dietary
recommendations. among African
among college students numerous articles on cigarette labels as
Americans in North
opposed to food labels. However, this J Nutr 140, 1832–1838.
and young adults. Carolina. J Am Diet
should not have precluded the included
Reported prevalence of studies from surveying stu- dents on 3. Campos S, Doxey J Assoc 105, 392–402.
nutrition label use smoking habits. In fact, one of the & Hammond D 11. Govindasamy R & Italia J
included studies did survey students on (2011) Nutrition (2000) The influence of
varied substantially
smoking habits, but these were not labels on pre- consumer
across studies, but a reported in relation to usage. demographic characteristics on
packaged foods: a
weighted average reveals systematic review. nutritional label usage.
that 36·5% of college Public Health Nutr 14,
students and young 1496–1506. J Food Prod Mark 5, 55–68.
adults said they use 4. Drichoutis A,
Lazaridis P & Nayga 12. Nelson MC, Story M,
nutrition labels always or Larson NI et al.
RM Jr (2006)
often. Female gender, Consumers’ use of (2008) Emerging
attitudes towards healthy nutritional labels: a adulthood and
diet, beliefs on the review of research college-aged youth:
studies and issues. an overlooked age for
importance of nutrition weight-related
Acad Mark Sci Rev
labels in guiding food 2006, 9. behavior change.
selection, self-efficacy, 5. Mhurchu CN & Obesity (Silver Spring) 16,
and nutrition knowledge Gorton D (2007) 2205–2211.
Nutrition labels and 13. Vella-Zarb RA & Elgar
and education were FJ (2009) The
found to predict nutrition claims in New
‘Freshman 5’: a meta-
Zealand and
label use. While Australia: a review
analysis of weight gain
providing nutrition in the freshman year of
of use and
college. J Am Coll Health
information at the point understanding. Aust
58, 161–166.
of purchase may nudge N Z J Public Health 31,
14. Harris K, Gordon-
consumers towards a 105–112.
Larsen P, Chantala
6. Cowburn G &
healthier diet(61), findings K et al. (2006)
Stockley L (2005)
from the present review Longitudinal trends
Consumer
in race/ethnic
indicate the potential understanding and
disparities in leading
heterogeneity in the use of nutrition
labelling: a health indicators
impact of nutrition systematic review. from adolescence to
labelling across Public Health Nutr 8, young adulthood.
population subgroups. 21–28. Arch Pediatr Adolesc
7. Grunert K & Wills J Med 160, 74–81.
Nutrition aware- ness 15. Mikkilä V, Räsänen L,
campaigns and (2007) A review of
Raitakari OT et al.
European research
education programmes (2007) Major dietary
on consumer
may be important patterns and
response to nutrition
cardiovascular risk
mechanisms to promote information on food factors from
nutrition label use labels. J Public Health
 Correlates of college students’ label use 214
childhood to adulthood. The Risk Develop- ment & Yeh M (2012)
Cardiovascular Risk in in Young Adults) Effectiveness of 37. Turner MM, Skubisz
study. Br J Nutr 112, exercise equivalents C, Pandya SP et al.
Young Finns Study. Br J
1654–1661. and calorie (2014) Predicting
Nutr 98, 218–225. visual attention to
17. Roy R, Kelly B, information on
Rangan A et al. vending machine nutrition information
16. Steffen LM, Van on food products: the
Horn L, Daviglus (2015) Food usage among college
environment students. J Acad Nutr influence of
ML et al. (2014) A motivation and
modified interventions to Diet 112, A64.
31. Roberto CA, Haynos ability. J Health
Mediterranean diet improve the dietary
AF, Schwartz MB et al. Commun 19, 1017–
score is associated behavior of young
(2013) Calorie 1029.
with a lower risk of adults in tertiary
estimation accuracy and 38. Visschers VH, Hess R
incident metabolic education settings: a & Siegrist M (2010)
syndrome over 25 systematic literature menu labeling
perceptions among Health motivation and
years among young review. J Acad Nutr product design
adults: the CARDIA Diet 115, 1647– individuals with and
without binge eating determine consumers’
(Coronary Artery 1681. visual attention to
and/or purging
18. Zhu H & An R Adolesc Health 44, 84– disorders. Eat Weight nutrition information
(2013) Impact of 86. Disord 18, 255–261. on food products.
home-delivered 24. Girz L, Polivy J, 32. Schuldt JP (2013) Does Public Health Nutr 13,
meal programs on Herman CP et al. green mean healthy? 1099–1106.
diet and nutrition (2012) The effects of Nutrition label color 39. Walters A & Long M
among older adults: calorie information on affects perceptions of (2012) The effect of
a review. Nutr Health food selection and healthfulness. Health food label cues on
22, 89–103. intake. Int J Obes (Lond) Commun 28, 814–821. perceptions of
19. National Institutes of 36, 1340–1345. 33. Temple JL, Johnson quality and purchase
Health (2014) Quality 25. Hieke S & Wilczynski KM, Archer K et al. intentions among high-
Assessment Tool for P (2012) Colour Me In (2011) Influence of involvement consumers
Observational Cohort – an empirical study simplified nutrition with varying levels of
and Cross-Sectional on consumer responses labeling and taxation nutrition knowledge.
Studies. https:// to the traffic light on laboratory energy J Nutr Educ Behav 44,
www.nhlbi.nih.gov/heal signposting system in intake in adults. 350–354.
nutrition labelling. Appetite 57, 184–192.
th-pro/guidelines/in- 40. van der Merwe D,
Public Health Nutr 15, 34. Hoefkens C, Pieniak
develop/cardio vascular- Kempen EL, Breedt S
773–782. Z, Van Camp J et al.
risk-reduction/tools/cohort et al. (2010) Food
26. Koenigstorfer J, (2012) Explaining the
(accessed March choice: student
Groeppel-Klein A, effects of a point-of-
2015). Kettenbaum M et al. consumers’ decision-
purchase nutrition-
20. Draper AK, Adamson (2013) Eat fit. Get making process
information
AJ, Clegg S et al. big? How fitness cues regarding food
intervention in
(2013) Front-of-pack influence food products with limited
university canteens: a
nutrition labelling: consumption label information. Int J
structural equation
are multiple formats volumes. Appetite 65, Consum Stud 34, 11–
modelling analysis.
a problem for 165–169. 18.
Int J Behav Nutr Phys Act
consumers? Eur J 27. Levi A, Chan KK & 9, 111. 41. Shriver CB & Scott-
Public Health 23, 517– Pence D (2006) Real Stiles A (2000) Health
35. Sharf M, Sela R,
521. men do not read Zentner G et al. (2012) habits of nursing
21. Ellison B, Lusk JL & labels: the effects of Figuring out food versus non-nursing
Davis D (2013) masculinity and labels. Young adults’ students: a
Looking at the label involvement on understanding of longitudinal study. J
and beyond: the college students’ nutritional informa- Nurs Educ 39, 308–314.
effects of calorie food decisions. J Am tion presented on 42. Cha E, Kim KH,
labels, health Coll Health 55, 91–98. food labels is Lerner HM et al.
consciousness, and 28. Mayfield K, Tang L & (2014) Health
inadequate. Appetite
demographics on Bosselman R (2014) literacy, self-efficacy,
58, 531–534.
caloric intake in Nutrition labeling for food label use, and
36. Turconi G, Bazzano
restaurants. Int J restaurant menu items: R, Roggi C et al. diet in young adults.
Behav Nutr Phys Act 10, college students’ (2012) Helping Am J Health Behav 38,
21. preferences for consumers make a 331–339.
22. Papakonstantinou E, nutrition information 43. Conklin MT, Cranage
more conscious
Hargrove JL, Huang and its influence on DA & Lambert C
nutritional choice:
CL et al. (2002) purchase intention. J (2005) College
Qual Assur Hospit acceptability of
Assessment of nutrition information students’ use of point
perceptions of Tourism 15, 310–325. of selection nutrition
29. McClain AD, Hekler EB at a cafeteria. Public
nutrition knowledge Health Nutr 15, 792– information. Top Clin
and disease using a & Gardner CD (2013) Nutr 20, 97–108.
Incorporating 801.
group interactive 44. Driskell JA, Schake MC
system: the Perception prototyping and & Detter HA (2008)
Analyzer (R). J Am iteration into Using nutrition
Diet Assoc 102, 1663– intervention labeling as a potential tool for
1668. development: a case changing eating habits of
23. Gerend MA (2009) study of a dining university dining hall
Does calorie hall-based
patrons. J Am Diet Assoc
information promote intervention. J Am
108, 2071–2076.
lower calorie fast Coll Health 61, 122–
food choices among 131.
30. Platkin C, Kelvin EA 45. Graham DJ & Laska
college students? J
Correlates of college students’ label use 214
MN (2012) Nutrition Welshimer KJ & Educ 38, 76–82. A (2014) Nutrition
label use partially Anderson SL (1999) 54. Wie S & Giebler K label use
mediates the Knowledge, (2014) College mediates the positive relationship
relationship between attitudes, and students’ perceptions between nutrition
attitude toward behaviors of college and knowledge and attitudes
healthy eating and students regarding behaviors toward calorie counts towards healthy eating with
overall dietary the 1990 Nutrition on menu. J Foodserv Bus dietary quality among
quality among Labeling Education Res 17, 56–65. university students in the UK.
college students. J Act food labels. J Am Appetite 83, 297–303.
Acad Nutr Diet 112, Diet Assoc 99, 445–
414–418. 449.

55. Cooke R & Papadaki

56. Smith SC, Taylor JG &

46. Jasti S & Kovacs S 50. Martinez OD, Roberto Stephen AM (2000) 59. Azman N & Sahak S
(2010) Use of trans CA, Kim JH et al. Use of food labels and (2014) Nutritional
fat information on (2012) A survey of beliefs about diet– label and consumer
food labels and its undergraduate student disease relationships buying decision: a
determinants in a perceptions and use of among university preliminary review.
multiethnic college nutrition information students. Public Health Procedia Soc Behav Sci
student labels in a university Nutr 3, 175–182. 130, 490–498.
population. J Nutr dining hall. Health Educ J 57. Lim HJ, Kim MJ & Kim 60. Swartz J, Braxton D &
Educ Behav 42, 307– 72, 319–325. KW (2015) Factors Viera A (2011) Calorie
314. associated with menu
47. Krukowski RA, 51. McLean-Meyinsse nutrition label use labeling on quick-service
Harvey-Berino J, PE, Gager JV & Cole among female college restaurant menus: an
Kolodinsky J et al. DN (2011) students applying the updated systematic review
(2006) Consumers Examining the theory of planned of the literature. Int J
may not use or prevalence of food- behavior. Nutr Res Pract Behav Nutr Phys Act 8,
understand calorie label use by 9, 63–70. 135.
labeling in university students. 58. Guthrie JF, Fox JJ, 61. Cioffi CE, Levitsky
restaurants. J Am J Food Distrib Res 42, Cleveland LE et al. DA, Pacanowski CR et
Diet Assoc 106, 917– (1995) Who uses al. (2015) A nudge in
84–89.
920. nutrition labeling, and a healthy direction.
52. Misra R (2007) what effects does label
48. Li K-K, Concepcion Knowledge, attitudes, The effect of nutrition
RY, Lee H et al. use have on diet labels on food
and label use among quality? J Nutr Educ 27,
(2012) An purchasing behaviors
college students. J 163–172.
examination of sex in university dining
differences in Am Diet Assoc 107,
facilities. Appetite 92,
relation to the 2130–2134.
7–14.
eating habits and 53. Rasberry CN,
Chaney BH, Deliens et al. BMC Public Health
nutrient intakes of
university Housman JM et al. (2015) 15:201
students. J Nutr (2007) Determinants DOI 10.1186/s12889-015-1553-4
Educ Behav 44, 246– of nutrition label use
250. among college
49. Marietta AB, students. Am J Health
 Correlates of college students’ label use 215
Determin qualitativ that university students
spent eight hours per
an important influence on
stu- dents’ weight and
ants of e study day on sedentary activities overall health [1,3]. A
physical using such as studying, watching great body of literature
television, gaming, points out that higher
activity focus computer activities, physical activity levels
and group sitting and talking, shop- are associated with lower
ping and hanging out [5]. health risks (incl.
sedentar discussio Physical activity overweight and obesity
y ns (including active related diseases) [6,7].
behavio Tom Deliens1*,
transportation)
sedentary behaviour have
and There is also growing

ur in Benedicte Deforche1,2,
evidence that excessive participation in sedentary behav-
universit Ilse De Bourdeaudhuij2 * Correspondence:
iours, such as watching
Tom.Deliens@vub.ac.be
y and Peter Clarys1 television, computer use
students
1
Department of Human Biometry and and sit- ting for
Biomechanics, Vrije Universiteit
work/study purposes, is
:a Brussel, Pleinlaan 2, 1050 Brussels,
Belgium associated with higher
Full list of author information is
Background every day [4]. available at the end of the article
The transition from Concerning sedentary
secondary school to behaviour, a UK study © 2015 Deliens et al.; licensee
BioMed Central. This is an Open
university is often revealed Access article distributed under
accompanied by unhealthy the terms of the Creative
Commons Attribution License
behaviour changes such as (http://creativecommons.org/licens
decreasing physical es/by/4.0), which permits
unrestricted use, distribution, and
activity and increasing reproduction in any medium,
sedentary behaviour provided the original work is
properly credited. The Creative
[1,2]. According to Commons Public Domain
Keating’s review [3], 40- Dedication waiver
(http://creativecommons.org/publicdomain/zer
50% of college students o/1.0/) applies to the data made
are physically inactive. A available in this article, unless
otherwise stated.
more recent study in
Czech university
students reported that
only 9% met the

nd university students’ physical activity and sedentary behaviour seemed to be moderated by university characteristics, su

criterion of 10,000 steps

Deliens et al. BMC Public Health (2015) Page 2 of
15:201

kind of strategy stu- dents are prepared/willing to

risk of obesity, independent of diet and physical participate. A review on
activity behaviour [8-11]. Despite the fact that internet
use as a specific sedentary behaviour may induce
mental health benefits due to its use for social
connection and support [10,12], research shows that
higher levels of sedentary behaviour are associated
with indicators of poorer well- being, increased risk of
depression, and weaker cognitive functioning [10,12-
14]. It is important to investigate physical activity and
sedentary behaviour as two distinct modes of
behaviour influencing weight and health inde-
pendently [15]. Although Rouse et al. [5] concluded
that physical activity and sedentary behaviours in
university students appear largely uncorrelated, other
studies showed a negative correlation between
sedentary behav- iours and physical activity in college
students [16,17]. In order to prevent weight and fat gains
in students, interven- tions should aim to increase
physical activity and decrease sedentary behaviours,
whether or not independently from each other [9].
Understanding why young people (do not) engage
in
physical activity and sedentary behaviour is
important for intervention efforts encouraging more
active life- styles [18]. According to Swinburn et al. [19]
individuals interact in a variety of micro-
environments or settings (e.g. schools, workplaces,
homes, (fast food) restaurants) which, in turn, are
influenced by the macro-environments or sectors (e.g.
food industry, government, society’s atti- tudes and
beliefs). Ecological models consider the connec- tions
and the continuous interactions between people
(intrapersonal) and their (sociocultural, policy and phys-
ical) environments [20-23].
In the literature, demographics (e.g. age, gender), psy-
chological factors (e.g. self-efficacy, perceived enjoy-
ment), social factors (social support from family
and friends), and physical environmental factors (e.g.
living/ built environment, access to facilities) were
reported to be possible influencing factors of
college/university stu- dents’ physical activity
(including active transportation) behaviour [3,24-26].
According to Haase et al. [24], physical activity is
also related to some macro environ- mental factors,
such as cultural factors and stage of na- tional
economic development. Although some studies
investigated determinants of physical activity in
univer- sity or college students, there is a lack of
information on how to change these determinants
and how to increase physical activity in this specific
population. Asking par- ticipants what strategies may
be effective to increase their physical activity may
contribute to develop tailored and effective
intervention programs. Moreover, the im-
plementation of interventions based on ideas from
stu- dents themselves may be more feasible and
sustainable on a university campus. A participatory
approach as such gives us an idea if and in what
Deliens et al. BMC Public Health (2015) Page 3 of
15:201
factors affecting program implementation concluded
that ‘shared decision-making’ (i.e. community
involve- ment or participation) not only consistently
led to better implementation, but also to higher
sustainability of the program [27]. Thus,
intervention programs taking the target group’s
opinions into account may be more likely to
succeed in the long run. Furthermore, there are
no studies investigating determinants of sedentary
behaviour in university or college students and no
studies investigating possible intervention
strategies. Previous studies investigating factors
related to seden- tary behaviours in young people
were limited to TV viewing [28]. There is a need to
examine determinants and correlates of total
sedentary behaviour rather than TV viewing alone
[28]. In addition, information on how to decrease
sedentary behaviours in university students is lacking.
In summary, research in university students is
still needed to better understand energy expenditure
behav- iours in order to develop effective strategies
aiming to in- crease physical activity and decrease
sedentary time [10]. Therefore, this study used a
qualitative research design to explore which factors
influence Belgian university stu- dents’ physical
activity (including active transportation) and total
sedentary behaviour (including screen behaviour, school
work, socialising, and passive transportation). Fur-
thermore, we collected ideas and recommendations
in order to facilitate the development of tailored
intervention programs aiming to increase physical
activity and decrease sedentary behaviours in
university students.

Methods
Participants
In this qualitative study focus group discussions
were used for data collection. To ensure sufficient
diversity of opinion, students from the second
through the fifth year of university from different
study disciplines were re- cruited using snowball
sampling. Possible participants were approached
face-to-face, by telephone or by email. No first year
students were included because of their ‘limited’
experience as a university student. The aim was to
recruit between six and ten participants per focus
group [29].

Procedure

Focus groups were held until saturation of new

informa- tion was reached, as in qualitative research
sample size can never be pre-determined [29]. To be
sure we did not miss any ‘new’ information, one
additional focus group session was held after
theoretical saturation was esti- mated. All focus
groups were organised in a conference room with an
oval table at the Faculty of Physical Educa- tion and
Physiotherapy of the Vrije Universiteit Brussel
(Brussels, Belgium) at a time and date convenient
for the
students and researchers. Before each focus group
all participants were asked to complete a short
question- naire, including demographics, height,
weight and per- ceived health. Each focus group was
facilitated by a moderator and an assistant
moderator (observer). The moderator (male PhD
student and first author of this manuscript) was
trained and prepared through partici- pating in
workshops organised and guided by experi- enced
focus group researchers. The assistant moderators
(Master students, who were trained by the
moderator) took notes during the discussions and
made sure the moderator did not overlook any
participants trying to add comments. Each focus group
discussion started with an introductory round in which
the researchers as well as the participants presented
themselves and in which the researchers explained
the purpose of the study. All focus group discussions
were audiotaped with permis- sion of the
participants. Drinks and snacks were pro- vided
during the focus group discussions. Afterwards, all
students received an incentive (a lunch voucher).
Ethics statement
Before each focus group explanation about the aim of
the study was given and an informed consent (in
which participants’ anonymity and confidentiality was
assured) was signed by each participant. The study
was approved by the Medical Ethical Committee of
the university hos- pital (Vrije Universiteit Brussel,
Brussels, Belgium). All procedures followed were in
accordance with the ethical standards of the
responsible committee on human ex- perimentation
(institutional and national) and with the Helsinki
Declaration of 1975, as revised in 2000.
Question guide
According to recommended focus group methodology
[30], a semi-structured question guide (see Table 1)
was
developed by the research team, aiming to identify
fac- tors influencing university students’ health and
weight related behaviours (including eating (and
drinking) be- haviour, physical activity and sedentary
behaviour). As mentioned, this paper will only focus
on determinants of students’ physical activity and
sedentary behaviour, as the results on determinants of
students’ eating behaviour have already been published
elsewhere [31]. After inten- sive collaboration with
experts with ample focus group experience, the
questions were carefully developed using appropriate
literature [30]. When development was com- pleted, the
question guide was tested within and revised by the
research team as well as pilot-tested in a group of
ten university students. Because no major changes
had to be made, ‘pilot’ discussion results were
included in later analysis [29]. After some
introductory questions, key questions focused on the
main purpose of this study,
i.e. identifying factors influencing students’ physical ac-
tivity and sedentary behaviour. Finally, students were
asked to share ideas concerning health promotion as
well as intervention strategies to increase physical activ-
ity and decrease sedentary behaviour in university
stu- dents. By using side questions, the moderator
could guide the focus group discussions in the right
direction and deviate the conversation when
distracted from the main issue, or when participants
were not talking about the appropriate behaviour.
Because the pilot focus group discussion revealed
that participants were not fully aware of the
meaning of sedentary behaviour, a hand-out power point
presentation explaining sedentary behaviour was given
to all participants before the start of each focus
group discussion. Sedentary behaviour was defined as
activities requiring low levels of energy expenditure
that occur while sitting or lying down [32]. In
addition, following examples of sedentary activities
were given: television watching, computer activities
(screen time),

Table 1 Focus group question guide

Question type Question
Opening 1. Where are you from and what’s your
name? Introduction 2. Describe a healthy person.
Transition 3. Thinking of ‘health in university students’, what comes to your mind?
4. Think back of the last year(s) being a university student. Did your body weight and/or body composition
change since you entered university?
5. Did your weight related behaviours (including eating, physical activity and sedentary behaviour) change since
you entered university?
Key 6. Which factors have caused these changes? Which factors influence current weight related behaviours (including eating, physical
activity and sedentary behaviour)? What barriers and enablers of weight related behaviour (including eating, physical activity
and sedentary behaviour) can you identify?
7.Which of the previous mentioned factors have had the greatest influence?
8. Soon, we will try to help students make healthier choices. Can you give us some advice on how to
promote healthy weight related behaviours (including eating, physical activity and sedentary behaviour) in
 students?
Ending 9. Do you have any remarks, suggestions, additions?
classes, studying, socialising, and passive Table 2 Characteristics of focus group participants
transportation. More detailed methodological (Mean ± SD, %, n = 46)
information can be found in Deliens et al. [31].
Gender (% of females) 63.0
Data analysis Age (yrs) 20.7 ± 1.6
Body mass index (BMI) (kg/m ) 2
22.4 ± 3.6
SPSS Statistics 20 was used to analyse data obtained
from the questionnaire and to calculate descriptive Underweight (%) 10.9
sta- tistics of the focus group sample. Data Normal weight (%) 67.4
obtained from the audio tapes where transcribed Overweight (%) 21.8
verbatim in Microsoft Word using Express Scribe and
Study career (yrs) 3.2 ± 1.0
Windows Media Player. All quotes were encoded
using the qualitative software program Nvivo9. Data Study discipline
were analysed using an inductive content analysis Human sciences (%) 67.4
approach. In a first step, data (quotes) were
Exact and applied sciences (%) 13.0
examined for recurrent instances of some kind, which
were then systematically identified across the data set, Biomedical sciences (%) 19.6
and grouped together by means of an open coding Residency (% living in student residence) 54.3
system (= content analysis) [33]. In a second and Smoking (% smokers) 8.7
third
step, themes were derived from the data, i.e. similar Self-reported health (% reporting poor to very 13.1
codes were grouped together into more general poor health status)
concepts (subcategories) and further categorised into Perceived physical activity level (% reporting little to 54.3
main cat- egories. This approach also allowed us to no physical activity)
identify mod- erating factors that influence the strength Perceived eating pattern quality (% reporting poor 19.5
of the relation between the determinant and the to very poor eating pattern)
independent variable [34]. More detailed information
about inductive content

analysis can be found elsewhere [33,35]. To ensure appropriate quotes were chosen to illustrate each
reli- ability of coding and data interpretations, analyses (sub)category.
were carried out independently by two researchers.
Doubts or disagreements were discussed with two other
researchers until consensus was reached.

Results
In this study, the estimated point of saturation was
ob- served after the sixth focus group session. One
add- itional focus group discussion was conducted to
be sure true saturation was established. In total,
seven focus group discussions were conducted
consisting of five to ten participants per group. The
sample (n = 46) con- sisted of 17 male and 29 female
students with a mean age of 20.7 ± 1.6 yrs (range =
18–26 yrs) and a mean study career of 3.2 ± 1.0 yrs.
Each focus group discussion lasted between 90 and 120
minutes (including questions about eating behaviour
which were not included in this paper). Additional
sample characteristics are described in Table 2.
According to the ecological principles two frameworks
of factors influencing physical activity (Figure 1) and
sedentary behaviour (Figure 2) in university students
were developed based on content analysis of the
focus group discussions. Both frameworks consist of
four major levels, i.e. individual (intrapersonal), social
envir- onment (interpersonal), physical environment
(commu- nity settings), macro environment, and an
additional level of university characteristics. The most
Suggestions for physical activity and sedentary
behaviour interventions
Students indicated that the university offers a lot of on- campus
facilities and sports lessons, but at the same time they felt they
were not well informed. Therefore, students suggested to
improve communication and pro- motion strategies: “I am
sure there are lots of activities on campus, but unless you
go ask and inform yourself, you don’t have a clue of
what kind of activities are being offered”. According to
participants, first year university students could be informed
during university’s open house days, or guided campus tours
during the first week of university. They also suggested to
promote all activities using media tools students are familiar with,
such as Facebook. One student came up with the idea to organise
a sports day (including initiation courses of all kinds of different
sports) for all university students, giv- ing them the possibility to
get to know the university’s sports program and to subscribe
themselves in one of the numerous sports activities. Students also
explained that cheaper and/or more flexible sports subscriptions
and formulas would lower the barriers of participation: “One
should offer cheaper sports activities and/or flexible
subscriptions for 10 lessons giving you the freedom of
choosing which lesson you want to participate in (e.g.
swimming, Zumba, …)”. Another student mentioned that such
sports lessons should be accessible for beginners: “if your
condition is in bad shape, you should be able to join
without any trouble”. Other students suggested that some
‘sports time’ could be incorporated as part of the
 Figure 1 Factors influencing physical activity behaviour of university students.

Figure 2 Factors influencing sedentary behaviour of university students.

 beneficial weight control or weight loss results [40].
curricula: “One could incorporate non-obligatory sports Hence, it
activities in students’ schedules. Also, it enables you to
get to know your class mates, so it improves social cohe-
sion as well”. With regard to active transportation, students
believed that providing ‘university bicycles’ (similar to the city
bicycles available in large cities all around the world) could
increase the use of bicycles for active transportation purposes:
“It would be great if you could subscribe into a bicycle rental
program allowing you to rent a bicycle for a certain
amount of time”. Con- cerning sedentary behaviour, students
believed that “when making sports activities more easy
accessible and more pleasant, students would spend less
time on the couch”. Participants also added that cultural
activities (like exhibitions and museums) should therefore be pro-
moted as well.

Discussion
The purpose of this explorative study was to identify
de- terminants of physical activity and sedentary
behaviour in Belgian university students. Furthermore
we collected ideas and recommendations in order to
facilitate the de- velopment of tailored intervention
programs aiming to increase physical activity and
decrease sedentary behav- iours in university students.
Similar to Story’s framework
[22] combining Bandura’s Social Cognitive Theory
[36] with Sallis’ ecological model [21] explaining
health be- haviour, we identified four major levels of
determinants: individual, social environment, physical
environment and macro environment. In turn,
these determinant levels were found to be
influenced by some university specific
characteristics.

Individual

Many psychological factors such as perceived

enjoyment, self-discipline, values, norms and beliefs,
and time man- agement were found to influence
physical activity and sedentary behaviour at the
same time. In Keating’s re- view [3] ‘having fun’ has
been addressed as one of the primary reasons for
college students to participate in physical activity or
enrol in elective physical activity courses. With
regard to time management, previous US studies
using focus group discussions revealed that stu-
dents feel like they lack time to be physically
active [37-39]. Students spend a lot of time on
study related sedentary activities (e.g. sitting in
class, studying, or sit- ting in front of their computer
for academic purposes), which makes it difficult to be
physically active [37,38]. To counter time
constraints, participants in the present study
suggested to incorporate ‘sports time’ as part of
their curriculum. With regard to time spent seated
in classes, previous research has shown that taking a
five minute walking break every hour could yield
should be the task of university policy makers to
inte- grate sufficient break-time during prolonged
classes. Moreover, class schedules can be arranged in
such a way that students have to relocate by foot or
by bike between classes.
Due to the lack of interest in physical activity
students often replace time they should ideally
spend on physical activity with sedentary activities.
On the other hand, when students are very
physically active throughout a certain part of the
day, physical fatigue might cause them to be
more sedentary during the rest of the day.
Although relaxation was not mentioned to be a
reason to be physically active, students felt they
rather needed to engage in sedentary activities
(such as TV watching) to clear their heads. This
might indicate that university students still choose
sedentary over physical activities in terms of relaxation
and recreation. Hence, physical activ- ity promoters
and policy makers are challenged to con- vince
students to engage in physical activities for
relaxation purposes. Finally, students revealed that
there is an absorbing quality to some sedentary
activities, such as spending time on social media.
Therefore, the com- pulsive nature of certain
sedentary behaviours, such as computer use (incl.
social media access), should be taken into account
with regard to intervention efforts.
Concerning future interventions, the present
study’s
findings support LaCaille’s [39] suggestion to
strengthen students’ self-regulation skills (e.g. self-
discipline, time management) around exercise as part
of the transition from secondary school to
university. McArthur and col- leagues [41] showed
that self-management strategies were strongly
associated with physical activity level. Moreover,
our results suggest that the same self- regulation
skills should be addressed when aiming to de- crease
sedentary behaviour during this transition period. A
randomised trial in college students showed that a
30- minute single session of one-on-one motivational
inter- viewing (including discussing perceived benefits
and bar- riers, personalised feedback, goal setting and
strategies for increasing physical activity levels)
increased moder- ate and vigorous physical activity
levels after one month [42]. Although no long-term
effects were evaluated, a single session intervention
as such may be more appeal- ing to college students
and easier to implement on col- lege campuses in
comparison to more intensive models [42]. Maybe
even more appealing to college students is the use
of smartphone applications. Recent research in
primary care patients showed that combined goal
setting with an assisting smartphone application
(based on self- monitoring and personalised
feedback) significantly in- creased the amount of
steps per day in comparison to a goal setting non-
application control group [43]. More- over, Bond and
colleagues [44] showed that prompting small physical
activity breaks after excessive sedentary
 been neglected so far [3]. Hence, experimental research
time through a smartphone application increased investigating the relative importance of physical envi-
phys- ical activity and decreased sedentary time in ronmental factors on physical activity, but also on seden-
overweight and obese individuals. To the best of our tary behaviour in university students is needed. The
knowledge, no intervention efforts have been made
so far to decrease sedentary behaviour in a
university or college student population. Future
(smartphone-based) experimental studies should
investigate if similar motivational and be- haviour
change techniques are also effective in decreas- ing the
amount of time university students spend in
sedentary mode.

Social environment

Although at the interpersonal level focus group

litera- ture in US university students only
mentioned social support from friends to be
influencing physical activity [37,39], the present study
demonstrated that the social environment influencing
students’ physical activity also included parental control,
modelling and peer pressure. At the same time, these
factors were found to influence students’ sedentary
behaviour as well. The fact that pre- vious US studies
did not find parental influences on stu- dents’
physical activity behaviour might be explained by the
longer home-university distances, forcing US univer- sity
students to reside away from home more often than
Belgian students, resulting in less parental
influences.

Physical environment

In accordance with previous US research

investigating determinants of physical activity [37],
current study re- sults showed that university students
are very susceptible to monetary costs. Moreover,
results revealed that price can be a barrier to
participate in healthy exercise behav- iour, but at the
same time be an enabler to choose other non-active
or sedentary (like TV viewing) behaviour. Therefore,
students proposed to make (on-campus) sports
activities cheaper and/or more flexible which
would lower the barriers of participation, resulting in
opting for more physical and less sedentary
activities.
Availability and accessibility of sports lessons and
facil- ities as well as TV or computer were found to
influence university students’ physical and sedentary
activities. Despite the abundantly available on-campus
sports facil- ities (at our university), participants of the
present study did not automatically engage in more
physical activities and/or less sedentary time. In
Keating’s review [3] it was concluded that the
influence of campus exercise or fit- ness facilities on
university students’ physical activity be- haviours was
still unclear. The same review also revealed that
research on the impact of campus size and overall
physical layout and structure on physical activity has
abovementioned might also indicate that physical
activ- ity and sedentary behaviour in university
students is not only influenced by the physical
environment, but also the social environment and
individual factors at the same time. The continuous
interaction between deter- minant levels suggests
that intervention strategies using multilevel
approaches may be most effective [21].

University specific characteristics

Some student characteristics (e.g. residency, exams,

etc.) seemed to be moderating relationships between
determi- nants and physical activity and sedentary
behaviour. For example, living in a student residence
might affect the strength of the relation between
modelling and physical activity and/or sedentary
behaviour. Students may ex- perience less parental
modelling but more peer model- ling when residing
away from home and vice versa.

Physical (in)activity versus sedentary behaviour

Despite the introduction given on sedentary

behaviour and its distinction from inactivity, it
was hard to keep participants focused on sedentary
behaviour as such. When asking them which factors
influenced their seden- tary behaviour, participants
tended to deviate and talk about physically inactive
behaviour instead. Therefore, the moderator had to be
very alert and redirect discus- sions when necessary.
Consequently, suggestions for in- terventions mainly
focus on strategies to be more physically active,
whereas little to no specific recommen- dations were
made to target actual sedentary behaviour. This means
that ‘sedentary behaviour’ is still a relatively unknown
concept among university students, indicating that
researchers along with policy makers still need to
work on familiarising students with this concept and
its association with overall health.
Although previous studies have shown that physical
inactivity should be investigated independently from
sed- entary behaviour [5,15], many factors in the
present study were found to influence physical
activity and sed- entary behaviour simultaneously.
Students also believed that the lack of physical
activity may increase the likeli- hood of spending
more time in sedentary mode, suggest- ing an
undeniable connection between both behaviours.
Previous research in college students showed that
com- puter use for men and television watching for
women were negatively correlated with exercise and
physical ac- tivity [16,17]. In accordance, Owen et al.
[15] highlighted that sedentary behaviour can coexist
with but also com- pete with physical activity.
Therefore, as students sug- gested, intervention
efforts aiming to increase university students’ physical
activity might decrease time spent sedentary as well.
It should be mentioned, however, that a recent review
and meta-analysis of controlled trials in adults
found that interventions aiming to promote
 a larger representative sample, should deter- mine the
physical activity (with no sedentary behaviour importance and value of each determinant,
compo- nent) were least effective in reducing
sedentary behav- iour, compared to those studies that
specifically targeted sedentary behaviour [45]. Thus, a
component focusing on reducing sedentary
behaviour may be needed to gen- erate meaningful
reductions in sedentary time [45].

