Title: Exercise adherence in employee exercise programs
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00100756/00001
 Material Information
Title: Exercise adherence in employee exercise programs implementation of a health education intervention
Physical Description: Book
Language: English
Creator: Stouffer, Kristine
Publisher: State University System of Florida
Place of Publication: Florida
Florida
Publication Date: 2000
Copyright Date: 2000
 Subjects
Subject: Health Science Education thesis, Ph. D   ( lcsh )
Dissertations, Academic -- Health Science Education -- UF   ( lcsh )
Genre: government publication (state, provincial, terriorial, dependent)   ( marcgt )
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )
 Notes
Abstract: ABSTRACT: The purpose of this study was to determine if a behavioral health education intervention would increase exercise self-efficacy thereby increasing exercise adherence among elementary school employees. In addition, social support observations were made to determine what types of social support existed for the treatment intervention groups. The design of this study was a quasi-experimental matched-pair design using four elementary school worksites (three treatment schools and one control school). The treatment intervention consisted of health promotion consultants using self-efficacy techniques through a "Personal Exercise Plan" or "PEP" during monthly onsite participant sessions. Self-efficacy and social support were assessed using the Causal Dimension Scale II and Social Support for Exercise Habits Scale, respectively
Abstract: Exercise adherence was assessed by use of self-reported exercise logs used to track adherence to a cardiovascular exercise program of at least three 20-minute sessions per week of aerobic activity. Exercise-related outcomes measures (taken during onsite pre and post program fitness assessments) included resting heart rate, blood pressure, body composition percentage, and estimated VO subscript 2max. The study sample included 45 treatment and 15 control group participants comprising 88% females; 77% Caucasians and 18% African Americans; 88% between the ages of 31 and 60 years of age; and 75% teachers and 25% office and custodial employees. Additionally, 93% of the control group and 60% of the treatment group were exercising before the start of the program.
Abstract: The health education intervention increased self-efficacy in the intervention group at midpoint, but decreased somewhat at program end. Coworker support seemed to be the greatest form of social support for the intervention group throughout the program. The intervention groups did not have greater adherence than the control group in terms of frequency and duration of cardiovascular exercise, however more participants in the treatment group were exercising at the program goal compared with control participants. The intervention group increased significantly in resting heart rate and diastolic blood pressure compared with the control group, however, both treatment and control groups failed to significantly improve in weight or body composition but did improve in systolic blood pressure.
Thesis: Thesis (Ph. D.)--University of Florida, 2000.
Bibliography: Includes bibliographical references (p. 114-133).
System Details: System requirements: World Wide Web browser and PDF reader.
System Details: Mode of access: World Wide Web.
Statement of Responsibility: by Kristine Stouffer.
General Note: Title from first page of PDF file.
General Note: Document formatted into pages; contains ix, 135 p.; also contains graphics.
General Note: Vita.
 Record Information
Bibliographic ID: UF00100756
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: oclc - 70786458
alephbibnum - 002640017
notis - ANA6848

Downloads

This item has the following downloads:

finaldissertation ( PDF )


Full Text












EXERCISE ADHERENCE IN EMPLOYEE EXERCISE PROGRAMS:
IMPLEMENTATION OF A HEALTH EDUCATION INTERVENTION
















By


KRISTINE STOUFFER


A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE
UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE
REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY

UNIVERSITY OF FLORIDA


2000

















ACKNOWLEDGEMENTS

The author wishes to acknowledge the Program Coordinator/Research Assistant, Julie

Gibbard; Assessment Team Members, Yvonne Dumas, Chris Eaddy, Amy Hagan, Kristin

Lauria, Robert Trimarche, and Roseanne Vullo; Health Promotion Consultants, Emily

Christian, Chris Eaddy, Jenny Griggs, Kristin Lauria, Laura Pople, and Robert

Trimarche; and Pilot Study Assistant, Jenny Griggs; whose hard work and dedication

made this research program possible.

