Professional Documents
Culture Documents
Physical Activity
Critical Review of the Literature
by Thanee Kaewthummanukul, PhD, RN, and Kathleen C. Brown, PhD, RN
Research Abstract
The purpose of this article is to provide a review of research literature related to factors that influence employee partici-
pation in physical activity. Eleven published studies investigating physical activity in samples of adult employees were
included in this review. Across these studies, self-efficacy, or belief in personal ability to perform this health behavior,
was the best predictor of physical activity among employees. Perceived benefits of physical activity and perceived health
status were also found to influence participation in physical activity. Occupational health nurses could offer self-care
classes and intervention programs that assist employees in improving their perceptions of self-efficacy to perform physi-
cal activity. Additionally, occupational health nurses should continue to reinforce the benefits of physical activity in all
health promotion venues.
R
egular physical activity is an important compo- activity, much of the U.S. population remains sedentary
nent of a healthy lifestyle that improves not only (Marcus, King, Clark, Pinto, & Bock, 1996; Sallis & Owen,
physical but also psychological health (Sullum, 1999). The U.S. Department of Health and Human Services
Clark, & King, 2000). Physical health benefits of regu- (2000) reported that only 15% of adults in 1997 engaged in
lar physical activity include increased muscle and bone the recommended type, duration, and intensity of physical
strength, decreased body fat, improved weight control, activity (occupational, lifestyle, and leisure-time activities)
and aerobic fitness (Volek, Vanheest, & Forsythe, 2005; and 40% of adults did not engage in leisure-time physi-
Vuori, 1998). Regular physical activity also enhances a cal activity. The 1990 National Health Interview Survey of
sense of well-being, reduces the risk of developing de- adult employees found that approximately half reported no
pression and anxiety, and improves mood (U.S. Depart- leisure-time physical activity and more than half reported
ment of Health and Human Services, 2000). Physical no strenuous physical activity during the workday (Centers
activity has been demonstrated to improve the health for Disease Control and Prevention, 2000).
status of individuals with diabetes, hypertension, con- Determinants of employee participation in physical
gestive heart failure, obesity, and depression (Brunet, activity have not been the focus of previous literature re-
Plotnikoff, Raine, & Courneya, 2005; Coats, 2005; Sal- views. Understanding the factors related to physical ac-
lis & Owen, 1999; Stewart et al., 2005). Physical activ- tivity (occupational, lifestyle, and leisure-time activity)
ity is defined as all movements involving large muscle among employees is essential to establishing effective
groups of the body; exercise is considered a subset of intervention programs. The design of intervention pro-
physical activity (Sallis & Owen, 1999). grams must be based on knowledge of factors that influ-
Despite the enormous benefits of regular physical ence employee adoption of physical activity.
This article provides a detailed review of the latest
About the Authors research literature related to factors that influence the lev-
Dr. Kaewthummanukul is Instructor, Faculty of Nursing, Chiang Mai Uni-
versity, Chiang Mai, Thailand. Dr. Brown is Professor and Director, Oc-
el of employee participation in physical activity. Through
cupational Health Nursing, School of Nursing, University of Alabama at an analysis of studies and a discussion of implications for
Birmingham, Birmingham, AL. occupational health research and practice, the review will
251
252
Table
Studies of Factors Influencing Physical Activity Participation Among Employees (Continued)
Sample Theoretical
Study Characteristics Study Design Instruments Framework Major Findings
Desmond et al. N = 325; male employees Descriptive, cross-sec- Demographics: age, education, Health Promotion Approximately 30% of the
(1993) at a large Midwestern tional study; question- income, and job category; three Model (Pender, variance in overall physical
public utilities company; naires completed cognitive–perceptual factors: 1987) activity was explained by
mean age = 37.5 years on-site perceived health status (TR and IC job category and self-
(range = 20 to 62 years); not reported); physical self-efficacy efficacy; 56% of the
fewer than 1% with < 12 scale, IC = 0.61 to 0.84; and per- variance in occupational
years of education; 83% ceived barriers, IC = 0.66 to 0.85 (V activity was explained by
annual income > $40,000; not reported for any variable); three income and job cat-
70% physical job and 25% types of physical activity: work in- egory; 6% of the variance
management dex, leisure-time index, and sports in sports activity was
activity index, TR = 0.74 to 0.88 explained by perceived
health status; approxi-
mately 14% of the variance
in leisure-time activity was
explained by self-efficacy
and job category.
Biddle et al. N = 131: 63 men, 68 Descriptive, cross-sec- Seven cognitive–psychosocial Theory of Planned 32% and 38% of variance
(1994) women; full-time employ- tional study; ques- variables: intention, IC = 0.94; Behavior in strenuous physical ac-
ees on university campus; tionnaires using mail perceived control, IC = 0.81; self- tivity were explained by all
age and occupational survey efficacy, IC = 0.91; attitude, social seven cognitive–psycho-
group not reported norms, benefits, and barriers, TR social variables for women
and IC not reported (V not reported and men, respectively;
for any variable); and self-report intention and self-efficacy
physical activity: work activity, were the best predictors of
home, and free-time activities in strenuous physical activity
average week (TR, IC, and V not for women and intention
reported) and attitude for men; at-
titude, perceived control,
benefits, and self-efficacy
were the best predictors of
intention for women; atti-
tude was the best predic-
tor for men.
