Factors Influencing Efficacy of Nutrition Education Interventions A Systematic Review

Systematic Review
Factors Influencing Efficacy of Nutrition Education

Interventions: A Systematic Review
Mary W. Murimi, PhD, RDN; Michael Kanyi, PhD; Tatenda Mupfudze, PhD;
Md. Ruhul Amin, MPH, MS; Teresia Mbogori, MS; Khalid Aldubayan, PhD
ABSTRACT
Objective: To examine systematically factors that contribute to the efficacy of nutrition education inter-
ventions in promoting behavior change for good health based on their stated objective. In a departure from
previous reviews, the researchers investigated factors that lead to success of various types of interventions.
Critical analysis of these factors constituted the outcome of this review.
Methods: This study followed Preferred Reporting Items for Systematic Reviews and Meta-analysis
criteria. A total of 246 original articles published between 2009 and 2015 in PubMed, Medline, Web of Sci-
ence, Academic Search Complete, Science Direct, Cochrane Reviews, ERIC, and PsychLIT were initially
considered. The number was screened and scaled down to 40 publications for the final analysis. Quality
assessment was based on the Cochrane Handbook for Systematic Reviews of Intervention. Studies were rated
as having low risk of bias, moderate risk, or high risk.
Results: Efficacy of nutrition education interventions depended on major factors: interventions that
lasted $5 months; having #3 focused objectives; appropriate design and use of theories; fidelity in inter-
ventions; and support from policy makers and management for worksite environmental interventions.
Conclusions and Implications: Intervention duration of $5 months, #3 focused objectives, random-
ization, use of theories, and fidelity are factors that enhance success of interventions based on the results of
this study.
Key Words: efficacy, interventions, nutrition education, systematic review (J Nutr Educ Behav. 2017;49:142-
165.)
Accepted September 8, 2016. Published online November 1, 2016.
INTRODUCTION edness of the study objectives, study likely to be effective than those conduct-
design and theory, and fidelity in inter- ed for a short period, <5 months, and do
Nutrition education can be viewed as vention. not offer personalized feedback. Other
any set of learning experiences designed The specific characteristics of the de- studies concluded that expert-led inter-
to facilitate the voluntary adoption of terminants of success of interventions ventions as well as studies that used
eating and other nutrition-related be- are still unclear.2 However, several behavioral theories, social support, and
haviors conducive to health and well- studies have been conducted to ascer- an educational approach to guide die-
being.1 Efficacy describes the ability tain determinants of efficacy of nutri- tary interventions were more likely to
to yield intended outcome; for the effi- tional education interventions. For be successful.4 Despite previous studies
cacy of an intervention to be evalu- example, another systematic review3 on the wider area of nutrition educa-
ated, it must be adequately described.2 concluded that educational intervention, there is still inadequate literature
Efficacy of nutrition education inter- tions that are sustained for a longer on the efficacies of the various nutrition
ventions depends on several factors time, >5 months, and offer personal- education interventions that were im-
including the duration and frequency ized feedback on dietary behavior and plemented in recent years. In a departure
of intervention, the number and relat- related health risk factors, are more from previous reviews that concen-
trated primarily on a single type of
intervention and its related outcome,
the current review investigated several
College of Human Sciences, Texas Tech University, Lubbock, TX factors that led to success of various
Conflict of Interest Disclosure: The authors’ conflict of interest disclosures can be found online types of interventions. The purpose
with this article on www.jneb.org. The first author of this article (M. W. Murimi) served on of this review was to examine system-
the JNEB staff as Associate Editor. Review of this article was handled, exclusively, by the atically the factors that contribute to
Editor-in-Chief to minimize conflict of interest. the efficacy of nutrition education in-
Address for correspondence: Mary W. Murimi, PhD, RDN, Department of Nutritional terventions in promoting behavior change
Sciences, College of Human Sciences, Texas Tech University, PO Box 41240, Lubbock, for good health and well-being based on
TX 79409; Phone: (806) 834-1812; Fax: (806) 742-3042; E-mail: mary.muimi@ttu.edu their stated objective. To achieve this
Ó2016 Society for Nutrition Education and Behavior. Published by Elsevier, Inc. All rights purpose, the researchers used population,
reserved. intervention, comparison, and outcomes
http://dx.doi.org/10.1016/j.jneb.2016.09.003 criteria to frame the research questions.5
142 Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017
Downloaded for Anonymous User (n/a) at University of Indonesia from ClinicalKey.com by Elsevier on November 26, 2019.
For personal use only. No other uses without permission. Copyright ©2019. Elsevier Inc. All rights reserved.
Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017 Murimi et al 143
METHODS Inclusion/Exclusion Criteria scores) (Tables 1 and 2). Fidelity as a

Literature Search factor in this systematic review was
The authors included in the review assessed from authors' declaration of
This systematic review was conducted research articles published in English limitation in their respective studies.
in accordance with recommendations that examined nutrition education in- Reviewers completed a detailed data
and criteria outlined in the Preferred Re- terventions in adults aged >18 years. extraction form. Extracted data were
porting Items for Systematic Reviews Studies were excluded if they were re- transferred to a spreadsheet (Tables 1
and Meta-analysis statement.6,7 Articles view articles, poster abstracts, or quali- and 2).
on studies that conducted nutrition tative, cross-sectional studies, or if the
education interventions on dietary be- target population had special nutri-
haviors were identified by performing tional needs (eating disorders, dia- Analysis Approach
literature searches in: PubMed, Medline, betic, hospitalized, etc). In addition,
Web of Science, Academic Search Comp- studies that failed to achieve any of The primary analytic goal was to deter-
lete, Science Direct, Cochrane Reviews, their objectives were excluded. In the mine the overall effectiveness of nutri-
ERIC and PsychLIT. The search was cases where multiple studies were con- tion education interventions to modify
limited to articles published between ducted on the same data set, only the dietary and exercise behaviors. To deter-
2009 and 2015. Key search words were most recently published study was mine whether an intervention was suc-
nutrition education, nutrition education included. There were 2 reviewers per cessful, the outcome of the study was
interventions, dietary behavior, food, database. Trained reviewers evaluated compared with the stated purpose
and health living. References of all whether articles met inclusion criteria and/or objectives of the study. Once a
retrieved studies were used to determine and determined the quality of the study was classified as having achieved
the source of information, whether they study. All researchers except the lead its intended purpose, the contributing
were primary, secondary, or website researcher went through group training, factors were assessed. Assessed factors
based, and to understand better the conducted by a systematic review and included: (1) the design of the study
basis for conclusions of the studies meta-analysis expert, which also involved including randomness, (2) the type of
that were reviewed. watching a webcast. intervention and activities imple-
All 6 members of the research team mented, (3) the duration and dosage of
were independently involved in re- the interventions, (4) number of objec-
viewing the references. Inclusion and Assessment of Study Quality/ tives in a study, (5) fidelity in interven-
quality measures were determined by Risk of Bias tion implementation, and (6) the use
the 3 senior researchers who conduct- of theory in directing the studies. These
ed an independent evaluation of each In the initial part of work of the current factors were identified through a thor-
article; afterward, several discussions review, researchers worked in pairs in ough review of published nutrition edu-
were held to reach a consensus, hence which data were extracted by 1 cation interventions. They were found
monitoring bias. A total of 246 original reviewer and verified by a second to be common in almost all published
studies published since 2009 and tar- reviewer. The risk of bias in any re- studies. The duration of intervention
geting healthy individuals without ported evidence should be at mini- was categorized as short if it had a cumu-
preexisting medical conditions were mum and evidence that is likely to lative length of >5 months and long if a
reviewed. This initial number was have high risk of bias serves a negli- study lasted for an accumulated period
screened and scaled down to 40 publi- gible purpose and thus should not be of $5 months. This classification of
cations for the final analysis. Screening included in a systematic review even duration was deemed appropriate based
criteria for inclusion and elimination if there is no better evidence.8 In this on the descriptions authors used of the
are illustrated in the Figure. review, determination of the quality respective original studies. The reviewed
of studies was guided by the Grading studies rarely reported the dosage and
of Recommendations Assessment, frequency of interventions. Therefore
Members of the Research Team Development, and Evaluation system it was reasonable to report the total
of rating quality of evidence.9 A thor- amount of time spent in intervention
The research team was composed of 6 ough assessment of the study's fidelity, in months.
members, 3 of whom held doctoral de- perceived conflict of interest regarding Another factor that emerged during
gree; the others had a master's degree outcome owing to sponsorship, study the review was worksite environment
in nutrition. The lead researcher was a design, imprecision, inconsistency, interventions. Worksite environments
full professor of nutrition and a regis- appropriate use of theories, reasonable differ from one site to another. There
tered dietitian. Two other researchers duration of intervention, and whether are various worksite environment in-
were faculty members in recognized in- a study achieved the stated objectives terventions for health living. These
ternational universities with wide expe- formed criteria for quality assessment. include the provision of health mes-
rience in nutrition, education, and Rating scores ranged from 1 to 6. Any sages around cafeterias, the provision
research. Each of the 3 senior researchers discrepancies were discussed until an of healthy food in cafeterias, encour-
paired with 1 junior researcher in each agreement was reached. Based on aging and providing walking space as
database for article search and retrieval. these criteria of assessment of study part of exercise for healthy living,
All 6 members were independently quality, studies were rated as having a and schedules and amounts regarding
involved in reviewing the articles and low risk of bias (5–6 scores), moderate eating, among others. The analysis of
initially screening them. risk (3–4 scores), or high risk (1–2 worksite environment interventions
144 Murimi et al Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017
test–posttest/quasi-experimental design
accounted for 23% (n ¼ 9), and the
non-experimental design constituted
8% (n ¼ 3) of the total articles included
in the analysis. Over half of the total
number of studies (58%; n ¼ 23) were
based on theory.
Interventions in the Reviewed