The author would also like to acknowledge the Living Well Employee Wellness

Center at the University of Florida and the Orion Fitness Center, Gainesville, Florida, for

lending needed assessment equipment, and the American Red Cross, Reebok, and United

Way for donating participant program completion awards.

In addition, the author would like to acknowledge her dissertation committee

members, Jill Varnes, Ed.D., Steve Dorman, Ph.D., Dan Connaughton, Ed.D., and Dave

Miller, Ph.D., her parents Robert and Patricia Stouffer, and her fiancee Florencio

Calderon, for their contributions and support of this project.















TABLE OF CONTENTS

page
ACKNOWLEDGEMENTS ........ ................... .... ii

LIST O F TA B LE S ............................................. .... ...... vi

L IST O F FIG U R E S ..................................................... ... vii

ABSTRACT ......... ... ........................................... ....... viii

CHAPTERS

1 INTRODUCTION ........................................ .................... 1

Introductory Background Statement ....................................... 1
Statement of the Problem ......... .............................. .... 7
Research Questions ................. ................. ................. ......... 8
Significance of the Study ......................................... ...... 9
D elim stations ...................................................... .. .. 10
Limitations ............. ............................................. ...... 11
Assumptions ...................................................... 12
Definition of term s .......................... ..................... ....... 12

2 REVIEW OF LITERATURE ....................................... ...... 15

In tro d u ctio n ................. ................. ................. 1............... .. 15
The Problem ...................................... ..... ..... 15
Benefits of Exercise ...................................... .... ......... 18
Exercise Adherence ............................................. 23
General Exercise Program s................. ............... ................. .. .... 24
General Worksite Programs .... ..................... ............ 25
Worksite Programs in Schools ........................................ 29
Self Efficacy ........ .................................................... 30
Self-Efficacy and Exercise Adherence ...................................... 32
Social Support ............................ ................... ......... 36
Social Support and Exercise Adherence ................. ............... 38
The Self-Efficacy/Social Support Relationship to Exercise Adherence ......... 39
Use of Behavioral Techniques to Enhance Social Support and Self Efficacy
in Health Education Interventions ................................... .. 42
Sum m ary .... .................... ............... ... ............ ....... 4 5











3 PROCEDURES AND METHODS OF ANALYSIS ............................ 47

Population and Sam ple ............................................ ........ 47
Settings ...................... ....... ............ ....... ........ 47
Form (design) and Sources of Data .................................. ....... 48
Treatm ent ... ................. ...... ......................................... 52
Instrum ents ................... .................. ................................... 54
Statistical A analysis ................................. ............................. ....... 58

4 RESULTS ................................................ 60

Introduction ................... .................. .................. ................. 60
Descriptive Results ................... ................... ........................ ... 60
Self-Efficacy and Social Support Results .................................... ..... 63
Assessment Measures Results .............................................. 68
Adherence Results ......................................... ...................70
Correlational Analyses ........................................ ............. 74
Summary ........................................ 75

5 DISCUSSION AND CONCLUSIONS ................................... ..... 76

Introduction ................... .................. .................. ................. 76
Discussion of Results ........................................ ............. 76
Discussion of Programming Aspects ............................ ........... 88
Strengths .................. ....... .................................... . . .... 92
W weaknesses .................................................................. ............ 92
Conclusions ................ ......... ........ .................. ................. 93
Implications for Health Education .......... ... ........ ............ ..... ........... 95
R ecom m endations .... ......... ......................... ............ 98

APPENDICES

A INFORMED CONSENT .................. ........... 100

B PRE-PARTICIPATION QUESTIONNAIRE .................................. 102

C CAUSAL DIMENSION SCALE II ............................................ 104

D SOCIAL SUPPORT FOR EXERCISE HABITS SCALE.......................... 106

E A SSE SSM EN T FORM ........................................... ........ 108

F EXERCISE VERIFICATION SHEET ................................ ....... 110

G PERSONAL EXERCISE PLAN ............................ ......................... 111











R E FE R E N C E S ................................................ ......... 114

BIOGRAPHICAL SKETCH .................................................... 134















LIST OF TABLES


Table page


I. Demographic Profile for Program Participants and Representative County ...... 61

II. Reported Participant Exercise and Sports/Leisure Status ..................... 62

III. Summation of Personal Exercise Plan (PEP) Information Reported
by Participants ................... .................. ................... ....... ... 64