AAOHN Journal
Duffy et al. N = 397 full-time (91%) Descriptive, cross-sec- Demographics: age, education, mari- Not stated Participants who had high-
(1996) or part-time employed tional study; ques- tal status, annual household income, er levels of personal belief
Mexican American tionnaires using mail and occupational status; three cogni- in their competency, good
women (working outside survey tive–psychosocial variables: Multidi- to excellent current health,
the home); mean age = mensional Health Locus of Control and personal control of
36 years (range = 19 to 70 Scale–Form A, IC = 0.67 to 0.84; their health (regardless
years); 85% high school current health (health perception of age, education, marital
253
254
Table
Studies of Factors Influencing Physical Activity Participation Among Employees (Continued)
Sample Theoretical
Study Characteristics Study Design Instruments Framework Major Findings
Burn et al. N = 709: 249 men, 460 Descriptive, cross-sec- Stage-of-exercise behavior scale, TR Transtheoretical Significant differences were
(1999) (65%) women; employ- tional study; question- = 0.78, V reported with a Seven-Day Model of Change found among stages of
ees from a government naires completed Recall Physical Activity Question- change for BMI, aerobic
office in Britain; mean age on-site naire; exercise self-efficacy, TR = exercise, dietary habits,
= 39 years (SD = 10.97 0.90; physical activity index, TR = exercise self-efficacy, and
years); mean BMI = 24.27 0.96; dietary habits (5-item standard stress when controlling
kg/m2 (SD = 3.69 kg/m2); instrument); smoking and alcohol for age; participants in the
adjusted means and SDs: use (single item); stress (anxiety and precontemplation stage re-
age (precontemplation = depression from Goldberg inventory) ported a significantly lower
45 ± 10 years, contem- (TR, IC, and V not reported); BMI; level of aerobic activity
plation = 41 ± 11 years, blood pressure (sphygmomanometer); than did those in the action
preparation = 39 ± 10 blood cholesterol (finger-prick test stage. Similarly, a signifi-
years, action = 34 ± 11 using a handheld Accutrend meter cant difference was found
years, maintenance = 38 ± [Boehringer Ingelheim, Ingelheim, between the contemplation
11 years) Germany]); urine tested for the pres- and maintenance stages;
ence of sugar, blood, and protein with dietary habits and exercise
urine test strip self-efficacy differed signifi-
cantly across the stages.
Boutelle et al. N = 9,043: 4,136 men, Descriptive, cross-sec- Demographics: weight status (de- Not stated Both gender participants
(2000) 4,907 women; employees tional study; question- fined by BMI), age, education, job in the highest activity
in 24 worksites in the Min- naires completed category (professional, white collar, quartiles were more edu-
neapolis–St. Paul metro- on-site or blue collar), marital status, and cated and younger. Men
politan area; demograph- ethnicity; other health behaviors: with high levels of activ-
ics of men and women fat consumption (Block Fat ity were more likely to be
described by quartile of Screener), stress (Perceived unmarried; higher levels
leisure-time exercise Stress Scale), smoking behavior, of leisure-time exercise
daily alcohol use, seat belt use, were positively associ-
and health; leisure-time physical ated with seat belt use
exercise (Leisure Time Exercise and negatively related to
Questionnaire) (TR, IC, and V not smoking, dietary fat intake,
reported for any variable) reported stress, and obe-
sity; leisure-time exercise
was positively associated
with daily alcohol use for
women.
AAOHN Journal
Burton & Turrell N = 24,454: 55% men, Descriptive, cross-sec- Demographics: occupation (profes- Not stated Most participants were clas-
(2000) 45% women (Australian); tional study; secondary sional, white collar, or blue collar) sified as having insufficient
age range, 18 to 64 years; analysis of data col- and hours worked; confounding and activity (67.7%); the rate
blue collar: 18% women, lected during face-to- mediating variables: living arrange- of insufficient activity was
48% men; white collar: face interviews using ment, smoking status, BMI, self-re- higher for women than for
52% women, 16% men; questionnaires ported health, and age; leisure-time men; those in blue collar
professional: 30% women, physical activity: walking, moderate occupations were 71%
255
tivity. Three studies used investigator-developed instruments for determining the role of personal goals and perceived
without reliability and validity assessments. The Leisure Time barriers to physical activity participation. The Health Be-
Exercise Questionnaire, developed by Godin and Shephard lief Model proposes that health behaviors are influenced
(1985), was the most frequently used measure, although it by the individual’s perceived susceptibility to developing
was used in only three of the studies. health problems, the perceived seriousness of the health
problems in relation to the individual’s quality of life, the
Theoretical Framework Contributing to Determinants individual’s belief in the benefits of the health behavior,
of Physical Activity and the individual’s perception of barriers to the behavior.