Strengths and limitations

This study adds important evidence to the limited

litera- ture investigating determinants of sedentary
behaviour in university students and in general.
Moreover, this is the first study collecting ideas and
recommendations to in- crease physical activity and
decrease sedentary behaviours in university students. This
should facilitate the develop- ment of effective and
tailored intervention programs aim- ing to improve
physical activity and sedentary behaviour
simultaneously. Secondly, as highlighted by Rouse et al.
[5], sedentary behaviour is multifaceted and should not
be limited to television viewing. Hence, participants
of the present study were given a priori information on
sedentary behaviour, making sure not only determinants
of TV view- ing but also other sedentary behaviours
(e.g. computer use, studying, socialising) were
explored. Finally, the re- search team chose focus
group discussions over e.g. in- depth interviews,
because the dynamic group interactions allowed us to
get better insight into the mechanisms be- hind
university students’ eating behaviours [33]. On the
other hand, the group setting might have intimidated
some participants which, in turn, might have limited
a greater sharing of their thoughts.
Our study has some limitations as well. Although we
might expect that behavioural differences according
to gender may be found [28,46], we chose to use
mixed- gender focus groups including students of
different study years and disciplines, allowing us to
create greater diversity of opinion within each focus
group. Secondly, participants were recruited using
snowball sampling, which is a purposive
nonprobability approach that is often used in
qualitative research, especially when the study is
explorative in nature. This approach allowed us to
generate rich and lively discussions, which may not
happen in a more random collection of participants
[29]. Using subjects who know one another may be a
limita- tion to the generalizability of the findings
beyond the group assessed. However, the purpose of
this study was to generate a rich understanding of
participants’ experi- ences and beliefs [47,48] and not
to generalize results [49]. Finally, no quantification
was used because the issue raised most frequently is
not necessarily the most important, even when it is
raised by a larger number of people [49]. In other
words, each idea or opinion should be equally
appreciated. Hence, future quantitative stud- ies, using
 References
making differentiation according to gender or other
1. Crombie AP, Ilich JZ, Dutton GR, Panton LB, Abood DA. The
stu- dent characteristics, as well as generalization
freshman weight gain phenomenon revisited. Nutr Rev.
possible. 2009;67(2):83–94.
2. Vella-Zarb RA, Elgar FJ. The ‘freshman 5’: a meta-analysis of weight
gain in the freshman year of college. J Am Coll Heal.
Conclusions 2009;58(2):161–6.
After entering university, students are continuously
chal- lenged by choosing between all kinds of
activities (in- cluding physical and sedentary
activities). Depending on their time schedule, a
certain amount of self-discipline is needed to be
physically active and not to engage in con- venient
sedentary activities. During this decision making
process (= individual factors), students are influenced
by their family and friends (= social environment),
as well as by the availability and accessibility and
prices of these activities (= physical environment). In
turn, choices have to be made within a university
specific setting, such as living in a student
residence, having exams (= modera- tors).
Recommendations for future physical activity in-
terventions include improving information strategies
regarding on-campus sports activities, cheaper
and/or more flexible sports subscriptions and
formulas, includ- ing ‘sports time’ into the curricula,
and providing univer- sity bicycles around campus.
Finally, students felt that the abovementioned
recommendations to increase phys- ical activity may
decrease students’ sedentary behaviour at the same
time. Our results should be considered a first step
towards the development of tailored and effect- ive
intervention programs aiming to improve university
students’ energy expenditure behaviours.
Abbreviations
BMI: Body mass index; SD: Standard deviation; SPSS: Statistical package
for the social sciences; TV: Television; UK: United Kingdom; US: United
States; yrs: Years.

Competing interests

The authors declare that they have no competing interests.

Authors’ contributions

TD participated in the design of the study, collected all data, performed

the data analyses and drafted the manuscript. IDB participated in the
design of the study and revised the manuscript critically. PC and BD
participated in the design of the study, contributed to the interpretation
of data and revised the manuscript critically. All authors read and
approved the final manuscript.

Acknowledgements

The authors thank all students participating in this study. Special thanks to all
Master thesis students for their efforts regarding data collection and
transcription.

Author details

1
Department of Human Biometry and Biomechanics, Vrije Universiteit
Brussel, Pleinlaan 2, 1050 Brussels, Belgium. 2Department of Movement
and Sports Sciences, Ghent University, Watersportlaan 2, 9000 Ghent,
Belgium.

Received: 5 December 2014 Accepted: 16 February 2015

 the factors affecting implementation. Am J Community Psychol. 2008;41(3–
3. Keating XFD, Guan JM, Pinero JC, Bridges DM. A meta-analysis of 4):327–50.
college students’ physical activity behaviors. J Am Coll Heal. 28. Salmon J, Tremblay MS, Marshall SJ, Hume C. Health risks, correlates, and
2005;54(2):116–25. interventions to reduce sedentary behavior in young people. Am J Prev
4. Sigmundova D, Chmelik F, Sigmund E, Feltlova D, Frömel K. Physical Med. 2011;41(2):197–206.
activity in the lifestyle of Czech university students: Meeting
health recommendations. European Journal of Sport Science.
2013;13(6):744–50.
5. Rouse PC, Biddle SJH. An ecological momentary assessment of the
physical activity and sedentary behaviour patterns of university
students. Health Educ J. 2010;69(1):116–25.
6. Warburton DE, Nicol CW, Bredin SS. Health benefits of physical
activity: the evidence. Can Med Assoc J. 2006;174(6):801–9.
7. Jakicic JM, Davis KK. Obesity and physical activity. Psychiatr Clin
North Am. 2011;34(4):829–40.
8. Hu FB. Sedentary lifestyle and risk of obesity and type 2 diabetes.
Lipids. 2003;38(2):103–8.
9. Must A, Tybor DJ. Physical activity and sedentary behavior: a
review of longitudinal studies of weight and adiposity in
youth. Int J Obes. 2005;29:S84–96.
10. Miller YD. Psychological determinants and outcomes of
sedentary and physical activity behaviours. Int J Behav Med.
2010;17(4):243–5.
11. Owen N, Healy GN, Matthews CE, Dunstan DW. Too much
sitting: the population health science of sedentary behavior. Exerc
Sport Sci Rev. 2010;38(3):105–13.
12. Teychenne M, Ball K, Salmon J. Sedentary behavior and depression
among adults: a review. Int J Behav Med. 2010;17(4):246–54.
13. Iannotti RJ, Janssen I, Haug E, Kololo H, Annaheim B,
Borraccino A. Interrelationships of adolescent physical activity,
screen-based sedentary behaviour, and social and psychological
health. Int J Public Health. 2009;54 Suppl 2:191–8.
14. Syvaoja HJ, Tammelin TH, Ahonen T, Kankaanpaa A, Kantomaa
MT. The associations of objectively measured physical activity
and sedentary time with cognitive functions in school-aged
children. Plos One. 2014;9(7):e103559.
15. Owen N, Leslie E, Salmon J, Fotheringham MJ. Environmental
determinants of physical activity and sedentary behavior. Exerc
Sport Sci Rev. 2000;28(4):153–8.
16. Buckworth J, Nigg C. Physical activity, exercise, and sedentary behavior
in college students. J Am Coll Heal. 2004;53(1):28–34.
17. Romaguera D, Tauler P, Bennasar M, Pericas J, Moreno C, Martinez
S, et al. Determinants and patterns of physical activity practice
among Spanish university students. J Sports Sci.
2011;29(9):989–97.
18. Biddle SJ, Gorely T, Stensel DJ. Health-enhancing physical activity
and sedentary behaviour in children and adolescents. J
Sports Sci. 2004;22(8):679–701.
19. Swinburn BA, Caterson I, Seidell JC, James WP. Diet, nutrition
and the prevention of excess weight gain and obesity. Public
Health Nutr. 2004;7(1A):123–46.
20. Mcleroy KR, Bibeau D, Steckler A, Glanz K. An ecological
perspective on health promotion programs. Health Educ Q.
1988;15(4):351–77.
21. Sallis JF, Owen N. Ecological Models of Health Behavior.
3rd ed. San Francisco: Jossey-Bass; 2002.
22. Story M, Neumark-Sztainer D, French S. Individual and
environmental influences on adolescent eating behaviors. J
Am Diet Assoc. 2002;102(3):S40–51.
23. Biddle SJH, Mutrie N. Psychology of Physical Activity:
Determinants, Well-Being and Interventions. 2nd ed. New
York: Routledge; 2008.
24. Haase A, Steptoe A, Sallis JF, Wardle J. Leisure-time physical
activity in university students from 23 countries: associations with
health beliefs, risk awareness, and national economic development.
Prev Med. 2004;39(1):182–90.
25. Reed JA, Phillips DA. Relationships between physical activity
and the proximity of exercise facilities and home exercise
equipment used by undergraduate university students. J Am Coll
Heal. 2005;53(6):285–90.
26. Simons D, Clarys P, De Bourdeaudhuij I, de Geus B,
Vandelanotte C, Deforche B. Why do Young Adults Choose
Different Transport Modes? Transport Policy: A focus group
study; 2014.
27. Durlak JA, DuPre EP. Implementation matters: a review of research
on the influence of implementation on program outcomes and
 29. Morgan DL, Scannell AU. Planning Focus Groups. Thousand Oaks,
California: Sage Publications; 1998.
30. Krueger RA. Developing Questions for Focus Groups. Thousand
Oaks, California: Sage Publications; 1998.
31. Deliens T, Clarys P, De Bourdeaudhuij I, Deforche B. Determinants

of eating behaviour in university students: a qualitative study using

focus group discussions. BMC Public Health. 2014;14:53. ISSN: 0735-9683
32. Atkin AJ, Gorely T, Clemes SA, Yates T, Edwardson C, Brage S, et al. (Print) 1545-0864
Methods of measurement in epidemiology: sedentary behaviour. Int (Online) Journal
J Epidemiol. 2012;41(5):1460–71. homepage:
33. Silverman D. Qualitative Research: Theory, Method and Practice.
Secondth ed. London, Thousand Oaks, New Delhi: Sage
https://www.tandfonli
Publications; 2004. ne.com/loi/whmq20
34. MacKinnon DP, Fairchild AJ, Fritz MS. Mediation analysis. Annu
Rev Psychol. 2007;58:593–614.
35. Elo S, Kyngas H. The qualitative content analysis process. J
Adv Nurs. 2008;62(1):107–15. Factors Influencing Healthy
36. Bandura A. Social Foundations of Thought and Action: A Social
Cognitive Theory. Prentice Hall: Englewood Cliffs, NJ; 1986. Eating Habits Among
37. Greaney ML, Less FD, White AA, Dayton SF, Riebe D, Blissmer
B, et al. College Students’ barriers and enablers for healthful weight College Students: An
management: a qualitative study. J Nutr Educ Behav.
2009;41(4):281–6. Application of the Health
38. Nelson MC, Kocos R, Lytle LA, Perry CL. Understanding the
perceived determinants of weight-related behaviors in late
adolescence: a qualitative analysis among college youth. J Nutr Educ
Belief Model
Behav. 2009;41(4):287–92.
39. LaCaille LJ, Dauner KN, Krambeer RJ, Pedersen J.
Psychosocial and environmental determinants of eating
behaviors, physical activity, and weight change among college Sameer Deshpande ,
students: a qualitative analysis. J Am Coll Heal.
2011;59(6):531–8.
Michael D. Basil & Debra
40. Swartz AM, Squires L, Strath SJ. Energy expenditure of
interruptions to sedentary behavior. Int J Behav Nutr Phy.
Z. Basil
2011;8:69.
41. McArthur LH, Raedeke TD. Race and Sex differences in college
student physical activity correlates. Am J Health Behav.
2009;33(1):80–90.
42. Martens MP, Buscemi J, Smith AE, Murphy JG. The short-term
To cite this
efficacy of a brief motivational intervention designed to increase article:
physical activity among college students. J Phys Act Health. Sameer
2012;9(4):525–32. Deshpande ,
43. Glynn LG, Hayes PS, Casey M, Glynn F, Alvarez-Iglesias A, Newell
J, et al. Effectiveness of a smartphone application to promote Michael D. Basil
physical activity in primary care: the SMART MOVE randomised & Debra Z. Basil
controlled trial. Br J Gen Pract. 2014;64(624):e384–91. (2009) Factors
44. Bond DS, Thomas JG, Raynor HA, Moon J, Sieling J, Trautvetter J, Influencing
et al.
B-MOBILE–a smartphone-based intervention to reduce sedentary time in Healthy Eating
overweight/obese individuals: a within-subjects experimental trial. Habits Among
Plos One. 2014;9(6):e100821. College Students:
An Application
45. Prince SA, Saunders TJ, Gresty K, Reid RD. A comparison of the
effectiveness of physical activity and sedentary behaviour of the Health
interventions in reducing sedentary time in adults: a systematic Belief Model,
review and meta-analysis of controlled trials. Obes Rev. Health Marketing
2014;15(11):905–19. Quarterly, 26:2,
46. Uijtdewilligen L, Nauta J, Singh AS, van Mechelen W, Twisk
JW, van der Horst K, et al. Determinants of physical activity and 145-164, DOI:
sedentary behaviour in young people: a review and quality 10.1080/0735968
synthesis of prospective studies. Brit J Sport Med. 0802619834
2011;45(11):896–905.
47. Morgan DL. The Focus Group Guidebook. Thousand Oaks,
California: Sage Publications; 1998.
To link to this article:
48. Maxwell J. Qualitative Research Design: An Interactive Approach. 2nd https://doi.org/10.1080/0735968
ed. Thousand Oaks, CA: Sage; 2005.
49. Krueger RA. Analyzing & Reporting Focus Group Results. Thousand
0802619834
Oaks, California: Sage Publications; 1998.

Health
Marketing
P
Quarterly
u
b

i
S
s
u
h
b
e
m
d
i

n
y
l
o
i
u
n
r
e

r
3
t
0
i

A l

p e

2 o

0
t
9
h
.
i
s

Article views: 151262

View related articles

Citing articles: 106 View

citing articles

Full Terms &

Conditions of access
and use can be found
at
https://www.tandfonline.
com/action/journalInfor
mation?
journalCode=whmq20
Health Marketing Quarterly, 26:145–164, 2009 Copyright #
Taylor & Francis Group, LLC ISSN: 0735-9683 print=1545-
0864 online
DOI: 10.1080/07359680802619834

Factors Influencing Healthy Eating Habits

Among College Students: An
Application of the Health Belief Model

SAMEER DESHPANDE, MICHAEL D. BASIL, and DEBRA Z. BASIL

University of Lethbridge, Lethbridge, Alberta, Canada

Poor eating habits are an important public health issue that has large
health and economic implications. Many food preferences are established
early, but because people make more and more independent eating
decisions as they move through adolescence, the transition to independent
living during the university days is an important event. To study the
phenomenon of food selection, the heath belief model was applied to
predict the likelihood of healthy eating among university students.
Structural equation modeling was used to investigate the validity of the
health belief model (HBM) among 194 students, followed by gender-
based analyses. The data strongly supported the HBM. Social change
campaign implications are discussed.
KEYWORDS health belief model, healthy eating, university students

Poor nutrition and obesity are among the most important health issues facing society
today, not only in terms of health, but also health care expenses (Goel, 2006;
Rashad & Grossman, 2004). There are a variety of predictors of obesity including
genetics, physical activity, and food consumption (Goel, 2006). There are other
outcomes of food choice and nutrition that also have an independent effect on
health including some types of cancer, cardi- ovascular disease, and diabetes
(Nicklas, Baranowski, Baranowski, Cullen,

The authors wish to acknowledge financial support from the University of Lethbridge’s Faculty of
Management and the University of Lethbridge Centre for Socially Responsible Marketing. The authors also
thank the three anonymous reviewers of the 2007 Society for Consumer Psychology Conference for their
valuable comments.
Address correspondence to Sameer Deshpande, University of Lethbridge, 4401 University Dr., Lethbridge,
AB T1K 3M4, Canada. E-mail: sameer.deshpande@uleth.ca

145
 Healthy Eating Habits Among College Students 14
6
Rittenberry, & Olvera, 2001). For these reasons, food selection is an important
consumer behavior with many long-term consequences to the individual in the
form of health and longevity and to society in the form of health costs.
Some research has shown that the most important factors predicting food
selection among adults are: taste, cost, nutrition, convenience, pleasure, and weight
control, in that order (Glanz, Basil, Maibach, Goldberg, & Snyder 1998). Many
studies have shown that people often establish these tastes and habits while they
are relatively young (Birch, 1999). Evidence suggests early establishment of habits and
preferences occurs for a variety of behaviors including media use (Basil, 1990) and
music listening (Holbrook & Schindler, 1994), as well as food choice (Birch, 1999).
Therefore it is advisable to begin establishing good eating habits when people are as
young as possible. Impor- tantly, however, for the very young many food decisions
are controlled by parents and preschools (Nicklas et al., 2001). Therefore, food choice
for the youngest age groups may be constrained by a number of factors.
An especially important time of life for food choice is when people step out
independently for the first time and begin to make all of their own food decisions.
For many people, this is the transition to college life. The transition to college or
university is a critical period for young adults, who are often facing their first
opportunity to make their own food decisions (Baker, 1991; Marquis, 2005) and
this could have a negative impact on students’ eating behaviors (Marquis, 2005;
Rappoport, 2003).
Previous literature has extensively discussed factors that influence eating
behaviors among college students. However, application of a behavioral model
such as the health belief model (HBM) has received less attention. Only three studies
were found that applied HBM in the college eating context (Garcia & Mann, 2003;
Von Ah, Ebert, Ngamvitroj, Park, & Kang, 2004; Wdowik, Kendall, Harris, & Auld,
2001). These studies examined avoiding dieting, a combination of eating and
exercise, and diabetic students, respec- tively. The present study provides valuable
insights into how health beliefs impact eating behaviors for college students—a
population at the crucial stage of transitioning into independent nutritional
practices.

EATING BEHAVIOR OF COLLEGE STUDENTS

Numerous studies have shown that college students often have poor eating habits.
Students tend to eat fewer fruits and vegetables on a daily basis and report high
intake of high-fat, high-calorie foods (Brevard & Ricketts, 1996; Driskell, Kim, &
Goebel, 2005; Racette, Deusinger, Strube, Highstein, & Deusinger, 2005). According to
the American College Health Association (2006), a 2004 study revealed that only
7.3% of students ate five or more servings of fruits and vegetables daily. The
transition to college life often
worsens dietary habits among students (Grace, 1997) which could contribute to
weight problems especially during the first year of college or university (Anderson
et al., 2003) and continue during later years of life (Centers for Disease Control,
1997; Racette et al., 2005).

Determinants of Eating Behavior

Previous studies have shown a link between demographic and psycho- graphic
characteristics with dietary behavior of college students. Driskell et al. (2005) revealed
few differences among lower and upper level students in terms of their dietary habits,
suggesting that habits established in the first year or two likely carry forward into
later college years. However, where a student lives seems to affect his or her dietary
habits and diet-related health (Brevard & Ricketts, 1996). Students living off-
campus reported a higher percentage of energy from protein. Similarly, serum
triglyceride level and the ratio of total cholesterol to high-density lipoprotein were
also higher among students living off-campus. The authors conclude that students
living off campus are choosing different foods than those living on campus.
Gender differences also exist (Racette et al., 2005). Female college students
tend to eat more fatty foods than male students, although their fruit and vegetable
consumption tends to remain similar. As discussed earlier, according to Brevard
& Ricketts (1996), residence on or off campus made a difference, but it also
interacted with gender. Higher energy from protein was more prevalent among men
living off campus than on campus. For women, higher serum triglyceride and ratio
of total cholesterol to high- density lipoprotein was found among those who lived
off campus. Horacek & Betts (1998) clustered male and female college students by
dietary intake differences. Four clusters were found: students influenced by
internal (hunger and taste) and external cues (friends and media), by budget, by
health, and neither of the factors. Males tended to be equally represented in all
the four clusters with a somewhat higher percentage in the cues group, while female
students tended to cluster in the cues group (55%) followed by health factors
(28%). In a study by Mooney & Walbourn (2001), females avoided certain foods for
their concern for weight, health and ethical reasons (especially when avoiding meat)
more significantly than males. Marquis (2005) similarly reported that females were
more significantly motivated by convenience, pleasure, price, and weight concerns
than male students. We can thus conclude that the dietary intake of male and female
college students is influenced by different factors.
Motives influencing eating behaviors among college students have been studied
as well. House, Su, and Levy-Milne (2006) investigated what benefits college students
believed result from a healthy diet. In this study, students at a Canadian university
reported healthy eating to be helpful in providing a healthy appearance (in terms
of weight, skin, physique, and so forth),
providing positive feelings, and preventing disease. Although the results in this study
were based on a focus group finding with 15 students (9 students were studying to
be dieticians) there are nonetheless similarities with studies conducted among general
adults (Steptoe, Pollard, & Wardle, 1995). Horacek & Betts (1998) found that taste,
time sufficiency, convenience, and budget influenced students’ eating habits in
that order. These seem to act more as barriers to healthy eating as revealed from
the focus group (House et al., 2006). One could assume that these barriers may
be more influential than benefits given the prevalence of eating habits among
college students.
Other factors associated with poor eating habits among college students include
a higher perception of stress (Cartwright, Wardle, Steggles, Simon, Croker, & Jarvis,
2003), and low self-esteem (Huntsinger & Luecken, 2004). Previous studies have also
reported a low level of nutrition knowledge (Barr, 1984; Van den Reek & Keith, 1984).
Lack of indepth nutrition knowledge has been attributed to reliance on sources that
provide inadequate information on nutrition (Thomsen, Terry, & Amos, 1987).

The Health Belief Model

Although studies have investigated demographic and psychographic charac- teristics
of healthy eating among college students, research is lacking in terms of
comparing the effectiveness of these predictors in a single model such as the
HBM. The HBM, developed in the 1950s (Rosenstock, 1974), is a expectancy-
value model. It has been employed in a variety of public health settings over the
years. The HBM postulates that when an individual per- ceives a threat from a
disease (measured by susceptibility to the disease and the severity of disease),
and perceived benefits from preventive action exceed barriers then the individual
is likely to take preventive action. In the HBM, demographic characteristics and
cues to action moderate the effects of the above mentioned predictors.
While three studies were found to apply the HBM to eating among college
students, their findings were less useful to the current study because of differences in
the nature of the sample or the dependent variables. One study (Wdowik et al.,
2001) applied HBM to understand how diabetic students manage their problem
and did not address the general student population. The two student groups may
differ from each other in their perceptions of healthy diet. For example, the
diabetic student population may perceive benefits and barriers of healthy diet
differently from general student population. A second study (Garcia & Mann,
2003) employed the model to understand how students resist dieting, not how
they approach healthy eating. Finally, a third study (Von et al., 2004) investigated the
influ- ence of HBM variables on physical activity and nutrition behavior among
other behaviors. Unfortunately, they combined physical and nutrition beha- viors as a
single measure, although the two behaviors conceptually differ.
 The applicability of the HBM has been reported in predicting healthy eating
among general adults (for example, Kloeblen & Batish, 1999; Sapp & Jensen,
1998), especially nutrition behaviors (see Chew, Palmer, & Kim, 1998). However,
these findings may be less generalizable to the college student population given the
differences between students and the general adult population in terms of lifestyle,
income, social environment, and food choices available on the campus. The current
study will attempt to extend the HBM’s applicability to predicting nutrition behavior
among college students. In the present study, the HBM has been extended in
accordance with the Sapp & Jensen (1998) study to include additional variables
relevant to the eating behaviors of adults. These variables include perception of
current dietary quality, perceived importance of eating a healthy diet, and
environmental variables.

Self-Efficacy
Self-efficacy has been added to the model as well. Self-efficacy is the individual’s
perception that he or she is able to perform the advocated behavior (Bandura,
1977). Self-efficacy has been shown to impact behavior in several consumer behavior
contexts such as response to fear appeals (Witte, 1992) and charitable donations
(Basil & Ridgway, 2002). Self- efficacy has played an important role in predicting
nutrition behavior among college students as well (Garcia & Mann, 2003; Von Ah et
al., 2004).
The central purpose of this study was to test the predictive ability of the
HBM on the likelihood of eating healthy in the next two-week period among
college students. As discussed earlier, since dietary behavior and motives differ
by gender, we also intend to test HBM independently for male and female college
students. The period of two weeks was chosen in an effort to assure that subjects
would be able to remember their eating habits. Healthy eating was defined as
‘‘Eating a diet that is low in bad fat, sodium, bad cholesterol and sugar and, high
in fiber, fresh fruits and vegetables.’’
HBM as modified for an eating context was hypothesized to predict college
students’ intentions to eat a healthy diet. Based on the above discussion, we
propose the following hypotheses:

H1: The importance of eating healthy will be positively influenced by dietary status,
perceived severity, perceived susceptibility, and cues.

H2: The importance of eating healthy will show a negative relationship with
barriers.

H3: The importance of eating healthy will show a positive relationship with eating
intentions.

H4: Food features will positively influence barriers. H5: Food

features will negatively influence benefits. H6: Food features will
positively influence efficacy.
 FIGURE 1 Model predicting likelihood of eating healthy.

H7: Barriers will negatively influence efficacy.

H8: Barriers will negatively influence behavior intention. H9: Benefits will
positively influence behavior intention. H10: Self-efficacy will positively
influence behavior intention.

See Figure 1 for a depiction of the modified HBM.

METHOD

Students were recruited at a Canadian university. The study was conducted by

advertising $5 for those who were willing to participate in a 20-minute survey. The
final convenience sample consisted of 194 Canadian university undergraduate
students. The gender distribution was 45% female and 55% male (n 88 and 106
respectively). Respondents were predominantly in the 18–25 year old range,
¼ 84% of the sample (n 162), predo- minantly Caucasian (75% of the
representing
sample, n 146) followed by Asians (20% of the sample, n 38), and¼predominantly
earning less than $24,000 (68% of the sample, ¼ n 133). Responses to height and
weight questions were used to calculate each respondent’s body mass index
¼
(BMI). The average BMI was 23 (for reference, a BMI of 25 or more indicates an
overweight ¼person).
This research was part of a larger study containing an experiment that
manipulated participants’ hypothetical health conditions and asked them to make food
product choices. The effects of the experimental manipulations
were controlled while calculating the effect of HBM variables for the present analysis
to assure that they did not impact the results.

Procedure
The study was first approved by the university’s Human Subjects Committee.
Respondents were invited to a computer lab where they first took part in the
manipulated health concern portion of the research. They then completed the
HBM survey which is the focus of the present analysis. After completing the study
they were given $5 as promised compensation and a feedback sheet informing
them of the purpose of the study.

Independent Measures
All items were measured with a 7-point bipolar scale ( 3 to 3)—unless otherwise
noted. Questions for most variables were derived from Sapp & Jensen’s (1998)
study. The ‘dietary status’ variable was measured with four questions (In your
opinion, was your diet in the previous two-week period lower or higher in
nutrition or was it just about right compared with Health Canada’s recommended
dietary guidelines? How healthy have your food choices been in the last two
weeks: Every meal=Never? In the course of the previous two-week period, how
often have you made good food choices? In the course of the previous two-week
period, how often have you made good food choices: Never=Always?). These
items produced an alpha of 0.80.
Perceived severity was measured using seven items (I will miss more than two
months of school or work; I will have long-lasting effects; I will be bed-ridden for
a long time, I will have medical expenses; I will harm my career; My social
relationships will suffer; I will hurt my family life) producing an alpha of .86.
Perceived susceptibility was a single item measure (Do you think some day you will get
seriously ill if you do not make good food choices?). The cues to action variable was
measured with three items (I would pay more attention to the quality of my food
choices if I read infor- mation in the mass media (news stories, ads, other
programs); I would pay more attention to my food choices if recommended by
a doctor; I would pay more attention to my food choices if friends or family
members suggested it) producing an alpha of .66. Importance of eating a healthy diet
was measured with two items (How important is it to you to eat a diet high in
nutrition? How important is nutrition to you when you shop for food?)
producing an alpha of .80 (r .67).
The food features variable was measured with four constructs: price, taste, ease
of preparation, and convenience (price; taste; ease of preparation; convenience is
important to me when I shop for food) producing an alpha of
0.61. Benefits of healthy eating was¼measured with five items (For me to eat a
nutritious diet most of the time in the next two-week period would be
harmful=beneficial; unpleasant=pleasant; bad=good; worthless=valuable;
unenjoyable=enjoyable) producing an alpha of .84. Barriers to healthy eating was measured
with three items (I don’t like the taste of most foods that are high in nutrition; I
think it would take too much time to change my diet most of the time in the next
two-week period to include more foods high in nutri- tion; Over the next two
weeks, I think it would be too hard to change my diet to include more foods high
in nutrition) producing an alpha of .79. Efficacy to healthy eating was measured with
two items (If I tried, I am confident that I could maintain a diet high in nutrition
most of the time in the next two-week period; If I wanted to, I feel that I
would be able to follow a diet high in nutrition most of the time in the next two-
week period) producing an alpha of .88 (r ¼ .80).

Dependent Measure
Likelihood of eating healthy’ was measured with three items, each measured on a
seven-point scale: (a) I intend to eat a nutritious diet most of the time in the next
two-weeks, Extremely Unlikely=Extremely Likely; (b) In the course of the next two-
week period, how often will you make good food choices? Never=Every Meal; (c) In
the course of the next two-week period, how often will you make good food choices?
Never=Always] producing an alpha of .82. Discriminant validity analysis was
conducted to check for the extent to which a construct was truly distinct from other
constructs. Pearson correla- tions between independent variables were observed. None
of the correlation scores were strong enough (maximum r .52, p < .01) to
warrant a merging or dropping of constructs. Finally, to test for common method bias
arising out of having a single source of information from students, we conducted
¼
Harman’s (1976) single-factor test (EFA) selecting varimax rotation and principle
components method with all the items across constructs. The results revealed
several factors with eigenvalues more than 1. We thus concluded that there are
many constructs available and that the items are measuring distinct constructs,
and found no evidence of a significant
common method bias.

RESULTS AND ANALYSIS

A Pearson correlation analysis was conducted with the independent and the dependent
variables. The correlation scores ranged from .60 to .62. Later,—a structural equation
model was constructed to test the modified HBM among all three samples (all,
male, and female students) using AMOS 5.0. As mentioned earlier, the effect of
the experimental manipulations were controlled in order to get a better
understanding of the effect of the modified HBM variables. All subjects participated in
the experimental study and were
assigned to one of the health concern manipulations (diabetes, heart disease, and
generally good health). The control variable was allowed to covary with every
independent variable and a regression path on every dependent variable among
all three samples. The influence of the control variable largely lacked significance.
In the all student sample, the variable had a significant influence on the likelihood to
eat healthy, and had a significant covariance with the adequacy of current diet
variable. There was no signifi- cant relationship of the control variable in the female
student sample. In the male student sample, the control variable had significant
covariance only with adequacy of current diet and the severity construct. In order
to reduce the clutter, paths of the control variable are not shown in the
figures.

Total Sample
The HBM for the all students sample was identified given the good fit indices (N 194,
CFI¼.99, RMSEA ¼ .05). The four¼ independent variables, dietary status (M 4.41, SD
1.11, B 0.44, p < .01), perceived severity (M 4.85,
¼ ¼ ¼ ¼
¼ B 0.17,¼p < .01), perceived susceptibility (M 5.18, SD 1.56,
SD 1.22, ¼ ¼

B ¼0.16, p < .01), and cues (M 5.22,¼ SD 1.12, ¼ B 0.22, p¼ < .01)
significantly predicted the importance of eating a healthy diet. In turn, the
importance variable (M 5.42,
¼ SD 1.16)¼predicted barriers to eating healthy (B 0.39, p
< .01) and likelihood to eat healthy with the regression weight of 0.57 (p < .01).
¼—
Food features (M 5.53, SD 0.90) had a significant influ- ence on barriers (B 0.15,
p < .05) and benefits (B 0.14, p < .05)¼but no influence
¼ on efficacy (B 0.09, n. s.).
In turn, barriers (M ¼ 3.35, SD 1.50) had a significant ¼ —
influence on likelihood to
eating healthy (B 0.14, ¼ p < .05), while benefits (M 5.98, SD ¼ 0.91) did¼not (B
0.05, n. s.). Barriers had a strong influence on efficacy (B 0.74, p < .01).¼In
— turn,
efficacy (M 5.58, SD 1.30) had a significant influence on the likelihood variable
¼ ¼ ¼
(M 5.28, SD 1.10, B 0.25, p < .01). Based on these results, we conclude that the
¼ —
data supported all hypotheses except for H6 and H9. See Figure 2.
¼ ¼
¼ ¼ ¼
Female Sample
The HBM model was later tested for each gender. The HBM for the female students
sample was identified given the good fit indices (N ¼ 88, CFI ¼ .98, RMSEA ¼ .
07). As expected, the model was supported by the data. The three independent
variables, dietary status (M ¼ 4.55, SD ¼ 1.15, B ¼ 0.52, p < .01), perceived severity (M ¼
4.77, SD ¼ 1.25, B ¼ 0.27, p < .01), and cues (M ¼ 5.41, SD ¼ 1.14, B ¼ 0.17, p < .05)
significantly predicted the importance of eating a healthy diet. But the influence of
perceived susceptibility (M ¼ 5.27, SD ¼ 1.54) on importance of eating a healthy
diet was not significant (B ¼ 0.10, n. s.). In turn, the importance variable (M ¼
5.64, SD ¼ 1.15) predicted barriers to eating healthy (B ¼ —0.73, p < .01) and
the likelihood
 FIGURE 2 Model predicting likelihood of eating healthy (all students).

to eat healthy with the regression weight of 0.82 (p < .01). Food features (M ¼
5.59, SD ¼ 1.01) did not have a significant influence on barriers (B ¼ 0.15, n. s.),
nor on efficacy (B ¼ 0.02, n. s.) but a significant influence on benefits (B ¼ —0.25,
p < .01). In turn, barriers (M ¼ 3.28, SD ¼ 1.62) had a significant influence on
likelihood to eating healthy (B ¼ —0.46, p < 0.1), while benefits (M ¼ 6.15, SD ¼
0.84) did not (B ¼ —0.31, n. s.). Barriers had a strong influence on efficacy (B ¼
—0.55, p < .01). In turn, efficacy (M ¼ 5.73, SD ¼ 1.19) had a significant
influence on the likelihood variable (B ¼ 0.33, p < .05). See Figure 3.

Male Sample
The HBM for the male students sample was identified with good fit indices (N 106,
CFI¼.97, RMSEA ¼ .06). As expected,
¼ the model was supported by the data. The
three independent variables, dietary status (M 4.30, SD 1.07, B 0.37, p < .01),
perceived susceptibility (M 5.09, SD 1.57, ¼
¼ ¼
B ¼0.23, p < .01), and cues (M 5.06, ¼ SD 1.10,¼B 0.16, p ¼< .05) signifi- ¼
¼
cantly predicted the importance of eating a healthy diet. But perceived
 FIGURE 3 Model predicting likelihood of eating healthy (female students).

severity (M ¼4.93, SD ¼1.20) failed to significantly influence the importance

variable (B 0.12,
¼ n. s.). In turn, the importance variable (M 5.25, SD¼ 1.14)
significantly predicted barriers (B 0.30, p < .01) and the likeli- hood to eating
¼ with the regression weight of 0.45 (p <¼.05).
healthy — Food features (M 5.47, SD
0.79) did not statistically influence barriers (B 0.07, n. s.), efficacy (B 0.14, n.
¼ (B 9.71, n.¼s.). Simi- larly, barriers (M 3.41, SD 1.39, B 0.04, n.
s.), and benefits
s.) ¼
and benefits (M 5.85, SD¼0.94, B 0.01, n. s.) did not statistically
¼ influence
likelihood to eating¼healthy. However,
¼ barriers
¼ — had a strong influence on efficacy
¼
(B 1.31, p < .01). In turn, efficacy (M 5.45, SD 1.38) had a significant influence on
¼ ¼—
the likelihood variable (B 0.39, p < .01) (See Figure 4). Comparing the chi-square
value and support for the hypotheses for the male and female samples,¼it —seems
the model fits the female group¼data slightly ¼ better (Chi-Sq. for females 18.89,
Chi-Sq. for males 23.94) despite¼a smaller sample size (Females 88, Males 106).
In addition, one-way ANOVA was conducted to compare the means of the
HBM variables by gender. Females had a higher intention (M 5.52, SD 1.06) ¼ to
intake a nutritious diet than males (M 5.08, SD 1.09, F 7.80, p < .01). Among
¼ ¼ ¼
independent variables,¼females similarly reported higher¼ means than males among
¼ ¼ ¼
all variables except for severity and barriers. However, there was a statistical
difference only among three variables.
 FIGURE 4 Model predicting likelihood of eating healthy (male students).