IV. Reported Reasons for Not Engaging in Regular Exercise Indicated on the
CD SII ...................................... .................... .... ......... 63

V. Self-Efficacy and Social Support Descriptive Statistics ........ ........... 65

VI. Assessment Measures Descriptive Statistics ...............................69

VII. Adherence Descriptive Statistics for 14 Week Program .....................71

VIII. Levels of Adherence for Cardiovascular Frequency ............................72

IX. Levels of Adherence for Cardiovascular Duration ..............................72
















LIST OF FIGURES


Figure


1. O'Donnell's Health Promotion Behavior Model ........... ............. 2

2. Methodology Design ............................... ......... 48


page















Abstract of Dissertation Presented to the Graduate School of the
University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy

EXERCISE ADHERENCE IN EMPLOYEE EXERCISE PROGRAMS:
IMPLEMENTATION OF A HEALTH EDUCATION INTERVENTION

By

KRISTINE STOUFFER

August 2000




Chairperson: Dr. Jill Varnes
Major Department: Health Science Education

The purpose of this study was to determine if a behavioral health education

intervention would increase exercise self-efficacy thereby increasing exercise adherence

among elementary school employees. In addition, social support observations were made

to determine what types of social support existed for the treatment intervention groups.

The design of this study was a quasi-experimental matched-pair design using four

elementary school worksites (three treatment schools and one control school). The

treatment intervention consisted of health promotion consultants using self-efficacy

techniques through a "Personal Exercise Plan" or "PEP" during monthly onsite

participant sessions. Self-efficacy and social support were assessed using the Causal

Dimension Scale II and Social Support for Exercise Habits Scale, respectively. Exercise

adherence was assessed by use of self-reported exercise logs used to track adherence to a

cardiovascular exercise program of at least three 20-minute sessions per week of aerobic









activity. Exercise-related outcomes measures (taken during onsite pre and post program

fitness assessments) included resting heart rate, blood pressure, body composition

percentage, and estimated VO2max.

The study sample included 45 treatment and 15 control group participants comprising

88% females; 77% Caucasians and 18% African Americans; 88% between the ages of

3 land 60 years of age; and 75% teachers and 25% office and custodial employees.

Additionally, 93% of the control group and 60% of the treatment group were exercising

before the start of the program.

The health education intervention increased self-efficacy in the intervention group at

midpoint, but decreased somewhat at program end. Coworker support seemed to be the

greatest form of social support for the intervention group throughout the program. The

intervention groups did not have greater adherence than the control group in terms of

frequency and duration of cardiovascular exercise, however more participants in the

treatment group were exercising at the program goal compared with control participants.

The intervention group increased significantly in resting heart rate and diastolic blood

pressure compared with the control group, however, both treatment and control groups

failed to significantly improve in weight or body composition but did improve in systolic

blood pressure.















CHAPTER 1
INTRODUCTION

This chapter will introduce the main concepts of investigation in this study including,

but not limited to, exercise adherence, self-efficacy, social support, and worksite exercise

programs. It also will state the main research problem to be investigated as well as the

primary and secondary research questions to be answered through the study. The

assumptions, delimitations, limitations, and main definitions also will be described.

Introductory Background Statement

Heart disease is considered to be the most prevalent, costly, and chronic

disease of lifestyle. Even though it is well known that regular participation in exercise

will decrease the risk of heart disease, low exercise adherence and the high incidence of

sedentary lifestyle continue to plague society. Better and more effective physical activity

interventions are needed, which encompass social, psychological, and behavioral

influences, to improve exercise adherence and long-term compliance to physical activity.