Eight of the studies identified a theory or model as an Rabinowitz et al. (1992) used Social Cognitive The-
organizing framework. The models or frameworks were ory (Bandura, 1977) to examine personal determinants
Pender’s Health Promotion Model (3 studies), Theory of of leisure-time exercise. The Social Cognitive Theory
Planned Behavior (2 studies), the Transtheoretical Model suggests that behavior change and maintenance of that
of Change (1 study), the Health Belief Model (1 study), behavior are a function of efficacy expectation or self-ef-
and Social Cognitive Theory (1 study). ficacy (i.e., individuals’ beliefs in their capability to per-
In three studies (Desmond et al., 1993; Pender et al., form certain behaviors) and outcome expectations (i.e.,
1990; Piazza et al., 2001), the Health Promotion Model individuals’ beliefs in the behavior–outcome relation-
(Pender, 1987) was specified as the theoretical framework. ship) (Bandura, 1977). Exercise self-efficacy was a major
The 1987 version of the Health Promotion Model identi- variable studied by Rabinowitz et al. (1992).
fied cognitive perceptual factors in the individual that, The models or theoretical frameworks used most fre-
when modified by situational, personal, and interpersonal quently in the studies were the Health Promotion Model
characteristics, result in participation in health-promoting and the Theory of Planned Behavior. Although different
behaviors, including physical activity. The reviewed stud- theoretical frameworks have been applied in the reviewed
ies based on the Health Promotion Model focused on dif- studies, no theory or model is sufficient to thoroughly
ferent variables selected from the model as determinants explain physical activity behavior. Further research is
of physical activity participation, including importance of needed to contribute to development of a comprehensive
health, perceived control over health, perceived self-effi- theory or model to predict physical activity behavior.
cacy, definition of health, perceived health status, behav-
ioral factors, and perceived barriers. However, revisions Determinants of Physical Activity
in the Health Promotion Model in 1996, after the 1987 Participation
version was tested and the results were analyzed, resulted Demographics
in the deletion of importance of health and perceived con- Several studies examined the association between the
trol over health (Pender, 1996). Definition of health and level of physical activity and demographics such as age
perceived health status were repositioned in the model (Boutelle et al., 2000; Burn et al., 1999; Burton & Turrell,
and included in a category of personal factors. 2000; Desmond et al., 1993; Duffy et al., 1996; Pender et al.,
Biddle et al. (1994) and Blue et al. (2001) used the 1990; Piazza et al., 2001), gender (Boutelle et al., 2000; Bur-
Theory of Planned Behavior (Ajzen, 1988) as a theoreti- ton & Turrell, 2000; Pender et al., 1990), ethnicity (Boutelle
cal framework. The Theory of Planned Behavior propos- et al., 2000), marital status (Boutelle et al., 2000; Burton &
es that participation in physical activity is predicted by a Turrell, 2000; Duffy et al., 1996; Pender et al., 1990), in-
measure of intention to engage in physical activity that, come (Desmond et al., 1993; Duffy et al., 1996; Pender et
in turn, is predicted by attitudes toward physical activity, al., 1990), education (Boutelle et al., 2000; Desmond et al.,
social norms toward engaging in physical activity, and 1993; Duffy et al., 1996; Pender et al., 1990), hours worked
perceived behavior control (Blue, 1995). (Burton & Turrell, 2000), and job category (Boutelle et al.,
Burn et al. (1999) used the Transtheoretical Model 2000; Burton & Turrell, 2000; Duffy et al., 1996).
of Change (Prochaska & Velicer, 1997), which suggests Five of the seven studies that examined age as a pre-
that an individual’s current behavioral status is associated dictor variable found that the level of physical activity de-
with a personal intention to change or maintain behav- creased with age (Boutelle et al., 2000; Burn et al., 1999;
ioral patterns (Marcus et al., 1996). The Transtheoretical Burton & Turrell, 2000; Duffy et al., 1996; Piazza et al.,
Model of Change describes the following stages of moti- 2001). These studies included male and female workers,
vational readiness to adopt exercise or other behaviors: except the studies by Duffy et al. (1996) and Piazza et al.
l Precontemplation (not thinking about changing). (2001), which included only female workers. Burn et al.
l Contemplation (thinking about changing). (1999) examined the relationship between age and an in-
l Preparation (ready to change). dividual’s motivational readiness to adopt exercise among
l Action (actively engaged in change). British government office employees and found the stage
l Maintenance (sustaining change over time) (Marcus, of motivational readiness to exercise varied significantly
Rossi, Selby, Niaura, & Abrams, 1992). by age. Individuals who did not think about adopting exer-
Burn et al. (1999) explored the relationship between in- cise were more likely to be older, and those who actively
dividuals’ motivational readiness to adopt exercise and engaged in exercise tended to be younger compared with
coronary heart disease risk factors. those in other stages. Pender et al. (1990), however, re-
The Health Belief Model (Becker, 1974) was used by ported that adoption of a health-promoting lifestyle among
Gebhardt and Maes (1998) as an explanatory framework female workers increased with age. In contrast to the pre-