Females reported statistically higher means for cues to action (M diff. 0.34, F¼4.54, p
< ¼.05), importance of eating a healthy diet (M diff. 0.39, F 5.50, ¼ p < .05)
¼ and
benefits from a healthy diet (M diff. 0.30, F 5.43, p < .05).
To summarize, for females, the intention to ¼consume ¼a healthy diet is
indirectly influenced by dietary status, severity, and cues through their effect on
importance of a healthy diet. The importance of a healthy diet also influ- ences
behavior intention through its influence on barriers. Food features influences
benefits, but benefits failed to influence behavior, while barriers influence behavior
directly as well as via its effect on efficacy. For males, the intention to consume a
healthy diet is indirectly influenced by dietary status, susceptibility, and cues
through their effect on importance of a healthy diet, by importance through its effect
on barriers, and by barriers through its effect on efficacy.
In general, the relationship of dietary status and cues on behavior intention
via importance of a healthy diet, the relationship of importance with behavior
intention through its relationship with barriers, and the relationship of barriers
with behavior intention via the influence of efficacy seem to hold true among both
males and females. Similarly, food features failed to predict barriers and benefits
failed to predict behavior intention
among both the samples. But males and females differed on other variables. The
severity construct among males and susceptibility among females failed to influence
importance of eating a healthy diet. Food features of price, taste, ease of
preparation, and convenience while shopping for healthy food negatively
influenced benefits among females, but not among males.
Finally, barriers significantly and negatively influenced behavior intention
among females but not among males. The barriers construct included two
important items of taste and time sufficiency. As Marquis (2005) suggested, these
issues matter to females.

DISCUSSION

Summary of Findings
These results make an important contribution to our understanding of student
health behavior. Previous research is extended by incorporating important
additional variables into the HBM. This research focused on an important
population which is at a key crossroads in nutritional health. The HBM was
helpful in predicting the variance in likelihood of healthy eating. Since eating
behaviors and reasons for eating vary by gender (Dutta-Bergman, 2005), we also
investigated how these factors varied by gender. The results of this study
demonstrate potential mechanisms by which these gender differences occur. In a
study conducted among the general population, Kristal, Hedderson, Patterson, &
Neuhauser (2001) argue that efforts need to target men. Since male students
reported a lower intention to consume a healthy diet than did females, our
research supports this contention.
Our findings were similar to other college studies in terms of influence of
variables such as perceived severity (Mooney & Walbourn, 2001), and effi- cacy to eat
healthy (Garcia & Mann, 1003; Von Ah et al., 2004). Some find- ings bear a close
resemblance with the Sapp & Jensen (1998) study conducted with the general
population. The influence of gender, dietary status, importance of a healthy diet,
food features, and cues to action on the intention to eat healthy in our study was
similar to their study. Despite similarities with earlier research this study found
no significant effects of the food features such as price, taste, ease of preparation,
and convenience on barriers, efficacy, and thus the likelihood to eat a healthy diet.
Although the mean was strongly positive (M 5.53 among all student sample), the
food features variable did not have a significant influence. Given the student lifestyle
(Belaski, 2001), preference, and findings reported by Horacek & Betts (1998) &
¼
Marquis (2005), food features were expected to have a strong influence on healthy
eating. But surprisingly, that was not the case. It is likely that other factors may
after all play a more important role in influen- cing healthy diet. HBM has allowed
us to understand the comparative influence of various factors.
Recommendations for Social Change Strategies to Promote Healthy
Eating Among College Students
A variety of social change campaigns could be employed to promote healthy eating
among university students. Since the HBM revealed a few important differences by
gender, males and females deserve tailored campaigns. For females, the
campaign should highlight the severity of not eating a healthy diet, while the one
targeting males should focus on increasing their perceptions of susceptibility. To
achieve these objectives, an education campaign would be an important tool to
employ. In terms of appeals, a com- bination of fear (highlighting negative
consequences among women and susceptibility among men) and efficacy could be
appropriate. A combination of fear and solution approach has been shown to be
effective (Witte, 1992). Use of food labels, another form of communication-only
tool, does influence dietary habits, but in a conditional manner. According to
Kristal et al. (2001), individuals who use labels report reductions in fat intake but
not an increase in fruits and vegetables simply because one can read
nutritional facts of processed foods but not fruits and vegetables. When used as point-
of-purchase (POP) messages, labels have shown less effectiveness among students,
especially when competing with more attractive commercial messages in the university
cafeteria (Mayer, Dubbert, & Elder, 1989). POP interventions have been used in the
past to promote a healthy diet, but with mixed results (Buscher, Martin, &
Crocker, 2001). POP interventions increased yogurt, pretzel, and whole fruit
sales but had no effect on fruit basket and vegetable basket sales. According to the
authors, effective POP displays emphasize budget friendliness, sensory=taste,
convenience, and energizing properties of food. Placement of POPs is also
critical. The researchers found that POP messages placed at the entrance of the
food- service facility were more effective than keeping them near the healthy food
items. POPs can also be used as reminders to ensure sustainability of
interventions (Maddux, 1993).
Successful attempts have been made to influence healthy eating when offered in
the form of a nutrition course (Matvienko, Lewis, & Schafer, 2001). But in many
cases, awareness creation by itself may not be an effective technique (Brown, Dresen,
& Eggett, 2005; Schnoll & Zimmerman, 2001), especially while addressing the
influence of efficacy to eat healthy. For exam- ple, a study conducted by Schnoll &
Zimmerman (2001) incorporated the self-regulation strategies of goal-setting and
self-monitoring in a nutrition course. Goal-setting by itself and in combination with
self-monitoring had a positive impact on self-efficacy to consume and consumption
of dietary fiber. In other campaigns, when promotion (that is newsletters and e-mails)
was combined with motivational interviews, the campaign seemed to have a positive
impact on fruit and vegetable consumption among college students (Richards,
Kattelmann, & Ren, 2006). Additionally, while reaching women,
studies have shown the effectiveness of a counseling approach (Leslie & St.
Pierre, 1999).
To address the influence of barriers and efficacy by promoting tasty but healthy
food with convenience on and off campus, employing the social marketing
approach could be effective. Social marketing applies marketing principles to
encourage voluntary behavior change, and a key element of social marketing is to
offer attractive alternatives in the environment. Since barriers influenced behavior
intention among females, a survey identifying those barriers should be conducted
followed by employing a social marketing campaign that would reduce these
barriers by altering the environmental factors on the campus. These efforts will
hopefully also increase efficacy among female students. For males, a social marketing
campaign would also be desirable but the focus ought to be explicitly on increasing
efficacy.
We found very few attempts that used a social marketing approach to
promote healthy diet behavior among college students. Buscher et al. (2001) was
one of them. They tested the effectiveness of product and place intervention
(ensuring easy availability of vegetables, fruits, pretzels, and yogurt) along with
promotional POP messages in the cafeteria. As discussed earlier, the study found
that sales of yogurt, pretzels, and whole fruit increased as a result of the
intervention. We commend the Buscher et al. study because they considered
promotion of products, a variety of products, products that are desirable and those
that compete with junk food options. The intervention also ensured availability of
intervention products, pro- motional tools that emphasized positive qualities of
products rather than focusing on creating fear, and in general a strong consumer
orientation. This may have resulted in a successful intervention.
Previous studies have recommended the use of prepaid meal plan (Brown et al.,
2005) and seeking cooperation of food service administrators (Leslie & St. Pierre,
1999) especially for on-campus food offerings. Buscher et al. (2001) has also
recommended dieticians to work with food service administra- tors for better offering
of meals on campus. Availability of healthy food will reduce the perception of
barriers of eating a healthy diet. According to Marquis (2005), convenience has two
dimensions, time and energy, in preparing healthy food. Both these dimensions
may warrant differing strategies among college students. Strategies such as meal
sharing and meal preparation sharing as suggested by Sobal & Nelson (2003) could
address these barriers.
A meal plan intervention study (Beth Brown et al., 2005) influenced students’
intake of food groups. Participants reported more favorable intake of vegetables, fruits,
and meat. This study also revealed significant interaction of meal plan participation
with gender. Males more favorably responded to the meal plan than females in
relation to intake of meat and vegetables. Other features such as attractive
packaging and branding of healthy food to deal with similar strategies utilized by
the competitors would be critical (Belaski, 2001).
 Tailored education campaigns should use student input in order to be
effective (Cousineau, Goldstein, & Franko, 2004), use computer and Inter- net
technology (Cousineau et al., 2004; Hanauer, Dibble, Fortin, & Col, 2004) and peer
networks (Kessler, Gilham, & Vickers, 1992) to reach the intended audience. Ohio
State University, for example, has utilized the Internet to pro- vide nutritional
information about food on campus (Belaski, 2001). While addressing students, use
of appropriate media (mass as well as non-mass media options) and the
messenger should be borne in mind since external cues played an important role
in our HBM model.

LIMITATIONS AND FUTURE RESEARCH

The main limitation of this study is that the experimental manipulation may have
influenced students’ perceptions and heightened the importance of eating
healthy. However, the relationship of the control variable (experi- mental
manipulation) was significant with very few variables among all three samples,
reducing the concern of this extraneous variable. Second, this study was conducted
only on one campus, using convenience sample, and within one culture. Testing
the influence of these factors on other campuses and a more representative sample
would increase the generalizability of the study. Nutrition habits and predictors may
vary by geographical location and this could influence the tendency to eat healthy.
Third, our dependent variable on intention to eat healthy was asked at a global
level. This may have prevented us from understanding the predictors unique to each
food group. Studies for example by Brown et al. (2005) showed how a prepaid
meal plan had a differential effect on each food group.
The moderate to high averages of many variables could be a result of social
desirability, nevertheless many of these results are consistent with other studies.
The zero-order correlation score of likelihood to eat healthy with a few variables
such as barriers and benefits are much higher than their standard regression weights.
This could be because there is a third variable moderating the influence. Future
research could be conducted to investigate this. Finally, in relation to factor analysis,
two limitations are noted. First, the alpha level for cues to action and food features
were under the .70 ceiling typically regarded as the minimum acceptable level.
Second, importance of eating healthy, and perceived susceptibility, scales were formed
on the basis of one or two items. Future studies should investigate these constructs
with multiple items to provide higher reliability.
The Kristal et al. (2001) study conducted among an adult population showed
that stage of change was one of the strongest predictors of decreased fat intake.
Individuals in the maintenance stage were highly likely to change their diet. Richards
et al. (2006) study showed favorable influence of tailoring messages by stages of
change of students on their consumption of fruits and
vegetables. It might be worthwhile to understand the predictors using HBM at various
stages. This would also enable campaign managers to better tailor their campaigns
appropriately.
As a follow up study, it would be useful to compare the influence of each
social change strategy (education only, education combined with coun- seling, and
social marketing) in influencing students’ healthy eating habits. This will enable
campaign managers to effectively appropriate funds. Two surprising results were
that food features failed to influence barriers and, sec- ondly benefits failed to
influence behavior intention. Future studies should investigate the nature of these
relationships, or lack thereof. Since efficacy played an important role in influencing
the behavior intention and showed a significant relationship with the barrier
variable, future studies should investigate the role of two types of efficacy as
discussed by Witte (1992): self-efficacy and response-efficacy.

REFERENCES

American College Health Association. (2006). American College Health Association- National College
Health Assessment (ACHA-NCHA) spring 2004 reference group data report (abridged). Journal of
American College Health, 54, 201–211. Anderson, D. A., Shapiro, J. R., & Lundgren, J. D. (2003).
The freshman year of college as a critical period for weight gain: An initial evaluation. Eating

Behaviors, 4, 363–367.

Baker, S. (1991). College cuisine makes mother cringe. American Demographics, 13(9), 10–13.
Bandura, A. (1977). Social Learning Theory. New York: General Learning Press. Basil, D. Z., &
Ridgway, N. M., 2002, February. Efficacy and Guilt Motivation: A

Model for Guilt Appeals. Austin, Texas: Society for Consumer Psychology Winter Conference.

Barr, S. I. (1984). Nutrition knowledge of female varsity athletes and university students. Journal
of the American Dietetic Association, 87, 1660–1664.
Basil, M. D. (1990). Primary news source changes: Question wording, availability, and cohort
effects. Journalism Quarterly, 67, 708–722.

Belaski, A. (2001). Practice points: College and university students present a challenge for
nutritionists. Journal of the American Dietetic Association, 101(8), 913.
Birch, L. L. (1999). Development of food preferences. Annual Review of Nutrition, 19, 41–62.

Brevard, P. B., & Ricketts, C. D. (1996). Residence of college students affects dietary intake, physical
activity, and serum lipid levels. Journal of the American Dietetic Association, 96(1), 35–38.
Brown, L. B., Dresen, R. K., & Eggett, D. L. (2005). College students can benefit by participating in a
prepaid meal plan. Journal of the American Dietetic Association, 105, 445–448.
Buscher, L. A., Martin, K. A., & Crocker, S. (2001). Point-of-purchase messages framed in terms of
cost, convenience, taste, and energy improve healthful snack
 selection in a college foodservice setting. Journal of the American Dietetic Association,
101, 909–913.
Cartwright, M., Wardle, J., Steggles, N., Simon, A. E., Croker, H., & Jarvis, M. J. (2003). Stress and
dietary practices in adolescents. Health Psychology, 22, 362–369.
Centers for Disease Control. (1997). Youth risk behavior surveillance: National college health risk
behavior survey – United States. Mortality and Morbidity Weekly Report, 46(SS-6), 1–54.
Chew, F., Palmer, S., & Kim, S. (1998). Testing the influence of the health belief model and a
television program on nutrition behavior. Health Communication, 10, 227–245.
Cousineau, T. M., Goldstein, M., & Franko, D. L. (2004). A collaborative approach to nutrition
education for college students. Journal of American College Health, 53, 79–84.
Driskell, J. A., Kim, Y.-N., & Goebel, K. J. (2005). Few differences found in the typical eating and
physical activity habits of lower-level and upper-level university students. Journal of the
American Dietetic Association, 105, 798–801.

Dutta-Bergman, M. J. (2005). Psychographic profiling of fruit and vegetable consumption: The

role of health orientation. Social Marketing Quarterly, XI(1), 19–35.

Garcia, K., & Mann, T. (2003). From ‘I wish’ to ‘I will’: Social-cognitive predictors of behavioral
intentions. Journal of Health Psychology, 8, 347–360.

Glanz, K., Basil, M., Maibach, E., Goldberg, J., & Snyder, D. (1998). Why Americans eat what they
do: Taste, nutrition, cost, convenience, and weight control concerns as influences on food
consumption. Journal of the American Dietetic Association, 98, 1118–1126.

Goel, R. K. (2006). Obesity: An economic and financial perspective. Journal of Economics

and Finance, 30, 317–324.

Grace, T. W. (1997). Health problems of college students. Journal of American College

Health, 45, 243–250.

Hanauer, D., Dibble, E., Fortin, J., & Col, N. F. (2004). Internet use among community college
students: Implications in designing healthcare interventions. Journal of American College
Health, 52, 197–202.

Harman, H. H. (1976). Modern factor analysis. Chicago, IL: University of Chicago Press.

Holbrook, M. B., & Schindler, R. M. (1994). Age, sex, and attitude toward the past as predictors of
consumers’ aesthetic tastes for cultural products. Journal of Marketing Research, 31, 412–
422.

Horacek, T. M., & Betts, N. M. (1998). Students cluster into 4 groups according to the factors
influencing their dietary intake. Journal of the American Dietetic Association, 98, 1464–
1467.

House, J., Su, J., & Levy-Milne, R. (2006). Definitions of healthy eating among university students.
Canadian Journal of Dietetic Practice and Research, 67, 14–18.

Huntsinger, E. T., & Luecken, L. J. (2004). Attachment relationships and health behavior: The
mediation role of self-esteem. Psychology and Health, 19, 515–526.
 Kessler, L. A., Gilham, M. B., & Vickers, J. (1992). Peer involvement in the nutrition education of
college students. Journal of the American Dietetic Association, 92, 989–991.
Kloeblen, A. S., & Batish, S. S. (1999). Understanding the intention to perma- nently follow a high
folate diet among a sample of low-income pregnant women according to the health belief
model. Health Education Research, 14, 327–338.

Kristal, A. R., Hedderson, M. M., Patterson, R. E., & Neuhauser, M. L. (2001). Predic- tor of self-
initiated, healthful dietary change. Journal of the American Dietetic Association, 101, 762–
766.
Leslie, M., & Pierre, St. R. W. (1999). Osteoporosis: Implications for risk reduction in the college
setting. Journal of American College Health, 48, 67–71.

Maddux, J. E. (1993). Social cognitive models of health and exercise behavior: An introduction and
review of conceptual issues. Journal of Applied Sports Psychology, 5, 116–140.

Marquis, M. (2005). Exploring convenience orientation as a food motivation for college students
living in residence halls. International Journal of Consumer Studies, 29, 55–63.

Matvienko, O., Lewis, D. S., & Schafer, E. (2001). A college nutrition science course as an intervention to
prevent weight gain in female college freshmen. Journal of Nutrition Education, 33, 95–101.
Mayer, J., Dubbert, P., & Elder, J. (1989). Promoting nutrition at the point of choice: A review. Health
Education Quarterly, 16, 31–43.

Mooney, K. M., & Walbourn, L. (2001). When college students reject food: Not just a matter of taste.
Appetite, 36, 41–50.

Nicklas, T. A., Baranowski, T., Baranowski, J. C., Cullen, C., Rittenberry, L., & Olvera, N. (2001). Family
and child-care provider influences on preschool chil- dren’s fruit, juice, and vegetable consumption.
Nutrition Reviews, 59, 224–235. Racette, S. B., Deusinger, S. S., Strube, M. J., Highstein, G. R., &
Deusinger, R. H. (2005). Weight changes, exercise, and dietary patterns during freshman

and sophomore years of college. Journal of American College Health, 53, 245–251.

Rappoport, L. (2003). How we eat. Appetite, culture, and the psychology of food.

Toronto, Canada: ECW Press.

Rashad, I., & Grossman, M. (2004). The economics of obesity. The Public Interest, 156, 104–112.

Richards, A., Kattelmann, K. K., & Ren, C. (2006). Motivating 18- to 24-year-olds to increase their
fruit and vegetable consumption. Journal of American Dietetic Association, 106, 1405–
1411.
Rosenstock, I. M. (1974). Historical origins of the health belief model. Health Education
Monographs, 2, 328–335.

Sapp, S. G., & Jensen, H. H. (1998). An evaluation of the health belief model for predicting
perceived and actual dietary quality. Journal of Applied Social Psychology, 28, 235–
248.
Schnoll, R., & Zimmerman, B. J. (2001). Self-regulation training enhances dietary self-efficacy and
dietary fiber consumption. Journal of American Dietetic Association, 101, 1006–1011.
Steptoe, A., Pollard, T. M., & Wardle, J.
Sobal, J., & Nelson, M. K. (2003). Commensal (1995). Development of a measure of
eating patterns: A community study, the motives underlying the selection of
Appetite, 41, 181–190.
 food: The food choice questionnaire. management in college students. Journal
Appetite, 25, 267–284. of Nutrition Education, 33, 17–23.

Thomsen, P. A., Terry, R. D., & Amos, R. J.
(1987). Adolescents’ beliefs about and
reasons for using vitamin=mineral
supplements. Journal of the American
Dietetic Association, 87, 1063–1065.
Witte, K. (1992). Putting the fear back into fear
appeals: The extended parallel process
model. Communications Monograph,
59, 329–349.

Van den Reek, M., & Keith, R. (1984).

Knowledge and use of United Stated
dietary goals by university students.
Nutrition Reports International, 29,
559–564.
Von Ah, D., Ebert, S., Ngamvitroj, A., Park,
N., & Kang, D.-H. (2004). Predictors of
health behaviours in college
students. Journal of Advanced
Nursing, 48, 463–474.
Wdowik, M. J., Kendall, P. A., Harris, M. A., &
Auld, G. (2001). Expanded health belief
See discussions, stats, and author profiles for this publication
at: https://www.researchgate.net/publication/8500147
Leisure-time physical activity in
university students from 23
countries: Associations with
health beliefs, risk awareness,
and national economic
development
Article in Preventive Medicine · August 2004

model predicts diabetes self- DOI: 10.1016/j.ypmed.2004.01.028 · Source: PubMed

C
IT
A
T
IO
N
S READS

3 1,982
6
4 authors, including:

Anne M Haase
Victoria University of Wellington
121 PUBLICATIONS 3,325 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Anxiety and the development and maintenance of anorexia nervosa: protocol for a systematic review
View project

TREAD - exercise and depression trial View project

All content following this page was uploaded by Anne M Haase on 18 August 2020.

The user has requested enhancement of the downloaded file.

 Preventive Medicine 39 (2004) 182 – 190
www.elsevier.com/locate/ypmed

Leisure-time physical activity in university students from 23 countries: associations with

health beliefs, risk awareness, and national economic development
Anne Haase, Ph.D.,a Andrew Steptoe, D.Phil.,b,* James F.
Sallis, Ph.D.,c and Jane Wardle, Ph.D.a
a
Health Behaviour Unit, Department of Epidemiology and Public Health, University College London, London WC1E 6BT, UK
b
Psychobiology Group, Department of Epidemiology and Public Health, University College London, London WC1E 6BT, UK
c
Department of Psychology, San Diego State University, San Diego, CA 92182, USA

Available online 16 April 2004

Abstract

Background. Physical inactivity has been linked with chronic disease and obesity in most western populations. However, prevalence of inactivity, health beliefs,
and knowledge of the risks of inactivity have rarely been assessed across a wide range of developed and developing countries.
Methods. A cross-sectional survey was carried out with 19,298 university students from 23 countries varying in culture and level of economic development.
Data concerning leisure-time physical activity, health beliefs, and health knowledge were collected.
Results. The prevalence of inactivity in leisure time varied with cultural and economic developmental factors, averaging 23% (North- Western Europe and the
United States), 30% (Central and Eastern Europe), 39% (Mediterranean), 42% (Pacific Asian), and 44% (developing countries). The likelihood of leisure-time physical
activity was positively associated with the strength of beliefs in the health benefits of activity and with national economic development (per capita gross
domestic product). Knowledge about activity and health was disappointing, with only 40 – 60% being aware that physical activity was relevant to risk of heart
disease.
Conclusions. Leisure-time physical activity is below recommended levels in a substantial proportion of students, and is related to cultural factors and stage of national
economic development. The relationship between health beliefs and behavior is robust across cultures, but health knowledge remains deficient.

D 2004 The Institute for Cancer Prevention and Elsevier Inc. All rights reserved.

Keywords: Coronary heart disease; Psychosocial; International; Prevalence; Exercise

Introduction individual countries indicate that the prevalence of adequate

The health benefits of leisure-time physical activity are
widely recognized, as inactivity is associated with increased
risk of coronary heart disease, various cancers, obesity, and
other health problems [1,2]. The recommendation to increase
physical activity is a key element of health-promotion strate-
gies in many countries, and of international initiatives targeted at
developed and developing countries such as the WHO Mega
Country Health Promotion Network [3,4]. Surveys from
* Corresponding author. Department of Epidemiology and Public Health, University
College London, 1-19 Torrington Place, London WC1E 6BT, UK. Fax: +44-20-7916-8542.
E-mail address: a.steptoe@ucl.ac.uk (A. Steptoe).
physical activity is relatively high in children and adolescents
[5,6] but substantially lower in adults, suggesting that late
adolescence and early adult life may be a critical period
of transition [7,8]. It is important therefore to monitor trends
in physical activity in young adults, and to understand factors
such as attitudes and knowledge of health benefits that may
be associated with activity levels.
International comparisons are valuable from the perspec-
tive of global public health since they help define variations
in physical activity in different cultures, point to common
determinants across countries, suggest countries or regions
in most need of intervention, and highlight good practice in
health promotion. Unfortunately, the measurement of lei-
sure-time physical activity is complex, and surveys using

0091-7435/$ - see front matter D 2004 The Institute for Cancer Prevention and Elsevier Inc. All rights reserved.
doi:10.1016/j.ypmed.2004.01.028
 A. Haase et al. / Preventive Medicine 39 (2004) 182–190
different methods are difficult to compare [9]. There have
been relatively few substantial international studies of
comparable samples from different countries [5,10 – 12],
and these have predominately focused on Europe. Informa-
tion concerning leisure-time physical activity in developing
and Pacific Asian countries is sparse. Consequently, it has
not yet been possible to formulate models of the evolution
of leisure physical activity in relation to levels of national
economic development, as have been proposed for other
health-related behaviors such as smoking [13].
This study assessed leisure-time physical activity in
university students from 23 countries, analyzing the preva-
lence of inactivity and activity at recommended levels of
three or more times a week [14]. We investigated whether
there were systematic differences in physical activity across
countries related to national economic development, and to
geographical, cultural, and political groupings. The strength
of individual attitudes concerning the benefits of physical
activity and knowledge about the role of physical activity in
preventing chronic disease was also evaluated. These psy-
chological variables are relevant for planning health educa-
tion, and may be associated with increased frequency of
leisure-time physical activity [15]. We used multilevel
modeling to assess the associations between leisure-time
physical activity and individual-level factors (attitudes, risk
awareness) and ecological-level factors (national economic
development) simultaneously. Because data were collected
from university students rather than representative samples
of young adults in each country, we cross-validated inter-
national differences in reported activity levels by compari-
son with existing data sets that included several of the
countries sampled [10].
Methods
These analyses were based on data from the International
Health and Behaviour Survey (IHBS), a cross-sectional
questionnaire study of health behaviors and attitudes in
19,298 university students from 23 countries, carried out
between 1999 and 2001. Assessments of physical activity
were identical to those included in the European Health
Behaviour Survey (EHBS), conducted between 1989 and
1991 [12,16]. The study was carried out with a network of
collaborators in participating countries (see Acknowledge-
ments). The countries involved were Belgium, Bulgaria,
Colombia, France, Germany, Greece, Hungary, Iceland,
Ireland, Italy, Japan, South Korea, the Netherlands, Poland,
Portugal, Romania, Slovakia, South Africa, Spain, Thailand,
United Kingdom, United States, and Venezuela. The ques-
tionnaire used for data collection was developed in English,
then translated and back-translated into 17 languages (Bul-
garian, Dutch, Flemish, French, German, Greek, Hungarian,
Icelandic, Italian, Japanese, Korean, Polish, Portuguese,
Romanian, Slovak, Spanish, and Thai). Collaborators col-
lected data from university students aged 17 – 30 years who
18
3
were not studying medicine or health-related topics. Be-
tween one and three universities were sampled in each
country. A variety of students were involved, including
those studying physical sciences, engineering, law, social
science, languages, geography, history, and economics. Data
collection was standardized, with the questionnaire being
administered to large groups of students usually at the end of
classes, so participation rates in most countries exceeded
90%. The sample size ranged from 471 to 2,028 respondents
per country. Respondents were told that the survey
concerned activities related to health and that an interna-
tional comparison was being carried out, but no further
details were given. Data from the United Kingdom were
collected in England; the German sample came from uni-
versities in the former Federal Republic (West Germany);
the South African data from a university with predominantly
black students; and the United States’ sample from univer-
sities in California and Colorado.
Two items were used to measure leisure-time physical
activity. The first item asked whether the individual had
taken any exercise (e.g., sport, physically active pastime) in
the past 2 weeks, and the second item asked for the number
of physical activity sessions over that time period. These
items are similar to those used in several behavioral risk
factor surveillance systems, such as the STEPS (Stepwise
Approach to Surveillance of Noncommunicable Disease
Risk Factors) and CARMEN (Conjunto de Acciones para la
Reduccion Multifactorial de Enfermedades No transmi-
sibles) programs [4]. Reliability data for these measures are
published elsewhere [17,18]. Participants were asked what
type of physical activity they carried out. These data are not
detailed in this paper, but the dominant forms of activity
were jogging/running, swimming, football (soccer), and
aerobics. A belief rating was completed of how important
physical activity was for the maintenance of health, ranging
from 1 = not at all important to 10 = very important.
Knowledge about the specific health benefits of physical
activity was assessed in a section of the questionnaire
inquiring about the influence of a range of lifestyle factors
such as smoking and alcohol on specific medical conditions
[19]. We analyzed the proportion of respondents who stated
that physical activity was related to the development of heart
disease. The economic development level of participating
countries was assessed by per capita gross domestic product
(GDP), abstracted from the World Bank development indi-
cators [20].
There are no data sets describing levels of leisure
physical activity in samples from all the countries included in
this study. The WHO cross-national study of health
behavior in school children aged 11 – 15 years [5] is not
comparable, because physical activity levels at this age are
strongly determined by school policies [21]. The largest
number of countries are represented in a pan-European
study of adults [10], so prevalence levels from 10 of the
countries in the present study were compared with these
results.
 The statistical analyses involved comparing leisure
physical activity, beliefs, and risk awareness in each
country sample, and in groups of countries defined by
cultural, geographical, and political criteria. Five groups of
countries were created: North-Western Europe and the
United States (Belgium, England, France, Germany, Ice-
land, Ireland, the Netherlands, and the United States), the
former socialist states of Central and Eastern Europe
(Bulgaria, Hungary, Poland, Romania, and Slovakia),
Mediterranean countries (Greece, Italy, Portugal, and
Spain), countries of the Pacific-Asian rim (Japan, Korea,
and Thailand), and developing countries (Colombia,
South Africa, and Venezuela).
Physical activity data were analyzed by dividing the
sample into three groups: inactive, low-frequency activity,
and recommended frequency activity. The ACSM 1990
guideline of exercising three or more times a week (six
times in 2 weeks) was used as a reference for determin-
ing whether people reached recommended levels of
activity [14]. This recommendation was embodied in
Healthy People 2000 [22]. Although public health guide- lines
for physical activity have been revised, this guide- line is
widely supported as identifying levels of physical activity
producing health benefits, and was used in the analyses of
physical activity carried out with the earlier EHBS [12].
The prevalence of inactivity, low-frequency activity, and
recommended activity was computed for men and women
in each country sample. There were small differences in
average age between countries, so data are presented as
age-adjusted prevalence with 95% confidence intervals
(CI).
The belief ratings were skewed toward positive scores.
These data were therefore analyzed by dichotomizing scores
as 9 or 10 (indicating strong beliefs) or 8 or less
(indicating weaker beliefs). The proportion of respondents
in each country or group of countries who had strong
beliefs in the importance of activity for health was then
analyzed.
Fig. 1. Age-adjusted prevalence (%) of physical inactivity in university students from 23 countries. Black bars: men. Shaded bars: women.
Awareness of the role of physical activity in heart disease
was analyzed as a binary variable. Associations between
beliefs, health knowledge, national economic development
(GDP), and physical activity were analyzed using multilevel
logistic regressions, assessing the odds of any leisure
physical activity, and the odds of recommended frequency
activity. Gender (with women coded 1 and men coded 0)
and age were included in these models. The multilevel
regressions were carried out using a random intercepts
maximum likelihood logit model in STATA. All analyses
involving groups of countries involved clustered data, so
confidence intervals were adjusted to take account of intra-
cluster effects using STATA.
Results
Prevalence of leisure-time physical activity
The age-adjusted prevalence of physical inactivity is
shown in Fig. 1, and the prevalence of activity at recom-
mended levels in Fig. 2. Overall, more women than men
reported no leisure-time physical activity (38%, CI 37 – 39
vs. 27%, CI 26 – 28, P < 0.0001). The differences were
significant ( P < 0.01 or less) in 16 of the 23 countries, but
were not reliable in Germany, Hungary, Iceland, Ireland,
Japan, the Netherlands, or the United States. The prevalence
of inactivity varied markedly across country samples, rang-
ing from 11% (Belgium) to 41% (Portugal and South
Africa) among men, and from 15% (USA) to 65% (Portu-
gal) among women. Leisure-time physical activity at rec-
ommended levels was more common in men (28%, CI 27 –
29) than women (19%, CI 18 – 20, P < 0.0001), and
differed widely between country samples. The lowest
prevalence of leisure-time physical activity at recommended
levels was in Japan and Germany for men, and Japan for
women, while the highest levels were reported by
students from Poland
Fig. 2. Age-adjusted prevalence of recommended frequency (six times in the past 2 weeks) leisure-time physical activity in university students from 23 countries. Black bars: men.
Shaded bars: women.

(men) and Iceland and Ireland (women). Interestingly, there 0.72 ( P = 0.02) with the European Union study, providing
was no overall difference in the proportion of men and external corroboration for the differences between countries
women active at low frequency (45% CI 44 – 46 for men, reported here.
43% CI 42 – 44 for women), indicating that the gender
difference was due to the proportions wholly inactive or
Table 1
active at recommended levels.
The prevalence of leisure-time physical activity by coun-
Prevalence of inactivity, low-frequency leisure-time activity, and recom- mended
frequency activity in five groups of countries
try group is shown in Table 1. Significant trends in Inactive Low- Recommended
inactiv- ity across groups of countries were present for both frequency frequency activity
men and women. Levels of inactivity were lowest in North- activity
Western Europe and the United States, and highest in Men
North-Western 21% (16 – 49% (43 – 30% (22 –
developing countries. There were less-consistent differences 26) 55) 39)
between groups of countries in activity at recommended Europe + United
levels; in neither men nor women did the prevalence of States (n = 2617) 25% (17 – 43% (35 – 32% (25 –
recommended activity differ between North-Western Europe Central and Eastern 35) 52) 40)
and the Unit- ed States, Central and Eastern Europe, and Europe (n = 1890)
Mediterranean 30% (22 – 40% (35 – 30% (19 –
Mediterranean countries. A significant trend across groups of 40) 45) 44)
countries was present in low-frequency activity in women (n = 1859)
but not men ( P Pacific Asian (n = 827) 34% (33 – 48% (42 – 18% (14 –
34) 54) 24)
< 0.001). A higher prevalence of low-frequency physical Developing countries 35% (28 – 43% (35 – 23% (22 –
activity was reported from North-Western Europe and the 42) 51) 24)
United States than in Mediterranean and developing (n = 1196)
P for trend across 0.05 n.s. n.s.
countries.
The per capita GDP in 2001 ranged in these countries groups of countries
for men
from more than $35,000 (USA) to less than $2,000
(Bulgaria, Colombia, South Africa, Thailand). The prev- Women
North-Western 24% (17 – 54% (47 – 22% (17 –
alence of leisure-time physical activity at any level was 32) 60) 29)
positively correlated with economic development (r = Europe + United
0.49, P = 0.02), so activity was more prevalent in more States (n = 33% (25 – 47% (39 – 19% (18 –
developed countries. 3618) Central and 43) 55) 21)
Eastern
Europe (n = 2276)
Comparison with European Union data Mediterranean 46% (34 – 32% (26 – 22% (15 –
59) 38) 32)
(n = 2370)
The prevalence of leisure-time activity at any level in the (n = 1148)
10 countries that participated in the comparison European P for trend across groups of 0.001 0.001 n.s.
Union study was 69%, not significantly different from the countries for women
67% reported by Martinez-Gonzalez et al. [10]. In addition, Note. Prevalence levels are age-adjusted. Ninety-five percent confidence intervals are adjusted for
the prevalence of activity in the present study correlated r = intracluster effects.
 Table 2
Knowledge of the influence of physical inactivity on heart disease in five groups of
countries
Men Women
North-Western Europe + United States 48% (42 – 52% (47 –
54) 57)
Central and Eastern Europe 43% (22 – 43% (24 –
67) 65)
Mediterranean 44% (33 – 45% (36 –
56) 54)
Pacific Asian 46% (24 – 44% (21 –
70) 70)
Developing countries 30% (16 – 26% (12 –
48) 47)
Note. Prevalence levels are age-adjusted. Ninety-five percent confidence intervals are adjusted
for intracluster effects.