Two main social, psychological, and behavioral-related concepts influencing exercise

behavior are social support and self-efficacy. The health education intervention tested in

this study involved improvement of self-efficacy towards exercise and a measurement of

types of social support that occur during the intervention. The intervention is part of a

worksite health promotion program provided to elementary school employees. To better

understand the roles of self-efficacy and social support in a worksite health










promotion exercise program, the subject areas of self-efficacy, social support, and

employee exercise programs must be examined.

Social Support, Self-efficacy and Exercise Adherence

To better understand the role of social support and of self-efficacy, one must first

understand the concept of self-efficacy and how it fits into Michael O'Donnell's Theory

of Health Promotion Behavior (first presented in Wallston, 1994). Self-efficacy,

introduced in Albert Bandura's social learning theory (Bandura, 1986), assesses the

person's belief that he/she can practice the behavior to achieve a desired outcome. Self-

efficacy has been shown to predict health behavior if the person both values health as an

outcome and believes that personal actions play a role in determining personal health

status (Wallston, 1994). In O'Donnell's Model of Health Promotion Behavior, social

support and self-efficacy are included within the right-hand portion of the model that also

includes components of the Theory of Planned Behavior (see Figure 1).

In O'Donnell's model, self-efficacy beliefs play a central role in predicting both

intentions and behavior. This yet untested, model also incorporates the construct of social

support as an important determinant of whether one's intentions do, indeed, get translated

into action.


-allth benefits V-Ile of
of behavior l Health I Prior experience M
Non-health_ upr Barriers
benefitnd of I- s


behavior I
Behiio s


S.ith r~eerents
Definition of







Figure 1
O'Donnell's Health Promotion Behavior Model
O'Donnell's Health Promotion Behavior Model










This model serves as a broad picture of how social support must be utilized as a function

of self-efficacy to promote positive health behaviors. The implementation of the health

education intervention in this study will test the latter half of O'Donnell's model in terms

of self-efficacy issues, social support and barriers, and prior experience for the behavior.

Whereas it appears that no single variable solely determines adherence to either

prescribed or self-initiated exercise regimens, self-efficacy has been consistently

identified as playing an important role in health and exercise behaviors (Bandura, 1986).

Numerous studies have linked social support to health enhancing and health impairing

behaviors. From the perspective of exercise behavior, one might expect socially

supportive behavior to play a particularly important role, especially in neophyte

exercisers or at-risk populations. Buddy systems, spousal participation and

encouragement, and positive feedback from exercise leaders and fellow participants all

have been suggested to play a role in continued exercise participation (Duncan &

McAuley, 1993).

Using positive reactions to an exercise program, goals for initially joining a program,

and social support, Wankel (1985) differentiated between participants and dropouts in an

exercise program. Dishman (1988) stated that social reinforcement in the form of group

or family support may be the most influential factor in adherence. Wallace, Raglin, and

Jastremski (1995) found that married pairs had significantly higher attendance and lower

dropout than married singles in a 12-month fitness program that appeared to be primarily

influenced by spousal support rather than by self-motivation. Oesterle (1988) found

social support to be significant in the indirect role as a modifier of behavior. Duncan,

Duncan, and McAuley (1993) found adherers perceived themselves to be socially









integrated recipients of adequate levels of guidance from members of their social network

within the exercise program. Nonadherers, however, perceived themselves to be less

socially integrated than adherers and recipients of higher levels of guidance.

Courneya and McAuley (1995) examined cognitive constructs as mediators of the

relationship between selected social influence constructs and adherence to structured

exercise classes. They found social support appeared to be mediated by perceived

behavioral control and intention to exercise. In a meta-analysis of social influence and

exercise, Carron, Hausenblas, and Mack (1996) supported an overall conclusion that

social influence has a positive influence on exercise behavior (both adherence and

compliance), cognitions about exercise involvement (both intentions to exercise and

efficacy for exercise), and attitudes associated with the exercise experience. The results

of this study also indicated that family did not represent the strongest source of social

influence for adherence behavior, rather the influence of important others had a stronger

impact. These researchers suggest that if important others can provide support without

exerting control over behavior, the individual can retain a perception of self-

determination, and adherence should be enhanced. Duncan, McAuley, Stoolmiller, and

Duncan (1993) endorsed the utility of a multidimensional view of social support and

found that specific provisions of support may play a significant role in an individual's

decision to adhere to a prescribed exercise regimen.