Studies
This review considered studies con-
ducted in person or face-to-face inter-
ventions at the individual or group
level. A majority of the studies (68%;
n ¼ 27) used a single type of interven-
tion. For example, a study whose
objective was to create education
materials using the target audience's
preferences and to implement a
heart-healthy diet education program
used a single intervention in which 2
registered dietitians led in-home edu-
cation sessions.49 Table 1 lists more ex-
amples of studies that employed a
single intervention. Another 33% of
studies (n ¼ 13) employed multiple in-
terventions including cooking, watching
videos, attending exercise classes, and
gardening. An example is a study
whose objective was to examine the
extent to which participants in a com-
bined physical activity and dietary
intervention achieved changes in mul-
Figure. Flow diagram illustrating the article filtering process as part of the systematic tiple health behaviors.36 Multiple in-
review. GEMs indicates Great Educational Materials. terventions involved the provision of
opportunities for physical activity and
is therefore case specific. Data on work- the current review, the duration of healthy eating before, during, and af-
site environment interventions were intervention, number of objectives, fi- ter church services. Table 2 provides
analyzed and reported along with other delity in intervention, use of theories, more examples on studies that used
initially identified factors. and use of worksite environment multiple interventions.
A semiquantitative approach was were identified as determinants of the
used to summarize findings from nutri- efficacy of nutrition education inter-
tion education interventions. Results ventions. This review provides the re- Effect of Duration of
from nutrition education interventions sults and a discussion of these factors. Intervention
were dichotomized based on whether Slightly over half of the nutrition ed-
they reported a statistically significant ucation interventions (53%; n ¼ 21) For the purposes of this systematic re-
(P < .05) improvement in diet intake, were successful in modifying knowl- view, the duration of the intervention
exercise, or other related risk factors edge, behavior, or physiologic outcomes was categorized as short if it had a cu-
for obesity and diet-related chronic based on their primary objectives. Of mulative length of <5 months and
diseases. The researchers used this the 21 successful studies, 14 used the- long if a study lasted for an accumu-
approach to allow for the diverse range ories and 7 did not. Another 48% of lated period of $5 months. This was
of reported statistics, outcomes, and the interventions (n ¼ 19) met some informed by the way studies reported
measurement units.2 but not all of their primary objectives; their time. Therefore, for convenience
9 of them used theories and 10 did not. in reporting of the results, the terms
short duration and long duration were
RESULTS adopted. The amount of time spent in
Study Designs in the Reviewed intervention was reported in months.
A majority of studies (68%; n ¼ 27) Studies The reviewed studies largely left out
were conducted in the US; the remain- dosage and frequency of intervention.
ing studies (33%; n ¼ 13) were con- Randomized control trials accounted In this review, two thirds (n ¼ 12) of
ducted in other parts of the world. In for 70% of the studies (n ¼ 28), the pre- nutrition education interventions that
Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017
Table 1. Face-to-Face, Individual, Group, and Peer Counseling/Nutrition Education Studies (n ¼ 27)
Risk of
Behavioral Bias
Study Design/ Length and Theory/ Achieved (Quality
Authors Study Population Study Objective Intervention Frequency Construct Objectives Measure) Major Findings
Arrebola 60 patients with grade To evaluate effects of Pre–post. 11 sessions No theory All Low Intervention was associated
et al, 201110 II overweight and lifestyle Intervention: conducted with significant

non-morbid grade modification physical activity every 2 wk for improvements in physical
I–II obesity (age program focused recommendations 6 mo functioning (80.37 18.90
range, 18–50 y) on diet, exercise, and psychological vs 89.40 13.95;
and psychological support. Group P < .001) and role of
support on health- sessions led by physical (20.37 9.10 vs
related quality of life doctor, nurse, or 23.14 6.67; P < .05),
dietitian. vitality (58.71 21.98 vs
No control 70.91 26.56; P < .01),
social functioning
(79.62 27.76 vs
86.57 25.45; P < .03),
and general health
(61.03 19.13 vs
69.42 18.80; P < .001)
factors.
Auld et al, 723 adults To analyze impact of Pre–post. 9 lessons taught SCT and Some Moderate ESBA elicited mean positive
201511 weight Intervention: ESBA for 8–12 wk Adult behavior change for food
management curriculum Learning resource management
intervention on delivered in group Theory (P < .01), food safety
physical activity/ settings. (P < .001), nutrition
exercise and body No control (P < .001), and physical
weight and activity level in participating
composition states (P < .01) except
New York. There was an
increase in dairy, fruit, and
vegetable intake in
Arkansas and California
(P < .05) but not in
Colorado, New York, and
Murimi et al 145
Ohio.
Babatunde 110 African American To assess RCT. 6 sessions (30– Health Belief All Low Overall, an educational
et al, 201112 adults aged effectiveness of Sessions of #15 45 min/wk) for Model program developed with a
50–93 y osteoporosis people; short 6 wk theoretical background
education program presentations/ was associated with an
to improve calcium lectures, hands-on improvement in calcium
(continued)
146 Murimi et al
Table 1. Continued
Risk of
Behavioral Bias
intake, knowledge, activities, and intake (mean increase,
and self-efficacy demonstrations to 556 mg dietary calcium;

help participants P < .001), knowledge
increase self- (P < .001), and self-

efficacy. efficacy (P < .001).
Control: Delayed NE
Backman 327 participants (156 To evaluate Quasi-experimental Intervention was SCT All Low Women in the treatment
et al, 201113 treatment; 171 effectiveness of design with 1-h nutrition and group reported significant
control), 75% of fruit, vegetable, and treatment and physicalactivity changes in 9 measures of
whom were low- physical activity control groups. education attitude, compared with 1

income African toolbox for Control group did classes per measure in the control
American women community not receive NE. week for 6 wk group (P < .05).
aged 18–54 y educators in Compared with those in
changing the control group, women
knowledge, in the treatment group
attitudes, and were also more likely to
behavior among make behavioral changes
women of low- to meet recommendations
income for fruit and vegetable
consumption (P < .001)
and physical activity
(P < .001).
Brennen and 16 African American To evaluate effects of Quasi-experiment. 10 individual No theory All Moderate Both systolic and diastolic
Williams, women aged 25– culturally sensitive Intervention: sessions, blood pressure
201314 63 y lifestyle intervention Counseling and 30 min each, decreased from a mean of
on blood pressure education on and 11 group 151/90 pre-intervention
and weight increasing physical sessions, to 131/76 post-
activity and dietary 60 min each, for intervention. There was a
intake of fruits and 12 wk significant decrease from
vegetables while pre- to post-intervention
decreasing dietary systolic (P ¼ .03) and
intake of salt and diastolic (P ¼ .001) blood
fat. pressures. There was a
No control statistically significant
improvement in self-
efficacy to exercise if
bored (P ¼ .02) or if busy
with other activities
(P ¼ .008). There was a

statistically significant
improvement in self-
perception as valuable/
worthless (P ¼ .02) and
choice as superficial/
profound (P ¼ .04).
Clifford et al, 101 college students To determine whether RCT. Subjects in 4 weekly 15-min SCT Some Moderate There were significant
200915
a series of 4 15- intervention group episodes improvements in knowledge

min, theory-driven viewed 4 15-min of fruit and vegetable
(SCT) cooking cooking programs recommendations in the
programs aimed at over 4 wk. Subjects intervention group
college students in control group compared with the control
living off campus viewed 4 5-min group post¼intervention
improved cooking programs on sleep and at 4-mo follow-up
self-efficacy, disorders. (P < .05). There were no
knowledge, significant changes in fruit
attitudes, and and vegetable motivators,
behaviors barriers, self-efficacy, or
regarding fruit and intake.
vegetable intake
Craigie et al, 75 adults aged $40 y To assess the RCT. Lifestyle 12 wk No theory Some Low 82% successfully
201116 feasibility of a intervention reported maintained or lost weight
lifestyle intervention composed of 3 (mean loss 1.1 kg, and
focusing on diet personalized 2.6 cm waist
and activity, counseling circumference) and 85%
participating in sessions plus reported eating 5 portions
cardiovascular telephone contact. of fruits and vegetables
screening compared with 56% at
baseline. No behavior
changes were detected in
control group.
Davis et al, 46 individuals aged Assessment of peer- Quasi-experiment. Length not No theory All Moderate There was increased intake
200917 >16 y led approach to Intervention: 10 taster specified but of low-fat dairy products
improving diet of sessions and 28 follow up was and reduced fat and salt
South Asians in cookery club conducted after intake. 80% and 75%
Southampton sessions. 1y made positive changes to
No control cooking practices and
eating patterns,
Murimi et al 147
respectively.
Duncan 286 adults enrolled in To conduct pre–post Pre–post design. At least 2 h/wk for Social All Low There was a significant
et al, 201318 English as a feasibility trial of Intervention: Healthy 12 wk Learning increase in fruit, vegetable
Second Language, Healthy Eating for Eating for Life Theories intake, nutrition
aged 18–73 y Life, which curriculum for at knowledge, action
(continued)
148 Murimi et al
Table 1. Continued
Risk of
Behavioral Bias
integrates content least 2 h/wk of planning, and coping
about healthy classroom planning among

nutrition to instruction for participants (P < .05 for all).
decrease cancer 12 wk.