Knowledge of health benefits

The proportion of male and female students who were

Fig. 3. Proportion of students with strong beliefs in the importance of physical activity
for health, in relation to reported frequency of leisure- time physical activity—Inactive:
no leisure-time physical activity in the past 2 weeks. Low freq PA: five or fewer
episodes in the past 2 weeks. Recommended PA: six or more episodes in the past
2 weeks. Black bars: men. Shaded bars: women. Error bars are standard errors of the
mean.
Beliefs in the importance of physical activity for health
Strong beliefs (ratings of 9 or 10) about the
importance of physical activity for health were related to
behavior in a dose-dependent fashion ( P < 0.0001). This
is illustrated in Fig. 3, where it is evident that the
proportion of respond- ents with strong beliefs was lowest
among inactive stu- dents, and greatest in those active at
recommended levels. This pattern did not differ by gender.
Nor were there differences in the prevalence of strong
beliefs about the importance of physical activity for health
across groups of countries. The correlation between
economic development and prevalence of strong beliefs
was not significant (r = 0.19, P = 0.41).
aware of the association between heart disease and physical
activity is shown in Fig. 4. In general, the levels of
knowledge were disappointing; even in Western developed
countries with established traditions of health education,
only 40 – 60% of students were aware that physical
activity was relevant to risk of heart disease. Overall, the
proportion of men (43%, CI 42 – 44) and women (45%
(CI 44 – 46) aware of the association was similar, with
no marked differences in any country sample. The
proportions of men and women in each country aware of
the association between physical activity and heart
disease was strongly correlated (r = 0.97, P < 0.0001).
Levels of knowledge of health benefits varied consider-
ably between country samples, with high levels in the
Netherlands, Slovakia, Greece, and Japan, and poor levels of
knowledge in Romania, Thailand, and South Africa. The
levels of knowledge in the five groups of countries com-
pared in this study are summarized in Table 2. With the
exception of the low levels of awareness in developing
countries, there were no systematic variations across the
other groups of countries. This suggests that knowledge

Fig. 4. Age-adjusted proportion of university students from each country who knew of the association between physical activity and heart disease. Black bars: men. Shaded bars: women.
 Table 3
Logistic regressions on leisure-time physical activity
Odds of any activity
(vs. inactivity)
Odds ratio P Odds ratio
(95% CI)
Age 0.99 (0.98 – 0.088
1.001)
Gender 0.57 (0.53 – 0.001
0.61)
Beliefs in 2.82 (2.62 – 0.001
3.03)
health
benefits
Knowledge 1.07 (0.99 – 0.055
1.15)
of physical
activity/heart
The reference category for odds ratios was inactivity in both analyses.
about the specific health benefits of physical activity is more
strongly associated with economic development than with
cultural and political factors. This impression was confirmed
by the positive correlations across countries between prev-
alence of health knowledge and economic development (r =
0.50, P = 0.019).
Associations between physical activity, beliefs, health
knowledge, and economic development
The multilevel regression analyses on any leisure-time
physical activity, and on physical activity at recommended
frequency, are summarized in Table 3. The likelihood of
being physically active at any level was greater in men than
women, in those with stronger health beliefs (odds ratio
2.82, CI 2.62 – 3.03), and in respondents from more eco-
nomically developed nations (odds ratio 1.38, CI 1.33 –
1.43). A similar pattern emerged in the analyses of leisure-
time physical activity at recommended levels. Gender,
health beliefs, and national economic development were
significantly associated with activity, while age and knowl-
edge of links between physical activity and heart disease
were not.
Discussion
Analysis of the prevalence of physical activity confirmed
our expectation that there would be wide variations between
country samples. Similar variations in Europe have been
observed in other studies [10 – 12]. We used two overlapping
strategies to try to understand the pattern across countries.
The first was to group countries by political, cultural, and
geographical criteria. The countries included in the North-
Western European and the United States category are all
mature democracies and market economies. The five Central
and Eastern European countries are emerging market econ-
omies that were socialist states until the collapse of com-
munism. The countries of the Mediterranean region share
cultural characteristics in terms of diet and lifestyle. The
Odds of activity at countries allocated to the Pacific Asian group were more
(vs. inactivity) heterogeneous, including Japan (an economically advanced
recommended frequency
P country with high per capita GDP), less affluent Thailand,
(95% CI) and intermediate Korea. The remaining three countries from
0.99 (0.99 – 0.88 Africa and South America were classified as being at a
1.00) relatively early stage of economic development. This group-
0.58 (0.53 – 0.001 ing of countries proved to have explanatory power in
0.63)
3.07 (2.83 – 0.001 relation to the prevalence of inactivity (Table 1). Inactivity
3.34) was least prevalent in both men and women from North-
Western Europe and the United States, increasing progres-
sively across country groups, so that inactivity was most
1.04 (0.96 – 0.30 prevalent in the developing country group.
1.13)
The second method of classifying countries was by using
an index of economic development. The two strategies
overlapped, since the countries of North-Western Europe and
the United States were the most affluent, while the
developing countries were by definition least affluent. But
the other countries in this survey were differently ranked
from their position in the five country grouping. For
example, economic development in terms of per capita
GDP was low in Central and Eastern European participating
countries, while being high in Mediterranean countries. The
classification of countries in terms of economic develop-
ment also showed significant associations with the preva-
lence of physical inactivity among students, independently of
age, gender, and health beliefs.
These findings suggest that there is a broad association
between the economic development stage of countries and
leisure-time physical activity, coupled with specific cultural
and geopolitical determinants. Levels of leisure physical
activity in young adults are generally higher in more
economically developed countries, with the exception of
Mediterranean countries. However, an interesting result is
that this pattern applied more strongly to the presence or
absence of any leisure physical activity, rather than to
physical activity at recommended levels. In the regression
analyses (Table 3), the odds associated with higher GDP
were reduced by 45% in the analysis of recommended
frequency GDP compared with any physical activity. Thus,
cultural and economic factors are less relevant to the
likelihood that university students exercise at a high fre-
quency. It may be that this behavior is more strongly
influenced by other factors (such as involvement in sports)
apart from broad economic and cultural influences.
Exercising three or more times a week was used as the
criterion for attaining recommended levels of activity. This
criterion established by the ACSM in 1990 has been
superseded in countries such as the United States with
public health recommendations based on the accumulation of
30-min, moderate-intensity physical activity on most days
of the week [1,23]. This level is more difficult to quantify
in surveys of the type used here, and the recom-
mendation has not currently been adopted in many
countries, so the older criterion was applied.
 There were strong associations between reported leisure
physical activity and the strength of beliefs in health
benefits, which emerged both in bivariate (Fig. 3) and
multivariate (Table 3) analyses, consistent with previous
findings with adults [15]. The causal sequence cannot be
determined in this study. Beliefs in health benefits may
stimulate physical activity, but might also emerge in people
who are currently active. The positive relationship was
evident in a wide range of cultural groups, climatic con-
ditions, and political settings, though it is interesting that the
associations were greater in North-Western Europe and the
United States than in other groups of countries. However, the
proportion of participants with strong beliefs in health
benefits did not vary by gender or across country samples.
Thus, it was not the case that students from the
developed countries such as the United States, Japan, and
North- Western Europe held stronger beliefs in health
benefits on average. It may be therefore that other motives
apart from health benefits determine the higher prevalence
of leisure physical activity in economically developed
countries. Fac- tors such as enjoyment, social support, and
trying to lose weight have been identified as correlates of
leisure-time physical activity in different samples [24,25], so
it would be valuable to evaluate a broader range of
physical activity correlates in future international studies.
Increasing the proportion of young people with strong
beliefs in health benefits might be a valuable component
of international health promotion.
A cornerstone of health education is transmitting infor-
mation about the diseases associated with lifestyle choices,
and this was indexed in the present study by knowledge
of the connection between physical activity and heart
disease. The levels of knowledge of links with heart disease
were not high in the study (Fig. 4), and exceeded 50% of
both men and women in only 8 of 23 country samples. In
an analysis of a subset of countries in this study, we have
shown that levels of knowledge changed little between 1990
and 2000 [16], so the penetration of health education
remains limited. Higher levels of knowledge were not related
to the country grouping (Table 2), but were associated
with greater eco- nomic development. More
economically developed countries may expend greater
resources on health education, while access to mass media
is also broader. Nonetheless, knowledge was not
associated with behavior. As shown in Table 3, knowledge
of the relationship between physical activity and heart
disease was not associated with leisure- time physical
activity. These null findings are consistent with previous
research [15]. Thus, although national eco- nomic
development may relate to health knowledge, this does
not translate at an individual level into participation in
leisure-time physical activity. Improving knowledge about
health effects should not be expected to be an effective
physical activity promotion strategy, even in less developed
countries.
The explanation for the gradient in physical activity
across students from countries varying in economic devel-
opment and cultural background is not certain. The pattern is
unlikely to be due to compensatory differences in non-
leisure physical activity since all the participants in this
sample were university students and not engaged in physical
labor. We investigated the issue in detail in a previous study
[12], and found that the contribution to the overall level of
physical activity of walking or bicycling for the purposes
of transportation amongst students was very small. Several
other factors might be relevant. Engaging in leisure-time
physical activity for health reasons in young people is
predicated on a desire to increase longevity in middle and
old age. Less economically developed countries have lower
overall life expectancies than affluent countries, so motiva-
tion for actions that extend life in older age may still
be limited. It can be argued that leisure-time physical activity
is strongly influenced by the cultural significance of
inactivity; in countries in which manual labor remains
common, inactivity may be a marker of prestige. Health
promotion targeted at chronic noncommunicable diseases
has until recently been a low priority in developing
countries [4], resulting in less encouragement of leisure-
time physical activity. The facilities for leisure-time
physical activity may be more limited in universities
from less affluent countries, leading to environmental
constraints on partici- pation. Students at universities in less
economically devel- oped countries may place greater
value on academic achievement, and may commit less
time in nonacademic leisure pursuits than individuals
from more privileged societies. There are probably wide
cultural variations in interest in sport and in sport
participation that have yet to be quantified systematically,
but which may also contribute to the differences across
country samples reported here.
This survey was not carried out with representative
samples of young adults from the countries involved.
University students are better educated and generally health-
ier than other sectors of society [26]. University educated
men and women engage in more leisure-time physical
activity then less well educated individuals in many
countries [27]. Levels of leisure physical activity are there-
fore likely to have been higher than would have been
reported from the population in general. There were several
reasons for carrying out the IHBS with this methodology
[19]. When making comparisons across countries, it is
necessary to compare like with like, and university students
are an easily identifiable, accessible, and homogenous
group. Students represent a significant sector of young
adults, and there is considerable concern about their health
behavior [25,28]. In most societies, graduates from univer-
sity take up more prestigious professional and managerial
occupations than less well-educated individuals, so are
likely to hold influential positions in education and policy
making in the future. In addition, conducting the study in
classes meant that response rates were high.
Our confidence in the accuracy of the differences in
leisure-time physical activity between country samples was
strengthened by the positive correlation (r = 0.72) between
levels reported in this survey and the representative samples
assessed by Martinez-Gonzalez et al. [10]. The differences
between countries were also similar to those reported in the
smaller selection of countries assessed by Stahl et al. [11].
Thus, although the overall levels of leisure physical activity
may not be representative, the relative levels reported by
different country samples appear reliable. In a previous
analysis, we have found that levels of physical activity
remained rather stable in students assessed in the EHBS
between 1989 and 1991 and this study in 1999 – 2001 in
13 countries included in both studies [16]. In that
comparison, the prevalence of physical activity at any
level in each country correlated r = 0.72 in men and r =
0.89 in women (both P < 0.001) across the 10-year
period.
Other limitations to this study should also be acknowl-
edged. The study was cross-sectional and based on self-
report measures, and causal conclusions cannot be drawn.
The universities involved in this study were not selected at
random, and there may be important variations within
countries in physical activity levels relating to geographical
location and the culture of the university. Physical activity
was assessed with simple items that did not discriminate
variations in the intensity or duration of each bout of
activity. Some individuals may have carried out activity at
a modest intensity for a short time, while others could have
been active at high intensities for long periods. Only simple
health beliefs were measured in this study, but many
different cognitive and attitudinal factors are known to
correlate with physical activity, and examining a broader
range of psychosocial variables may help explain the
variation in frequency of leisure-time physical activity
across countries. The measure of health knowledge was
also limited, and more comprehensive assessments may be
informative [29]. Nevertheless, the present study indicates
that between one fifth and one half of university students do
not engage in leisure physical activity, and that knowledge of
health beliefs is limited. Further investigation of interna-
tional differences may highlight particular mediators of
physical activity that could be targeted by health promotion
in less economically developed countries, to help these
states avoid the high levels of obesity and chronic non-
communicable disease prevalent in the most affluent
countries of the world.
Acknowledgments
This research was supported by Cancer Research UK and
by the Economic and Social Research Council, UK. The
following colleagues participated in data collection for the
International Study of Health and Behaviour—Diepenbeek,
Belgium: Jan Vinck; Sophia, Bulgaria: Irina Todorova;
Bogota´, Colombia: Pablo Sanabria-Ferrand and Diana
Urrego; Paris, France: France Bellisle and Anne Marie
Dalix; Wuppertal and Marburg, Germany: Claus Vo¨gele and
Gudrun Sartory; Athens, Greece: Bettina Davou and
Antonis Armenakis; Budapest, Hungary: Maria Kopp and
Reka Baranyai; Reykjavik, Iceland: Sigurlina Davidsdottir;
Dublin, Ireland: Ray Fuller; Kurume, Japan: Akira Tsuda;
Turin, Italy: Anna Maria Zotti, Gabriella Pravettoni and
Massimo Miglioretti; Groningen, The Netherlands: Robbert
Sanderman; Poznan, Poland: Helena S`ık and Micha3 Ziarko;
Lisbon, Portugal: Joao Justo; Cluj-Napoca, Romania:
Adriana Baban; Bratislava, Slovakia: Gabriel Gulisˇ;
Sovenga, South Africa: Karl Peltzer; Granada, Spain: Jaime
Vila, Nieves Perez, Humbelina Robles, Nieves Vera
Guerrero; Bangkok, Thailand: Kiriboon Jongwutiwes and
Maream Nillapun; San Diego, CA and Greeley, CO, USA:
Kelli Glass and Sacha Pampalone; Caracas, Venezuela: Nuri
Bages.
References
[1] U.S. Department of Health and Human Services. Physical activity
and health: a report of the surgeon general. U.S. Department of
Health and Human Services, Centers for Disease Control and
Prevention. Atlanta, GA; 1996.
[2] Vainio H, Bianchini F. IARC handbooks of cancer prevention: weight
control and physical activity. IARC. Lyons; 2002.
[3] U.S. Department of Health and Human Services. Healthy
people 2010. US Department of Health and Human Services.
Washington, DC; 2000.
[4] World Health Organization. WHO mega country health
promotion network: behavioural risk factor surveillance guide.
World Health Organization. Geneva; 2002.
[5] Currie C, Hurrelmann K, Settertobulte W, Smith R, Todd J.
Health and health behaviour among young people. World Health
Organiza- tion. Copenhagen; 2000.
[6] Pate RR, Freedson PS, Sallis JF, Taylor WC, Sirard J, Trost SG,
et al. Compliance with physical activity guidelines: prevalence in a
popu- lation of children and youth. Ann Epidemiol
2002;12:303 – 8.
[7] Caspersen CJ, Pereira MA, Curran KM. Changes in physical activity
patterns in the United States, by sex and cross-sectional age. Med
Sci Sports Exerc 2000;32:1601 – 9.
[8] Leslie E, Fotheringham MJ, Owen N, Bauman A. Age-related
differ- ences in physical activity levels of young adults. Med Sci
Sports Exerc 2001;33:255 – 8.
[9] Sarkin JA, Nichols JF, Sallis JF, Calfas KJ. Self-report measures
and scoring protocols affect prevalence estimates of meeting
physical ac- tivity guidelines. Med Sci Sports Exerc
2000;32:149 – 56.
[10] Martinez-Gonzalez MA, Varo JJ, Santos JL, DeIrala J, Gibney
M, Kearney J, et al. Prevalence of physical activity during
leisure time in the European Union. Med Sci Sports Exerc
2001;33:1142 – 6.
[11] Stahl T, Rutten A, Nutbeam D, Bauman A, Kannas L, Abel T, et
al. The importance of the social environment for physically active
lifestyle— results from an international study. Soc Sci Med
2001;52:1 – 10.
[12] Steptoe A, Wardle J, Fuller R, Holte A, Justo J, Sanderman R, et
al. Leisure-time physical exercise: prevalence, attitudinal and
behavioral correlates in young Europeans from 21 countries.
Prev Med 1997; 26:845 – 54.
[13] Lopez AD, Collishaw NE, Piha T. A descriptive model of the
ciga- rette epidemic in developed countries. Tob Control
1994;3:242 – 7.
[14] American College of Sports Medicine Position Stand. The
recommen- ded quantity and quality of exercise for developing and
maintaining cardiorespiratory and muscular fitness in adults. Med
Sci Sports Exerc 1990;22:265 – 74.
[15] Trost SG, Owen N, Bauman AE, Sallis JF, Brown W. Correlates
of adults’ participation in physical activity: review and update. Med
Sci Sports Exerc 2002;34:1996 – 2001.
[16] Steptoe A, Wardle J, Cui W, Bellisle F, Zotti A.M, Baranyai R, et
al. Trends in smoking, diet, physical exercise, and attitudes toward
health in European university students from 13 countries, 1990 –
2000. Prev Med 2002;35:97 – 104.
[17] Wardle J, Steptoe A. The European health and behaviour
survey: rationale, methods and initial results from the United
Kingdom. Soc Sci Med 1991;33:925 – 36.
[18] Steptoe A, Sanderman R, Wardle J. Stability and changes in
health behaviours in young adults over a one year period.
Psychol Health 1995;10:155 – 69.
[19] Steptoe A, Wardle J. The European health and behaviour survey:
the development of an international study in health psychology.
Psychol Health 1996;11:49 – 73.
[20] World Bank. World Bank development report. New York: World
Bank and Oxford Univ Press; 2003.
[21] Sallis JF, Conway TL, Prochaska JJ, McKenzie TL, Marshall
SJ, Brown M. The association of school environments with
youth phys- ical activity. Am J Public Health 2001;91:618 –
20.
[22] US Department of Health and Human Services. Healthy people
2000. US Department of Health and Human Services.
Washington, DC; 1991.
[23] American College of Sports Medicine Position Stand. The
recommen- ded quantity and quality of exercise for developing and
maintaining cardiorespiratory and muscular fitness, and
flexibility in healthy adults. Med Sci Sports Exerc 1998;30:975
– 91.
[24] Leslie E, Owen N, Salmon J, Bauman A, Sallis JF, Lo SK.
Insuffi- ciently active Australian college students: perceived
personal, social, and environmental influences. Prev Med
1999;28:20 – 7.
[25] Lowry R, Galuska DA, Fulton JE, Wechsler H, Kann L,
Collins JL. Physical activity, food choice, and weight
management goals and practices among US college students. Am
J Prev Med 2000;18:18 – 27.
[26] Patrick K, Grace TW, Lovato CY. Health issues for college
students. Annu Rev Public Health 1992;13:253 – 68.
[27] Pratt M, Macera CA, Blanton C. Levels of physical activity
and inactivity in children and adults in the United States: current
evidence and research issues. Med Sci Sports Exerc
1999;31:S526 – 33.
[28] Wechsler H, Rigotti NA, Gledhill-Hoyt J, Lee H. Increased levels
of cigarette use among college students: a cause for national
concern. JAMA 1998;280:1673 – 78.
[29] Beier ME, Ackerman PL. Determinants of health knowledge: an
in- vestigation of age, gender, abilities, personality, and interests. J
Pers Soc Psychol 2003;84:439 – 48.

View publication stats

Yahia et al. BMC Public Health (2016) 16:1047

DOI 10.1186/s12889-016-3728-z

RESEARCH ARTICLE Open Access

Level of nutrition knowledge

and its association with fat
consumption among college
students
Najat Yahia1*, Carrie A. Brown2,4, Melyssa Rapley1 and
Mei Chung3
 using one- way ANOVA, chi-square, and student’s t-test.
0 mg of cholesterol daily had lower mean nutrition scores than those students with lower fat or cholesterol intake (8 po
on intervention studies to assess the efficacy of nutrition knowledge on pattern of fat consumption

Background by year 2030 [4]. Recent estimates from the 2013 Behav-
Cardiovascular disease (CVD) is the leading cause of
ioral Risk Factor Surveillance System (BRFSS) indicated that
death in the United States [1, 2]. According to the
40.6 % of Michigan adults aged 18 years and older
Centers for Disease Control and Prevention (CDC),
reported having high blood cholesterol levels, with higher
about 600,000 American adults die annually of heart dis-
prevalence among white non-Hispanic adults (42.2 %)
ease, or 1 of every 4 deaths [3]. According to a report
than Black, non-Hispanic adults (36.1 %) [5].
published by the Michigan Department of Community
A diet high in saturated fat, trans fat, and cholesterol is
Health in 2015, the number of cases of heart disease in
known to raise levels of serum blood cholesterol and can
Michigan is projected to rise from 600,000 to 2.9 million negatively impact cardiovascular health [6]. Accord- ingly,
the American Heart Association recommends that
healthy adults should limit their intake of saturated fat
* Correspondence: yahia1n@cmich.edu
to less than 7 % of total daily calories, trans fat to less
1
Department of Human Environmental Studies, Central Michigan University, than 1 % of total daily calories, and cholesterol to less
Wightman 108, Mt. Pleasant, MI 48859, USA
Full list of author information is available at the end of the article

© 2016 The Author(s). Open Access This article is distributed under the terms of the Creative Commons Attribution 4.0 International
License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium,
provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and
indicate if changes were made. The Creative Commons Public Domain Dedication waiver
(http://creativecommons.org/publicdomain/zero/1.0/) applies to the data made available in this article, unless otherwise stated.
Yahia et al. BMC Public Health (2016) Page 2 of 10
16:1047

Michigan University (CMU)

than 300 mg a day [7, 8]. Likewise, the 2010
Dietary Guidelines for Americans call for reductions
in the con- sumption of saturated and trans fat such as
cream, but- ter, beef fat (tallow, suet), chicken fat,
pork fat (lard), stick margarine, and shortening [9].
College is a critical period in life since many
lifestyle habits are formed and may persist into
adulthood, thereby impacting health [10–13]. Several
investigators have reported unhealthy dietary practices
among college students such as increasing
consumption of fast food and high-fat foods among
college students [9–21]. Butler et al. reported
significant increases in the percentage of fat and the
number of alcoholic beverages consumed daily in a
sample of female students at a Midwestern
university in the U.S. during their first year of college
[20]. Other authors reported a significantly higher
intake of total and saturated fat and a lower intake of
polyun- saturated and monounsaturated fat in a
sample of uni- versity students, when compared to
the American Heart Association recommendations
[21]. The U.S. Surgeon General, in two previous
reports, titled “Nutritional Prevention
Strategy/Health Eating” and “The Surgeon General's
Vision for a Healthy and Fit Nation”, identified
universities and schools as sites to raise awareness
and educate students about nutrition to help them
under- stand and apply the Dietary Guidelines for
Americans recommendations to reduce their risk for
heart disease [22, 23].
Literature on the impact of nutrition education
on
dietary practices has been mixed. Several investigators
have reported that nutrition education may change
stu- dents’ dietary habits and food choices [24–28]
while others reported no significant correlation
between nutri- tion knowledge and food choices [29,
30]. Indeed, there could be many factors that impinge
on a students’ diet- ary behaviors, but some basic
understanding of nutrition is necessary for a diet
change to occur [31]. Whether nu- trition education
can provide students with the power to select
healthier food choices and how well students apply
food-related knowledge into their daily dietary
practices deserve further attention. Thus, this pilot
study aimed to explore whether higher nutrition
knowledge level is associated with a lower level of fat
consumption in a sample of university students
from Central Michigan University.

Methods
Design and sample
This study was a cross-sectional survey. A sample of
231 students (71 % females and 29 % males), with
a mean age of 20 years, was drawn from the
university campus. Students were recruited in the
classroom (Foods and Nutrition classes) and online
(Blackboard announ- cements) by a Central
Yahia et al. BMC Public Health (2016) Page 3 of 10
16:1047 participants were
Nutrition and Dietetics professor during fall 2011-
spring 2012 semesters. The Food and Nutrition
classes were introductory nutrition classes offered
to all stu- dents from any majors and are considered
as University Program (UP) - General Education
courses. Students who enroll in these classes have
no prior academic nu- trition knowledge since these
classes are introductory nutrition classes offered to
all undergraduate students from any majors as part
of UP/General Education courses irrespective of
students’ year in school. Students agreeing to
participate were asked to sign a consent form, in
harmony with the Helsinki declaration, and to come
to a laboratory classroom for anthropometric
measurements and to receive a numerical code for
com- pleting a self-administered online
questionnaire that in- cluded questions related to
students’ demographic, nutrition knowledge, and
daily fat consumption. The questionnaire was
available online via SurveyMonkey Pro
(SurveyMonkey.com, LLC, Palo Alto, CA) for about
10 weeks to accommodate students’ response times.
Students were not offered any incentives for their
par- ticipation and were informed that they could
withdraw from the study at any time.
Two hundred and fifty students participated in
this study, however 19 students were excluded as
they did not complete the entire questionnaire. The
study proto- col was approved by the CMU
Institutional Review Board (IRB).

Data collection

Data were collected in two steps. First,

anthropometric parameters, including weight, height,
percentage body fat, visceral fat score and waist
circumference, were mea- sured by a CMU Nutrition
and Dietetics professor.
Weight, percentage body fat, visceral fat, and body
mass index were measured using a Tanita
bioelectrical impedance analyzer (BIA) SC-331S
(Tanita Corporation, Chicago, USA). For BIA
measurements, the student’s height, age, and
gender were entered into the machine before
testing. Then, the student stepped into the BIA’s
footpads with bare feet (both feet touched the
elec- trodes). Weight, percentage body fat, visceral
fat, and body mass index were recorded from the
BIA readings. Since severe dehydration or over-
hydration can affect BIA readings, participants’
measurements were taken in the morning after
overnight fasting and without con- suming alcohol
or any stimulant, on an empty bladder, and with no
intense exercise. Height was measured while a
student was standing erect and without shoes using
a stadiometer (Seca 217 Height Measuring Stadi-
ometer, Quick Medical, Issaquah, WA, USA).
Body mass index (BMI) was used to assess students’
body
weight category. According to the BMI guidelines
published by the National Institutes of Health,
 aspect of general nutrition knowledge as follows: section 1-
categorized into four groups as follows: knowledge of recommendations regarding increasing
underweight (BMI ≤18.5), normal weight (18.5 ≤ and decreasing intake of different food groups; section
BMI < 24.9), over- weight (25 ≤ BMI < 30), and 2- nutrient knowledge; section 3-
obese (BMI ≥ 30) [32]. The healthy range for body
fat percentage was considered as 8−19 % for
males and 17−32 % for females (Tanita BIA SC-
331S body fat ranges for healthy adults). A
visceral fat rating from 1 to 12 was considered
healthy for all, and 13 − 59 indicated an excess
level of visceral fat. Waist circumference was
recorded with a flexible, non-stretchable
measuring tape according to the National
Institutes of Health guidelines [32].

Online questionnaire

In the second step, students were asked to complete

an online questionnaire administered via
SurveyMonkey (https:// www.surveymonkey.com).
Students were told that they could withdraw from
the survey at any time and were given an option to
skip questions. The ques- tionnaire was pilot-tested
on 20 students before it was administered to the
students in this study. It consisted of questions
related to students’ demographics, dietary fat intake,
and nutritional knowledge, and was divided into 3
main parts as follows:

Part I – Demographics: 8 questions related to age,

gender, major of study (Science (such as Nutrition and
Dietetics, Health and Fitness, Exercise Science,
Kinesiology, Community Health) or non-Science major
(such as Music, Religion, Theater, Philosophy), year in
school, ethnicity, place of residence (on-campus/off-
campus), dieting, and smoking status.
Part II – Block Dietary Fat Screener: a validated 17-
item food frequency questionnaire (FFQ) was used
to assess students’ usual fat intake (Block Fat
Screener, NutritionQuest, Berkeley, CA, USA) [33]. A
description of the questionnaire is presented in
reference 33. The fat screener includes questions
about commonly con- sumed high-fat foods (41 food
items) and designed to rank individuals with
regard to their usual total fat intake. Data
collected from the fat screener were analyzed
using prediction equations to generate point
estimates of total fat (grams), saturated fat
(grams), percent calories from fat, and
cholesterol (mg).
Part III – Nutrition Knowledge: 50 questions related
to nutritional knowledge. This part was taken
from a previously published study by Parmenter
and Wardle [34], whose approval was obtained
prior to the study. This part includes four
sections (dietary recommendations, sources of
nutrients, everyday food choices, and diet-disease
relationship), each section assessing a different
 percentage body fat, the mean value was 23.9 ± 8.6
food choice; and the last section (section 4) about %. The mean values of visceral fat scores for males
the relationships between diet and disease. The and females were 1.9 ± 1.8 and 3.7 ± 3.4,
questions were in the form of multiple choice, fill-in- respectively (Table 1). Based on BMI, the majority of
the-blank, and check mark. Each question carries one students (68 %) were within the healthy
point for a correct answer. Section 1 (dietary
recommendations) has a
maximum score of 11; Section 2 (sources of
foods/ nutrients) has a maximum score of 69;
Section 3
(choosing everyday foods) has a maximum score of
10;
Section 4 (diet-disease relationships) has a
maximum score of 20. Students’ responses to
these sections
were scored according to the nutrition knowledge
questionnaire’s scoring scale [34], and the scores
were summed to obtain the nutrition knowledge
score. A
detailed description of the nutrition knowledge
questionnaire and the scoring scale are available
at:
https://www.ucl.ac.uk/hbrc/resources/resources_e
b and described elsewhere [34].

Data analysis
Power calculation was not performed because this
study is exploratory in nature to inform the design
of future studies. Statistical analyses were
performed using the SAS (9.3, U.S.A) software.
Students’ results on the t-test and Chi-square test
for independence were used to examine differences
in the anthropometric characteris- tics between male
and female students. Results are expressed as
means ± SD (standard deviation). One-way ANOVA
was used to test differences in fat, saturated fat, and
cholesterol intake for various demographic variables.
Student's t-test results were used to examine
differences in nutrition score by gender, high fat
consumption (>35 % calories), and high cholesterol
(>300 mg/day). Sep- arate linear regression for
nutrition scores on estimated saturated fat intake and
cholesterol intake (controlling for gender, height,
weight, age, and self-reported dieting), were used to
analyze the association between nutrition scores and
saturated fat or cholesterol consumption. Par- tial
correlation coefficients (controlling for gender, height,
weight, and age) between nutrition scores and
saturated fat intake and cholesterol intake were also
calculated. All reported p values were based on 2-
sided tests with a sig- nificance level of 5 %
(Additional files 1, 2, 3, 4 and 5.

Results
Participants’ characteristics
Two hundred and thirty one students (71 % females
and 29 % males), with a mean age of 20.6 ± 2.0
years, partici- pated in this study. The average
weight was 69.8 ±
16.7 kg, with an average height of 166.8 ± 9.1 cm.
The mean BMI was 24.2 ± 4.4 kg/m2. As for
 Table 1 Students’ characteristics
Male N = 67 Female N = 164 Total N = 231 P Value
Age 21.3 ± 2.4 20.3 ± 1.6 20.6 ± 2.0 0.0002
Weight 83.9 ± 16.9 64.0 ± 12.7 69.8 ± 16.7 <.0001
Height 176.7 ± 6.6 162.9 ± 6.5 166.8 ± 9.1 <.0001
Waist circumference 91.1 ± 11.1 81.1 ± 9.6 83.6 ± 11.0 <.0001
BMI 26.0 ± 4.9 23.4 ± 4.0 24.2 ± 4.4 <.0001
% body fat 16.7 ± 8.1 26.8 ± 6.9 23.9 ± 8.6 <.0001
Visceral fat score 3.7 ± 4.2 1.9 ± 1.8 2.4 ± 2.8 <.0001
Ethnicity 0.0905
White 85 % 92 % 90 %
Black/other 15 % 8% 10 %
Body mass index 0.0003
BMI <18.5 3% 1.7 % 2.6 %

BMI 18.5− 24.9 47.8 % 76.2 % 68.0 %

BMI > 25 (overweight) 35.8 % 17.1 % 22.5 %
BMI > 30 (obese) 13.4 % 4.3 % 6.9 %
Academic level of study 0.6660
First year 14 % 13 % 14 %
Second year 20 % 27 % 25 %
Third year 26 % 27 % 27 %
Fourth year 20 % 19 % 19 %
Fifth year 20 % 13 % 15 %
Major of study
Health science 37 % 56 % 51 % 0.0078
Non- health science 63 % 44 % 49 %
Current place of residence 0.6493
On-campus 30.0 % 32.9 % 32.0 %
Off-campus 70.0 % 67.1 % 68.0 %
Currently dieting
No 80.6 % 82.3 % 81.8 % 0.7584
Yes 19.4 % 17.7 % 18.18 %
Smoking
Non-smokers 81 % 93 % 90 % 0.0163
Smokers 11 % 4% 5%
Former smokers 9% 3% 5%

weight category, particularly female students, and about

one-third were either overweight (22.5 %) or obese
(6.9 %). Of the participating students, 90 % reported
as white, and 10 % were African American or other
ethnic origin, reflecting the composition of ethnic
groups at CMU. Fifty-one percent of students were
science majors with a higher percentage for females
than males (56 % vs. 37 %, P = 0.0088). More than
two-thirds of the stu- dents (70 %) lived outside of
campus, and 80.9 % re- ported not following any
special diet. The percentages of 2nd and 3rd year
students were 25 % and 27 %,
respectively, while the other 3 levels (1st, 4th, and
5th year) were all under 20 %. The majority of the
students were non-smokers (90 %); 5 % were current
smokers; and 5 % were former smokers.
Daily fat consumption of participating students
Daily mean intake of total fat, saturated fat, and
choles- terol based on students’ characteristics is
presented in Table 2. Results indicate that there was
a significant dif- ference in the consumption of total
fat between male and female students. Daily mean
intake of total fat was
Table 2 Daily mean intake of dietary fats by students’ characteristics

Total fat (g) Saturated fat (g) Cholesterol (g)

Variables
Gender

Males 85.9 ± 24.8 28.9 ± 9.1 298.2 ± 79.3

Females 92.0 ± 18.8 23.5 ± 6.9 224.5 ± 61.7

Body mass P-value* 0.0427 <.0001 <.0001

index

BMI <18.5 81 ± 19.7 22 ± 8.3 222.3 ± 69

BMI 18.5- 24.9 91.6 ± 20.8 25 ± 7.9 242.1 ± 74.6
BMI > 25 (overweight) 85.3 ± 19.3 24.6 ± 7.7 248.1 ± 73.9
BMI > 30 (obese) 95.8 ± 25 29.2 ± 8.7 285.2 ± 79.2
P-value** 0.1120 0.1431 0.1414

Year in school
1st-year undergraduate 90.9 ± 24.2 25.3 ± 9.5 250.4 ± 89.5
2nd -year undergraduate 93.7 ± 21.2 25.9 ± 7.8 250.1 ± 70.6
3rd -year undergraduate 87.2 ± 18.2 23.9 ± 7.5 236.7 ± 75
4th -year undergraduate 91.8 ± 20.1 25.7 ± 7.9 249.6 ± 73
5th -year undergraduate 87.1 ± 23.1 24.6 ± 8 242.8 ± 74.1
P-value** 0.4175 0.6716 0.8536
Major of study

Science 85.6 ± 19.7 22.8 ± 6.9 224.1 ± 62.7

Non-Science 95.1 ± 21.4 27.5 ± 8.4 268 ± 80.5
P-value* 0.0006 <.0001 <.0001
Ethnicity White 91 ± 20.2 25.2 ± 7.8 242.6 ± 73.6
Black 83.4 ± 26.4 23.7 ± 9.4 269.8 ± 85.5
P-value* 0.0996 0.3916 0.0996
Residential
status
On-Campus 90.7 ± 19.7 25.3 ± 7.6 248.5 ± 72.7
Off-Campus 89.1 ± 23.4 24.5 ± 8.7 240.3 ± 80
P-value* 0.5804 0.4669 0.4381
Dieting No 91.6 ± 21 25.6 ± 8 250.2 ± 75.4
Yes 83.8 ± 19.5 22.9 ± 7.4 226.7 ± 70.8
P-value* 0.0277 0.0478 0.0666
Smoking status

Non-smoker 90.2 ± 20.5 24.9 ± 7.9 243.2 ± 74.4

Current smoker 96.6 ± 18.7 29.3 ± 7 285.6 ± 66.8
Former smoker 83.2 ± 29.3 24.4 ± 10.2 249.3 ± 88.8
P-value** 0.2940 0.1427 0.1409
* T-test for independent samples
**1-way ANOVA F-statistic

higher among females (92 g/day) compared to males (23.5 g/day) than males (28.9 g/day).
(85.9 g/day) (P <0.001). However, saturated fat and Regarding the year in school, students’ daily mean
choles- terol consumptions were lower for females in- take of total fat did not significantly differ
among
students, ranging between 87.1 g (lowest) for the 5th
year students and 93.7 g (highest) for 3rd year
students. Likewise, mean saturated fat intake and
cholesterol ranged between 23.9 g and 236.7 g
(lowest) for the 3rd year students to 25.9 g and 250.1
g (highest) for the 2nd year students. There was also
no significant difference
Male 67 62.5 ± 15.7 0.01
Female 164 67.4 ± 12.0
Percent fat
<35 % 139 69.8 ± 13.0 <.0001
between white and African American students in
daily total mean intake of total fat, saturated fat, and >35 % 92 61.2 ± 12.4
choles- terol. Also, students’ residential status and Cholesterol
smoking habit were not associated with daily mean <300 mg 181 67.7 ± 12.7 0.0002
intake of total fat, saturated fat, and cholesterol. >300 mg 50 59.8 ± 13.9
On the other hand, students’ daily mean values of
total fat intake, saturated fat, and cholesterol were *T-Test for independent Samples

sig- nificantly different based on the major of study

and dieting. Students majoring in science consumed
lower amounts of total fat (85.6 g), saturated fat
(22.8 g), and cholesterol (224.1 g) compared to non-
science majors (95.1 g, 27.5 g, and 268 g,
respectively) (P <0.0001). Likewise, students who
were on a diet consumed less total fat (P <0.0277),
saturated fat (P <0.0478) and choles- terol compared to
students who were not following any diet (Table 2).
While the mean cholesterol intake was lower for
students on a diet (225.7 g vs 250.2 g), this dif-
ference did not achieve statistical significance (P =
0.0666).