Self-efficacy generally has its greatest impact during the action phase of behavior

change (Wallston, 1994). According to Bandura (1986), there are four principle ways to

change a person's efficacy beliefs. The first and most effective way is to help the person

have an authentic mastery experience. Successful exercise attainments enhance









perceptions of physical efficacy, whereas failures debilitate perceived capabilities

(McAuley & Courneya, 1993). Social modeling provides a second source of information

from which self-efficacy can be enhanced. When participants are sedentary or older or

have little experience in the activity of interest, such a social modeling source may be

particularly salient. A third approach toward enhancing self-efficacy is through social

persuasion or social support. In a combination of social modeling and persuasion, one

can implement a "buddy exercise system," in which two or three individuals exercise

together, serve as models and social supports, and take on a sense of responsibility for

each other (McAuley & Courneya, 1993). A fourth means of enhancing self-efficacy

beliefs, according to Bandura, is to modify one's physiological reactions and/or how one

interpret signals from the body. Participants should be taught to interpret gradual change

in the degree of physical symptoms (i.e. fatigue, cardiorespiratory responses, and

muscular tension) as markers of improved conditioning and, thus, increased physical

capabilities.

In summary, focusing on methods for boosting self-efficacy can safeguard against

discouragement, feelings of displeasure and incompetence, and a proclivity to give up in

the face of any real or perceived adversity and challenge of physical activity.

Implementing a self-efficacy modeled exercise program should lead to increased exercise

adherence and overall health benefits. However, before a self-efficacy modeled exercise

program can be implemented within the worksite, exercise adherence issues within

worksite exercise programs must be examined.









Exercise Adherence in Worksite Exercise Programs

About 30% of employees within an organization will actually start participating when

a program is offered to them. Furthermore, 3 to 6 months after the start of a program,

only 50-60% of the original participants are still participating in the program (Lechner &

De Vries, 1995). Attitude toward physical activity, personality, type of exercise program

(frequency, intensity, and duration of training and mode of activity), body weight and

composition, medical problems (injuries, level of fitness, group supervision), staff,

spouse influence, age, sex, socioeconomic status, cost/method of payment, and time-

related factors are cited most frequently as factors affecting adherence to worksite

exercise programs (Pollock, 1988). Self-motivation and whether one is a smoker or a

blue-collar worker also have been shown to be significant in determining exercise

adherence (Oldridge, 1984). The most commonly cited reasons for not participating in a

corporate exercise class are "lack of time" and "lack of suitable facilities" (Shepard,

1988).

A literature review on the role of behavioral models in worksite exercise promotion

has revealed that the habit of exercising, the attitude towards regular exercise, the

perceived barriers to exercising, and the perceived self-efficacy to exercise were the

strongest variables associated with an intention to exercise (Godin & Shepard, 1990).

Godin and Gionet (1991) examined the psychosocial factors explaining an employee

population's intention to exercise and found significant support for the roles of habit of

exercise, attitude towards regular exercise, and perceived barriers to exercise, noting that

perceived self-efficacy was not measured in the actual study. Kimiecik (1992) examined

the theory of planned behavior in predicting the exercise intentions and behavior of









corporate employees and found subjects had the strongest intentions to exercise when

they had a favorable attitude toward exercise and believed they could successfully

perform the behavior. Also, a significant interaction between perceived control and

intention indicates that the more the attainment of a behavioral goal is viewed as being

under one's control, the stronger one's intention to perform the behavior.

In summary, the behaviors and needs of those exercise participants in the worksite are

the same as or quite similar to those within the community or general population. The

research suggests several guidelines when implementing exercise programs. First,

educational efforts should be focused on enhancing physical activity-related knowledge,

attitudes, and beliefs. Second, training in motivational and relapse prevention strategies

should be offered as well as social support for exercise. Also, gender and age are factors

that should be considered. Lastly, the individual's perceived motivation prior to

beginning an exercise program must be considered as well as perception of having

sufficient time to exercise (King, Taylor, Haskell, & DeBusk, 1990; Wilson, Wagner, &

Dwyer, 1991). Hence, the need for implementation of a health education intervention

enhancing self-efficacy and promoting social support among employees is clearly

indicated.