health disparities No control group
Endevelt 127 older adults aged To determine impact Partial RCT. 5 visits for 6 mo No theory All Low DIT group showed
et al, $75 y of intensive DIT: Intervention led significant improvement in
201119 nutritional by dietitian or cognitive function and
intervention medical treatment. depression score
program on health A physician led a compared with the

and nutritional standard care change in the other 2
status of group with an groups. DIT group
malnourished educational showed a significant
community booklet. improvement in intake of
Nonrandomized carbohydrates, protein,
‘‘untreated vitamin B6, and vitamin
nutrition’’ group. B1 and had a significantly
lower cost of physician
visits than did the other 2
groups (P < .05 for all).
Francis 58 women aged To create education RCT. 3 mo Social All Low Intervention and control mini-
and Taylor, 54–83 y materials using Intervention: 2 Marketing nutritional assessment
200949 target audience’s individual Theory scores improved
preferences and registered dietitian– (P < .001). Intervention
implement heart- led in-home subjects consumed more
healthy diet education sessions fiber than did control
education program Control: 2 education subjects (P ¼ .01) and
designed using material mailings reduced sodium intake
needs and (P ¼ .02). Controls
preferences reduced energy (P ¼ .01)
and cholesterol intakes
(P ¼ .03), likely because of
decreased food intake.
Ha and 80 college students To determine whether Pre–post. 3 times/wk for SCT Some Moderate At baseline, total mean grain
Caine-Bish, aged 18–24 y there would be an Intervention: General 50 min for consumption was 3.07oz
201120 increase in whole- nutrition class. semester and mean whole-grain
grain consumption No control consumption was 0.37 oz.
after students After the study, mean
completed an consumption of whole-
interactive grain products significantly
introductory increased to 1.16 oz
nutrition course (P < .001) but total mean
focusing on grain intake remained the
disease prevention same (3.06 oz).
Hsu et al, 25 adults aged >18 y To examine feasibility, RCT. Project CHANGE SDT All Low Both interventions showed
201321 acceptability, and Intervention: Exercise was 8-wk large effect sizes on
preliminary results training and weekly randomized, changes in weekly energy

of exercise behavioral controlled trial expenditure, moderate PA,
intervention with a intervention. with follow-up at and brisk walking. Both
Healthy at Every Control: Exercise only 4 wk interventions showed
Size orientation small effect sizes for all
fitness variables, including
body mass index, waist–
hip ratio, predicted
VO2max, 1RM machine
chest press, and leg press.
Adherence to PA goal was
better for the intervention
group at follow-up. The
Self-Determination
Theory based exercise
intervention with a Healthy
at Every Size resulted in
larger effect sizes for
changes in key
motivational variables,
including self-
determination, autonomy,
and goal-setting,
planning and scheduling
self-efficacy (P not listed).
Ireland, et al, 43 healthy adults To investigate RCT. 8 wk No theory Some Low After 8 wk, urinary sodium
201022 aged 20–75 y whether dietary 2 different education excretion decreased from
education enabled methods using 121 50 to
reduction in salt either Australia’s 106 47 mmol/24 h
consumption National Heart (7.3 3.0 to 6.4 2.8 g
Foundation Tick salt/24 h) in the Tick group
Murimi et al 149
symbol or Food and from 132 44 to
Standards Australia 98 50 mmol/24 h
and New Zealand’s (7.9 2.6 to 6.0 3.0 g
low-salt guideline of salt/24 h) in the Food
120 mg sodium/ Standards Australia New
100 g food. Zealand group (P < .05,
(continued)
150 Murimi et al
Table 1. Continued
Risk of
Behavioral Bias
with no between-group
difference).
Kannan et al, 102 low-income To reduce nutrition Pre–post. 13 lessons taught PEN-3 All Low 77% reported adopting at
201023 African American risks and enhance Intervention: peer-led at 1-wk intervals model least 1 healthy eating
women aged protective nutrition nutrition for 13 wk and TTM behavior (moderating
18–45 y and analyze curriculum. sodium or serving more
changes in self- No control fruits and vegetables to
efficacy their families), 23%
adopted at least 2 such
behaviors, and 45%

adopted both dietary and
biomedical behaviors
(self-monitoring blood
pressure, and exercising).
Kontogianni 126 individuals aged To evaluate impact on Pre–post. 6 bimonthly No theory Some Moderate There was decreased
et al, 45–67 y dietary and activity Intervention: NE with sessions for 1 y consumption of whole-fat
201224 habits of non- dietician. dairy and processed meat
intensive, No control (P ¼ .02 and .02,
community-based respectively), sugar
lifestyle intervention (P ¼ .006), and refined
for type 2 diabetes cereals (P ¼ .05). There
prevention in high- was improved diet,
risk Greek decreased body weight
individuals (P ¼ .04), plasma
triglycerides (P ¼ .02), and
2-h post-load plasma
glucose (P ¼ .05)
compared with those who
had worsened dietary
habits. Total time spent
daily on physical activity
remained unchanged
throughout the
intervention.
Kreausukon 114 full-time To improve fruit and RCT. Not specified but SCT All Low A social-cognitive
et al, undergraduate vegetable Intervention group follow-up at intervention to improve
201225 students (aged 18– consumption received 6 wk fruit and vegetable
psychological consumption was
25 y) at Chiang Mai program that superior to a knowledge-
University, Thailand addressed self- based education session
efficacy and with significantly greater
strategic planning. intention, planning, and
Control group self-efficacy for fruit and
received handouts vegetable consumption
about general (P < .05 for all). Both the
nutrition guidelines intervention and control

groups demonstrated an
increase in fruit and

vegetable consumption
from baseline to 1 wk after
the intervention (P < .001
for both) but the increase
was greater in the
intervention group
(P ¼ .01).
Mendonca 167 adults aged To determine effects Quasi-experimental/ 4 sessions, 60 min No theory All Low There was a reduction in
Rde and 40–65 y of health pre–post each for 7 mo systolic blood pressure
Lopes, interventions on Intervention: Guided (P ¼ .02) and use of
201226 dietary habits and physical exercise, animal fats (P < .01) as
physical nutrition well as an increase in the
measurements intervention, percentage of individuals
nutritional with a normal waist
education groups, circumference and daily
individual nutritional consumption of greens/
care. vegetables and milk/dairy
No control products (P < .01)
Milliron et al, 153 adults aged To promote healthy RCT. 10 min No theory Some Moderate No significant differences
201227 20–65 y eating behavior and Intervention: 10 min between the 2 groups on
weight face-to-face. purchased total,
management Control: Received no saturated, or trans-fat and
education on Eat servings of total
Smart shelf tags vegetables. However, the
posted in store. intervention group
purchased significantly
more servings (per
1,000 kcal) of whole fruit
and dark green/bright
Murimi et al 151
yellow vegetables
compared with the
control group.
(continued)
152 Murimi et al
Table 1. Continued
Risk of
Behavioral Bias
Nakade et al, 226 overweight/ To evaluate effects of RCT. 5 counseling No theory All Low The intervention group lost
201228 obese adults in behavioral Intervention: 30 min sessions for 1 y significantly more weight
Japan, aged approach that individual and follow-up than the control group
40–65 y emphasized counseling and 1 y after (– 5.0 kg vs 0.1 kg for men

tailored behavior 20 min group intervention and –3.9 kg vs –0.2 kg for
counseling, diet, sessions about women). Dietary intake
weight loss, and effective exercise and number of walking
weight provided by steps improved in the
maintenance registered dietitians intervention group. After
and exercise 1-y follow-up, the