Nutrition knowledge survey

Table 3 shows the results of the students’ responses

to the nutrition knowledge survey, indicating that
fe- male students scored higher than male students
(67.4 ± 12.0 vs. 62.5 ± 15.7, respectively, out of
pos- sible 105 points, P = 0.01). Students who
consumed less than 35 % of daily calories from fat
had a higher nutrition score (P <0.0001) (Fig. 1) and,
similarly, students consuming less than 300 mg of
cholesterol had a higher nutrition score (P = 0.002)
(Table 3) (Fig. 2).
Results of the students’ responses to the four parts of the
nutrition knowledge survey showed that students scored highest
on the “Dietary Recommendations” sec- tion compared to the
other two sections (Source of Foods/Nutrients and Diet-Disease
Relationship). Female students scored higher than male students
on all 3 parts of the nutrition knowledge survey, particularly on the
“Dietary Recommendations” and the “Diet-Disease

Table 3 Mean scores of nutrition knowledge survey

n Mean NK score P-value*

Total population 231 66.0 ± 13.4
Gender
 knowledge scores (Figs. 1 and 2). Proper nutrition
knowledge is useful in improv- ing dietary habits, and
students are empowered when they have the
necessary nutrition knowledge and skills needed to
Relationships” sections, where scores were statistically make healthful lifestyle choices [39–41].
different between genders (Table 4).
Controlling for gender, height, weight, age and
dieting, there was a significant association between the
students’ intake of total fat, saturated fat, and
cholesterol and nu- trition knowledge score (P
<0.001) (Table 5).

Discussion
The 2010 Dietary Guidelines for Americans
recommend that consumption of saturated fat
should be limited to
<10 % of daily calories, dietary cholesterol to <300
mg per day, and trans fat to <1 % of daily calories or
as little as possible, primarily to reduce risk of
CVD (9). This study looked at the association
between nutrition know- ledge and fat
consumption in a sample of CMU stu- dents. Our
findings support the general assumption that
students’ nutrition knowledge is associated with im-
proved food choices pertaining to types of dietary
fats, and the findings concur with results of previous
studies [35–37].
In this study, nutrition knowledge was negatively
cor- related with fat and cholesterol intake. Students
who consumed more than 35 % of calories from
fat or
>300 mg of cholesterol daily had lower mean
nutrition scores than those students with lower fat
or cholesterol intake (8 points lower and 7.9 points
lower, respectively) (P <0.001). Using linear
regression for nutrition scores on estimated saturated
fat intake and cholesterol intake (controlling for
gender, height, weight, age, and dieting), nutrition
scores were negatively associated with satu- rated
fat intake (-0.15, P <0.0001) (Fig. 1) and cholesterol
intake (-1.38, P <0.0001) (Fig. 2). The results are
correl- ational and therefore cannot show
directionality. How- ever, it is likely that nutrition
education/ knowledge can lead to improvement of
students’ eating habits and healthier food choices.
In general, to be able to un- derstand and apply
the Dietary Guidelines recom- mendations for
healthful eating practices, some basic understanding
of nutrition is necessary [36]. Mazier and McLeod
reported that a single course in nutrition was
effective at improving the nutrition knowledge of
their undergraduate students [37]. A previous
study con- ducted among 269 undergraduate
Canadian students to examine the impact of
nutrition education on fat con- sumption found that
students who had taken a nutrition course
consumed less fat than those with no nutrition
education [38]. In this study, students with greater
nutri- tion knowledge consumed lower amounts of
total fat, saturated fat, and cholesterol per day
compared with students with lower nutrition
 Fig. 1 Association of Nutrition Knowledge Score (Total) vs. Total Saturated Fat Intake by Gender

Regarding differences in gender, in this study,
female students had greater nutrition knowledge than
male stu- dents (the mean nutrition score for women
was 5 points higher than that of men (P = 0.01)). This
finding was not surprising. In general, women are
more likely than men to be interested in diet,
nutrition, and body weight, par- ticularly during
college years [42–45]. A previous study conducted
among 479 Swedish university students found that
female students had healthier habits than male stu-
dents, despite being stressed, whereas male students
showed a high level of overweight and obesity and
were less interested in nutrition advice and health-
enhancing
activities [44]. Also, the authors reported that female
students were more interested in changing their dietary
habits than male students were [44]. Another
study evaluating the health knowledge of 428 African
American university students found that 75 % of
students exhibited high levels of health knowledge, and
female students dis- played higher levels of knowledge
than male students [45]. In our study, female students
consumed less saturated fat and cholesterol than male
students. This finding is con- sistent with a previous
study among 184 Iranian female students aged
18−35 years showing that female stu- dents who
perceived themselves as being in the healthy

Fig. 2 Association of Nutrition Knowledge Score (Total) vs. Cholesterol Intake by Gender
Table 4 Mean scores of nutrition knowledge survey and fat screening questionnaire by gender
Total score
Nutrition knowledge sections (total points)
Dietary recommendations (11)
Sources of foods/nutrients (69)
Choosing everyday foods (10)
Diet-disease relationships (15)
*T-Test for independent Samples
weight range had a significantly lower intake of satu-
rated fat and higher intake of monounsaturated fat
compared to other students [46].
The results also showed that students’ daily mean
in- take of total fat, saturated fat, and cholesterol
were sig- nificantly different based on major of
study and whether they were dieting. Students
majoring in sci- ence consumed lower amounts of fat
compared to non- science majors (P <0.0001).
Likewise, students who were on a diet consumed
less total fat, saturated fat, and cholesterol than
students who were not following any diet (P
<0.0277). It is likely that students majoring in
science would have covered more coursework related
to nutrition than non-science majors. Also,
students following a diet would be more likely to
restrict their daily fat intake [42, 47]. Nevertheless,
in this study, stu- dents’ daily mean intake of total
fat, saturated fat, and cholesterol did not differ by
body mass index, year-in- school, ethnicity,
residential living condition, or smok- ing status
among students. These findings were not
surprising [38, 48, 49]. Regarding fat intake, fat
intake per se is not the only factor that affect body
weight since a clustering of factors including
heredity, a low level of physical activity, poor diet,
and smoking can contribute to weight gain [50].
Likewise, students’ nu- trition knowledge will likely
not increase without stu- dents taking a nutrition
class, irrespective of how long students stay in
college. Emrich and Mazier looked at the impact of
nutrition knowledge on year-in-school and reported
no significant differences in fat intake be- tween
first-year students and fourth-year students.
Table 5 Association between fat intake and
nutrition knowledge scorea
Partial Regression
correlationa parameterb
Total fat intake (g) -0.27 -0.42
Saturated fat (g) -0.27 -0.15
Cholesterol (mg) -0.27 -1.38
a
Partial Correlation with Nutrition Score; controlling for gender, height, weight
and age
Male Female P
value*
8.1 ± 1.7 7.6 ± 2 8.3 ± 1.5 0.002
43.4 ± 8.6 41.8 ± 9.9 44.1 ± 8 0.06
5.5 ± 1.8 5.2 ± 1.9 5.6 ± 1.8 0.10
8.9 ± 3.8 7.9 ± 4.5 9.4 ± 3.5 0.01
However, the authors reported that there were
signifi- cant differences in fat intake between first-
year science students with some nutrition education
and those with- out [38]. This would suggest that
nutrition education rather year-in-school is
influencing students’ fat intake.
In contrast to our results, which did not reveal
any significant differences between on- and off-
campus stu- dents, Emrich and Mazier reported
differences in fat in- take between students living on
campus and off campus [38]. This could be due to
the fact that the residential dining facility at CMU
has a program called “Just4U Nutrition”, which
offers students all types of healthy and balanced meal
plans including vegetarian, vegan, Kosher, low-fat,
cholesterol-free, and gluten-free meals. Students also
have the option to choose an individualized meal plan
that matches their lifestyle. However, our results
were in agreement with a previous study
conducted among 210 Iranian students to determine
factors associ- ated with nutrition knowledge and
body weight, which found no significant correlation
between nutritional knowledge, body mass index, and
smoking status [49].
Study’s limitations
This study is limited in its small sample size and in
that most of the participating students were female
(71 % fe- male vs. 29 % male). It is possible that
female students may have been more interested in
research related to health issues than male students
since students voluntar- ily entered into the study.
Nevertheless, the dominance of female participants
reflects the university’s student body data and is
consistent with the gender composition of previous
studies [51, 52]. The limitations of any food frequency
questionnaire (FFQ) and FFQ screeners (such
P as Block Fat Screener) are well recognized [53]. Al-
value*** though Block Fat Screener cannot estimate dietary fat
b
<.0001 Multiple linear regression of nutrition scores on Total Fat, saturated fat intake
or cholesterol intake (controlling for gender, height, weight, age, and
<.0001 self-reported dieting)
***P-value for correlation and regression parameter
<.0001
intake accurately, the purpose of this study was not
to
estimate the amount of dietary fat intake in college
stu- dents but rather to examine the associations
between dietary fat intake level and nutrition
knowledge, which is an appropriate use of the FFQ
screener in research. Be- cause a FFQ screener is
composed of a pre-specified food list or set of
behavioral questions, any single screener may not
reflect the eating patterns of a given
 population. Therefore the generalizability of our findings
to other populations is limited.
Conclusion
This pilot study found that students with more
nutritional knowledge consumed less unhealthy fats and
cholesterol. Students are empowered when they have
the necessary knowledge and skills needed to make
healthful lifestyle choices [54]. Given the importance of
healthy eating in re- ducing CVD risk factors among
students, future research on this topic is needed
among this vulnerable age group. Factors such as food
cost, food preparation, and cooking methods should also
be included in future research since these factors can
impact how effectively students can apply nutritional
knowledge into their everyday eating habits. In
conclusion, the study’s results suggest that stu- dents
were able to translate nutrition knowledge into their
daily diet by reducing their saturated fat and cholesterol
intake. This finding magnifies the role of nutrition educa-
tion as a potential tool in health campaigns to
promote healthy eating patterns among college
students. We be- lieve that the results of this pilot
study can inform the de- sign of future nutrition
education intervention studies to assess the efficacy of
nutrition knowledge on pattern of fat consumption
among college students.
Additional files
file
1: Consent Form. A sample of the consent form that was given to participants in
le 2: A copy of the study’s questionnaire. (XML 412 kb)
le 3: A copy of the raw data file. (CSV 157 kb)
le 4: Copy of Regression and Correlation Analysis (CSV 996 bytes)
le 5: SAS data analyses file. (DOCX 19 kb)
Abbreviations
BRFSS: Behavioral risk factor surveillance system; CDC: Centers for
disease control and prevention; CVD: Cardiovascular disease
Acknowledgments
A special note of appreciation goes to Dr. Allan Geliebter and
Professor Steven Couture for reviewing the draft. Also, I would like to
extend my sincere appreciation to all CMU students who participated
in this study. In addition, a thank you note goes to Tanita
Corporation for providing the body fat analyzer scale for this
research.
Funding
This work was partially supported by the Faculty Research and Creative
Endeavors (FRCE) Premier Display grant at CMU.
Availability of data and materials
Competing interests
The authors declare that they have no competing interests.
Consent to publish
Not applicable.
Ethics approval and consent to participate
Students agreeing to participate in the study were asked to sign a
consent form, in harmony with the Helsinki declaration, and were
informed that they could withdraw from the study at any time
without penalty. The study protocol was approved by the Central
Michigan University Institutional Review Board (IRB). IRB can be reached
at: cmuirb@cmich.edu and the project ID is: IRBNet ID: 278950-2.
Author details
1
Department of Human Environmental Studies, Central Michigan University,
Wightman 108, Mt. Pleasant, MI 48859, USA. 2Jean Mayer USDA Human
Nutrition Center on Aging at Tufts University, Boston, MA, USA.
3
Department of Public Health and Family Medicine, Tufts University
School of Medicine, 136 Harrison Avenue, Jaharis 264, Boston, MA 02111,
USA. 4Department of Biostatistics, Boston University, 801 Massachusetts
Avenue, Boston, MA 02118, USA.
Received: 1 April 2016 Accepted: 22 September 2016
References
1. Go AS, Mozaffarian D, Roger VL, Benjamin EJ, Berry JD, Blaha MJ,
et al. American Heart Association Statistics Committee and Stroke
Statistics Subcommittee. Heart disease and stroke statistics-2014
update: a report from the American Heart Association. Circulation.
2014;129:e28–e292. doi:10.1161/01.cir.0000441139.02102.80.
2. Murphy SL, Xu JQ, Kochanek KD. Deaths: Final data for 2010. Natl
Vital Stat Rep. 2013;61(4):1–117.
3. CDC. Vital Signs: avoidable deaths from heart disease, stroke,
and hypertensive disease-United States, 2001-2010. MMWR 2013;
62(35):721-7
4. this
Michigan Department
study. (TXT 5 kb) of Community Health. February 2015
Michigan Health Statistics. Division for Vital Records and Health
Statistics. Available at :
https://www.michigan.gov/documents/mdch/CVH_fact_sheet_update-_
Final_3.4.15 483077_7.pdf
Data are available at cbrown03@bu.edu. Attached are copies of the statistical
analysis.
Authors’ contributions
NY carried out questionnaire design, manuscript preparation, data collection
and study coordination. CB performed all the statistical analysis. MR
contributed in data collection and data entry. MC reviewed all the
statistical analysis. All authors read and approved the final
manuscript.
5. Centers for Disease Control and Prevention, National Center for
Chronic Disease Prevention and Health Promotion, Division of
Population Health. BRFSS Prevalence & Trends Data [online]. 2015.
[Accessed Oct 12, 2015]. URL:
http://www.cdc.gov/brfss/brfssprevalence/.
6. Centers for Disease Control and Prevention. Chronic Diseases and
Health Promotion. May 2014
7. Lichtenstein AH, Appel LJ, Brands M, et al. Diet and lifestyle
recommendations revision 2006: a scientific statement from the
American Heart Association Nutrition Committee. Circulation.
2006;114:82–96.
8. Mozaffarian D, Benjamin EJ, Go AS, Arnett DK, Blaha MJ, Cushman
M, de Ferranti S, Després J-P, Fullerton HJ, Howard VJ, Huffman
MD, Judd SE, Kissela BM, Lackland DT, Lichtman JH, Lisabeth LD,
Liu S, Mackey RH, Matchar DB, McGuire DK, Mohler ER 3rd, Moy
CS, Muntner P, Mussolino ME, Nasir K, Neumar RW, Nichol G,
Palaniappan L, Pandey DK, Reeves MJ, Rodriguez CJ, Sorlie PD,
Stein J, Towfighi A, Turan TN, Virani SS, Willey JZ, Woo D, Yeh RW,
Turner MB; on behalf of the American Heart Association Statistics
Committee and Stroke Statistics Subcommittee. Heart disease and stroke
statistics-2015 update: a report from the American Heart Association
[published online ahead of print December 17, 2014]. Circulation
2014. doi:10.1161/CIR.0000000000000152.
9. U.S. Department of Agriculture, U.S. Department of Health and
Human Services. Dietary Guidelines for Americans, 2010.
Washington, D.C: U.S. Government Printing Office; 2010.
10. Kotler LA, Cohen P, Davies M, Pine DS, Walsh BT. Longitudinal
relationships between childhood, adolescent, and adult eating
disorders. J Am Acad Child Adolesc Psychiatry. 2001;40:1434–
40.
11. Neumark-Sztainer D, Wall M, Larson NI, Eisenberg ME, Loth K.
Dieting and disordered eating behaviors from adolescence to
young adulthood: findings from a 10-year longitudinal study.
J Am Diet Assoc. 2011;111(7):1004–11.
 33. Block G, Gillespie C, Rosenbaum EH, Jenson C. A rapid food
12. Nelson M, Story M, Larson N, Neumark-Sztainer D, Lytle L. screener to assess fat and fruit and vegetable intake. Am J Prev Med.
Emerging adulthood and college-aged youth: An overlooked age 2000;18(4):284–8. Available at: http://www.nutritionquest.com/wellness/free-
for weight-related behavior change. Obesity. 2008;16:2205–11. assessment-tools- for-individuals/fat-intake-screener/.
13. Lien N, Lytle LA, Klepp KI. Stability in consumption of fruit, 34. Parmenter K, Wardle J. Development of a Nutrition
vegetables, and sugary foods in a cohort from age 14 to age 21. Prev Knowledge Questionnaire. Eur J Clin Nutr.
Med. 2001;33:217–26. 1999;53:298–308.
14. Racette SB, Deusinger SS, Strube MJ, Highstein GR, Deusinger RH.
Weight changes, exercise, and dietary patterns during freshman and
sophomore years of college. J Am Coll Health. 2005;53:245–
51.
15. Sira N, Pawlak R. Prevalence of overweight and obesity, and
dieting attitudes among Caucasian and African American
college students in Eastern North Carolina: A cross-sectional
survey. Nutr Res Pract. 2010;4(1):36–42.
16. Sulliman K, Rodas-Fortier K, Neyman M. A survey of dietary and
exercise habits and perceived barriers to following a healthy
lifestyle in a college population. Californian J Health Promot.
2004;2(2):10–9.
17. Anding JD, Suminski RR, Boss L. Dietary intake, body mass index,
exercise, and alcohol: Are college women following the dietary
guidelines for Americans? J Am Coll Health. 2001;49(4):167–
71.
18. Papadaki G, Hondros JA, Scott M. Kapsokefalou. Eating habits of
university students living at, or away from home in Greece. Appetite.
2007;49(1):169– 76. doi:10.1016/j.appet.2007.01.008.
19. Deliens T, Clarys P, De Bourdeaudhuij I, Deforche B. Determinants of
eating behaviour in university students. A qualitative study using
focus group discussions. BMC Public Health. 2014;14(1):53.
doi:10.1186/1471-2458-14-53.
20. Butler SM, Black DR, Blue CL, Gretebeck RJ. Change in diet, physical
activity, and body weight in female college freshman. Am J
Health Behav. 2004;28(1):24–32.
21. Chourdakis M, Tzellos T, Pourzitaki C, Toulis KA, Papazisis G,
Kouvelas D. Evaluation of dietary habits and assessment of
cardiovascular disease risk factors among Greek university students.
Appetite. 2011;57(2):377–83.
22. The U.S. Surgeon General. National Prevention Strategy/Healthy
Eating. 2014. Available at
http://www.surgeongeneral.gov/priorities/prevention/ strategy/healthy-
eating.html. Accessed 25 Mar 2016.
23. U.S. Department of Health and Human Services: The Surgeon
General’s Vision for a Healthy and Fit Nation. Rockville, MD. U.S.
Department of Health and Human Services, Office of the Surgeon
General, 2010. [Available at
http://www.surgeongeneral.gov/library/obesityvision/obesityvision2010.pd
f. Accessed 25 Mar 2016.
24. Ha EJ, Caine-Bish N. Effect of nutrition intervention using a general
nutrition course for promoting fruit and vegetable consumption
among college students. J Nutr Educ Behav. 2009;41(2):103–
9.
25. Ha EJ, Caine-Bish N. Interactive introductory nutrition course
focusing on disease prevention increased whole-grain consumption by
college students. J Nutr Educ Behav. 2011;43(4):263–7.
26. Lin LP, Dali WP. The impact of nutrition education interventions
on the dietary habits of college students in developed nations: a
brief review. Malays J Med Sci. 2012;19(1):4–14.
27. Brown ON, O'Connor LE, Savaiano D. Mobile MyPlate: a pilot study
using text messaging to provide nutrition education and promote
better dietary choices in college students. J Am Coll Health.
2014;62(5):320-7.
28. Shahril MR, Wan Dali WPE, Lua PL. A 10-Week Multimodal
Nutrition Education Intervention Improves Dietary Intake among
University Students: Cluster Randomised Controlled Trial. J Nutr
Metab. 2013;2013:658642.
https://www.hindawi.com/journals/jnme/2013/658642/.
29. Douglas PD, Douglas JG. Nutritional knowledge and food practices
of high school athletes. J Am Diet Assoc. 1984;84:1198–202.
30. Perron DK, Endres J. Knowledge, attitudes and dietary practices of
female athletes. J Am Diet Assoc. 1985;85:573–7.
31. Ozdogan Y, Ozcelik AO. Evaluation of the nutrition knowledge of
sports department students of universities. J Int Soc Sports Nutr.
2011;8:11. doi:10.1186/1550-2783-8-11.
32. Clinical Guidelines on the Identification, Evaluation, and Treatment
of Overweight and Obesity in Adults–The Evidence Report. National
Institutes of Health. Obes Res. 1998;6 Suppl 2:51S-209S.
 NIH Public Access
Author Manuscript
35. Mitchell SJ. Changes after taking a college basic nutrition course. J
Nutrition label use
J Acad Nutr Diet. Author manuscript; available in PMC 2013 March 01.
Am Diet Assoc. 1990;90(7):955–61.
36. Worsley A. Nutrition knowledge and food consumption: can partially mediates the
nutrition knowledge change food behaviour? Asia
NI Pacific J Clin Nutr. 2002;11:S579–85. relationship between
H- 37. Mazier MJ, McLeod SL. University science students' knowledge of

P
fats. Can J Diet Pract Res. 2007;68(3):154–9.
38. Emrich TE, Mazier MJ. Impact of nutrition education on university
attitude toward healthy
A
students' fat consumption. Can J Diet Pract Res.
2009;70(4):187–92. eating and overall
39. Contento P, Koch A, Lee H, Calabrese-Barton A. Adolescents
A demonstrate improvement in obesity risk behaviors after dietary quality among
ut completion of choice, control & change, a curriculum addressing
personal agency and autonomous motivation. J Am Diet Assoc. college students
ho 2010;110(12):1830–9.
40. Brown LB, Dresen RK, Eggett DL. College students can
r benefit by participating in a prepaid meal plan. J Am Diet Assoc. Dan J. Graham, Ph.D. [Research Associate] and
2005;105:445–8.
M 41. Wardle J, Haase AM, Steptoe A, Nillapun M, Jonwwutiwes K,
Division of Epidemiology and
Community Health, University of
an Bellisle F. Gender differences in food choice: The contribution of
health beliefs and dieting. Ann Behav Med. 2004;27(2):107–16. Minnesota, Twin Cities, 1300 S 2nd
42. Connor-Greene PA. Gender Differences in Body Weight Perception
Street, Suite 300, Minneapolis MN
and Weight-Loss Strategies of College Students. Women Health.
1988;14(2):27– 42. doi:10.1300/J013v14n02_03. 55454, Phone: 612-624-1818, Fax:
43. Yahia N, El-Ghazale H, Achkar A, Rizk S. Dieting Practices and 612-624-4173, djgraham@umn.edu
Body Image Perception among Lebanese University Students.
Asia Pac J Clin Nutr. 2011;20(1):21–8. Melissa N. Laska, Ph.D., R.D. [Assistant Professor]
NI 44. von Bothmer MI, Fridlund B. Gender differences in health habits Division of Epidemiology and
and in motivation for a healthy lifestyle among Swedish university
H- students. Nurs Health Sci. 2005;7:107–18. Community Health, University of
45. Livingston IL, Saafir BD, Manuel RC. Health knowledge among Minnesota, Twin Cities, 1300 S 2nd
P historically black college and university students: an exploratory Street, Suite 300, Minneapolis MN
study. Coll Student J. 2012;46(3):581–8.
A 46. Alipour B, Farhangi AM, Dehghan P, Alipour M. Body image 55454, Phone: 612-624-8832, Fax:
A perception and its association with body mass index and nutrient 612-624-4173, mnlaska@umn.edu
intakes among female college students aged 18–35 years from
ut Tabriz, Iran. Eat Weight Disord Stud Anorexia Bulimia Obes.
Abstract
2015;1:1–7.
ho 47. Lowry R. Galuska, D A, Fulton, J E, Wechsler, H, Kann, L and Individuals who frequently
Collins, J L. 2000. Physical activity, food choice, and weight
r management goals and practices among US college students.
read nutrition labels tend
to both value healthy
M Am J Prev Med. 2000;18:18–27.
48. Wong JE, Parnell WR, Anna S, Howe AS, Lubransky AC, Katherine eating and engage in
an E, Black KE, Skidmore P. Diet quality is associated with measures healthy dietary practices
of body fat in adolescents from Otago, New Zealand. Public Health
Nutr. 2015;18:1453–60. doi:10.1017/S1368980014001645.
more than individuals
49. Zarei M, Taib MNM, Zarei F. Lifestyle factors and dietary intake who read labels
of Iranian postgraduate students in Universiti Putra Malaysia infrequently. However, the
(UPM). Electronic Physician. 2013;5(3):687–97. relationship between label
doi:10.14661/2013.687-697.
50. Arnaud C, Faeh D, Paccaud F, Cornuz J. Consequences of Smoking
use, attitude toward
NI for Body Weight, Body Fat Distribution, and Insulin Resistance. healthy eating, and dietary
Am J Clin Nutr. 2008;87(4):801–9. quality remains unclear,
H- 51. Association ACH. American College Health Association-National particularly among young
College Health Assessment II: Reference Group Executive Summary
P Spring 2013. American College Health Association: Hanover, adults, about whom little
A MD; 2013. is known with regard to
52. Yahia N, Brown C, Rapley M, Chung M. Assessment of college nutrition label use.
A students’ awareness and knowledge about conditions relevant The present study
to metabolic syndrome. Diabetology & Metabolic Syndrome.
ut 2014;6:111. doi:10.1186/ 1758-5996-6-111. investigated whether
53. National Cancer Institute. National Institutes of Health. Dietary nutrition label use
ho Assessment Primer, Screeners at a Glance; 2016 [Accessed on May 28, mediates the relationship
r 2016]. Available from:
between eating-related
https://dietassessmentprimer.cancer.gov/profiles/screeners/index.html
M 54. Taras H. Nutrition and student performance at school. J Sch attitudes and dietary
Health. 2005; 75(6):199–213. behaviors among young
an adult college students.
Using cross- sectional
online survey data
collected in 2010 from a
convenience sample in
Published in final edited form as: Minneapolis/St. Paul, MN
J Acad Nutr Diet. 2012 March ; 112(3): 414–418. (598 attending a 2-year
community college; 603
attending a public 4-year
university; mean age
 =21.5 years; 53.4% reserved.
non-white; 52.5% Corresponding
female), study findings Author: Dan J.
indicate that students Graham (will
who reported also handle
frequently reading requests for
nutrition labels were
reprints).
more likely to have
Publisher's DisclaiTmhiesri:s a PDF
healthier dietary
file of an unedited manuscript that has been
intakes (e.g., less fast accepted for publication. As a service to our
food and added sugar; customers we are providing this early version of the
more fiber, fruits and manuscript. The manuscript will undergo copyediting,
vegetables), compared typesetting, and review of the resulting proof before it
to those who read is published in its final citable form. Please note that
labels sometimes or during the productionprocess errors may be discovered
which could affect the content, and all legal
rarely (p<0.001).
disclaimers that apply to the journal pertain.
Further, frequent
nutrition label use
was a significant
partial mediator of the
relationship between
eating-related attitude
(i.e., feeling that it is
important to prepare
healthy meals) and
dietary quality,
indicating that label
use may be one
means by which
individuals who value
healthy eating translate
their attitude into
healthy eating
behaviors. Even among
those who did not
believe it was
important to prepare
healthy meals, frequent
nutrition label use was
significantly associated
with healthier dietary
intake, suggesting that
label use may operate
independently of
nutrition-related
attitude in contributing
to healthful diet.

Keywords
nutrition label; young adult; dietary
quality

©2011 The
American
Dietetic
Association.
Published by
Elsevier Inc.
All rights
 Graham and Laska Page 2

Introduction
Previous research has revealed health-relevant differences between individuals who report
frequently reading nutrition labels on food packages and those who do not (1-3). Some of
NI these differences are attitudinal (e.g., individuals who frequently read labels tend to
H- perceive health and healthy eating as more important than those who do not) (1); other
P differences are behavioral (individuals who frequently read labels tend to eat more healthfully
compared with those who read labels infrequently or never) (2-3). Despite these differences,
A and although nutrition label use has been identified as a mediator of the relationship between
A income and dietary quality (3), no published research has yet examined nutrition label use
as a potential mediator between health/diet-relevant attitudes and corresponding behaviors.
ut
ho Young adulthood is a time when individuals, having only recently left the family home,
r are beginning to make more independent choices about food and a time when many
individuals adopt unhealthy dietary behaviors and gain excess weight (4). It is important not
M only to achieve a better understanding of nutrition label use among this age group, but also
an the extent to which it may act as a mediator between individual-level psychosocial factors
(such as attitudes about healthy foods) and dietary outcomes. However, with a few recent
exceptions (5-6), existing research examining nutrition label use has focused on the
general adult population, with participant ages averaging in the 40s and 50s (7-8), while label
use among younger individuals has been studied far less. Indeed, a recent review of
nutrition label use reports that, of the 129 extant papers addressing this topic, “most studies
NI used convenience samples of the general adult population… [and only] one study looked at
adolescents” (9, p.23). Despite being traditionally overlooked in diet research, the eating-
H- related behaviors of individuals exiting adolescence and entering adulthood are important
P to understand.
A The present study examined dietary attitudes, behaviors, and nutrition label use among a
A diverse sample of college students. It was hypothesized that those who reported reading
nutrition labels frequently would be more likely to engage in healthy dietary practices
ut and would demonstrate greater nutrition knowledge than would those who read labels less
ho frequently. In addition, it was hypothesized that self-reported nutrition label use would
r mediate the relationship between healthy eating attitude and healthy dietary behavior,
such that reading labels would be a mechanism by which individuals holding healthy
M attitudes toward eating translate their beliefs into healthy eating behaviors.
an
Methods
Recruitment
In the spring of 2010, a sample of 1201 college students (598 attending a 2-year
community college; 603 attending a public 4-year university; mean age =21.5; 53.4% non-
NI white; 52.5% female) in the Twin Cities area of Minnesota participated in a study of
H- nutrition- and weight- related issues. Data collectors approached students on campus, inviting
them to participate. Students were given a weblink and access code to enter an online
P survey. The survey took approximately 30-35 minutes to complete, after which participants
A had their height, weight, and body composition measured on campus and received a $50 gift
card for their participation. Participants were also entered in a lottery to win an Apple
A iTouch™ device.
ut All study procedures were approved by the University of Minnesota Institutional Review
ho Board.

r
M
an

J Acad Nutr Diet. Author manuscript; available in PMC 2013 March 01.
 Graham and Laska Page 3

Assessments of Attitude and Behavior

Nutrition Label Use—Participants completed a survey item, adapted from previous research
(10), in which they self-reported the frequency with which they read the nutrition
NI information on food labels before purchasing foods/beverages (i.e., “How often do you
H- read the nutrition labels on food labels before purchasing foods or beverages?”) with
response options including “never or rarely,” “sometimes,” “often,” and “always or almost
P always.” Students who selected often and always or almost always were categorized as
A frequent label users.
A
Importance of Preparing Healthy Meals—Participants also self-reported their agreement
ut with the following statement: “It is important to me to prepare healthy meals,” by
ho endorsing a response option ranging from 1 (strongly disagree) to 5 (strongly agree). One-
week test-retest reliability for importance of preparing healthy meals was 0.64 (p<0.0001).
r
M Dietary Assessment—Participants self-reported dietary behaviors for the previous 30 days
using the Five Factor Screener from the 2005 NHIS Cancer Control Supplement, the
an Percentage Energy from Fat Screener, and the All-Day Fruit and Vegetable Screeners
from the Eating at America’s Table Study (all available at http://riskfactor.cancer.gov/diet/
screeners/; for validity results, see references 11-13). The Percentage Energy from Fat
Screener was modified slightly due to the inadvertent omission of one of the items,
mayonnaise, included on the original instrument. From these dietary screeners, summary
variables (i.e., percent of calories from fat, and dietary fiber (grams), calcium (mg), added
NI sugar (teaspoons), fruits and vegetables (servings), and dairy (servings)) were calculated.
H- Participants also self-reported whether or not they followed a vegetarian diet (14) and
how many times per week they consumed fast food (including any restaurant where food is
P ordered at a counter or drive-through window; (15)).
A
A For these analyses, a healthy eating behavior composite score was computed as a sum (0 –
8) of the number of these eight eating practices the individual followed, with higher
ut scores indicating more of the healthy eating practices were followed. Each of the 8
ho behaviors has been previously linked to positive health outcomes, and include:
consuming <35% of daily calories from fat (16); fiber intake of between 21 and 38
r grams/day, based on age and sex (17); calcium intake between 1000 and 1300 mg/day,
M based on age and sex (18); dairy intake of ≥3 servings per day (19); added sugar intake
an of between 3 and 18 tsp/day, based on age, sex, and physical activity level (20); fruit and
vegetable intake between 7 and 13 servings/day, based on age, sex, and physical activity
level (17); <3 times per week eating fast food (21-22); and following a vegetarian diet
(23-24).