Statement of the Problem

This study assessed differences in exercise adherence, self-efficacy and types of social

support between employee groups receiving and not receiving a health education exercise

promotion intervention.









Research Questions

The primary research questions focus on the main areas of interest in the study (i.e.,

self-efficacy, social support, and the relationship to exercise adherence). The secondary

research questions focus on supporting areas of interest including additional outcome

measures (blood pressure, body composition, and estimated VO2max) and other

categorical variables.

Primary Research Questions

1. Does exercise self-efficacy increase during and after implementation of a health

education intervention to promote exercise adherence?

2. What types of social support for exercise are received by subjects during and after

implementation of a health education intervention to promote exercise adherence?

3. Do intervention groups have greater adherence to the exercise prescription than the

non-intervention group?

Secondary Research Questions

1. Does the intervention group have a greater improvement in blood pressure than the

non-intervention group?

2. Does the intervention group have a greater improvement in resting heart rate than the

non-intervention group?

3. Does the intervention group have a greater weight loss than the non-intervention

group?

4. Does the intervention group have a greater improvement in body composition than

the non-intervention group?









5. Does the intervention group have a greater improvement in cardiovascular fitness

(i.e., improved estimated VO2max) than the non-intervention group?

6. Does a relationship exist between previous exercise participation and adherence?

7. Does a relationship exist between previous sports or active leisure participation and

adherence?

Significance of the Study

The contribution of regular physical activity to such physical benefits as

cardiovascular health, strength, endurance and weight management is well established

(Bouchard, Shepard, Stephens, Sutton, & McPherson, 1990; Wankel, 1993). It also is

known that worksite exercise programs increase productivity; decrease absenteeism,

turnover rate and industrial injuries; enhance health practices of workers; facilitate

employee recruitment; and improve corporate image (Godin & Gionet, 1991). Many

health promotion advocates also emphasize the economic benefits of exercise, in terms of

cost-effectiveness, in addition to its contribution to better health (Hatziandreu, Koplan,

Weinstein, Caspersen, & Warner, 1988). Despite all the positive benefits of exercise,

adherence to exercise regimens continues to be a problem among the general population.

Well-documented statistics consistently indicate that the attrition rate from exercise

programs approximate 50% within the first six months (Duncan & McAuley, 1993).

It is clear that a gap exists between knowledge and behavior. There is a need for non-

traditional behavioral programs that integrate social, psychological, interpersonal and

cultural factors with knowledge to bridge the gap between information and behavior.

Health education interventions are key to unlocking the potential of this multi-factor

integrative approach that is needed to change health behavior. Health educators are









skilled in delivering interventions that consider the psychosocial, interpersonal and

cultural aspects of health behavior. These aspects must be addressed if a change in health

behavior, particularly adoption of exercise, is to take place.

Thus, the explanation and prediction of exercise behavior must be an important

objective among researchers in the areas of exercise science, behavioral and preventative

medicine (Dishman, 1987), and health education if any kind of impact on cardiovascular

health is to be obtained. Little systematic research has demonstrated use of social support

in helping people maintain exercise regimens. Even less research has analyzed types of

social support and the function of self-efficacy in exercise adherence.

This study's proposed health education intervention will test use of increasing self-

efficacy toward exercise through common methods of health education as well as assess

the types of social support formed. The results of this study will provide definitive

insight as to how a non-traditional modeled exercise program can effect exercise

adherence and perpetuate long-term exercise compliance. The results also will imply the

cost-effectiveness of a successful worksite exercise intervention program and the impact

of such a program on various types of demographic subgroups. It is hoped the self-

efficacy modeled health education intervention will serve as a benchmark intervention for

integrating behavioral techniques to bridge the gap between exercise knowledge and

actual exercise adherence and demonstrate that health educators are needed to implement

interventions such as these in the workplace to increase exercise adherence.