instructors. intervention group
Control: No NE maintained significantly
lower weight, lower energy
intake, and improvement
in irregular eating habits
(P < .05 for all variables).
Pimentel et al, 67 Brazilian adults To evaluate RCT. 36 sessions over No theory Some Low The intervention group
201029 aged 50–69 y, with effectiveness of Intervention: 12 mo showed a significant
impaired glucose nutritional individual and decline in body weight
tolerance and at education program group counseling (3.4%), body mass index
least 1 other risk on anthropometric, once and twice per (5.7%), cholesterol
factor for diabetes dietary, and month, respectively intake (49.5%), fasting
mellitus 2 metabolic with team of glycemia (14.0%), fasting
parameters with nutritionists insulin (9.0%),
impaired glucose Control: No NE postprandial glycemia
tolerance (21.0%), postprandial
insulin (71.0%), total
serum cholesterol
23.0%), and glycated
hemoglobin (24.0%). A
decrease in energy intake
(5%; P ¼ .06) and low
density lipoprotein
cholesterol (25%; P ¼ .07)
was observed in the
interventional group,
although it did not reach
statistical significance.
Plawecki and 69 adults aged To enhance physical RCT. 3 sessions Health Belief Some Low Comparison of week 1 and
Chapman- 55–75 y activity and Intervention: lectures (duration not Model and week 8 data indicated
Novakofski, nutritional and hands-on mentioned) for Theory of significant improvement
201330 behaviors active learning 8 wk Reasoned for the treatment but not
Control: intervention Action the control group for
was delayed calcium, and vitamin D
(P < .05). There was
limited response to the

exercise outcome
variables, with many not

participating in that
section of the program.
Ritchie et al, 3,015 and 3,004 To explore impact of Pre–post cross- 3 sessions over TTM All Low After nutrition education,
201031 before and after WIC on family sectional design. 6 mo women and caregivers
NE; pregnant or behavior regarding Caregivers reported increased
postpartum fruits and received education recognition of education
women/caregivers vegetables, whole intervention in a messages, positive
of children enrolled grains, and lower- group (class) or in movement in stage of
in WIC, aged fat milk individual (one-on- change for target food
22–36 y one counseling) items, increased family
format. consumption of fruits and
No control whole grains, and
replacement of whole milk
with lower-fat milk.
Shahnazari 84 US veterans aged To determine RCT. 9 NE sessions for Stage of All Low Multiple coaching contacts
et al, 25–80 y effectiveness of Intervention: 9 6 mo Change decreased intake of
201332 nutrition-based individualized NE Model energy, fat, and
wellness coaching sessions. Coaching carbohydrates by 31%
using multiple consisted of 15-min (P < .001). Weight loss of
contacts and sessions with final 5% from baseline (92.8 to
simple educational 60-min session at 88.2 kg; P < .01) was
tips on health end of 6 mo. Total of observed in the
eating and weight 3.75 h educational intervention group with
management contact for mean body mass index
intervention and 1 h decreasing from 30.4 to
for control group by 28.9 (P < .05). The control
same nutrition group showed a decrease
coach for each in fat intake by 20%
veteran throughout (P < .01) but no
Murimi et al 153
study statistically significant
Control: 1 h changes in intake of other
individualized NE nutrients or body weight
session. (88.7 to 87.4 kg).
Veterans’ readiness to
change eating behavior
(continued)
154 Murimi et al
Table 1. Continued
Risk of
Behavioral Bias
for weight loss improved
with nutrition coaching.

Silva et al, 239 women aged To analyze impact of RCT. 30 sessions, 2 h SDT All Low At 12 mo, the intervention
201033 30–45 y weight Intervention: NE each weekly or group showed increased

management sessions covering: bimonthly for 1 y weight loss (–7.29%,) and
intervention on physical activity, higher levels of physical
theory-based eating/nutrition, activity/exercise
psychosocial body image, and (þ138 26 min/d of
mediators, physical other cognitive and moderate plus vigorous
activity/exercise, behavioral contents exercise; þ2,049 571

and body weight Control: General steps/d) compared with
and composition health education control subjects (P < .001).
Sorensen 56 individuals aged To compare effect of RCT. 8 mo No theory Some Low The NLP therapy group lost
et al, 22–55 y behavior The first step was 12- 1.8 kg and the cooking
201134 modification wk weight loss group lost 0.2 kg during
consisting of either program. the 5 mo of weight
a gourmet cooking Participants maintenance. The
course or NLP achieving at least dropout rate was lower
therapy on weight 8% weight loss during the active cooking
regain were randomized treatment compared with
to 5 mo of either the NLP group. There was
NLP therapy or no difference in weight
course in gourmet maintenance after 2 and
cooking. 3 y of follow-up.
Wieland 34 women (Hispanic, To evaluate socio- Pre–post. 6 wk No theory Some Moderate After the intervention,
et al, Somali, and culturally Intervention. 6-wk participants were more
201235 Cambodian) aged appropriate program with 2 90- likely to exercise regularly
22–68 y physical activity min classes per (P < .001). They reported
and nutrition week. higher health-related
intervention in No control quality of life (P < .001).
community-based Self-efficacy for diet and
participatory exercise; weight loss,
research approach waist circumference, and
blood pressure were not
significantly different after
the intervention.
DIT indicates Dietetic Intensive Treatment; ESBA, Eating Smart Being Active; NE, nutrition education; NLP, Neurolinguistic Programming; PA, Physical Activity; RCT, ran-
domized control trials; SCM, Stage of Change Model; TRA, SCT, Social Cognitive Theory; SDT, Self-determination Theory; WIC, Special Supplemental Nutrition Program
for Women, Infants, and Children.
Table 2. Multicomponent Nutrition Education Intervention Studies (n ¼ 13)
Behavioral Risk of
Study Study Design/ Length and Theory/ Achieved Bias (Quality
Authors Population Study Objective Intervention Frequency Construct Objectives Measure) Major Findings
Baruth and 360 African To examine extent to RCT. 15 mo SEM All Low Up to 19% indicated no
Wilcox, Americans aged which participants Intervention: PA and change in health

201336 $18 y in combined healthy eating behavior, 31%
physical activity before, during and changed 1 health
and dietary after church behavior, 31%
intervention services changed 2 health
achieved changes Control: delayed behaviors, 13%
in multiple health intervention changed 3 health
behaviors behaviors, and 5%
changed all 4 targeted
health behaviors.
Combinations of
multiple behavior
change included PA
and dietary behaviors,
which suggests that
both behaviors can be
changed
simultaneously.
Cullen et al, 1,004 Texas To evaluate modified RCT. 6 classes No theory Some Low There was a significant
200937 EFNEP clients in curriculum for 6- Intervention: 6 short for 6 wk BMI decrease at
100 classes, session Texas videos with goal postintervention
mean age 35 y EFNEP promoting setting, problem compared with
healthful home solving, guided baseline only for the
food environments discussion, and intervention group.
and parenting skills handouts. Then a This change was not
related to obesity weekly goal sheet maintained at
prevention for recording was follow-up.
issued Participants
monitored daily goal
attainment and
return goal sheet
Murimi et al 155
the following week
Control: Traditional
EFNEP class
included brief
discussion and food
preparation
(continued)
156 Murimi et al
Table 2. Continued
Behavioral Risk of
Dirige et al, 673 Filipino- To evaluate 18-mo RCT. 18 mo TTM Some Low Intervention participants
201338 American adults nutrition and Intervention: showed significant
aged $18 physical activity Workshops and increases in PA

intervention (active activities (cooking (P < .05), adoption of a
life) conducted demonstrations, low-fat diet (P < .05),

through culturally recipe contests, and stage of change
specific supermarket tours, for fruits and
organizations group aerobic vegetables (P < .05),
classes, gardening) dietary fat intake
Control: Cancer (P < .01), and PA
screening, (P < .01). Intervention

alternative did not lead to
medicine, and increases in number of
stress management participants eating $5
servings/d of fruits and
vegetables.
French et al, 160 metropolitan To describe and Group RCT (4 18 mo worksite No theory Some Low Energy intake decreased
201039 transit workers report results from garages) intervention significantly and fruit
aged 20–79 y worksite obesity Intervention: and vegetable intake
prevention Enhancement of increased significantly
intervention that PA facilities, in intervention garages
targeted transit increased compared with control
employees availability of and garages. However,
lower prices for BMI and PA changes
healthy vending were not significant.
machine
choices, etc
Control: No
intervention
Iriyama and 57 male workers in To evaluate effects of RCT crossover. 6 mo TTM and All Low Mean BMI was
Murayama, Japan new worksite Intervention: Precede- significantly reduced
201440 weight-control Intervention group Proceed from baseline value of
program using received 6-mo model 25.6 kg/m2 to 25.3 kg/
combination of program consisting m2 at month 6 and to
nutrition education of nutrition 24.8 kg/m2 at year 1
environmental education and (P ¼ .008) and was
interventions provision of healthy significantly lower at
cafeteria meals and year 1 than at baseline
nutritional and month 6 in
information. 1-y multiple comparison
follow-up tests in intervention
Control received group; BMI increased
same intervention over time in control
as counterparts in group (P < .001).
crossover design
from 6 mo after
study entry
Johansen et al, 198 adults aged To present effect of RCT. 6 sessions, TTM Some Low Differences between
201041
25–63 y intervention study Intervention: 2 h each, intervention and control