In addition, participants reported the number of fruits, vegetables, and calories they
believed they should consume per day in order to be healthy. Actual recommended levels
NI of intake (for calories, fruit and vegetable servings) were calculated for each participant
H- individually based on age, sex, and physical activity level (17).
P
Demographic Information—Participants self reported age, sex, race and ethnicity.
A
A
Statistical Analyses
ut
Analyses investigating dietary differences between those individuals who reported frequently
ho reading nutrition labels and those who did not included independent samples t- tests
r comparing the percent of participants in each group who followed each of the eight
healthy dietary practices assessed.
M
an
 Graham and Laska Page 4

The differences between participant estimates of the number of calories, servings of

fruits, and servings of vegetables they should consume daily to meet public health
guidelines and actual recommended levels of intake were compared for frequent label
NI readers and infrequent label readers with the hypothesis that frequent label readers
would provide more accurate estimates.
H-
P Analyses testing for mediation included linear regression models fit to test whether self-
reported nutrition label use mediated the relationship between attitude toward healthy
A
eating and healthy eating behavior. Three linear regression models were fit in accordance
A with the recommendations of Baron and Kenny (25) and MacKinnon (26): 1) a model in
ut which attitude toward healthy meals predicted healthy eating behavior; 2) a model in
which attitude toward healthy meals predicted nutrition label use; and 3) a model in which
ho nutrition label use predicted healthy eating behavior in the presence of attitude toward
r healthy meals. All regression analyses included gender, race/ethnicity, and age as covariates.
M Additional regression analyses were conducted to test whether nutrition label use
an predicted the healthy dietary practices composite score among those individuals reporting
agreement with the statement “It is important to me to prepare healthy meals,” as well
as among those who did not agree with the statement. All analyses were conducted using
SPSS 19.0.0.1, release date January 21, 2011, IBM SPSS (Chicago, Illinois).

Results and Discussion

NI Participants in the study sample were more likely to represent racial and ethnic minorities and
H- were younger than the total student population at the two schools. At the 2-year community
college, 18%, 61%, and 21% of study participants were under 19 years old, 19 – 24, and
P
over 24, respectively, compared with 6%, 54%, and 41% of all enrolled students.
A Study participants from the 2-year college represented more racial/ethnic minorities (41%
A white, 34% African-American, 22% Asian, 12% other), compared to all enrolled students
(62% white, 20% African-American, 12% Asian, 4% other). Gender did not differ appreciably
ut (53% participants were female vs. 55% of all students). At the 4-year public university,
ho 11%, 85%, and 4% of study participants were under 19 years old, 19 – 24, and over 24,
respectively compared with 10%, 79%, and 11% for all enrolled undergraduate students.
r
Study participants from the 4-year university represented more racial/ethnic minorities (52%
M white, 8% African-American, 36% Asian, 10% other), compared to all enrolled students
an (70% white, 4% African-American, 8% Asian, 18% other. Gender did not differ (52%
participants were female vs. 52% of all students).

Thirty-five percent of participants were classified as frequent label readers (reporting nutrition
label viewing “always or almost always” or “often”). This percentage is on the low end
of the range of self-reported rates of label use (typically about 40-60%) presented in other
studies (see reference 27 for review). However, the mean response to the label use question
NI in the present study was 2.3 on a 1-4 scale, which is similar to the mean of 3.3 on a 5
H- point scale (1=never, 5=always) reported by Misra and colleagues in their 2007 study of
P label use among college students in Texas (5), both of which suggest that the average
college student uses nutrition labels sometimes. In that participants in our study tended to
A be somewhat younger than those included in most previous nutrition-label research, the
A slightly lower rate of frequent label use reported here relative to the rates summarized in
the Grunert and Wills review (27) is consistent with evidence that label use increases
ut with age (27).
ho
Consistent with our a priori hypotheses, independent samples t-tests revealed that
r frequent nutrition label readers showed higher rates of engaging in each of the 8 healthy
M dietary behaviors (significantly higher for eating fruits/vegetables, fiber, and a vegetarian
an diet, as well as limiting fast food and added sugar; p<0.001) compared with the
infrequent label
 Graham and Laska Page 5

readers (see Table 1). In addition, consistent with our hypotheses and with previous
research (5,28), frequent label readers demonstrated greater nutrition knowledge than
infrequent label readers. Estimates of how many calories and fruit and vegetable servings
NI they needed in order to be healthy provided by frequent label readers were significantly
nearer to the true recommendations (by more than 200 calories and approximately half of a
H- serving each of fruits and vegetables; all p’s <0.001) compared to the estimates provided
P by infrequent label readers.
A
Linear regression analyses (see Figure 1) revealed that 1) attitude toward preparing
A healthy meals was a statistically significant predictor of the healthy eating composite score;
ut 2) attitude toward preparing healthy meals significantly predicted nutrition label use; and 3)
when included in a regression model together, both attitude and nutrition label use
ho significantly predicted the healthy eating composite score, suggesting possible partial
r mediation by nutrition label use of the relationship between attitude toward healthy
eating and healthy dietary practices.
M
an
A Sobel’s test of partial mediation (25): produced a z-score of
4.80, indicating that nutrition label use was indeed a significant partial mediator of the
relationship between attitude toward healthy eating and healthy dietary practices.

In addition, nutrition label use significantly predicted the healthy eating composite among
NI both attitude-based groups of participants (i.e., those who agreed that it was important to
H- prepare healthy meals [b(se) =0.27(0.05), β =0.20, p<0.001 and those who did not (b(se) =
0.16(0.07), β =0.12, p=0.02)]).
P
A The role of nutrition label use as a partial mediator between attitude toward preparing
healthy meals and healthy dietary practices indicates that consumers with an inclination to
A make healthy dietary choices may be utilizing nutrition labels as one way of putting their
ut health preferences into action – a possibility supported by previous research demonstrating
ho that health concerns predicted frequent label use among college students (29). The present
analyses indicated that higher levels of label use were related to greater engagement in
r healthy eating behaviors not only among those participants with healthy eating attitudes,
M but also among those who did not believe it was important to prepare healthy meals.
Although provision of nutrition information on its own has not been shown to be highly
an successful in changing dietary behaviors (30-31), accurately understanding nutrition
information may be one of the many contributors to dietary intake, and may operate
independently of nutrition- related attitude. These findings underscore the need for accurate,
readily accessible, and understandable nutrition labeling on foods in the US.

Although the present study is the first of its kind to indicate that nutrition label use
NI partially mediates the relationship between attitude and diet among young adults, there are
important limitations to address. Participants comprised a diverse sample of both traditional
H-
and non- traditional college students; however, college students may differ in numerous ways
P from the larger population of young adults. Thus, these results may not generalize to
A young adults beyond 2- and 4-year college students. The participants were also recruited
via convenience sampling, and responders were more likely to represent racial and ethnic
A minorities and were younger than the total student population at their respective colleges;
ut although not assessed, it is possible that responders differed from non-responders in
additional ways. It is also important to note that this analysis is limited by a cross-
ho
sectional design; thus, it is not possible to determine from these data whether attitude
r toward healthy meals preceded nutrition label use and dietary behavior. Longitudinal research
M could help to clarify this temporal ordering, and a prime opportunity for prospective
research in this area is emerging
an
 Graham and Laska Page 6

as new nutrition labels are beginning to appear on the front of food packages in the
United States (32). Measuring consumers’ health-related attitudes before front-of-pack labels
are ubiquitous and assessing attitudes again after a period of naturalistic exposure to
NI the new labels could facilitate a clearer temporal understanding of the role of label use
in linking attitudes to behavior.
H-
P Conclusions
A Reading nutrition labels appears to be a mechanism through which college students who
A value healthy meal preparation make healthy dietary decisions. Even among those who do not
ut believe it is important to prepare healthy meals, nutrition label use is linked with healthier
dietary intake, suggesting that label use among college students relates to healthful dietary
ho intake independently of attitude toward healthy meals.
r
M Acknowledgments
an The authors would like to thank Stacey Moe, Katherine Lust, Rose Hilk, Lee Snyder, Dawn Ann Nelson and the entire Student Health
and Wellness Study research team for their assistance in data collection.

References
1. Drichoutis AC, Lazaridis P, Nayga RM. Nutrition knowledge and consumer use of
nutritional food labels. Eur Rev Agric Econ. 2005; 32(1):93–118.
NI 2. Kreuter MW, Brennan LK, Scharff DP, Lukwago SN. Do nutrition label readers eat
H- healthier diets? Behavioral correlates of adults’ use of food labels. Am J Prev Med. 1997;
13(4):277–283. [PubMed: 9236964]
P 3. Perez-Escamilla R, Haldeman L, Gray S. Assessment of nutrition education needs in an
A urban school district in Connecticut: establishing priorities through research. J Am Diet
Assoc. 2002; 102(4):559–562. [PubMed: 11985418]
A
4. Nelson MC, Story M, Larson NI, Neumark-Sztainer D, Lytle LA. Emerging adulthood and
ut college- aged youth: an overlooked age for weight-related behavior change. Obesity. 2008;
ho 16(10):2205– 2211. [PubMed: 18719665]
r 5. Misra R. Knowledge, attitudes, and label use among college students. J Am Diet
Assoc. 2007; 107(12):2130–2134. [PubMed: 18060900]
M 6. Visschers VHM, Hess R, Siegrist M. Health motivation and product design determine
an consumers’ visual attention to nutrition information on food products. Public Health Nutr.
2010; 13(7):1099– 1106. [PubMed: 20100390]
7. Grunert KG, Wills JM, Fernandez-Celemin L. Nutrition knowledge, and use and
understanding of nutrition information on food labels among consumers in the UK.
Appetite. 2010; 55(2):177–189. [PubMed: 20546813]
8. Satia JA, Galanko JA, Neuhouser ML. Food nutrition label use is associated with
NI demographic, behavioral, and psychosocial factors and dietary intake among African
Americans in North Carolina. J Am Diet Assoc. 2005; 105(3):392–402. [PubMed:
H- 15746826]
P 9. Cowburn G, Stockley L. Consumer understanding and use of nutrition labelling: a
A systematic review. Public Health Nutr. 2005; 8(1):21–28. [PubMed: 15705241]
10. Nayga RM, Lipinski D, Savur N. Consumers’ use of nutritional labels while food
A shopping and at home. J Consum Aff. 1998; 32(1):106–120.
ut 11. Thompson FE, Midthune D, Subar AF, Kipnis V, Kahle LL, Schatzkin A. Development
ho and evaluation of a short instrument to estimate usual dietary intake of percentage
energy from fat. J Am Diet Assoc. 2007; 107(5):760–767. [PubMed: 17467371]
r 12. Thompson FE, Midthune D, Williams GC, et al. Evaluation of a short dietary assessment
M instrument for percentage energy from fat in an intervention study. J Nutr. 2008;
an 138(1):193S– 199S. [PubMed: 18156424]
 Graham and Laska Page 7

13. Thompson FE, Subar AF, Smith AF, Midthune D, Radimer KL, Kahle LL, Kipnis V.
Fruit and vegetable assessment: Performance of 2 new short instruments and a food
frequency questionnaire. J Am Diet Assoc. 2002; 102(12):1764–1772. [PubMed:
NIH- 12487538]
14. Neumark-Sztainer D, Story M, Resnick MD, Blum RW. Adolescent vegetarians. A behavioral
P
profile of a school-based population in Minnesota. Arch Pediatr Adolesc Med. 1997;
A 151(8):833– 838. [PubMed: 9265888]
A 15. Nelson MC, Lytle LA. Development and evaluation of a brief screener to estimate fast-
food and beverage consumption among adolescents. J Am Diet Assoc. 2009; 109(4):730–
ut 734. [PubMed: 19328271]
ho 16.USDA. USDA; 2005. Dietary guidelines for Americans. Web site.
r http://www.health.gov/ dietaryguidelines/dga2005/document/pdf/Chapter6.pdf
17. Institute of Medicine Dietary Reference Intakes for Energy, Carbohydrate, Fiber, Fat, Fatty
M Acids Cholesterol, Protein, and Amino Acids. Washington, D.C.: National Academies
an Press; 2005.
us 18. Institue of Medicine. IOM; 2010. Dietary Reference Intakes for Calcium and Vitamin D.
Web site: http://www.iom.edu/Reports/2010/Dietary-Reference-Intakes-for-Calcium-and-
Vitamin-D.aspx
19. MyPyramid.gov. Mypyramid; 2008. Inside the pyramid: How much food from the milk
group is needed daily. Web site:
http://www.mypyramid.gov/pyramid/milk_amount_table.html
NIH- 20. Johnson RK, Appel LJ, Brands M, et al. Dietary sugars intake and cardiovascular health: a
scientific statement from the American Heart Association. Circulation. 2009; 120(11):1011–
P 1020. [PubMed: 19704096]
A 21.Boutelle KN, Fulkerson JA, Neumark-Sztainer D, Story M, French SA. Fast food for
family meals: relationships with parent and adolescent food intake, home food availability and
A
weight status. Public Health Nutr. 2007; 10(1):16–23. [PubMed: 17212838]
ut 22. Pereira MA, Kartashov AI, Ebbeling CB, et al. Fast-food habits, weight gain, and insulin
ho resistance (the CARDIA study): 15-year prospective analysis. Lancet. 2005; 365(9453):36–42.
[PubMed: 15639678]
r
23.Key TJ, Davey GK, Appleby PN. Health benefits of a vegetarian diet. Proc Nutr Soc.
M 1999; 58(2): 271–275. [PubMed: 10466166]
an 24.Key TJ, Fraser GE, Thorogood M, et al. Mortality in vegetarians and nonvegetarians: detailed
findings from a collaborative analysis of 5 prospective studies. Am J Clin Nutr. 1999; 70(3
us Suppl): 516S–524S. [PubMed: 10479225]
25. Baron RM, Kenny DA. The moderator-mediator variable distinction in social
psychological research: conceptual, strategic, and statistical considerations. J Pers Soc
Psychol. 1986; 51(6): 1173–1182. [PubMed: 3806354]
26. MacKinnon, DP. Introduction to Statistical Mediation Analysis. New York, NY: Psychology
Press; 2008.
NIH-
27. Grunert K, Wills J. A review of European research on consumer response to nutrition
P information on food labels. J Public Health. 2007; 15(5):385–399.
A 28. Guthrie JF, Fox JJ, Cleveland LE, Welsh S. Who uses nutrition labeling, and what
effects does label use have on diet quality. J Nutr Educ. 1995; 27(4):163–72.
A
29. Rasberry C, Chaney BH, Housman JM, Misra R, Miller PJ. Determinants of nutrition
ut label use among college students. Am J Health Educ. 2007; 38(2):76–82.
ho 30. Harnack LJ, French SA. Effect of point-of-purchase calorie labeling on restaurant and
cafeteria food choices: a review of the literature. Int J Behav Nutr Phys Act. 2008;
r 5(51):1–6. [PubMed: 18182102]
M 31. Harnack LJ, French SA, Oakes JM, Story MT, Jeffery RW, Rydell SA. Effects of calorie
an labeling and value size pricing on fast food meal choices: results from an experimental
trial. Int J Behav Nutr Phys Act. 2008; 5(63):1–13. [PubMed: 18182102]
us 32. Brownell KD, Koplan JP. Front-of-package nutrition labeling – an abuse of trust by
the food industry? N Engl J Med. 2011; 364(25):2372–2375.
 Graham and Laska Page 8

NIH-
NI
P
H-
A
P
A
A
ut
A
ho
ut
r
ho
M
ran
M
us
an

NIH-
NI
P
H-
A
P
A
A
ut
A
ho
Frequent Label Readers (o fItnefnr,equent Label
ut
r always/almost between
Figure 1. Test of Mediation of the Relationship always) Reade (never/rarely,
Attitude toward Healthy Eating an
sometime
ho
M Overall Dietary Quality by Nutrition Label Use
Note: All regression analyses included 418 775
ran Na the following covariates: gender, race/ethnicity, and
age
M
us a Overall Dietary Quality ranged from 0 to 8 with 1 point for each of the following:
DEMOGRAPHIC VARIABLE
an met dairy recommendation (≥3 servings/day); met 22.0 (N=418)
Age (years) 21.3 (N=773)
fat recommendation (<35% calories
BMI (kg/m 2
) from fat) 24.7 (N=418) 24.7 (N=775)
Female (based on
met fruit/vegetable recommendation 64%* age,
gender, 47% (N=773)
and physical activity)
(N=418)
* (N=418)
NIH- met calcium recommendationWhite 58%
(based on gender, age) 41% (N=775)

NI
P met added sugar recommendation (based on gender, age, and physical activity)
DIETARY VARIABLE
H-
A met fiber recommendation (based on gender, age)
Added Sugar (recommendation met)[20] 41.1%* (N=399) 24.7% (N=716)
P
A ate fast food <3 times per
24.0% (N=371) 20.2% (N=664)
A
[18]
Calcium (recommendation met)
ut week ate a vegetarian diet
Dairy (≥3 servings/day) [19] 16.3% (N=412) 14.9% (N=759)
A
ho
Graham and Laska Fast Food (<3 times/week)[21-22] 39.5%* (N=418) 26.6% (N=775) e
ut
r
Fat (<35% of calories)[16] 93.0% (N=415) 90.4% (N=762)
ho
M 1
Fiber (recommendation met) [17] 15.6%* (N=392) 6.1% (N=707)
ran
T [17]
Fruit/Vegetable (recommendation met) 5.3%* (N=400)
Comparison 2.1% (N=721)
of demographic and dietary variables
M
us a among frequent
*
9.0% (N=415) and infrequent
4.0% label readers
(N=765)
Vegetarian[23-24]
an b
Total Healthy Dietary Practices b (mean)
l 2.5* (n=417) 1.9 (N=754)
r

Absolute differences between participant estimated need and

recommended intake
Calories 636.0* (N=407) 868.5 (N=735)

Fruit (servings) 1.5* (N=404) 1.8 (N=710)

Vegetables (servings) 2.8* (N=404) 3.3 (N=714)

Importance of preparing healthy mealsc 4.3* (N=417) 3.6 (N=769)

 *
p<0.001
a
Variations in sample size for dietary variables due to removal of outliers 3 standard
deviations above the mean.

b
Total Healthy Dietary Practices was calculated as the total number of the 8 dietary recommendations the
participantmet; for those participants

who were missing up to 3 dietary variables, this variable was

calculated as the percentage of recommendations met for all valid
variables multiplied by 8 (participants missing more than 3 dietary
variables were also coded as missing the composite variable.)
c
This item, “It’s important to me to prepare healthy meals,” had response options ranging from 1 (“strongly
disagree”) to 5 (“strongly agree.”)

Nutrients 2010, 2, 1075-1085; doi:10.3390/nu2101075

OPEN ACCESS
nutrients
ISSN 2072-6643
www.mdpi.com/journal/nutrients

Article

Trends in Body Fat, Body Mass Index and Physical Fitness Among Male and
Female College Students
Peter Pribis 1,*, Carol A. Burtnack 1, Sonya O. McKenzie 1 and Jerome Thayer 2

1
Department of Nutrition and Wellness, Andrews University, 8475 University Boulevard - Marsh Hall 313, Berrien Springs,
MI 49104, USA; E-Mails: burtnack@andrews.edu (C.A.B.); mckenzis@andrews.edu (S.O.M.)
2
Center for Statistical Services, Andrews University, 4195 Administration Drive, Berrien Springs, MI 49104, USA; E-Mail:
thayerj@andrews.edu (J.T.)

* Author to whom correspondence should be addressed; E-Mail: pribis@andrews.edu; Tel.: +1-269-471-3386; Fax:
+1-269-471-3485.

Received: 10 September 2010; in revised form: 14 October 2010 / Accepted: 22 October 2010 /
Published: 25 October 2010

Abstract: There have been many publications in recent years reporting on the quantity of physical activity among
college students using indirect indicators such as steps walked per day or time spent on physical activities. The
purpose of this study was to describe the trends of physical fitness related to BMI and body fat among university
students between 1996 and 2008. The results showed a significant decline in the average fitness levels measured
as an estimation of VO2max for male and female students (p < 0.001 for both sexes). The linear trend for BMI by
years was not significant for both sexes (p for males = 0.772, p for females = 0.253). On average, in the last
13 years, % body fat was increasing 0.513%/year for males and 0.654%/year for females. There is a significant indirect
correlation between the student’s VO 2max levels and % body fat, r = −0.489; p < 0.001 for males; and r = −0.416, p
< 0.001 for females. Approximately 23.9% of the variance in the VO2max levels in males and 17.3% in females
can be explained by the variance in % body fat. The results support recent findings that physical fitness among
college students is declining and body fatness is increasing.

Keywords: students; obesity; BMI; % body fat; physical fitness; VO 2max levels
Nutrients 2010, 2 1076

1. Introduction

In the recent decade, a decline in physical activity among college students has been observed [1-3]. Regular physical
activity is an important part of a healthy lifestyle. It is associated with decreased risk of heart disease [4], obesity [5], and
cancer [6] and related to psychological well-being with lower levels of stress [7-9] and better cognitive functioning [10].
Recent studies indicate that almost half of the U.S. college student population does not participate in moderate or vigorous
physical activity [11]. There is an alarming decline in physical activity among college students compared with those in
high school [12]. Based on self-reported height and weight, approximately 35% of US college students are overweight or
obese [13].
There have been several publications in recent years reporting on the quantity of physical activity performed by college
students [3,11,14]. The primary purpose of our study is to describe changes in the levels of physical fitness measured as
estimation of maximum uptake of oxygen during a graded exercise (VO2max) in the last 13 years in relationship to
measurements of obesity (BMI, % body fat) in the male and female population. Our study was conducted at Andrews
University, which is a private Seventh-day Adventist (SDA) campus. SDAs represent a unique population to study
because of their emphasis on healthy lifestyle [15,16]. This religious group strongly recommends for its members to
follow a healthy lifestyle defined as proper use of “air, sunlight, temperance, rest, exercise, proper diet, water, and trust.”
[17].

2. Materials and Methods

This descriptive longitudinal study analyzed MicroFit tests data accumulated between years 1996 to 2008. Since MicroFit
testing was part of mandatory curricula for all physical activity classes, informed consent forms were not collected. The
study was approved by the Institutional Review Board of Andrews University (IRB Protocol #07-119).

2.1. Participants

Overall, 5101 students took the MicroFit test which was always done during the fall semester (September–December)
and spring semester (January–April) of each year. Every Andrews University student that participated in a physical activity
class was required to evaluate his/her physical fitness by taking the MicroFit test.

2.2. MicroFit testing

MicroFit is a Food and Drug Administration (FDA) approved medical assessment tool which measures several
parameters of physical fitness: body composition (weight, height and % body fat), blood pressure (systolic and diastolic),
muscular strength, flexibility, resting heart rate, aerobic fitness (estimation of VO2max), and calculates a total fitness
score [18]. In addition, the MicroFit data contained the name, age, sex, and the date the student took the test. For
statistical analysis the names of the students were removed. For the testing we used the MicroFit FAS-2 System which
consist of automated system for measuring resting blood pressure and heart rate, interactive skinfold calliper, flexometer for
flexibility testing, medical scale for measuring weight and biceps strength, a stationary
bike for cardiovascular fitness testing, chest strap transmitter and heart rate receiving unit and the
MicroFit Health Wizard software version 5.3.5.

2.3. Body composition

Body composition evaluation measured the major structural components of the body: muscles, and fat. The height was
measured in cm and the weight in kg. Body Mass Index (BMI) was calculated using the weight and height of the student
(BMI = W/H², where W is weight in kilograms and H2 is height in meters squared). The skinfold test, used to calculate
the percent body fat, was measured in millimeters and taken from three sites on the right side of the body: chest,
abdomen, and thigh for males, and triceps, suprailium and thigh for females. For males, the three skinfold measurements
were added up (MSF = chest + abdomen + thigh) and computed with the Jackson-Pollock method [19] to measure body
density: Db = 1.10938 − 0.0008267(MSF) + 0.0000016(MSF) 2 − 0.0002574(age). The following formula was then used to
compute % body fat: % Body Fat = [(4.95/Db) − 4.5] × 100. For females, the three skinfold measurements were added
up (FSF = triceps + suprailium + thigh) and computed with the Jackson-Pollock-Ward method [20] to measure
body density: Db = 1.10994921 − 0.0009929(FSF) + 0.0000023(FSF)2 − 0.0001392(age). The following formula
was then used to compute % body fat: % Body Fat = [(5.01/Db) − 4.57] × 100. All calculations were done automatically
by the computer.

2.4. Blood pressure and resting heart rate

Blood pressure was measured using the oscillometric method with an automated MicroFit FAS-2 blood pressure
system. Blood pressure was measured before any active test like biceps strength or aerobic fitness. The participant was
asked to sit quietly for 5 minutes before the measurement. During the measurement the participant was sitting upright, relaxed
with both feet flat on the floor, and their arm resting on the table. The cuff was wrapped around their upper arm so the
bottom edge of the cuff was 2–3 cm above the point where the upper arm joins the lower arm. The lower edge of the cuff
was at the level of the participant’s heart. The blood pressure cuff was connected to the computer and the measurement
was automatic. The systolic and diastolic blood pressure was reported in millimeters of mercury (mmHg).
Resting heart rate was determined automatically at the end of the blood pressure measurement. If the participants resting
heart rate was above 100 beats per minute (bpm) he/she was asked to wait another 5 minutes to repeat the measurement.
If the heart rate remained high the participant was told that their resting heart rate was out of range for the Fitness Profile, and
that he/she should mention the elevated heart rate during their next visit with their physician.

2.5. Muscle strength

The biceps strength test measures the maximum force generated in a single pull. This is a static test where the elbows are
fixed at 90 degrees and there is no bar movement during the pull. This measurement was done using the static strength-
testing device the MicroFit FAS-2 strength platform. Before the test the technician asked the participant: “Do you have a
back or arm injury or is there any
other reason you should not lift heavy objects?” If the participant’s answer was positive, the biceps test was skipped. The
participants were instructed to keep their back straight during the pull and to rotate their pelvis forward by squeezing their
buttocks together. Holding the bar with palms facing up and the strap adjusted so the elbow join was at 90 degrees, the
participants were asked to pull the bar up using biceps muscles only for 3 seconds. The program then calculated the final
strength score in kilograms.

2.6. Flexibility

Flexibility was assessed using the MicroFit FAS-2 flexometer to measure lower back and hamstring flexibility. The
participants sat on the floor, with their shoes off, their legs straight, and feet against the flexometer foot stop. Before the test
the technician asked the participant: “Do you have a back injury or is there any other reason you should not try to touch
your toes?” If the participant’s answer was positive, the flexibility test was skipped. When participant reached forward and
touched the flexometer for 3 seconds, a measurement was recorded in centimeters.

2.7. VO2max—the maximum uptake of oxygen during a graded exercise

Estimation of VO2max measured the maximum uptake of oxygen during a graded exercise. This study estimated the
VO2max using the Åstrand bike protocol. This protocol was recommended for young and middle-age adults. The Åstrand
bike protocol started with a three minute work stage and then used three additional two minutes work stages thereafter.
If the heart rate at the end of a work stage was below the Threshold Exercise Heart Rate (TEHR), the work load depending on
the heart rate was increased by 0.5 kp (25 Watt) or 1.0 kp (50 Watt). If heart rate at the end of a work stage was
above the TEHR, the work level was maintained for three more minutes and then the test was complete. The TEHR
for ages below 50 years is 120 bpm, for ages above 50 years is 125 bpm. The VO2max was then calculated based upon the
final work load and the average of the last two minute heart rates. The formula for VO2max estimation was: VO2max
(mL/kg/min) = VO2max (from table) × 1000/weight (kg) × age factor [21]. The age factor was computed using the following
equation: 100/(1.37 × age + 66.8). Before the test, male participants answered the following question: “During the past three
months did you engage in vigorous activities like running or cycling for at least 15 minutes/day on three days/week?”
If the answer was positive, the initial load was set automatically at 2.0 kp (100 Watt). If the answer was negative the initial
load was 1.5 kp (75 Watt). For females the initial load was constantly set at 1.0 kp (50 Watt). The participants wore a heart
rate monitor during the test. The computer responded to changes in heart rate by automatically adjusting the work load level of
the bicycle ergometer. The entire test takes ten minutes, including a two-minute cool down period at the end. The
computer then produces a printout with the test results and comparison of the results to the national norms for the
appropriate age and sex group.

2.8. Statistical analysis

The Statistical Package for the Social Sciences (SPSS; version 18.0) was used for the data analysis. A linear regression
was used to measure trends between VO2max, BMI, body composition, and years.
Pearson correlations were computed between VO2max, BMI and body composition. P values less than or equal to 0.05 on
two-sided tests were considered statistically significant.

3. Results and Discussion

3.1. Demographics

Table 1 describes the characteristics of the study population. Out of the 5101 Andrews University students who
participated in this study, 45% were males and 55% were females. Mean % body fat for males was 11.6% and 22.4%
for the females, respectively. The mean VO2max for males was
38.7 mL/kg/min and 34.2 mL/kg/min for females, respectively. The mean BMI was 24.1 kg/m2 for males and 24.0 kg/m2
for females, respectively.

Table 1. Characteristics of the study population.

Males Females
Gender (%, n) 44.5 (2273) 55.4 (2828)
Age (years; mean ± SD) 21.5 ± 4.6 21.9 ± 5.8
% Body Fat (mean ± SD) 11.6 ± 6.5 22.4 ± 6.7
Bicep Strength (kg; mean ± SD) 41.7 ± 12.0 23.8 ± 7.4
Flexibility (cm; mean ± SD) 37.8 ± 12.1 43.1 ± 11.6
Systolic BP (mmHg; mean ± SD) 129.4 ± 15.3 118.4 ± 14.1
Diastolic BP (mmHg; mean ± SD) 76.7 ± 10.2 73.5 ± 9.4
Resting Heart Rate (bpm; mean ± SD) 73.1 ± 13.5 78.7 ± 12.9
VO2max (mL/kg/min; mean ± SD) 38.7 ± 11.0 34.2 ± 10.2
Height (cm; mean ± SD) 175.0 ± 7.5 162.2 ± 7.0
Weight (kg; mean ± SD) 77.1 ± 15.8 65.6 ± 16.0
BMI (kg/m2; mean ± SD) 24.1 ± 4.5 24.0 ± 5.3
BMI: body mass index; BP: blood pressure; SD: standard deviation.

3.2. Trends in fitness levels

Figure 1 and Figure 2 represents the trends in physical fitness as expressed in estimates of VO2max and BMI and %
body fat for males and females between the years 1996 and 2008. There is a significant linear trend between the
VO2max and the years (r = −0.248, p < 0.001) for males as well for females (r = −0.135, p < 0.001). On average, in the
last 13 years the VO2max was decreasing 0.812 mL/kg/min a year for males and 0.414 mL/kg/min a year for
females.
There was a noticeable increase in VO2max in 1998 and 1999, followed by noticeable decreases in VO2max in males and
females in the years 2000, 2002 and 2007. Females step by step recovered back between 2003 and 2006. In the year
2007 we observed the largest decline in VO2max in males and females out of the last 13 years. In 2008, the level of
physical activity for the male population bounced back, with an increase in VO2max, while the female results showed
less improvement.
3.3. Trends in BMI

The linear trend for BMI by year was not significant for both sexes (for males p = 0.772, p for females = 0.253).
Combining data for all years there was a significant indirect correlation between the student’s VO2max levels and BMI, r =
−0.334; p < 0.001 for males; and r = −0.414, p < 0.001 for females. Approximately 11.1% of the variance in the
VO2max levels in males and 17.1% in females can be explained by the variance in BMI.

Figure 1. Trends in physical fitness, BMI and % body fat in males (1996–2008).

45

40

35

30

25

20

15

10

5
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Linear trends for Fitness and % Body Fat (p < 0.001)
Fitness BMI % Body Fat Linear (Fitness) Linear (% Body Fat)
VO2max (mL/kg/min) kg/m2

Figure 2. Trends in physical fitness, BMI and % body fat in females (1996–2008).

45

40

35

30

25

20

15

10

5
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Linear trends for Fitness and % Body Fat (p < 0.001)
Fitness BMI % Body Fat Linear (Fitness) Linear (% Body Fat)
VO2max (mL/kg/min) kg/m2
3.4. Trends in % body fat

One of the problems associated with BMI is that there is no way to know if the person is heavily muscled or
overweight. Therefore, body fat percentage has been recently recommended as a more accurate measurement of body
fatness [22-25]. There is a significant linear trend between the % body fat and the years (r = 0.264, p < 0.001) for males
as well as for females (r = 0.324, p < 0.001). On average, in the last 13 years, % body fat increased 0.513%/year for
males and 0.654%/year for females. There is a significant indirect correlation between the student’s VO2max levels and % body
fat, r = −0.489; p < 0.001 for males; and r = −0.416, p < 0.001 for females. Approximately 23.9% of the variance in the
VO2max levels in males and 17.3% in females can be explained by the variance in
% body fat.

3.5. Fitness categories

Using MicroFit’s own criteria for appropriate age and sex we have divided the students according to their aerobic fitness
VO2max results into four fitness categories—poor, fair, fit and excellent. Figures 3 and 4 show the percentage of males and
females who fell into these four fitness categories. The decline in physical fitness occurred differently in males and females.
Starting with the year 1999, males dramatically increased in the poor category. The trend peaked in the year 2007 and
there was a rebound in the year 2008. While the poor category increased disproportionally, there were fewer and fewer
males, who could be considered excellent, fit or fair (Figure 3). In contrast to males, the percentage of females in
the poor category fluctuated up and down. There were several rebounds, in 2003, 2006 and 2008; however, they were not
as pronounced as in males. In 2007 there was a dramatic increase of females in the poor category. As the poor fitness category
increased there were fewer losses from the fair, fit, and excellent categories in females in comparison to males (Figure 4).
Although both sexes have declined in their physical fitness levels, the results from our study show that there is a more
pronounced and deeper decline in the males then females.

Figure 3. Percentage of males falling into the four difference fitness categories (1996–2008).

100%
90%
80%
70%
60%
50%
40%
30%
20%
10%
0%
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008
Poor Fair Fit Excellent
<35.9 mL/kg/min 36.0–40.9 mL/kg/min 41.0–48.0 mL/kg/min >48.0 mL/kg/min
 Figure 4. Percentage of females falling into the four difference fitness categories (1996–2008).