Delimitations

1. Worksites used in the study consist of four Gainesville area elementary schools that

volunteered to participate in the study.









2. Subjects were recruited through mini health fairs held at the school sites and flyer

distribution.

3. Data were collected during Spring 2000. Participants had to be available during the

Spring 2000 term from January through May.

4. The CDSII (McAuley, Duncan, & Russell, 1992) was employed to assess subjects'

self-efficacy causal attributions for previous attrition from exercise programs.

5. The Social Support for Exercise Habits Scale was used to measure the types of social

support that occur during the exercise program.

6. The YMCA 3-Minute Step Test was used to estimate VO2max

7. Anthropometric measures included height, weight, and three-site body composition.

8. All subjects interested in participating in the exercise program needed to meet the

screening criteria of having no history of cardiac disease or orthopedic injuries that

would contraindicate participation in an exercise program.

Limitations

1. Any level of workers (i.e., blue collar, administration, etc.) may choose to not

participate in the study.

2. All classifications of race may not be equally represented in the study sample.

3. Both genders may not be equally represented in the study sample.

4. All employees dropping out of the program may not be available for final data

collection due to relocation, etc.

5. Employees may not turn in all exercise verification sheets causing incomplete

adherence data.









6. Findings depend on the ability of the CDSII and Social Support for Exercise Behavior

Scales to accurately assess exercise self-efficacy and sources of social support,

respectively.

7. The YMCA step test was only an estimation and not an actual measure of VO2max

8. Resting heart rate was measured at the end of the day, instead of the preferred time of

morning wake.

9. Blood pressure was only measured one time at the pre and post fitness assessments.

Assumptions

1. School worksites participating have similar demographics and characteristics.

2. Gainesville area elementary school employees participating in the study adequately

represent the population of elementary school employees at their respective schools.

3. The CDSII and Social Support for Exercise Behavior Scales are adequate for data

collection necessary for the study.

Definition of Terms

Self-efficacy: A person's belief that he/she can perform a behavior in order to achieve a

desired outcome. For purposes of this study, the intended behavior is engaging in regular

exercise. The desired outcome is adhering to a given aerobic fitness plan.

Social support: Any support received by another individual for the promotion of engaging

in regular exercise.

Exercise adherence: Those participants engaging in regular exercise (as defined below)

for the duration of the program.









Regular exercise: Exercising aerobically (within the given target heart rate), for three 20-

minute sessions per week for a total of approximately twelve 20-minute sessions per

month for five months.

Drop out rate: Subjects identified as "drop outs" will be defined as discontinuing exercise

or exercising below the program goal for at least seven weeks after the start of the

program.

Pre-Participation Questionnaire (PPQ): A screening questionnaire designed to determine

any contraindications for exercise reported by participants (see "Instruments" section in

Chapter 3).

Estimated VO2mx (or aerobic fitness): The greatest rate of oxygen uptake by the body

measured during dynamic (cardiovascular) exercise, via a step test (for purposes of this

study). VO2max is often used as a measure of aerobic fitness.

"Class A": A classification by the American Heart Association and American College of

Sports Medicine for apparently healthy exercise participants or participants with no

contraindicating risks for exercise.

Health Promotion Consultant: A health education or exercise physiology student who

underwent training to deliver the health education intervention sessions to treatment

program participants.

Personal Exercise Plan (PEP): A behavioral contract (in a worksheet format) designed to

promote mastery experience within the consulting sessions. In the PEP, participants can

set goals, work through barriers and strategies, create an activity schedule and have

supporters sign the contract.






14


Non-starter: A participant recruited for the program, who underwent pre-program

assessment, but for various reasons did not actually start the program.

Strength training: Any type of weight bearing exercise including, but not limited to

weight lifting, calistenics, resistance bands, etc.