intentions to Combination of for 7 mo after intervention were
change dietary group sessions, significant for sugar-rich
behavior and individual drinks and rapeseed oil
changes made in counseling, and (P < .05). Intention to
dietary intake organized exercise reduce dietary intake of
groups fat, sugar, and white
Control: General flour, and to change
advice participants type of fat and increase
would receive from intake of vegetables and
general practitioner legumes shifted for
intervention group
(P < .05) from pre-
action (pre-
contemplation,
contemplation, and
preparation) stages to
action stage in
intervention group but
not in control group.
Difference between
groups at follow-up was
significant (P < .05). No
significant differences
were found for intention
to increase fruit intake.
Linde et al, 1,672 participants To influence weight RCT (worksite level). 2 y, frequency No theory None Moderate No differences between
201242 aged 18–75 y gain positively Six worksites in US varied reported sites in key outcome of
among employees metropolitan area with center weight change over 2-
over 2 y were recruited and y study period.
randomized in pairs
Murimi et al 157
at worksite level to
2-y intervention or
no-contact control
Intervention: Posters
at worksites, link to
Web sites with
(continued)
158 Murimi et al
Table 2. Continued
Behavioral Risk of
useful information
on food availability
and price, PA
promotion, scale
access, and media

enhancements
Control: No contact
Racette et al, 123 participants To evaluate Cohort randomized Weekly TTM Some Low Improvements (P < .05)
200943 aged 36–54 y effectiveness of trials. Intervention for 1 y were observed at both
with BMI worksite health included worksites for fitness,
32.9 8.8 kg/ promotion program pedometers, blood pressure, and

m2, employed at on improving healthy snack cart, total, high-density
1 of 2 selected cardiovascular Weight Watchers lipoprotein, and LDL-C.
worksites within disease risk factors meetings, group Additional
a large medical exercise classes, improvements
center seminars, team occurred in intervention
competitions, and group in BMI, fat mass,
participation Framingham risk score,
rewards and prevalence of
Control: Personal metabolic syndrome;
health reports only changes in BMI
containing and fat mass were
assessment results different between
control and intervention
worksites.
Rustad and 118 ethnically To assess impact of Pre–post intervention. 3 sessions No theory All Moderate Postintervention
Smith, diverse, low- short-term nutrition Experiential and in 6 wk. increased nutrition
201344 income women intervention using interactive lectures, Each class knowledge and
aged 23–45 y education on activities, and lasted favorable nutrition
comprehensive demonstrations. 75–90 min behavior. Responses
array of nutrition Educational to 7 of 11 questions in
and health topics sessions on knowledge and 9 of 11
shopping and variable set changed
budgeting, healthy significantly at
cooking, improving postintervention
food security by (P < .01 for both).
growing foods. Women also
No control decreased
consumption of fast
foods and processed
snacks high in sugar,
salt, and fat, and fatty
cuts of meat, and
decreased addition of
sugar, salt, and butter
to foods (P < .01).
Sarrafzadegan 12,514 To evaluate RCT. Duration of Precede– Some Moderate Prevalence of abdominal
et al, 201345
participated at component of Intervention: Public intervention Proceed obesity, hypertension,

baseline and Healthy Heart education through activities model, Social hypercholesterolemia,
9,570 in post- Programs to mass media, varied: 3–4 y. Learning hypertriglyceridemia,
intervention assess feasibility intersectoral Theory, and and high LDL-C
survey, aged and outcomes of cooperation and innovation decreased significantly
24–54 y program on lifestyle collaboration, diffusion in intervention area vs
behaviors and risk professional approach control area in both
factors for chronic education and sexes. However,
non-communicable involvement, reduction in
diseases marketing, overweight and
organizational obesity was significant
development, only in females
legislation and (P < .05 for all). There
policy development, were no significant
as well as research changes in prevalence
and evaluation of diabetes mellitus.
Savoie et al, 203 participants To determine whether Retrospective post- Not indicated TPB All Low Mean responses of
201546 aged $18 y participation in then-pre design individual questions
selected Intervention: lecture, and mean lesson
Supplemental cooking scores increased
Nutrition Assistance demonstration, significantly from
Program–Education sample tasting of pretest to posttest in
lessons had an food prepared in the menu planning,
impact on intent to class, handout. shopping lesson, and
improve nutrition- No control the My Plate lesson
related behaviors of (P < .001).
participants
Wilcox et al, 74 African To report results of RCT. 15 mo SEM Some Low There was a significant
201347 Methodist intervention Intervention: Churches effect favoring
Episcopal targeting physical implemented; intervention group in
churches and activity and healthy sharing messages self-reported leisure
Murimi et al 159
1,257 adult eating from pulpit; passing time (d ¼ 0.18;
members within out educational P ¼ .02). No group
them materials (provided), differences were found
create Faith, for self-reported fruit
Activity, and and vegetable
(continued)
160 Murimi et al
Table 2. Continued
Behavioral Risk of
Nutrition Program consumption,
Control: delayed measured blood
intervention pressure, and self-

components at end reported fat- and fiber-
of 15 mo related behaviors.
Lowe et al, 96 adult To evaluate nutritional RCT. 3 mo No theory Some Low There was no difference
201048 participants and weight 3 mo of baseline data between groups in total
(BMI changes in collection, then 3- energy intake over
29.7 6.0 kg/ program that used mo intervention; 6- study period.
m2) hospital or worksite cafeterias and 12-mo Across groups, energy
university to reduce postintervention and percentage of