100%
90%

80%
70%
60%

50%
40%

30%
20%
10%

0%
1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008

Poor Fair Fit Excellent

<29.9 mL/kg/min 30.0–34.9 mL/kg/min 35.0–42.0 mL/kg/min >42.0 mL/kg/min

The major finding of this study was the slow, gradual decline in fitness levels during the last 13 years, measured
as an estimation of maximal oxygen consumption (VO 2max) on a bicycle ergometer in college men and women. At the same
time, we observed a slow increase in % body fat values for both sexes. The value of our study is the reported estimation
of VO2max, which is the most accurate measurement of physical fitness. In the last few years, several publications have
reported indirect indications of physical fitness, like steps walked per day, or time spent participating in physical
activities [12-13,26-28].
The Dietary Guidelines for Americans [29] indicate that adults should perform moderate activity for at least 30 minutes
daily, or most of the week, and participate in activities to strengthen muscles and improve flexibility. The 2008 Physical
Activity Guidelines for Americans [30] recommend that adolescents should engage in 60 minutes or more of physical
activity daily. Most of the 60 minutes or more should be either moderate- or vigorous-intensity aerobic physical activity,
and should include vigorous-intensity physical activity at least three days a week. As part of their 60 or more minutes of
daily physical activity, children and adolescent should include muscle-strengthening physical activity and bone-strengthening
physical activity on at least three days of the week. Students who follow these guidelines should be able to improve or at
least maintain their VO2max levels. Unfortunately, research indicates that college years are characterized by the
worsening of food habits [31], meal skipping [32,33], snacking [34], and frequent consumption of fast foods [32,34].
College years are also characterized by rapid reduction in physical activity [1,2,12,35] and the beginning of a sedentary
lifestyle [36].

3.6. Limitations

There are a few limitations to the study that need to be considered. First, the technicians who collected the data
changed over the time period in which students were taking the MicroFit test, and
therefore the techniques between technicians may have varied. Secondly, the software and testing equipment
experienced difficulty at random times which contributed to some missing data.
The students took the exam at different times throughout the semester as part of their physical education activity
course requirements. Students who took the MicroFit test at the beginning of the semester may have had worse
results that those who took the test later in the semester after participating in a physical activity. However, there is no
evidence that only students with a poor fitness score completed the MicroFit at the beginning of the semester and thus it
can be assumed that the fitness scores averaged over the course of the school semester. Because MicroFit does not contain
data on ethnicity, generalizability of the results may be limited.

4. Conclusions

Our findings show that there is a small and declining minority of male and female college students who are physically
in shape. The MicroFit data shows that the fitness levels measured as estimation of VO2max have gradually decreased among
males and females over the past 13 years. The BMI and
% body fat has fluctuated up and down, and is significantly indirectly correlated with VO2max levels. As VO2max decreased,
the BMI and % body fat increased for both sexes. The observed trends are unfortunate because it has been demonstrated
that physical activity and good nutrition can have a positive effect on the overall performance of students. Physical
activity can reduce stress levels and improve work-related time management. Strategies should be implemented to counter
the present trend and help young people improve their physical fitness.

Acknowledgements

I would like to express my gratitude to the many technicians who administered the MicroFit test during the period of
the study and to Gregory Morrow and Dominique Wakefield for supervising this endeavour. Thanks to Winston J. Craig,
Alice C. Williams, Carolyn J. Baltazar, and Rachelle M. Booth for their technical assistance.

References

1. American College Health Association. American College Health Association-National

College Health Assessment (ACHA-NCHA) Spring 2005 Reference Group Data Report
(Abridged). J. Am. Coll. Health 2006, 55, 5-16.
2. American College Health Association. National College Health Assessment Reference
Group Executive Summary Fall 2008. Available online: http://www.acha-ncha.org/docs/ACHA-
NCHA_ Reference_Group_ExecutiveSummary_Fall2008.pdf (accessed on 12 June
2010).
3. Sacheck, J.M.; Kuder, J.F.; Economos, C.D. Physical fitness, adiposity, and metabolic risk factors
in young college students. Med. Sci. Sports Exerc. 2010, 42, 1039-1044.
4. Powell, K.E. Habitual exercise and public health: An epidemiological view. In Exercise
Adherence: Its Impact on Public Health; Dishman, R.K., Ed.; Human Kinetics: Champaign, IL, USA, 1988; pp. 15-
40.
5. Shaw, K.; Gennat, H.; O’Rourke, P.; Del Mar, C. Exercise for overweight or obesity.
Cochrane Database Syst. Rev. 2006, 4, CD003817.
6. Coyle, Y.M. Lifestyle, genes, and cancer. Methods Mol. Biol. 2009, 472, 25-56.
7. Brown, J.D. Staying fit and staying well: physical fitness as moderator of life stress. J. Pers.
Soc. Psychol. 1991, 60, 555-561.
8. Pertruzello, S.J.; Landers, D.M.; Hatfield, B.D.; Kubitz, K.A.; Salazar, W. A meta-analysis on
the anxiety-reducing effects of acute and chronic exercise. Outcomes and mechanism.
Sports Med. 1991, 11, 143-182.
9. Crews, D.J.; Landers, D.M. A meta-analytic review of aerobic fitness and reactivity to
psychological stressors. Med. Sci. Sports Exerc. 1987, 19, S114-S120.
10. Etnier, J.L.; Salazar, W.; Landers, D.M.; Petruzzelo, S.J.; Myungwoo, H.; Nowell, P.
The influence of physical fitness and exercise upon cognitive functioning: a meta-analysis.
J. Sport Exerc. Psychol. 1997, 19, 249-277.
11. Douglas, K.A.; Collin, J.L.; Warren, C.; Kann, L.; Gold, R.; Clayton, S.; Ross, J.G.;
Kolbe, L.J. Results from the 1995 national health risk behaviour survey. J. Am. Coll. Health
1997, 46, 55-66.
12. Bray, S.R.; Born, H.A. Transition to university and vigorous physical activity: implications
for health and psychological well-being during transition to university life. J. Am. Coll. Health
2004, 52, 181-188.
13. Lowry, R.; Galuska, D.A.; Fluton, J.E.; Wechsler, H.; Kann, L.; Collins, J.L. Physical activity,
food, choice, and weight management goals and practices among US college students. Am.
J. Prev. Med. 2008, 18, 18-27.
14. Center for Disease Control. The Behavioral Risk Factor Surveillance System (BRFSS). Available
online: http://www.cdc.gov/brfss/index.htm (accessed on 1 May 2010).
15. Beeson, L.W.; Mills, P.K.; Phillips, R.L.; Andress, M.; Fraser, G.E. Chronic disease
among Seventh-day Adventists, a low risk group. Cancer 1989, 64, 570-581.
16. Fraser, G.E.; Shavlik, D.J. Ten Years of life—Is it a matter of choice? Arch. Intern. Med.
2001,
61, 1645-1652.
17. White, E.G. The Ministry of Healing; Pacific Press Publishing Association: Mountain View, CA,
USA, 1905.
18. MicroFit Health and Fitness Systems. Colleges & Universities. Available online:
http://www. microfit.com/site_colleges/index.html (accessed on 1 May 2010).
19. Jackson, A.S.; Pollock, M.L. Generalized equations for prediction body density of men. Br. J.
Nutr. 1978, 40, 497-504.
20. Jackson, A.S.; Pollock, M.L.; Ward, A. Generalized equations for predicting body density of
women. Med. Sci. Sports Exerc. 1980, 12, 175-182.
21. Åstrand, P.-O.; Ryhming, I. A nomogram for calculation of aerobic capacity (physical
fitness) from pulse rate during submaximal work. J. Appl. Physiol. 1954, 7, 218-221.
22. Wang, J.; Thornton, J.C.; Kolesnik, S.; Pierson, R.N., Jr. Anthropometry in body
composition. An overview. Ann. N. Y. Acad. Sci. 2000, 904, 317-326.
23. Powers, S.K.; Howley, E.T. Exercise Physiology: Theory and Application to Fitness and
Performance; McGraw Hill: Boston, MA, USA, 2009.
and Human Performance; Lea & Febiger:
24. McArdle, W.D.; Katch, F.I.; Katch, Philadelphia, PA, USA, 1991.
V.L. Exercise Physiology: Energy, Nutrition, 25. Wilmore, J.H.; Costill, D.L. Physiology of
 Sport and Exercise; Human Kinetics:
Champaign, IL, USA, 2004.
26. Mestek, M.L.; Paisance, E.; Grandjean, P.
The relationship between pedometer-
determined and self-reported physical
activity and body composition variables in
college-aged men and women. J. Am. Coll.
Health 2008, 57, 39-44.
27. McArthur, L.H.; Raedeke, T.D. Race and
sex in college students physical activity
correlates. Am.
J. Health Behav. 2009, 33, 80-90.
28. Taliaferro, L.A.; Rienzo, B.A.; Pigg,
R.M.; Miller, M.D.; Dodd, V.J.
Association between physical activity
and reduced rates of hopelessness,
depression, and suicidal behavior
among college students. J. Am. Coll.
Health 2008, 57, 427-435.
29. USDA/HHS. Nutrition and Your Health:
Dietary Guidelines for Americans, 5th ed.;
US
Department of Agriculture and Health and Human
Services: Washington, DC, USA, 2000.
30. USDA/HHS. 2008 Physical Activity Guidelines
for Americans; US Department of
Agriculture and Health and Human
Services: Washington, DC, USA,
2008.
31. Grace, T.W. Health problems of
college students. J. Am. Coll. Health 1997,
45, 243-250.
32. Nicklas, T.A.; Baranowski, T.; Cullen,
K.W.; Bergerson, G. Eating patterns,
dietary quality and obesity. J. Am. Coll.
Health 2001, 20, 599-608.
33. Center for Disease Control and
Prevention; National Center for Chronic
Disease Prevention and Health
Promotion. Physical Activity and Health: A
Report of the Surgeon General; US
Department of Health and Human
Services: Washington, DC, USA,
1996.
34. Hertzler, A.A.; Webb, R.; Frary, R.
148
Overconsumption of fat by college
students, the fast food connection. Ecol.
Food Nutr. 1995, 34, 49-57.
35. Grubbs, L.; Carter, J. The relationship of
perceived benefits and barriers to
reported exercise behaviours in college
undergraduates. Fam. Community Health
2002, 25, 76-84.
36. Pinto, B.M.; Marcus, B.H. A stage of
change approach to understanding
college students’ physical activity. J. Am.
Coll. Health 1995, 44, 27-31.
© 2010 by the authors; licensee MDPI, Basel, Switzerland.
This article is an open access article distributed under
the terms and conditions of the Creative Commons
Attribution license
(http://creativecommons.org/licenses/by/3.0/).
See discussions, stats, and author profiles for
this publication at:
https://www.researchgate.net/publication/84
42268
Physical
Activity,
Exercise,
and
Sedentary
Behavior in
College
Students
Article in Journal of American College Health
· July 2004
DOI: 10.3200/JACH.53.1.28-34 · Source: PubMed
CITATIONS
 RE DS
A 2,903

2 authors:

Janet Buckworth
Claudio R Nigg
The Ohio State
University of Hawai'i System
University
211 PUBLICATIONS 5,298 CITATIONS
27 PUBLICATIONS 728
CITATIONS

as uploaded by on 11 January 2017.

f the downloaded file. All in-text references underlined in blue are added to the original document and are linked to publications on ResearchGate, le
 JOURNAL OF AMERICAN COLLEGE HEALTH, VOL. 53, NO. 1

Physical Activity, Exercise, and

Sedentar Behavior in College
Students

Janet Buckworth, PhD; Claudio Nigg, PhD

AbstractT.he authors examined the relationship betweenthe previous 7 days. Many adult health behaviors are esta
phys-
ical activity, exercise, and sedentary behaviors in 493 lished during late adolescence and early6 asdoutlhtheood,
college
students who were enrolled in 10 conditioning activity classes
decline in physical activity in adolescence and young adul
he course.
ticipation
and had completed questionnaires at the beginning of t They analyzed les. Sur-
sedentary activities and indicators of par hood is a disturbing 3tr,7ePnhdy.sical inactivity, a serious
tivity and e
and
watch-
in exercise and physical activity by demographic variab pervasive public health concern, has been designated as o
tudents
younger
veyed men reported greater participation in physical ac of the priorities in the Healthy People 2010 Objectives (i exercise than did women, and
they also spent more tim Focus Area 22, Physical Activity and Fitness,ators.and OfObjective ing television/videos and using the computer. Older s 22-
nd televi-
1, 22-2, and 226 -a3n)d Healthy Campuses 2010 Cam- ith exer-
were more likely to report more computer use, whereas nd rein-
paign Objective 22-2/3. 8
-related
students scored higher on several physical activity indic
hould be se
the sedentary behaviors, only computer use for men a inExercise
this and physical activity have been described in
sion watching for women were negatively correlated w national samples (National College Health Risk Behavio cise and physical activity. In
addition to accessibility a Survey: NCHRBS, 199; 9N5ational College Health Assess- forcing properties of sedentary activities, the
entary
gender
ment: NCHA, 205)0,0and in various college populations
relationships between sedentary and active behaviors s
(eg, African Americand older stude)n.tHs owever,
10 11

considered in designing interventions to promote exerci fewin

bed
ctiv-
population. studies have examined the relationship between studen
opt
exercise and physical activity and their sedentary behavio
level of
Key Wordsc:ollege students, exercise, physical activity, sed Physical inactivity has been associated with the risk of se
chronic diseases and health conditions 1(2)e,g, obesity
eral are
and

T
and the role of sedentary behaviors in public health is no he
health benefits of physical activity descri gaining more atten13tiToenle. vision watching has been asso-
the US Surgeon General’s Report on Physical Aciated
hat theywith obesity in ch1i4l,d15raennd adul1t6s,and reduc-
1
ity and Hea ltphrovide compelling reasons to adeasttions in this sedentary behavior in children have been ass
3 of
and maintain an active lifestyle. Unfortunately, the ciated with decreases in body ma1s7s index.
physical activity declines during The contribution of sedentary behaviors to the risk o
ado2,l3essoctehnacteabout
70% of American adults are sedentary or inactive physical inactivity is not clear. In a sample of college stu
below the recommended level for dents, television watching was not strongly correlated wi
healt4hAbcceonredfiitnsg. to the 2000 National 14 physical activity variables (eg, total minutes spent i
College Health As5s5e7ss%moenf t,
male and 61% of female college students reported tleisure/moderate activity and number of days in the previo
performed no vigorous or moderate exercise on at l 7 days of participation in strength training) although it w
correlated with physiological indicators of fitness such a
resting heart rate and recovery pulse following a 3-minu
Janet Buckworits han associate professor of Sports and step tes18t.Other sedentary activities, including reading,
Exercises Sciences at The Ohio State University, Columbus, and studying, or computer use, can compete with exercise wh
Claudio Nigisgan associate professor in the Department of individuals are making choices about how to spend their d
Public Health Services, John A. Burns School of Medicine,
University of Hawaii at Manoa. cretionary time. For example, in a study of young Australia

28
 se were EXERCISE & SEDENTARY BEHAVIOR

adults, those in the highest tertile of computer u tered questionnaires, including items on age, gender
ing the
most likely to be physically in1a9 ctive. racial/ethnic group, and academic class, during the first le
Behavioral economics provides a model for study ture classon
ctivity of the 10-week academic quarter according to
decision to choose sedentary or active 2b0eInhdaivviiodr-s. motiva- protocol approved by the University Office of
Research college
uals base their decisions to engage in a particular a Risks heir acad-
Protection.
access (the work needed to obtain the activity) and etionary
tion (the reinforcing value of the activity). Although his time Measures
factors ng
students have specific time constraints related to t Exercise Behavior
emic schedules, they also have considerable discr value
Questions related to participation in exercise from th
time. The choices they make about how to spend t ol-related 9
influence their level of physical activity, and variousNCHRBS heystudents
provided data on the number of the past 7 days
near participated in vigorous exercise, stretchi
that
influence these choices. For example, the reinforci e more
exercises, activities to strengthen or tone muscles, and mo
and subsequent participation in studying and scho erate intensity exercise. We also asked 3 additional que
computer use may increase for college students as t was to
graduation, whereas organized exercise may tak tions about typical exercise duration (min), frequency i
hysical
days per week, and length of time in months exercising
lege stu-
effort and their participation can decrease. this level. One-week test-retest reliabilities for these 3 ite
ferences
Therefore, our purpose in conducting this study d gender
analyze the relationship between exercise and p were 0.93, 0.90, and 0.97, respeNct=ive5l6y2)1.(
activity and sedentary behaviors in a sample of col s among
Physical
ld Activity History
dents. We conducted analyses of gender and age dif
tudents
We used the CARDIA Physical Activity History Ques-
because researchers have found significant age an
nd phys-
23
differences in physical activity and its determinant tionnair nificant
eto measure participation in moderate intensity
1 1 ,2 1
college stude n t s .We hypothesized that (1) mensity (eg,
wouactivitieswalks or hikes, home maintenance) and vigorous inte
be more active than women would be; (2) older s exercise (eg, jogging or running, strenuous sports would
engage in studying and computer use more a physical activities
dentary over the previous 12 months. We com
ical activities less than younger students; (3) sig puted but notactivity scores based on reports of frequency of pa
positive relationships among different measures of ticipation per month and number of months of participatio
rrelated
and physical activity but not among different se Significant relationships have been reported between CA
activities were likely, and (4) television watching DIA activity scores and objective and physiological indica
2 3 2 1
studying or computer use would be negatively co tors of exerc is We.allace and assoc ia rtepsorted 1- week
with exercise. test-retest reliabilities of 0.81 (moderate), 0.89 (vigorous
and 0.88 (total) in a college s23ample.
METHOD western
Sedentary Behaviors
Participants n our
We used 3 questions to measure participation in sede
ffects of which were similar to procedures used i university
Students enrolled in the classes at a large mid tary activities,
in the autumn (6 cNla=ss2e1s,5) and spring (4 other studi change
2 4 ,2 5
e s . Participants reported the numbers of hours
ne
classesN, = 278) academic quarters participated i in a typical week they spent watching television and/o
study. The classes were part of a study to test the evideos, 73.8% studying, and using the computer. In an earlie
curriculum modifications on mediators of behavior study, were
Wallace and asso2c1iraetpesorted 1-week test-retest
and exercise adher2e2nOcuer. analyses focused on baseli cial/eth-
reliabilities of 0.72 (watching television and/or videos)
0 years,
data. The sample included 288 women (58.3%); 0.62 (studying), and 0.87 (using the computer) a sample
were White, 16.2% were African American, 3.8% college 18students.
and
Asian, 2.2% were Hispanic, and 4% listed other ra (28.4%
Statistical 1.0%
Analysis
nic groups. The students’ average age was 21± 4.
96.5%)
and 90.8% of the sample were between the ages of To analyze data, we used Statistical Package for the Soci
24 years. Class rank was consistently distributed Sciences (SPSS) for Windows,™ version 11.0 (SPSS Inc. first year,
24.3% sophomore, 24.1% junior, and 21989–2001) and we used descriptive statistics to compu senior/5th year; 2.3%
not reporting), and most ( demographic variables. We ve con-
used analysis of variance
were enrolled in school fulltime. lecture
(ANOVA) and multiple regression with Tukey’s Honestly
Procedure tudents
Significant Difference (HSD) post hoc comparisons to ana
raining
lyze differences in physical activity indicators and sedenta
dminis-
The classes we included in the study were electi behaviors by sex and by age. Effect size was represented
ditioning activity courses consisting of a 50-minute 2 with .01, .09, and .25 representing small, medium, and
and 45-minute exercise labs 3 times week. More slarge effects, respect2i6vaenlyd, we used Pearson correlation
were enrolled in aerobic dance (44%) and weight tcoefficients to determine the relationships between the stu
(39%), than in jogging exercise labs (16%). We a variables. We set the level of significancepa<p0ri.o0r5i.at

VOL 53, JULY/AUGUST 2004 29

BUCKWORTH & NIGG

RESULTS computerp <( .05). Time spent studying and time viewing
Students reported that they had spent almost 30 television/videos were not associated with age. The young
a typical week engaged in sedentary behaviors, hours in
students reported more days per week devoted to stretch
mostly
studying (13.25 ± 9.16 hr), but they also indicated tandhad
hey greater participation in vigorous intensity activities d
participated in some exercise at the beginning of t ing the past 12 months; they also indicated they had be (Table
he class
1). More of these students engaged in adeq exercising at their current level for significantly mor els of moderate
uate lev- stupde<nt.s05(). These relation-
acti≥vi5tyo(f the previous 7 days, 30.6 months than the older
vs 19.5%) and vigorous ac≥tiv3itoyf (the previous 7 % ships changed when we analyzed the data separately
days,
f
53.2% vs 37.6%) than students who were sample dmen
in theand women. For men, the only exercise or physica
9 xer-
activity variable associated with age was the number o
1995 NCHRB SM. ore students in this sample also e
cised, compared with those surveyed in the 25000 Nmonths CHA of regular exerrci=se = .008). For
0(.193p, ercise at
survey on participation in vigorous or moderate ex women, there were significant negative correlation
least 3 times in the past 7 days (73.1% vs 40.5%). en and
between age and days per week of moderarte=activity (
iors and
–.178,p
We found significant differences between the m ted more = .004), exercise durart=io–n.1(27p, = .043), and
the women on variables indicating sedentary behav vigorous intensity activities during the rp=as–t.y1e6a7r,
( deos and
physical activity (Table 2). The male students repor ents (
p = .008).
er levels
hours per week spent watching television and/or vi We also analyzed the relationships between exercise
indica-
using the computer compared with the femaple<stud physical activity, and sedentary behaviors for the whol
.05). At the same time, the men self-reported high gesample
and (Table 3) and for men and women separately. I
). Corre-
of exercise compared with women on most activity general, exercise and physical activity variables were si
tors p( < .05). hat the related to each other, with stronger correlatio
nificantly
sing the
We found significant relationships between a between the measures representing higher intensity phy
sedentary behaviors and activity variables (Table 3 cal activity and exercise. Among sedentary behaviors, th lations
of sedentary variables with age revealed t amount of television watching was not related to the tim older the
student, the more time he or she spent uspent using computers. For men only, a significant positiv
correlation existed between television watching and tim
spent studyinrg=(.226p, = .002).
Correlations between sedentary behaviors and exercis
TABLE 1. Sedentary Behaviors and and physical activity revealed different relationships fo
Physical
Activity Indicators
men and women. For male students, television watchin
was not related to any of the exercise or physical activit
Variable M SD indicators, but computer use was negatively correlated wi
N
days per week they engaged in vigorous exercise an
Sedentary behaviors† strength training, typical exercise frequency, and modera
TV/Videos 10.56 8.59 475 and vigorous physical activity during the past year. Com
Study 13.25 9.16 476 puter use was not related to any of the exercise or physic
Computer 5.96 5.83 474
activity indicators for the female students. For them, tele
Total sedentary 29.72 15.19 474 sion watching was negatively correlated with all items fro
the NCHRBS, typical frequency of exercise, and history o
Activity indicators vigorous physical activity during the past year. Time spe
NCHRBS‡
Vigorous 2.79 1.97 482
Stretch 2.21 1.98 482 studying was positively correlated with days per week o
Strength 2.16 1.98 482 CARDIA¶

Moderate 2.97 2.40 481 Moderate 137.67 118.20 461

Typical exercise

Frequency‡ 3.41 1.96 479

Duration§ 47.35 36.46 477

Length|| 23.39 39.35 463

strength training for females, and with typical duration COMMENT
exercise for both males and females. In this study, we analyzed the relationships among sede
tary behaviors and exercise according to demographic va
Vigorous 305.07 263.61 461 ables of students who were enrolled in college conditionin
Total 442.74 350.87 461 activity classes. These students were typically physicall
active at the beginning of the course, reporting some cons Note. NCHRBS =
vey ; CARDIA = Coronary Artery Risk Development in
National College Health Risk Behavior Surt-ent exercise for almost 2 years, on average, and therefo 9
Young 2 3
Adultscould
. provide insight into sedentary behaviors of physical
†hr/wk active college students. We found no hypothesized positiv
‡d/wk relationship between age and studying, but increasing a
§min was associated with more time using the computer. The po
||mo

¶activity score/yr itive association between computer use and age may refle
the different academic demands for upper-class studen

30 JOURNAL OF AMERICAN COLLEGE HEALTH

 EXERCISE & SEDENTARY BEHAVIOR

TABLE 2. Study Variables, by Sex, in an Examination of Exercise and Sedentary Behavior

Men Women

Variable Mean SD N Mean SD N F df p 2

Sedentary behaviors†
TV/Videos 12.02 9.06 192 9.57 8.11 283 9.50 473 0.00 0.02
Study 12.92 10.01 193 13.47 8.55 283 0.41 474 0.52 0.00
Computer 6.74 6.65 192 5.42 5.15 282 5.86 471 0.02 0.01
Total sedentary 31.62 16.88 192 28.43 13.80 282 5.07 472 0.02 0.01
Activity indicators
NCHRBS§
Vigorous 3.28 2.05 198 2.45 1.84 284 21.66 480 0.00 0.04
Stretch 2.18 2.02 198 2.23 1.95 284 0.07 480 0.79 0.00
Strength 2.69 2.01 198 1.80 1.88 284 24.78 480 0.00 0.05
Moderate 2.94 2.43 198 2.99 2.38 283 0.03 479 0.85 0.00
Typical exercise
Frequency‡ 3.84 1.94 197 3.10 1.92 282 17.19 477 0.00 0.03
Duration§ 61.89 43.19 196 37.21 26.63 281 59.38 475 0.00 0.11
Length|| 31.32 45.91 195 17.62 32.70 268 14.07 461 0.00 0.03
CARDIA¶
Moderate 150.28 123.00 198 128.18 113.78 263 3.97 459 0.05 0.01
Vigorous 376.67 279.56 198 251.17 237.61 263 27.05 459 0.00 0.06
Total 526.94 371.70 198 379.35 320.72 263 20.85 459 0.00 0.04

Note. NCHRBS = National College Health Risk Beha9v;ioCrASRuDrIvAey= Coronary Artery Risk Development in Yo2u3ng Adults.
†hr/wk
‡d/wk
§min
||mo
¶activity score/yr

When we compared computer use between academi from the National Longitudinal Study of Adolescent the
c classes,
24
senior/5th year students reported significantly mHealthore com-
. However, compared with the 18-to-22-year age puter
use (7.6 ± 6.9 h/wk) than the freshmen (4.8 ± 4group in.7that h/wk),
study, participants in this sample watche p = .003 and
juniors (5.1 ± 5.2p h=/w.0k1)7, , but not sopho- fewer hours of television (10.56 vs 13.4 h/wk). Men in ou mores
n time that they spent more time using t spent studying or
(6.3 ± 5.8 hrs/wk). We found no difference istudy also reported
watching television as a function ocomputer than their ffemale
acade-counterparts did.
mic year. We hypothesized that television watching would be th
As we had predicted, female students’ age was negonly atively
sedentary activity correlated with measures of exerci
correlated with exercise (days/wk of moderate i and ntensity
physical activity, but we found unexpected relationshi
exercise and typical exercise duration) and physical activitybetween specific sedentary activities for female and ma
(vigorous intensity activities during the past year ).students.
These For women only, television watching was nega
decreas-
correlations reflect the continuation of a pattern of tively correlated with many exercise and physical activit
ing physical activity observed in hig2h7 esvcehnoothl,ough indicators, whereas studying was positively
correlated wi ay have
fact, forof typical exercise and strength training. W
enrollment in the conditioning activity classes m the duration
introduced a bias toward exercise in this sample. In found exercis-
no significant relationship between computer use a men,
nds
increasing age was associated with more time physical activity is an indicators. These results suggest that tele
ing at their current level. Social support from frie sion watching is the sedentary behavior that competes w
ation in
important exercise correlate 2f1oarnmdevnig, orous activity active behaviors for these college women’s
discretionar may be fostered in college through men’s llment
particip intime and is a greater risk factor for physical
inactivity f group activities, such as intramural sports and enro women than for men. Typically, television is more
accessi al activ-
conditioning activity classes. lyble
active
than exercise and requires less effort than computer u
Associations between gender and level of physic dentary
In addition, the entertainment and distraction that televis
ity reflect the literature; men were more physical ater for is immediate, but the reinforcing value of exerci
provides
ndings loss, improved fitness) is usually delayed.
than women; they also reported more time in se (eg, weight
activities. Hours of television per week were gre Although male students watched more television and use
men than for women (Table 2), which is similar to fithe computer more than female students, only computer u

VOL 53, JULY/AUGUST 2004 31

 B
32
U
C
K
W
O
R
T
TABLE 3. Correlation Matrix of Age, Sedentary Behaviors, and Physical Activity Indicators

Variable 1 2 3 4 5 6 7 8 9 10 11 12 13 14

Age 0.04 0.06 0.20** 0.13** –0.06 –0.09* –0.06 –0.08 –0.04 –0.02 0.13** 0.04 –0.10 –0.06
1. Television† 1.00 0.10* 0.05 0.65** –0.15** 0.12* –0.06 –0.11* 0.10* –0.01 0.06 –0.06 –0.13* –0.12*

2. Study† — 1.00 0.15** 0.72** 0.03 0.08 0.13** –0.02 0.04 0.14** –0.04 0.04 0.01 0.02

3. Computer† — 1.00 0.51** –0.05 –0.04 –0.04 –0.04 0.01 –0.01 –0.06 –0.07 –0.08 –0.09

4. Total sedentary† — 1.00 –0.09 –0.04 0.03 –0.09 –0.03 0.07 –0.01 –0.04 –0.10* –0.09
NCHRBS‡
5. Vigorous — 1.00 0.56** 0.57** 0.32** 0.67** 0.49** 0.34** 0.30** 0.44**
0.44**
6. Stretch — 1.00 0.50** 0.34** 0.52** 0.33** 0.17** 0.17** 0.25**
0.25**
7. Strength — 1.00 0.24** 0.57** 0.43** 0.18** 0.26** 0.33**
0.34**

8. Moderate — 1.00 0.30** 0.20** 0.14** 0.28** 0.27** 0.30**

J Typical exercise
O 9. Frequency‡ — 1.00 0.60** 0.31** 0.28** 0.41** 0.40**
U
10. Duration§ — 1.00 0.26** 0.17** 0.36** 0.33**
R
N 11. Length|| — 1.00 0.20** 0.28** 0.28**
A
L CARDIA¶
O
12. Moderate — 1.00 0.64** 0.81**
F
A 13. Vigorous — 1.00 0.97**
M
E 14. Total — 1.00
RI
C
A Note. Each column represents a row variable with the corresponding number in the left-hand column. NCHRBSRi=skNBaethioanvaiol rCSoluler9vgCeeAyH.ReDaIlAth= Coronary Artery Risk Devel- opment in Young
Ad2u3lts.
N †hr/wk
‡d/wk

§min

||mo

¶activity score/yr

*p < .05. *p*< .01.

 EXERCISE & SEDENTARY BEHAVIOR

was negatively associated with exercise and physic ty,

al activi- REFERENCES
specifically, days of vigorous exercise and streng
th train-
1. US Dept Health and Human SePrhvyisciceasl . Activity
ing per week, and vigorous and moderate intensity a and Health: A Reportctivities
of the Surgeon General. Atlanta, GA: US
to those
during the past year. These findings are comparable Dept Health and lts
Human that
Services, Centers for Disease Control a
uter use
in a study of 697 18-to-30-year old Australian adu Prevention, Nationalbe Center
inac- for Chronic Disease Prevention an
Health Promotion: 1996.
mputer
reported that individuals with high levels of comp
women
2. Sallis JF. Epidemiology of physical activity and fitness in
eedeCto
(ie, > 8 hr/wk) were more likely than other adults to children and adolesc nritts.Rev Food Sci Nutr. 1993;33:
oportions
tive1.9 It may be that for college men, using the co in405–408.
acad-
provides reinforcement similar to that received by hysical
3. Stone EJ, McKenzie TL, Welk GJ, Booth M. Effects of
women
from watching television. However, researchers nphysical activity interventions in youth: review anAdmsynthesis.
ege stu-
J Prev Med. 1998;15:298–315.
develop more specific measures to determine the pr use are
discre-
4. Booth FW, Chakravarthy MV. Costs and consequences of
of computer time spent in recreational viewing and
sedentary living: new battleground for an oPlrdeseidnenetm’sy.
of an
emic pursuits. The lack of association between p Council on Physical Fitness and Sports: Research Digest.
edentary
activity and exercise and studying for both men and 2002;3:1–7.
may be explained by its utilitarian nature for coll 5. National College Health Assessment Web Summary.
edentary
dents, whereas exercise, television, and computer haviors. American College Health Association. October 17, 2002.
elevision
<http://www.acha.org/projects_programs/ncha_sampledata_pub
dying for
recreational activities that compete for students’ lic.cfm>
d differ-
tionary time. 6. US Dept Health and Human Services. Healthy People
discre-
Our findings from this study support a concept 2010:Understanding and Improving Health. DHHS Publication
eting
independent and interactive relationship between s No. 017-001-00543-6. 2000. Washington, DC, US Government
studies
was an
behaviors and exercise and physical Printing Office.
2 8 ,2 9
behav-
a c t iwvihtiiecsh, introduces the notion of motion 7. Calfas KJ, Sallis JF, Nichols JF, et al. Project GRAD: Two-
constructive, utilitarian s ons that
year outcomes of a randomized controlled physical activity inte
behaviors versus escapist/mindless sedentary be vention among young
moreadAumltsJ. Health Promot. 2000;
ents and
Recreational sedentary behaviors (eg, watching t 18:28–36. esearch
and videos) and those that are obligatory (eg, stu 8. Grizzell
etect a J, Moses K, Nelson G. Healthy campuses 2010:
al of col-
exams) have different levels of perceived choice an making it happAenm.erican College Health Association. 2002.
ent reinforcement value for college students. d<http://www.csupomona.edu/~jvgrizzell/hc2010/hc2010.ppt.>
choice
Participation in exercise during college students’ ity and
es while
9. Douglas KA, Collins JL, Warren CW, et al. Results from
the 1995
lues of National College Health Risk
BehaviJorAmSurvey.
ate and
tionary time may be positively modified by targCollege Health. 1997;46:55–66.

sedentary activities, as has been demonstrated in K e l l e y G A , L o w i n g L , K e l l e y K . P s y c h o l o g i c a l readiness

10.
28. The
with children in which increased physical activity of Black college student s to be physicalJlyAamctCiovlele.ge
attern,

outcome of strategies to reduce time in sedentary Health. 1998;47:83–87.

roject.
iors2.0 Based on our findings, college health pro 11. Wallace LS, Buckworth J. Application of the transtheoret-
specialists should consider designing interventi pose ical model to exercise behavior among nontraditional college st dentsA. m J
physical activities as more accessible and ommuni-
Health Ed. 2001;32:39–47.
xercise
12. Booth FW, Gordon SE, Carlson CJ, Hamilton MT. Waging
ervices,
rewarding than using the computer for male stud war on modern chronic diseases: primary prevention through ex
Ohio
watching television for female students. Future r cise biologJy.Appl Physiol. 2000;88:774–787.
should include the development of instruments to d Hu FB, Li TY, Colditz GA, Willett WC, Manson JE. Tele-
13.
more extensive range of sedentary activities typic vision watching and other sedentary behaviors in relation to risk

obesity and type 2 diabete s mellitus in JwAMomA.en.

lege students, as Lee and30 Khianvge done for older 2003;289:1785–1791.

adults, that can be evaluated in terms of perceive Andersen RE, Crespo CJ, Bartlett SJ, Cheskin LJ, Pratt M.
14.

and reinforcement value. Decreasing accessibil Relationship of physical activity and television watching with
rewards from major competing sedentary activiti body weight and level of fatness among children: results from t

Third National Health and Nutrition ExaminatJioAnMAS.urvey.

increasing the accessibility and reinforcement va 1998;279:938–942.
exercise and physical activity can have immedi Gortmaker SL, Must A, Sobol AM, Peterson K, Colditz
15.

long-term health benefits for college students. GA, Dietz WH. Television viewing as a cause of increasing obesi-
ty among children in the United States, 1A9r8c6h –P1ed99A0do. -
ACKNOWLEDGMENTS lesc Med. 1996;150:356–362.
This research was funded by NIH-NHLBI grant HL-608 16. Salmon J, Bauman A, Crawford D, Timperio A, Owen N.
among Australian adults Th e associatio
participating n betwee n tel
in varying levels of evision viewing and overweight
leisur
authors thank Sharon
contributions Baldwin,
to this p Camille O’Bryant, Craig M and Sean Walsh for their valuable
time physical activiItnyt. J Obes Relat Metab Disord.
2000;24:600–606.
NOTE
17. Robinson TN. Reducing children’s television viewing to
prevent obesity: a randomized controlledJAMtAri.al.
For comment and further information, please direct c
cations to Janet Buckworth, PhD, FACSM, Sport and E 1999;282:1561–1567.