employees aged employees’ calorie follow-ups. energy from fat
21–65y content of Participants were decreased and percent
purchased foods randomly assigned of energy from
and improve their to 1 of 2 carbohydrates
macronutrient intervention increased from baseline
intake groups: to intervention period (all
Environmental P < .01). Follow-up
Change or analyses, conducted by
Environmental averaging baseline
Change Plus months 1 and 2 and
Energy Density comparing them with
Education and intervention month 3 as
Incentives. a conservative estimate
Randomization of of overall impact of
participants intervention, indicated
occurred within that change in energy,
each worksite carbohydrate, and fat
intake remained
significant (P < .001).
Providing nutrition labels
and reducing energy-
density of selected foods
was associated with
improved dietary intake.
BMI indicates body mass index; EFNEP, Expanded Food and Nutrition Education Program; LDL-C, low-density lipoprotein cholesterol; PA, physical activity; RCT, random-
ized control trials; SEM, Structural Ecological Model; TPB, Theory of Planned Behavior; TTM, Trans-theoretical Model.
lasted longer than 5 months (n ¼ 18) waist circumference, and blood pres- initial phase of the study, to enhance
met their primary objectives.17,19,26,28, sure were not significantly different reliability of the results of follow-ups.
29,31,32,36,40,43,50
For example, a study from baseline after 2-hour classes
whose objective was to evaluate the each week for 6 weeks.
effectiveness of a nutritional education Lack of Fidelity in Delivery
program that involved bimonthly
Fidelity in intervention ensures that
group discussions for 12 months, and Effect of Number of Study all intervention activities are executed
that included written and oral didactic Objectives/Focus as planned in the methods. There
instructions on anthropometric, dietary,
were few reported cases of lack of
and metabolic parameters with impa- In addition to the long duration, this
intervention fidelity that could have
ired glucose tolerance, reported a systematic review found that studies
compromised the findings. A peer-
decrease in 2 risk factors related to with few or focused objectives were
led nutrition education intervention
diabetes mellitus.29 more successful in meeting all of their
that addressed maternal and infant
Similarly, a study that took 1 year stated objectives than were interven-
health through dietary patterns re-
and whose objective was to study the tions that had several unrelated objec-
ported that some facilitators neglected
impact of a weight management inter- tives. For example, a 4-month nutrition
to follow the complete lesson plans by
vention on theory-based psychosocial education intervention whose objec-
omitting parts of a lesson or failing to
mediators reported weight loss (–7.29%,) tive was to increase whole-grain con-
use the facilitator's guide, or they did
and increased levels of physical activ- sumption among students who completed
not show enthusiasm in promoting
ity/exercise (þ138 26 min/d of mod- an interactive introductory nutrition
the desired behavior among the peer-
erate plus vigorous exercise; þ2,049 571 course focusing on disease preven-
led cohorts. Although the interven-
steps/d) compared with control sub- tion20 indicated that student partici-
tion was the same by design and
jects (P < .001).33 Finally, a study that pants increased whole-grain consumption
content information, results of the
involved a physical activity and healthy from 0.37 to 1.16 oz (P < .001). Simi-
study varied among cohorts. As a
eating intervention over 15 months larly, another study22 with only 1
result, the success of the interventions
reported significant results and an 18% objective, to investigate whether die-
was affected by the human factors of
effect on the intervention group in tary education enabled a reduction in
the presenter.23
self-reported leisure time (P ¼ .02; salt consumption, indicated that after
d ¼ 0.18).47 8 weeks of dietary education interven-
On the contrary, this review found tion there was a reduction in salt con- Theory-Based Studies
that nutrition education interventions sumption and urinary sodium excretion.
that lasted for a short duration were In contrast, interventions with >3 Slightly over half of the studies
less likely to meet their stated objec- unrelated objectives were not success- (57.5%; n ¼ 23) reported being theory
tives. A relevant example here is a ful in meeting all of their objectives. based and used at least 1 theory. The
study whose purpose was to promote For instance, a study that had >3 ob- most common theories used to design
healthy eating behavior and weight jectives, conducted in 3 phases with and implement nutrition education
management.27 In this study, inter- different feeding regimes, reported interventions in studies selected in
vention involved 10 minutes of face- inconsistent results at the 3 phases.34 this review were the Trans-theoretical
to-face education on the printed In phase 1, 88% of participants (n ¼ Model and Social Cognitive Theory.
EatSmart shelf tags posted in the store. 49) completed 12 weeks of calorie re- The majority of the theory-based
The control group received no educa- striction and achieved 8% weight studies (61%; n ¼ 14 of 23) were
tion about the EatSmart shelf tags loss. Phase 2 involved 5 months of successful in achieving their stated ob-
posted in the store. Outcome measures weight maintenance; participants jectives, whereas the remaining
included purchases of total saturated were divided into 2 groups each with theory-based studies (39%; n ¼ 9)
and trans fat (grams per 1,000 kcal), different feeding regimes and hence achieved some but not all of their pri-
fruit, vegetables, and dark green or yel- different objectives. The 2 groups mary objectives.
low vegetables (servings per 1,000 kcal) experienced different numbers of This review considered studies that
derived through a nutritional analysis participant dropout, which affected provided information about how they
of participants' shopping baskets. Re- the final results. There was no differ- used theories in the design of the
sults showed no significant differences ence in weight maintenance after 2 study as a best practice, rather than
between the control and intervention and 3 years of follow-up. In general, just mentioning the theory casually
groups on total purchases as well as it was observed that follow-up studies in the introduction or methods: for
in the choice of health eating grocery did not yield many results. A lot of instance, a study conducted by Savoie
products except in fruit and dark green follow-up studies did not yield a et al46 to determine whether partici-
or yellow vegetables, thus indicating significant change from the initial re- pation in a selected Supplemental
minimal attainment of the study ob- sults. Therefore, it can be implied Nutrition Assistance Program–Education
jectives. that a majority of researchers could (SNAP-Ed) that clearly showed how
Similarly, a nutrition education have paid less attention to the the constructs of the Theory of
intervention on physical activity and follow-ups. The researchers in this re- Planned Behavior used in the design
nutrition35 revealed that self-efficacy view recommend stringent measures and implementation of the study les-
for diet and exercise, weight loss, in follow-up studies, just as in the sons showed an impact on the intent
to improve nutrition-related behav- environmental interventions did not featured in this discussion. Critical
iors of participants as stated in the ensure changes sufficient for the inter- analysis of these factors provides the
study objective. Results of this study vention, which affected the results. outcome of the current review.
showed that posttest scores were Such a failure occurred when the The randomized control trial (RCT)
significantly higher than pretest scores worksite management and collabora- design has a reputation of being robust.
related to menu planning, shopping les- tors resisted making sufficient envi- This design is therefore appropriate for
sons, and My Plate lessons (P < .001). ronmental changes to modify dietary baseline studies intended to inform a
However, this review observed that and exercise behavior in employees larger intervention project.51 This re-
although some studies indicated that at work despite promoting the view indicated that a majority (70%;
they were theory based, they failed to behavior.39,42 For example, a worksite n ¼ 28) of the nutrition education in-
describe explicitly how the theories guided environmental intervention study terventions used an RCT design. The
the studies. For example, some re- used posters at worksites, links to RCT design is robust and may be attrib-
searchers15 reported using Social Cogni- Web sites with useful information uted to the success of the interventions
tive Theory; others45 reported using about food availability and prices, in achieving their stated objectives.
several theories including the Precede– physical activity promotion, access Interventions that lasted for
Proceed Model, Social Learning Theory, to scales, and media enhancements >5 months reported a higher level of
and the Innovation Diffusion Approach. to promote weight gain control.42 success. They were mainly multiple-
However, the theory constructs and Intervention components were food component interventions. This finding
how they were used or measured were selection, promotion of walking and supported the results of a previous re-
not described in either article. stair use, weight self-monitoring, and view that reported that remarkably
Nevertheless, although 45% of health information at work. However, more studies on nutrition education
studies (n ¼ 18) were not informed the provision of a variety of healthy with long-term follow-up were associ-
by a theory, they were equally as suc- options and extra time for exercise ated with success.51 Another study
cessful as those that were. For instance, were not offered to employees. The re- noted that behavior change takes time
a study to describe and report the results of this study indicated that there and practice.52 Therefore, it may be
sults from a worksite obesity preven- were no differences between the inter- argued that the length of time taken
tion intervention that targeted transit vention and control sites in the key for intervention and the frequency of
employees was not guided by a theory, outcome of weight change over the exposure are important factors for the
and yet it indicated success.39 The re- 2-year study. success of a nutrition education inter-
sults indicated that energy intake vention. However, another study
decreased significantly and fruit and noted that interventions with long du-
vegetable intake increased significantly DISCUSSION rations are associated with a higher cost
in intervention garages compared of implementation and participants'
with control garages. A summary of The purpose of this review was to sys- attrition, which constrained some in-
the studies that used theories and tematically examine factors that terventions.24
those that did not is provided in contribute to the efficacy of nutrition This systematic review revealed
Tables 1 and 2. education interventions in promoting that studies with few and succinct ob-
behavior change for good health and jectives were more successful than
well-being based on their stated objec- were those with numerous and at
Environmental Interventions at tive. The main findings of this review times unrelated objectives. Other re-
the Worksite indicated that the efficacy of nutrition views acknowledged the effectiveness
education interventions depends on the of few objectives in nutrition educa-
Worksite environmental interventions duration of the intervention, having few tion interventions.51 The current re-
have an integral part in modifying die- focused objectives, the appropriate view noted that studies with #3
tary and weight management behav- use of theories, fidelity in interventions, objectives that were related were suc-
iors when executed appropriately. For and support from policy makers and cessful even when the duration of
example, a study to evaluate the effects management for the environmental intervention was <6 months.
of a new worksite weight control pro- interventions. These findings are largely It is important to report fidelity in
gram using nutrition education envi- congruent with the results of another interventions because it allows readers
ronmental interventions among male review conducted by Baird et al.4 and other researchers to judge the
adults in Japan reported significant re- Therefore, factors that were identi- quality of the intervention and how
sults. At the 1-year follow-up, the inter- fied as determinants of efficacy and various factors may have influenced
vention group had significantly greater which form the discussion of this the outcome.53 Lack of fidelity in the
reductions in body weight, body mass study are: (1) the types and use of de- delivery of a program has a counter-
index, and alanine aminotransferase signs, (2) the type of intervention that productive effect on the results of an
than the control group did (P ¼ .02, characterized the studies, (3) the dura- intervention. Fidelity in intervention
.02, and .86, respectively).40 tion and dosage of the interventions, is a critical element that is rarely re-
For worksite environmental inter- (4) the number of objectives in a ported in many studies.54 This review
ventions to be successful, sufficient study, (5) fidelity in intervention, found that the few studies that re-
appropriate changes must be imple- and (6) the use of theories in nutrition ported fidelity discovered that it nega-
mented at the right places. This re- education interventions. Worksite tively affected the results: Some sites
view discovered that some worksite environmental interventions are also achieved their objectives whereas
others failed to achieve their objec- mean body mass index from a base- by a theory have a higher chance of
tives despite a similar program. line value of 25.6 kg/m2 to 25.3 and achieving their purpose. Lack of fidel-
However, whereas peer educators 24.8 kg/m2 at 6 months and 1 year, ity in peer-led interventions, lack of
might have experienced some profes- respectively (P < .05). These worksite behavior support for environmental
sional challenges in the delivery of environmental interventions high- interventions, and too-short dura-
educational interventions, other studies light the need for effective collabora- tions were factors that contributed to
noted that they could provide a good tion among nutritionists, policy makers a lack of success in some of the inter-
form of social support associated with and stakeholders within institutions ventions. It was observed that various
successful behavior change.4 It is there- including schools and workplaces human factors affected the effective-
fore important for the training of peer- where the food environment has a ness of an intervention in cases where
led interventions to emphasize fidelity big role in the success of a nutrition peers and paraprofessionals imple-
in the implementation in an effort to education intervention. These find- mented the nutrition interventions.
achieve desired results. ings are in agreement with results of Although the lack of fidelity during
This systematic review revealed that another study regarding the effects interventions was not rampant, the
although the majority of the studies of environmental, policy, and social few reported incidences had profound
that were theory based were successful marketing interventions on physical ramifications on the results. The studies
in achieving all of their primary objec- activity and fat intake of middle school with control groups had better inter-
tives (53.8%), a good number (45%) of students. The study noted that envi- pretation of the results, which enhanced
well-designed, non-theory interventions ronmental and policy interventions the validity of the outcome. This im-
were equally successful in achieving were effective in increasing physical plies that studies that used RCTs had
their primary objectives. The current activity at school. Appropriate changes a better chance of replicability, followed
findings support a review by Baba- are necessary for the success of work- by those that employed a quasi-
tunde et al,12 who noted that overall, site environmental interventions.48,56 experimental design. The eligible studies
an educational program developed with This review has limitations. First, that were reviewed and whose inter-