Science, School of Physical Activity and Educational S 18. Prochaska JJ, Sallis JF, Sarkin JA, Calfas KJ. Examination

349 Larkins Hall, 377 West 17th Avenue, Columbus, of the factor structure of physical activity JbCehlinavEipoir-s.

43210-1284 (e-mail: buckworth.1@osu.edu). demiol. 2000;53:866–874.

VOL 53, JULY/AUGUST 2004 33

 BUCKWORTH & NIGG

19. Fotheringham MJ, Wonnacott RL, Owen N. Compu Longitudina l ter Stud y of Adolescent HJ ePaeldtiha.tr. 1999 ;
use
erils and
and physical inactivity in young adults: public health p potentials of 135:301–306.
new information technAonlon gBieehsa. v Med.
2000;22:269–275. 25. Wallace LS, Buckworth J. Longitudinal shifts in exercise stages of change in
havior:
college stJudSpeonrttss.Med Phys Fitness.
20. Epstein LH, Roemmich JN. Reducing sedentary be 2003;43:209–212.
har-
role in modifying physical actEivxeitryc . Sport Sci Rev. 26. Cohen SJt.atistical Power Analysis for the Behavioral Sci-
applica-
2001;29:103–108. nge.
ences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.
udents:
21. Wallace LS, Buckworth J, Kirby TE, Sherman WM. C 27. Kann L, Kinchen SA, Williams BI, et al. Youth Risk
acteristics of exercise behavior among college students: Behavior Surveillance—United StatesMM19W9R9. . 2000;
Valid- DIA
tion of social cognitive theory to predicting stageProevf cha 49:1–96.
and
Med. 2000;31:494–505. 28. Dietz WH. The role of lifestyle in health: the epidemiolo-
gy and consequence s of inacPtirvoictyN. utr Soc. 1996 ;
escent
22. Buckworth J. Exercise adherence in college st 55:829–840.
issues and preliminary reQsuuesltt.s.2001;53:335–345. 29. Owen N, Leslie E, Salmon J, Fotheringham MJ. Environ-
23. Jacobs DR, Hahn LP, Haskell WL, Pirie P, Sidney mental determinants of physical activity and sedentary behavi
S. ity and reliability of short physical activity history: Exerc Sport Sci Rev. 2000;28:153–158.
CAR the Minnesota Heart Health PrJogCraarmdp.ulm 30. Lee RE, King AC. Discretionary time among older adults:
Rehabil. 1989;9:448–459. how do physical activity promotion interventions affect sedenta
24. Gordon-Larsen P, McMurrayRG,PopkinBM. Adol
physical activity and inactivity vary by ethnicity: the Nand active behavioArnsn?Behav Med. 2003;25:112–119.
 34 JOURNAL OF AMERICAN COLLEGE HEALTH

View publication stats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/11593058

Sleep Habits and Patterns

of College Students: A
Preliminary Study
Article in Journal of American College Health · December
2001
DOI: 10.1080/07448480109596017 · Source: PubMed

CITATIONS
254 READS

42,141

3 authors, including:

Walter C Buboltz
Franklin C
Louisiana Tech
University Brown Yale
70 PUBLICATIONS 1,944 CITATIONS University
23 PUBLICATIONS 877 CITATIONS
All content following this page was uploaded by Franklin C Brown on 26 May 2014.
The user has requested enhancement of the downloaded file.
 View publication stats

See discussions, stats, and author profiles for this publication at:
https://www.researchgate.net/publication/5367394

Sleep patterns in college students: Gender

and grade differences

Article in Journal of Psychosomatic Research · March 2004

Source: PubMed

CITA
TIO
NS
READS
18
8 2,533

2 authors, including:

Ling-Ling Tsai
National Chung Cheng University

18 PUBLICATIONS 584 CITATIONS

All content following this page was uploaded by Ling-Ling Tsai on 14 August 2019.

The user has requested enhancement of the downloaded file.

 Journal of Psychosomatic Research 56 (2004) 231 – 237

Sleep patterns in college students

Gender and grade differences
Ling-Ling Tsai*Sheng-Ping Li
Department of Psychology, National Chung-Cheng University, 160 San-Hsing, Ming-Hsiung, Chia-Yi 621, Taiwan, ROC

Received 16 September 2002; accepted 3 July 2003

Abstract

Objective: Since gender effect is inconsistent and grade effect has not been
addressed in previous studies, we investigated both effects on the daily sleep
patterns in a group of young college students. Methods: The sample
consisted of 237 students aged 18 – 24 years. Each subject completed a 7-
day sleep log. Results: Gender differences were found in several sleep
variables and those were mostly not dependent on weekday/weekend differ-
ence. The female students went to bed and rose earlier and had longer sleep
latency, more awakenings, and poorer sleep quality than the male. Gender
differences were also shown in the relationship between sleep quality and
other sleep variables. The correlation between sleep quality and rise time,
time in bed,
and sleep efficiency was stronger in men than in women. In contrast, grade
differences were mostly dependent on weekday/ weekend difference. The
freshmen rose earlier and had shorter sleep time than did the other students
on weekdays only. Sleep latency was the longest in seniors on weekdays
only. Conclusion: This study showed that gender differences in sleep
patterns and sleep difficulties were remarkable in the group of young college
students. Alarmed by the high prevalence of sleep difficul- ties among
general college students, it is recommended that the students should be
informed of their sleep problems and the consequences.
D 2004 Elsevier Inc. All rights reserved.

Keywords: Gender differences; Sleep disturbances; Undergraduates; University students

Introduction
Previous studies in general adult population have shown
that sleep patterns are both quantitatively and qualitatively
related to the age and gender [1 – 4]. The prevalence of
insomnia is found to be more common among old people
and in women [2,5 – 11]. A recent study showed that the
prevalence of global dissatisfaction with sleep increased
with age and was higher in women [12]. Significant age
differences are also found in both habitual sleep patterns and
sleep disturbances in adolescents. Among both girls and
boys, bedtime [3,13 – 15] is delayed, total sleep time is
decreased [3,13 – 17], and the prevalence of sleep problems
increases with age [16]. Compared to older adolescents,
young adults go to bed even later and sleep even less
[3,13,18]. Disrupted sleep, daytime sleepiness, and dissat-
* Corresponding author. Tel.: +886-5-2720411x32201; fax: +886-5- 2720857.
E-mail address: psyllt@ccu.edu.tw (L.-L. Tsai).
isfaction with sleep are more common in young adults than in
adolescents [19].
In contrast to age difference, few gender differences in
sleep patterns and sleep disturbances have been reported in
adolescents [3,14 – 17]. Although girls go to bed later,
wake up earlier, and thus sleep less than boys, they do
not have sleep problems more frequently than boys [16].
Compared to adolescents, gender differences in young
adults seem to be more frequently reported. Women aged
between 18 and 24 years go earlier to bed and wake up
earlier than men [3] but have equivalent amounts of total
sleep time, number of awakenings, and daytime naps [3,9].
With a wider age range, a study found that women (age
17 to 30 years) were more likely to have nightmares,
delayed sleep onset, and frequent night awakenings [20].
Another study using Pittsburgh Sleep Quality Index (PSQI)
showed that women aged 20 – 29 years have poorer sleep
quality than men [1]. Reyner and Horne [4], using sleep
logs and actimetry, found no significant gender differences
in sleep latency, wake up time, total sleep time, or sleep
quality in the 20 – 34 years group, but they found that women had

0022-3999/04/$ – see front matter D 2004 Elsevier Inc. All rights reserved.
doi:10.1016/S0022-3999(03)00507-5
 L.-L. Tsai, S.-P. Li / Journal of Psychosomatic Research 56 (2004) 231–237
more nocturnal awakenings than men. The major cause of
awakenings in these young women is children. A home-
based electrophysiological sleep recording study in the
group of 20 – 34 years also failed to find significant
gender differences in total sleep time, sleep latency, wake
after sleep onset, and sleep efficiency. However, it did show
that young men had less slow wave sleep than young
women [21]. One epidemiological study found that the
prevalence of insomnia was higher but that of insufficient
sleep was lower in young women aged 20 – 39 years than
in young men [2]. Taken together, although gender
differences in sleep patterns and sleep disturbances are
global and substantial in general adult population, they are
inconsis- tent and, if any, relatively small effect size in
young adults. Furthermore, some sleep parameters,
which are more related to sleep disturbances, e.g., sleep
latency, number of nocturnal awakenings, and sleep
quality, show gender differences starting only in the age of
20s. Some biological maturational processes may
contribute to gender differ- ences in the sleep patterns and
cause sleep disturbances in young adults. However, factors
other than biological ones such as role in family,
response to stress, and life style (eating, drinking,
smoking, exercise, etc.) can also be involved in gender
differences. For example, it has been reported that child
caring and disturbances due to the bed partner are the two
major causes of nocturnal awakenings in young women
[4].
Many college students are single and at their late
adolescent and early adult ages. Their sleep patterns and
sleep disturbances could be different from those of non-
students at their ages [13]. Compared to nonstudent adults in
their 20s, college students have later bedtime and rise
time and show higher incidence of daytime sleepiness in
addition to physical and mental health complaints. Al-
though the student group has the lowest prevalence of
global sleep dissatisfaction of all occupations [12], 71% of
college students express dissatisfaction with their sleep
[22]. Previous studies in college students do not show
consistent gender differences in sleep patterns. One study
showed that gender differences were not associated with
time in bed, number of awakenings, or sleep quality [23].
Another study found no gender difference in sleep latency
[24]. In contrast, contradictory findings in gender differ-
ences have been reported. Both shorter [25] and longer
[26] sleep duration and both poorer [27] and better [26]
sleep quality have been reported in men than in women.
For sleep disturbances, both significantly lower [24] and
higher [27] incidence have been reported in men than in
women. These discrepancies cannot be simply attributed to
cultural differences of student samples since some studies in
the same country also showed inconsistent results, e.g., Refs.
[23 – 25]. Age, however, could be one of the con- founding
factors since gender difference effect sizes in most sleep
parameters are small and even smaller in younger ages.
Most studies included a wide age range of college
students, e.g., Refs. [23 – 26].
23
2
In this study, we intended to determine gender effect on
daily sleep patterns in college students aged 18 – 24 years. If
gender differences occur in such a narrow age range, it
would be concluded that gender did play a significant role
in sleep patterns in early adulthood. Another purpose of
this study is to investigate whether grade (educational
level), related to biological as well as academic age, is
involved in sleep patterns in college students.
Methods
Subjects
The sample consisted of 237 college students enrolled
in the course of ‘‘Sleep Management’’ offered in the
spring semester in 1998 and open to all the students in
the authors’ university. The students were distributed
evenly in three classes at three different time slots
(8:10 – 10:00 a.m., 2:10 – 4:00 p.m., 4:10 – 6:00 p.m.)
on
Wednesday. The students who took this course were
required to record 7-day sleep logs in the first month of
the class as the course assignment. However, they were
solicited to give their permission on the incorpora- tion of
their sleep log data in this study as courtesy and were
informed that no extra credit would be given for their
participation. Of the 362 students enrolled in the class,
314 completed the consent forms. Among them,
237 students were aged 18 – 24. Students from all aca-
demic fields of the university, i.e., arts, social sciences,
science, engineering, law, and business, were included in
the study.
Materials and measures
The daily sleep patterns were recorded in sleep logs. It
has been shown that subjective estimates of sleep time
and sleep latency in the mornings after the overnight
sleep recordings are positively correlated to recorded
sleep time and sleep latency, respectively [28,29]. The
sleep log form was designed for a 7-day recording period
and tended to show the daily sleep pattern of the users.
The form included a line space at the top for filling in
subject name, major, grade, and recording period. This
was followed by the instruction translated as ‘‘Please
record the daily sleep information in 5 minutes after
getting up. Record each nap in 5 minutes after waking up
as well.’’ A grid was placed below the instructions. The
grid included dates (listed as rows in the first column)
and headings for eight self-reported question variables
and four self-calculated variables (listed as columns in
the first row). The question variables included bedtime,
time falling asleep, number of awakenings during the
sleep period, time waking up, rise time, sleep quality
evaluation (from 1 = extremely awful to 10 = ex- tremely
great), naptime, and significant events on the
Table 1
Demographic characteristics of participating college students
Gender Freshman Sophomore Junior Senior
Woman Age, year (range) 18.9 F 0.8 (18 – 21) 20.3 F 1.05 (19 – 21.1 F 0.8 (20 – 24) 22.2 F 0.6 (21 –
23) 23)
Subject number 29 24 31 26
Man Age, year (range) 19.1 F 0.8 (18 – 21) 20.0 F 0.9 (19 – 22) 21.6 F 0.9 (20 – 24) 22.7 F 0.6 (22 –
24)
Subject number 48 19 48 12

previous day. The self-calculated variables included sleep
latency, time asleep, time in bed, and sleep efficiency.
Data analysis
A total of 12 variables were derived from sleep logs.
Bedtime, rise time, number of awakenings, sleep quality,
and naptime were recorded by subjects. Sleep latency (the
time difference between bedtime and time falling asleep),
time asleep (the time difference between time falling asleep
and time waking up), time in bed (the time difference
between bedtime and rise time), sleep efficiency (time asleep
× 100/ time in bed), total sleep (the sum of sleep time at
night and naptime during the day), bedtime regularity
(standard devia- tion of bedtime over the 7-day recording
period), and rise time regularity (standard deviation of rise
time over the 7-day recording period) were then calculated.
For bedtime, rise time, time asleep, time in bed, and total
sleep, we used 2 (genders) × 4 (grades: freshman,
sophomore, junior, and senior) × 2 (days: weekday,
weekend) analysis of variances (ANOVAs) with repeated
measures on day. The data dis- tributions of sleep latency
(skewness g1 = 1.123), number of awakenings (g1 = 2.886),
sleep efficiency (g1 = 1.433), sleep quality (g1 =— 0.893),
naptime (g1 = 1.045),— bedtime standard deviation (g1 =
1.365), and rise time standard deviation (0.797) were
significantly skewed. Since signifi- cantly skewed sample
violates the assumption of normality associated with
parametric statistical methods, nonpara- metric statistical
methods were used and medians/semi- interquartile ranges
were presented for those skewed data.
Gender effect was evaluated by Mann–Whitney test with
chi-square approximation, grade effect by Kruskal–Wallis
test, and day effect by Wilcoxon signed ranks test with
normal (z) approximation.
Based on the cutoff values with significantly increased
odds ratios of global sleep dissatisfaction in [12], we define
sleep difficulties as time in bed less than 7 h or mean
sleep latency longer than 30 min. We also arbitrarily
defined sleep difficulties as the number of awakenings
more than once, sleep efficiency less than 85%, rating on
sleep quality less than 6, or naptime longer than an hour.
Pearson chi- square tests were performed for evaluating the
dependency of the distribution of each sleep difficulty on
gender and on grade, separately.
Pearson correlation was calculated between sleep qual- ity
and other sleep variables for each subject. The Fisher’s r to z
transformation was then performed for each corre- lation
coefficient. One-group t test was used to test whether the
means of the z scores were zero and independent t test
was used for the gender differences. We performed post
hoc comparisons of means by using Tukey’s test for
ANOVAs and Mann–Whitney test for Kruskal–Wallis test.
All statistical analyses were performed using SYSTAT 7.0
for Windows.
Results
The demographic characteristics of the participating
college students are depicted in Table 1. No gender differ-

Table 2
Sleep variables for weekdays and weekends by gender
Woman (n = 110) Man (n = 127)
Sleep variable Weekda Weekend Weekday Weekend Statistical significance
y
Bedtime (hh:mm F min) 1:27 F 1:21 F 69 1:40 F 1:45 F 82 G*
57 55
Rise time (hh:mm F min) 8:27 F 9:12 F 79 8:39 F 9:39 F 89 G*, D***
61 57
Time asleep (min) 381 F 428 F 65 384 F 438 F 75 D***
57 52
Time in bed (min) 420 F 471 F 63 418 F 474 F 69 D***
57 51
Total sleep (min) 419 F 466 F 78 411 F 465 F 87 D***
60 60
Sleep latency (min) 17.5/6.4 15.0/5.7 14.0/5.7 12.0/6.5 G*, D**
Number of awakenings 0.8/0.5 1.0/0.5 0.6/0.4 0.5/0.5 G**
Sleep efficiency (%) 92.1/3.4 93.3/3.8 93.0/2.7 94.4/2.5 D**
Sleep quality 7.2/0.6 7.5/0.7 7.6/0.9 8.0/0.9 G*, D***
Naptime (min) 36.0/22.7 26.4/30.0 22.2/20.7 24.0/23.7 G*
Data are presented as means F S.D. or medians/semi-interquartile ranges. See the text for a detailed description. Subject numbers (n) are given in parentheses. Gender effect is abbreviated as
G and day effect as D.
* Significant difference P < .05.
** Significant difference P < .01.
*** Significant difference P < .001.
Table 3
senior years was less than that of freshmen and juniors.
Correlations between sleep quality and other sleep variables
However, we thought this would have, if any, few effect on
Sleep variable Woman (n = 110) Man (n = 127) gender and grade interaction since interaction effect of
Bedtime .061 F .562 — .029 F .545 gender and grade was not significant on any sleep variable.
Rise time* .042 F .546 .215 F .537yyy
Time asleep .241 F .615yyy .370 F .571yyy
Weekday/weekend differences
Time in bed* .064 F .560 .258 F .586yyy
Sleep latency (min) — .304 F .575yyy .397 F .644yyy As shown in Table 2, the rise time ( F = 84.36, df =
Number of awakenings — .405 F .682yyy .396 F .925yy 1.229, P < .001), but not bedtime, was significantly
Sleep efficiency (%)* .374 F .594yyy .532 F .624yyy earlier on weekdays than on weekends. The students
Naptime (min) — .042 F .498 — .139 F .917 spent signifi- cantly less time in bed ( F = 123.45, df =
Data are presented as means F S.D. The correlation coefficients of each subject were 1.229, P < .001), slept shorter ( F = 102.22, df = 1.229, P <
converted to Fisher’s z scores, and the means and standard deviations of the z scores .001), had less total sleep ( F = 74.40, df = 1.229, P < .001),
were then converted back to correlation coefficients. One-group t test was performed longer sleep latency (z = 3.274, P = .001) and lower sleep
to test whether the means of the z scores were zero and independent t test for gender efficiency (z = 3.237, P = .001), and gave significantly
differences. lower ratings on sleep quality (z = 4.201, P < .001) on
* Significance level for gender differences, P < .05. y weekdays as compared with those on weekends.
Significance level for one-group t test, P < .05. yy
Significance level for one-group t test, P < .01. yyy
Significance level for one-group t test, P < .001.
Gender differences

Table 2 also shows that the women had earlier bedtime

ence was found in the mean age. The mean age consistently
increased with grade ( F = 164.24, df = 3.233, P =.0001).
It was noted that the number of males in sophomore and
( F = 5.35, df = 1.229, P = .022) and rise time ( F =
4.14, df = 1.229, P = .043), longer sleep latency (v2 =
6.00, df = 1, N = 237, P = .014 on weekdays; v2 = 4.84,
df = 1, N = 237, P = .028 on weekends), more awakenings
during night sleep (v2 = 7.12, df = 1, N = 237, P = .008
on week- days; v2 = 12.54, df = 1, N = 237, P < .001 on
weekends), lower ratings on sleep quality (v2 = 3.52, df
= 1, N = 237, P = .061 on weekdays; v2 = 9.92, df = 1, N =
237, P = .002 on weekends), and longer naptime (v2 =
7.95, df = 1, N = 237, P = .005 on weekdays; v2 = 5.48, df =
1, N = 237, P = .019 on weekends) than the men. Gender
differences were not significant in the bedtime regularity
(median/semi-interquar- tile range: 50/16 min in women vs.
52/18 min in men) or the rise time regularity (66/18 in
women vs. 73/22 in men). Bedtime was more regular than
rise time (Wilcoxon signed ranks test, z approximation =
6.13, P < .001).
We define sleep difficulties as time in bed less than
7 h, mean sleep latency longer than 30 min [12], number
of awakenings more than once, sleep efficiency less than
85%, rating on sleep quality less than 6, or naptime
longer than an hour. The percentage of difficulties in
sleep latency was significantly higher in women (11.82%
on weekdays and 15.45% on weekends) than in men
(11.81% on weekdays and 6.30% on weekends) on
weekends. Sleep difficulties in the number of awakenings
were also higher in women (40.91% on weekdays and
37.27% on weekends) than in men (25.20% on weekdays
and 18.11% on weekends). Lastly, more women (21.82%
on weekdays and 23.64% on weekends) than men
(11.02% on weekdays and 21.82% on weekends) took
mean naptime longer than an hour on weekdays. In
contrast, gender differences were not found in the per-
centage of sleep difficulties in time in bed (48.94% of all
the students on weekdays and 19.83% on weekends),
sleep efficiency (12.66% on weekdays and 15.19% on

Fig. 1. Mean and standard deviation of rise time (top), time in bed (middle), and time
asleep (bottom) across grade on weekdays and weekends.
 Fig. 2. Medians and semi-interquartile range of sleep latency (top) and naptime (bottom) across grade on weekdays and weekends.
weekends), or sleep quality (14.77% on weekdays and
10.55% on weekends).
The relationship between sleep quality and other sleep
variables is presented in Table 3. Sleep quality was posi-
tively correlated with time asleep and sleep efficiency but
negatively with sleep latency and number of awakenings in
both men and women. In contrast, sleep quality was
correlated with rise time and time in bed in men but not
in women. Furthermore, the relationships between sleep
quality and rise time, time in bed, and sleep efficiency were
stronger in men than in women.
Grade differences
Interaction effect of grade and day was found on rise time
( F = 3.83, df = 3.229, P = .011), time in bed ( F =
4.19,
df = 3.229, P = .007), and time asleep ( F = 3.50, df =
3.229, P = .016). The freshmen had a shorter sleep time
at night than other students on weekdays, which was not due
to the bedtime difference but an earlier rise time on
weekdays (Fig. 1). Of all the students, the seniors had the
longest sleep latency (v2 = 8.83, df = 3, N = 237, P = .
032 on weekdays; v2 = 5.57, df = 3, N = 237, P = .134
on weekends) and nap- time (v2 = 18.76, df = 3, N =
237, P < .001 on weekdays; v2 = 9.35, df = 3, N = 237,
P= .025 on weekends; Fig. 2).
The percentage of sleep difficulties in time in bed was the
highest in the freshmen on weekdays (v2 = 16.84, df = 3,
N = 237, P < .001). Difficulties in sleep latency were the
highest in the seniors on weekdays (v2 = 9.14, df = 3,
N = 237, P = .027). Grade differences were not found
in the percentage of sleep difficulties in awakenings,
sleep efficiency, sleep quality, or naptime.
Daytime napping and nighttime sleep
To determine whether longer naptime was related to
worse sleep patterns at night [29], we compared the sleep
variables of the students whose naptime was longer than an
hour with those of the rest of the students. Indeed, students
with long naptime had more awakenings at night
(v2 = 11.13, df = 1, N = 237, P = .001 on weekdays;
v2 = 6.99, df = 1, N = 237, P = .008 on weekends) and
poorer sleep quality (v2 = 6.04, df = 1, N = 237, P = .
014 on week- days; v2 = 3.10, df = 1, N = 237, P = .078
on weekends).
Discussion
This study aimed to determine gender and grade effects
on daily sleep patterns in the college students aged 18 –
24 years. Significant gender and grade differences in several
sleep variables were found in this student group. Although
the student subjects were recruited exclusively from those
enrolled in a single course, the gender and grade effect
on sleep might not be specific to this particular group of
students. First, the means of all, except for naptime, sleep
variables in this study were comparable to those of a group
of undergraduates who did not take the course (n = 49,
unpublished data). Second, though the students enrolled in
the sleep management course could be somewhat more
concerned about their sleep, they probably did not have
more sleep problems than other students. The PSQI scores of
the students who took the course were not higher but
even lower than those of the students who did not took
the course (mean global PSQI score, 5.46 F 1.75, n = 279
vs.
6.52 F 2.42, n = 42; unpublished data). Nonetheless, we
could not definitively exclude the possibility that female
students and/or seniors tended to select the course because
of sleep difficulties, which might thereby be related to the
gender and grade effect shown in this study.
As previous studies have shown, the college students
woke up later, had longer time in bed and sleep time, and
reported better sleep quality on weekends [17,23,24]. In
contrast to the results in [24], this study showed that the
bedtime on weekends remained the same as on weekdays.
The discrepancy could be explained by much later bedtime
on weekdays shown in this study than previously shown.
The students went to bed on weekends at the time equiv-
alent to those in [24]. On the other hand, not reported
previously, the students had shorter sleep latency and better
sleep efficiency on weekends. Stress, especially the emo-
tional responses to it, is related to several sleep aspects,
including sleep latency [31]. It is possible that for some
students, less stress on weekends is one of the factors
related to shorter sleep latency on weekends. Furthermore,
based on their relationship to sleep quality (Table 2),
shorter sleep latency, longer sleep time, and higher sleep
efficiency all directed to better sleep quality on weekends.
In contrast to previous studies, which mostly failed to
find gender differences in daily sleep patterns, we did find
those in several sleep variables. Given that female college
students reported a higher level of stress than male students
[32] and, as mentioned above, that stress is related to
sleep [31], stress could be involved in the gender
difference in sleep. On the other hand, since gender
differences were also shown in the relationship between
sleep quality and other sleep variables, sleep perception
and cognition difference could be another factor.
However, future studies are still needed to clarify how
biological and psychological matura- tional processes,
stress, sleep perception, and cognition, as well as other
factors, relate to gender differences in self- reported
sleep patterns in college students.
One special aspect concerning gender differences is that
the poorer sleep patterns and the higher prevalence of sleep
difficulties in the female college students persist even on
weekends. It was further noted that the number of women
with sleep latency difficulty even increased on weekends
compared to weekdays. In contrast, the number of men with
sleep latency difficulty decreased on weekends. It seemed
that weekend was of benefit to men but not to women in
terms of reduction in sleep latency difficulty. We think
this finding is very important and merits further studies.
Week- ends are thought to be more relaxed and thus more
helpful to a good night sleep than weekdays. Why women
could not take the advantage of weekends like men needs
further studies to answer the question. Another thing to be
pointed out is that unexpected gender differences were also
found in the number of awakenings at night in these
young college students. One previous study in general young
adults found that the major causes of more awakenings in
young women were children and bed partners [4]. Of the
237 students
participating in this study, 235 students answered question
10 (‘‘Do you have a bed partner or roommate?’’) in PSQI
administered in class as a course activity. Only 4 of the
235 student reported having a partner in same bed. On the
other hand, married rate was below 16% and fertility rate
was below 7% in women aged 20 – 24 years in Taiwan
(MOI Statistical Information Service
http://www.moi.gov.tw/W3/ stat/, Department of Statistics,
Minister of the Interior, ROC). It is likely that most female
students in this study were single and had no children.
Even if most of them had no children and had no bed
partners, the women still had more night awakenings
than the men. The relationship between naptime and
awakenings suggests longer naptime could be attributed
to more awakenings in the female college students.
Future studies are still needed to clarify whether other
factors also cause the young female college students to
wake up more frequently at night.
To the best of our knowledge, this is the first study to
show an interaction effect between grade and weekday/
weekend on sleep among college students. Grade differ-
ences in sleep patterns were shown mostly on weekdays
only. The freshmen got up earlier and slept less compared to
other students on weekdays but not on weekends. The
seniors having the longest sleep latency were shown only
on weekdays. Over twice more seniors (26.31%) than other
students (9.03%) had sleep latency difficulty on weekdays
but not on weekends. It is likely that grade differences
shown only on weekdays were due to different activity
schedules and academic demands on weekdays in different
grades instead of biological age differences. Although those
grade differences subsided on weekends, the much higher
number of short sleep time in the freshmen (67.53%) and
that of sleep latency difficulty in the seniors on weekdays
still warrant a heightened attention to the grade-related
sleep problems among college students.
In summary, this study found significant gender and
grade differences in several sleep variables among college
students. The gender difference in most sleep variables did
not depend on weekday/weekend difference, but the grade
difference depended on weekday/weekend difference. The
poorer sleep patterns of the female students on both
weekdays and weekends warrant special attention to the
progress of their sleep problems in the long run. The
freshmen need to be especially aware of the consequences of
insufficient sleep and the seniors need to deal with long
sleep latency and naptime. On the other hand, this study
found that the percentage of sleep difficulties in general
college students was amazingly high, e.g., over 48% of
the college students had short sleep time, which suggests
a probably equivalent amount of students had insufficient
sleep on weekdays. Although taking naps during daytime
may be beneficial to make up for some insufficient sleep
at night and to reduce sleepiness in class, it was
limited because naptime over an hour would be related to
more awakenings at night. Since near one fifth of the
young college students had short sleep duration on
weekends,
sleep debt could be chronically accumulated in those
students. In addition to insufficient sleep, other sleep
difficulties are also common among the college students.
These findings warrant sleep education programs and
interventions for the students as previously suggested
[24,30]. It is conceivable that sleep education in college
students may help them be aware of their own sleep
problems and thereby willing to choose activity schedules
and sleep habits good for sleep.
Acknowledgments
The authors would like to thank teaching assistants Yi-
Ling Wang, Chung-Kang Shen, Chung-Shan Kao, Yu-Zhe
Tsai, and Yu-Tien Chang for helping in the administration
and collection of sleep logs. This study was supported by
National Science Council, R.O.C. College Student Grant
NSC88-2815-C194-024H.
References
[1] Doi Y, Minowa M, Uchiyama M, Okawa M. Subjective sleep
quality and sleep problems in the general Japanese adult
population. Psychia- try Clin Neurosci 2001;55:213 – 5.
[2] Liu X, Uchiyama M, Kim K, Okawa M, Shibui K, Kudo Y, Doi
Y, Minowa M, Ogihara R. Sleep loss and daytime sleepiness in
the general adult population of Japan. Psychiatry Res 2000;93:1
– 11.
[3] Park YM, Matsumoto K, Shinkoda H, Nagashima H, Kang MJ,
Seo YJ. Age and gender difference in habitual sleep– wake
rhythm. Psy- chiatry Clin Neurosci 2001;55:201 – 2.
[4] Reyner A, Horne JA. Gender- and age-related differences in
sleep determined by home-recorded sleep logs and actimetry
from 400 adults. Sleep 1995;18:127 – 34.
[5] Ford DE, Kamerow DB. Epidemiologic study of sleep
disturbances and psychiatric disorders. JAMA
1989;262:1479 – 84.
[6] Husby R, Lingjaerde O. Prevalence of reported sleeplessness in
north- ern Norway in relation to sex, age and season. Acta
Psychiatr Scand 1990;81:542 – 7.
[7] Janson C, Gislason T, Backer WD, Plaschke P, Bjornsson E,
Hetta J, Kristbjarnason H, Vermeire P, Boman G. Prevalence of
sleep distur- bances among young adults in three European
countries. Sleep 1995; 18:589 – 97.
[8] Kim K, Uchiyama M, Okawa M, Liu X, Ogihara R. An
epidemio- logical study of insomnia among Japanese general
population. Sleep 2000;23:41 – 7.
[9] Lindberg E, Janson C, Gislason T, Bjornsson E, Hetta J, Boman
G. Sleep disturbances in a young adult population: can gender
differ- ences be explained by differences in psychological
status? Sleep 1997;20:381 – 7.
[10]Ohayon M. Epidemiological study on insomnia in the general
pop- ulation. Sleep 1996;19:S7 –S15.
[11]Pallesen S, Nordhus IH, Nielsen GH, Havik OE, Kvale G,
Johnsen S, Skjotskift S. Prevalence of insomnia in the adult
Norwegian pop- ulation. Sleep 2001;24:771 – 9.
[12]Ohayon MM, Zulley J. Correlates of global sleep dissatisfaction in
the German population. Sleep 2001;24:780 – 7.
[13]Fukuda K, Ishihara K. Age-related changes of sleeping pattern
during adolescence. Psychiatry Clin Neurosci 2001;55:231 –
2.
[14]Gau S-F, Soong W-T. Sleep problems of junior high school
students in Taipei. Sleep 1995;18:667 – 73.
[15]Wolfson AR, Carskadon MA. Sleep schedules and daytime function-
ing in adolescents. Child Dev 1998;69:875 – 87.
[16]Liu X, Uchiyama M, Okawa M, Kurita H. Prevalence and
correlates of self-reported sleep problems among Chinese
adolescents. Sleep 2000;23:27 – 34.
[17]Strauch I, Meier B. Sleep need in adolescents: a longitudinal
ap- proach. Sleep 1988;11:378 – 86.
[18]Ohayon MM, Roberts RE, Zulley J, Smirne S, Priest RG.
Prevalence and patterns of problematic sleep among older
adolescents. J Am Acad Child Adolesc Psychiatry
2000;39:1549 – 56.
[19]Ohayon MM, Roberts RE. Comparability of sleep disorders
diagnoses using DSM-IV and ICSD classifications with
adolescents. Sleep 2001;24:920 – 5.
[20]Coren S. The prevalence of self-reported sleep disturbances in
young adults. Int J Neurosci 1994;79:67 – 73.
[21]Hume KI, Van F, Watson A. A field study of age and gender
differ- ences in habitual adult sleep. J Sleep Res 1998;7:85 –
94.
[22]Hicks RA, Fernandez C, Pellegrini RJ. Striking changes in the
sleep satisfaction of university students over the last two decades.
Percept Mot Skills 2001;93:660.
[23]Hawkins J, Shaw P. Self-reported sleep quality in college students:
a repeated measures approach. Sleep 1992;15:545 – 9.
[24]Buboltz WC, Brown F, Soper B. Sleep habits and patterns of
college students: a preliminary study. J Am Coll Health
2001;50:131 – 5.
[25] Jean-Louis G, von Gizycki H, Zizi F, Nunes J. Mood states
and sleepiness in college students: influences of age, sex, habitual
sleep, and substance use. Percept Mot Skills 1998;87:507 –
12.
[26]Ban DJ, Lee TJ. Sleep duration, subjective sleep disturbances
and associated factors among university students in Korea. J
Korean Med Sci 2001;16:475 – 80.
[27]Liu X, Tang M, Hu L, Wang A. Sleep quality and its
correlates in college students. Chin Ment Health J
1995;9:148 – 50.
[28]Carskadon MA, Dement WC, Mitler MM, Guilleminault C,
Zar- cone VP, Spiegel R. Self-reports versus sleep laboratory
findings in 122 drug-free subjects with complaints of chronic
insomnia. Am J Psychiatry 1976;133:1382 – 8.
[29]Rogers AE, Caruso CC, Aldrich MS. Reliability of sleep diaries
for assessment of sleep/wake patterns. Nurs Res 1993;42:368 –
72.
[30]Brown FC, Buboltz WC. Applying sleep research to university
stu- dents: recommendations for developing a student sleep
education program. J Coll Stud Dev 2002;43:411 – 6.
[31]Verlander LA, Benedict JO, Hanson DP. Stress and sleep
patterns of college students. Percept Mot Skills 1999;88:893
– 8.
[32]Bush HS, Thompson M, van Tubergen N. Personal assessment
of stress factors for college students. J Sch Health 1985;55:370
– 5.
View publication stats