a theoretical background was associ- only articles that were published in En- ventions were considered successful
ated with an improvement in calcium glish were considered. Therefore, there were considered largely sustainable
intake (mean increase, 556 mg dietary is a possibility that some recent and and could easily be replicated.
calcium; P < .001), knowledge (P < .001), important findings published in lan- The researchers concluded that the
and self-efficacy (P < .001). This indi- guages other than English were left use of theories in designing nutri-
cates that well-used theory is likely to out. Another limitation is that there tional education interventions was a
make an intervention successful. This was a potential that studies that did common practice in which 55% of
review observed that some of the not find a significant effect in their inter- the analyzed studies (n ¼ 22) were
studies that claimed to use a theory vention were not published, and there- informed by at least 1 theory. Many
failed to describe explicitly how the fore were not included in this review. studies that used theories indicated
theory was used in the study. This Finally, the review was limited by success in achieving their objectives.
finding agrees with the results of a re- articles that did not include adequate However, a lack of details regarding
view on experimentally based evi- information in their methods and re- how the behavior theories guided
dence of the theoretical mechanisms sults. This posed a challenge to eval- the studies made it difficult to assess
of dietary behavior change,55 which uate the contributions of specific the effect of the theories mentioned
concluded that future intervention tri- components of nutrition education in some studies. The researchers
als need to focus on identifying effec- interventions and their effectiveness concluded that the use of theories is
tive procedures for mediator change properly, including the use of theory a good practice in interventions and
and adopting a more rigorous and sys- and the dosage of intervention (fre- that worksite environmental inter-
tematic approach to theory testing. quency and duration). Despite the ventions provide an important oppor-
Finally, in worksite environmental limitations, the current review drew tunity for behavioral adjustments for
interventions, interventions without its strength from the fact that the re- better health, but that they need the
appropriate support from collabora- searchers investigated several factors cooperation of the policy makers.
tors to support the desired behavior that led to the success of various types The results of this study suggest that
were less likely to meet their objec- of interventions. This was a departure more focused, clearly defined, measur-
tives. For instance, an intervention from previous reviews that concen- able objectives are associated with
did not meet the primary objective, trated primarily on a single type of behavior change, whereas the more
which was to affect weight gain posi- intervention and the related outcome. ambitious use of many objectives
tively over 2 years, owing to weak may limit the effectiveness of nutrition
and inconsistent implementation of education by taking away from the
environmental changes in the work- IMPLICATIONS FOR main message and confusing partici-
place.42 In contrast, an intervention RESEARCH AND pants. The objective should have a
by Iriyama and Murayama40 on work- PRACTICE clear targeted behavior, followed by
site weight control was able to imple- adequate dosage or exposure to facili-
ment changes in the workplace The results of this review suggest that tate the desired behavior change. A
cafeteria and introduce healthy me- nutrition education interventions with purposeful selection of behavior the-
nus. The results of this intervention longer duration have few and focused ory that will guide the intervention
indicated a significant decrease in objectives, and that those that are guided based on the desired behavior change.
A clear use of the theory in designing ventions: explanation and elaboration. J 18. Duncan LR, Martinez JL, Rivers SE,
and implementing the intervention Clin Epidemiol. 2009;62:1-34. et al. Healthy eating for life English as
should be reflected in the results. For 7. Moher D, Tetzlaff J, Tricco AC, a second language curriculum: primary
studies that use multiple sites with Sampson M, Altman DG. Epidemiology outcomes from a nutrition education
different implementers, ensuring fidel- and reporting characteristics of system- intervention targeting cancer risk reduc-
ity is critical to the success of the inter- atic reviews. PLoS Med. 2007;4:78. tion. J Health Psychol. 2013;18:950-961.
vention. Training should emphasize 8. Higgins JPT, Altman DG, Sterne JAC. 19. Endevelt R, Lemberger J, Bregman J,
the important message and targeted Assessing risk of bias in included studies. et al. Intensive dietary intervention by
behavior. Researchers should also In: Higgins JPT, Green S, eds. Cochrane a dietitian as a case manager among
consider the Guide for Effective Nutri- Handbook for Systematic Reviews of Inter- community dwelling older adults: the
tion Interventions and Education as a ventions. Version 5.1.0. http://www. EDIT study. J Nutr Health Aging.
checklist to help design or improve cochrane-handbook.org. Accessed 2011;15:624-630.
their research methods for effective in- October 1, 2016. 20. Ha EJ, Caine-Bish N. Interactive intro-
terventions.57 9. GRADE Working Group. Grading ductory nutrition course focusing on
For worksite and other environ- quality of evidence and strength of rec- disease prevention increased whole-
mental interventions, it is important ommendations. BMJ. 2004;328:1490. grain consumption by college students.
for policy makers to make healthy 10. Arrebola E, Gomez-Candela C, J Nutr Educ Behav. 2011;43:263-267.
choices the easy ones by allowing Fernandez-Fernandez C, Loria V, 21. Hsu YT, Buckworth J, Focht BC,
time for the exercise. This can be im- Munoz-Perez E, Bermejo LM. Evalua- O’Connell AA. Feasibility of a self-
plemented in many ways depending tion of a lifestyle modification program determination theory-based exercise
on the worksite environment: for for treatment of overweight and non- intervention promoting healthy at
example, including markings indi- morbid obesity in primary healthcare every size with sedentary overweight
cating the distance covered along a and its influence on health-related quality women: Project CHANGE. Psychol
path between 2 points, such as build- of life. Nutr Clin Pract. 2011;26:316-321. Sport Exerc. 2013;14:283-292.
ings, or by providing healthy alterna- 11. Auld GS, Baker L, Conway J, 22. Ireland DM, Clifton PM, Keogh JB.
tives or choices at the worksite. Dollahite MC, Lambea MC, McGirr K. Achieving the salt intake target of 6 g/
Outcome effectiveness of the widely day in the current food supply in free-
adopted EFNEP curriculum Eating living adults using two dietary educa-
REFERENCES Smart-Being Active. J Nutr Educ Behav. tion strategies. J Am Diet Assoc. 2010;
2015;47:19-27. 110:763-767.
1. Contento IR, Balch GI, Bronner YL, 12. Babatunde OT, Himburg SP, 23. Kannan S, Sparks AV, Webster JD,
et al. The effectiveness of nutrition edu- Newman FL, Campa A, Dixon Z. The- Krishnakumar A, Lumeng J. Healthy
cation and implications for nutrition ory-driven intervention improves cal- Eating and Harambee: curriculum
education policy, programs, and cium intake, osteoporosis knowledge, development for a culturally-centered
research: a review of research. J Nutr and self-efficacy in community- bio-medically oriented nutrition educa-
Educ. 1995;27:312-328. dwelling older Black adults. J Nutr tion program to reach African Amer-
2. Ammerman AS, Lindquist CH, Educ Behav. 2011;43:434-440. ican women of childbearing age.
Lohr KN, Hersey J. The efficacy of 13. Backman D, Scruggs AA, Atiedu S, et al. Matern Child Health J. 2010;14:535-547.
behavioral interventions to modify die- Using a toolbox of tailored educational 24. Kontogianni MD, Liatis S,
tary fat and fruit and vegetable intake: a lessons to improve fruit, vegetable, and Grammatikou S, Perrea D,
review of the evidence. Prev Med. 2002; physical activity behaviors among Afri- Katsilambros N, Makrilakis K.
35:25-41. can American women in California. J Changes in dietary habits and their asso-
3. Pomerleau J, Lock K, Knai C, McKee M. Nutr Educ Behav. 2011;43:S75-S85. ciation with metabolic markers after a
Interventions designed to increase adult 14. Brennen M, Williams CL. Lifestyle non-intensive, community-based life-
fruit and vegetable intake can be effec- management of cardiovascular risk fac- style intervention to prevent type 2 dia-
tive: a systematic review of the literature. tors in African American women. betes, in Greece. The DEPLAN study.
J Nutr. 2005;135:2486-2495. ABNF J. 2013;24:92-97. Diabetes Res Clin Pract. 2012;95:
4. Baird J, Cooper C, Margetts BM, 15. Clifford D, Anderson J, Auld G, 207-214.
Barker M, Inskip HM. Changing health Champ J. Good Grubbin’: impact of a 25. Kreausukon PP, Gellert S, Lippke S,
behaviour of young women from TV cooking show for college students Schwarzer R. Planning and self-
disadvantaged backgrounds: evidence living off campus. J Nutr Educ Behav. efficacy can increase fruit and vegetable
from systematic reviews. Proc Nutr 2009;41:194-200. consumption: a randomized controlled
Soc. 2009;68:195-204. 16. Craigie AM, Barton KL, Macleod M, trial. J Behav Med. 2012;35:443-451.
5. Guyatt G, Drummond R, Meade M, et al. A feasibility study of a personal- 26. Mendonca Rde D, Lopes AC. The ef-
Cook D. The Evidence Based-Medicine ised lifestyle programme (HealthForce) fects of health interventions on dietary
Working Group Users’ Guides to the Med- for individuals who have participated in habits and physical measurements. Rev
ical Literature. 2nd ed. Chicago, IL: cardiovascular risk screening. Prev Med. Esc Enferm USP. 2012;46:573-579.
McGraw Hill; 2008. 2011;52:387-389. 27. Milliron BJ, Woolf K, Appelhans BM.
6. Liberati A, Altman DG, Tetzlaff J, et al. 17. Davies JA, Damani P, Margetts BM. A point-of-purchase intervention
The PRISMA statement for reporting Intervening to change the diets of featuring in-person supermarket educa-
systematic reviews and meta-analyses low-income women. Proc Nutr Soc. tion affects healthful food purchases. J
of studies that evaluate health care inter- 2009;68:210-215. Nutr Educ Behav. 2012;44:225-232.
28. Nakade M, Aiba N, Suda N, et al. 38. Dirige OV, Carlson JA, Alcaraz J, et al. gram: a randomized controlled trial in
Behavioral change during weight loss Siglang Buhay: nutrition and physical African-American churches. Am J Prev
program and one-year follow-up: activity promotion in Filipino- Med. 2013;44:122-131.
Saku Control Obesity Program Americans through community orga- 48. Lowe MR, Tappe KA, Butryn ML,
(SCOP) in Japan. Asia Pac J Clin Nutr. nizations. J Public Health Manag Pract. et al. An intervention study targeting
2012;21:22-34. 2013;19:162-168. nutritional intake in worksite cafeterias.
29. Pimentel GD, Portero-McLellan KC, 39. French SA, Harnack LJ, Hannan PJ, Eat Behav. 2010;11:144-151.
Oliveira EP, et al. Long-term nutrition Mitchell NR, Gerlach AF, Toomey TL. 49. Francis SL, Taylor ML. A social mar-
education reduces several risk factors Worksite environment intervention to keting theory-based diet-education
for type 2 diabetes mellitus in Brazilians prevent obesity among metropolitan program for women ages 54 to 83 years
with impaired glucose tolerance. Nutr transit workers. Prev Med. 2010;50: improved dietary status. J Am Diet As-
Res. 2010;30:186-190. 180-185. soc. 2009;109:2052-2056.
30. Plawecki K, Chapman-Novakofski K. 40. Iriyama Y, Murayama N. Effects of a 50. Moschonis G, Katsaroli I, Lyritis GP,
Effectiveness of community interven- worksite weight-control programme Manios Y. The effects of a 30-month di-
tion in improving bone health behav- in obese male workers: a randomized etary intervention on bone mineral
iors in older adults. J Nutr Gerontol controlled crossover trial. Health Educ density: the Postmenopausal Health
Geriatr. 2013;32:145-160. J. 2014;73:247-261. Study. Br J Nutr. 2010;104:100-107.
31. Ritchie LD, Whaley SE, Spector P, 41. Johansen KS, Bjorge B, Hjellset VT, 51. Broekhuizen K, Kroeze W, Poppel MN,
Gomez J, Crawford PB. Favorable Holmboe-Ottesen G, Raberg M, Oenema A, Brug J. A systematic review
impact of nutrition education on Cali- Wandel M. Changes in food habits of randomized controlled trials on the
fornia WIC families. J Nutr Educ Behav. and motivation for healthy eating effectiveness of computer-tailored phys-
2010;42:2-10. among Pakistani women living in Nor- ical activity and dietary behavior promo-
32. Shahnazari M, Ceresa C, Foley S, way: results from the InnvaDiab- tion programs: an update. Ann Behav
Fong A, Zidaru E, Moody S. Nutri- DEPLAN study. Public Health Nutr. Med. 2012;44:259-286.
tion-focused wellness coaching pro- 2010;13:858-867. 52. Olander C: Office of analysis, nutrition
motes a reduction in body weight in 42. Linde JA, Nygaard KE, MacLehose RF, and evaluation food and nutrition service.
overweight US veterans. J Acad Nutr et al. HealthWorks: results of a multi- Nutrition Education and the Role of Dosage.
Diet. 2013;113:928-935. component group-randomized worksite US Department of Agriculture; 2007.
33. Silva MN, Vieira PN, Coutinho SR, environmental intervention trial for 53. Gutman SA, Murphy SL. From the
et al. Using self-determination theory weight gain prevention. Int J Behav desk of the editor and associate edi-
to promote physical activity and weight Nutr Phys Act. 2012;9:14. tor—information commonly unre-
control: a randomized controlled trial 43. Racette SB, Deusinger SS, Inman CL, ported in intervention effectiveness
in women. J Behav Med. 2010;33: et al. Worksite Opportunities for Well- studies. Am J Occup Ther. 2012;66:7-10.
110-122. ness (WOW): effects on cardiovascular 54. Gearing RE, El-Bassel N, Ghesquiere A,
34. Sorensen LB, Greve T, Kreutzer M, disease risk factors after 1 year. Prev Baldwin S, Gillies J, Ngeow E. Major
et al. Weight maintenance through Med. 2009;49:108-114. ingredients of fidelity: a review and sci-
behaviour modification with a cooking 44. Rustad C, Smith C. Nutrition knowl- entific guide to improving quality of
course or neurolinguistic program- edge and associated behavior changes intervention research implementation.
ming. Can J Diet Pract Res. 2011;72: in a holistic, short-term nutrition edu- Clin Psychol Rev. 2011;31:79-88.
181-185. cation intervention with low-income 55. Cerin E, Barnett A, Baranowski T.
35. Wieland ML, Weis JA, Palmer T, et al. women. J Nutr Educ Behav. 2013;45: Testing theories of dietary behavior
Physical activity and nutrition among 490-498. change in youth using the mediating
immigrant and refugee women: a 45. Sarrafzadegan N, Kelishadi R, variable model with intervention pro-
community-based participatory Sadri G, et al. Outcomes of a grams. J Nutr Educ Behav. 2009;41:
research approach. Womens Health Is- comprehensive healthy lifestyle pro- 309-318.
sues. 2012;22:225-232. gram on cardiometabolic risk factors 56. Sallis JF, McKenzie TL, Conway TL,
36. Baruth M, Wilcox S. Multiple behavior in a developing country: the isfahan et al. Environmental interventions for
change among church members taking healthy heart program. Arch Iran eating and physical activity: a random-
part in the faith, activity, and nutrition Med. 2013;16:4-11. ized controlled trial in middle schools.
program. J Nutr Educ Behav. 2013;45: 46. Savoie MR, Mispireta M, Rankin LL, Am J Prev Med. 2003;24:209-217.
428-434. Neill K, LeBlanc H, Christofferson D. 57. Hand R, Abram J, Brown K, Ziegler P,
37. Cullen KW, Lara Smalling A, Intention to change nutrition-related Scott Parrott J, Steiber A. Development
Thompson D, Watson KB, Reed D, behaviors in adult participants of a sup- and Validation of the Guide for Effec-
Konzelmann K. Creating healthful plemental nutrition assistance program- tive Nutrition Interventions and Educa-
home food environments: results of a education. J Nutr Educ Behav. 2015;47: tion (GENIE): a tool for assessing the
study with participants in the expanded 81-85. quality of proposed nutrition education
food and nutrition education program. 47. Wilcox S, Parrott A, Baruth M, et al. programs. J Nutr Educ Behav. 2015;47:
J Nutr Educ Behav. 2009;41:380-388. The Faith, Activity, and Nutrition pro- 308-316.
165.e1 Murimi et al Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017
CONFLICT OF INTEREST
The authors have not stated any con-
flicts of interest.

Factors Influencing Efficacy of Nutrition Education Interventions A Systematic Review

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Factors Influencing Efficacy of Nutrition Education Interventions A Systematic Review

Uploaded by

Copyright:

Available Formats

Systematic Review

Factors Influencing Efficacy of Nutrition Education

METHODS Inclusion/Exclusion Criteria scores) (Tables 1 and 2). Fidelity as a

Interventions in the Reviewed

et al, 201110 II overweight and lifestyle Intervention: conducted with significant

and self-efficacy demonstrations to 556 mg dietary calcium;

increase self- (P < .001), and self-

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

a series of 4 15- intervention group episodes improvements in knowledge

about healthy classroom planning among

decrease cancer 12 wk.

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

preliminary results training and weekly randomized, changes in weekly energy

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

nutrition guidelines intervention and control

increase in fruit and

40–65 y emphasized counseling and 1 y after (– 5.0 kg vs 0.1 kg for men

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

limited response to the

variables, with many not

with nutrition coaching.

201033 30–45 y weight Intervention: NE each weekly or group showed increased

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

Wilcox, Americans aged which participants Intervention: PA and change in health

aged $18 physical activity Workshops and increases in PA

life) conducted demonstrations, low-fat diet (P < .05),

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

25–63 y intervention study Intervention: 2 h each, intervention and control

access, and media

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

participated at component of Intervention: Public intervention Proceed obesity, hypertension,

intervention pressure, and self-

Journal of Nutrition Education and Behavior Volume 49, Number 2, 2017

You might also like