CHAPTER 2
REVIEW OF LITERATURE

Introduction

This chapter will present a literature review of the following areas of research: an

introduction to physical inactivity as an epidemic; the benefits of exercise; the issues

surrounding exercise adherence; characteristics of worksite exercise programs; an in

depth look at self-efficacy and social support studies; research examining the relationship

of self-efficacy and social support to exercise adherence; and behavioral techniques

implemented in health education interventions. The chapter will conclude with a

summary linking these areas together to set a research rationale for the present study.

The Problem

The message from the nation's scientists is clear, unequivocal, and unified: physical

inactivity is a risk factor for cardiovascular disease (Fletcher, et al., 1996; Pate, et al.,

1995) and its prevalence is an important public health issue. New scientific evidence

based on epidemiological observational studies, cohort studies, controlled trials, and

basic research has led to an unprecedented focus on physical activity and exercise

(AHA/ACSM Scientific Statement, 1998). The promotion of physical activity is at the

top of our national public health agenda, as seen in the publication of the 1996 report of

the US Surgeon General on physical activity and health (USDHHS, 1996).









The promotion of regular exercise has become an important priority for public health

interventions because of the documented physiological and psychological health benefits

(Courneya & McAuley, 1995). These health benefits include reduced risks for all-cause

mortality, coronary heart disease, hypertension, colorectal cancer, obesity, osteoporosis,

depression, anxiety, and stress (Bouchard, Shepard, Stephens, Sutton, & McPherson,

1990). Regular aerobic physical activity increases exercise capacity and plays a role in

both primary and secondary prevention of cardiovascular disease (Chandrashekhar &

Anand, 1991; Morris & Froelicher 1991; Paffenbarger, Hyde, Wing, Hsieh, 1986; Smith

et al., 1995; Wenger et al., 1995).

Exercise training increases cardiovascular functional capacity and decreases

myocardial oxygen demand at any level of physical activity in apparently healthy persons

as well as in most subjects with cardiovascular disease. Regular physical activity is

required to maintain these training effects. The potential risk of physical activity can be

reduced by medical evaluation, risk stratification, supervision, and education (Wenger et

al., 1995). Exercise can help control blood lipid abnormalities, diabetes, and obesity. In

addition, aerobic exercise adds an independent blood pressure-lowering effect in certain

hypertensive groups with a decrease of 8 to 10 mm Hg in both systolic and diastolic

blood pressure measurements (Braith, Pollock, Lowenthal, Graves, & Limacher, 1994;

Hagberg, 1990; Hagberg, Montain, Martin, & Ehsani, 1989; Jennings, Deakin, Dewar,

Laufer, & Nelson, 1989).

There is a direct relationship between physical inactivity and cardiovascular

mortality, with physical inactivity also serving as an independent risk factor for the

development of coronary artery disease (Blair, Kohl, Paffenbarger, et al., 1989; Lee,









Hsieh, & Paffenbarger, 1995; Morris, Clayton, Everitt, Semmence, & Burgess, 1990;

Powell, Thompson, Caspersen, & Kendrick, 1987). There is a dose-response relation

between the amount of exercise performed from approximately 700 to 2000 kcal of

energy expenditure per week and all-cause mortality and cardiovascular disease mortality

in middle-aged and elderly populations (Blair, Kohl, Barlow, et al., 1995; Lee, Hsieh, &

Paffenbarger, 1995;). The greatest potential for reduced mortality is in the sedentary who

become moderately active (Blair, Kohl, Barlow, et al., 1995). Most beneficial effects of

physical activity on cardiovascular disease mortality can be attained through moderate-

intensity activity (40% to 60% of maximal oxygen uptake, depending on age) (Blair,

Kohl, Barlow, et al., 1995; Lee, Hsieh, & Paffenbarger, 1995; Pate et al., 1995). The

activity can be accrued through formal training programs or leisure-time physical

activities.

Although most of the supporting data are based on studies in men, more recent

findings show similar results for women (Blair, Kohl, Paffenbarger, et al., 1989;

Lemaitre, Heckbert, Psaty, & Siscovick, 1995). Results of pooled studies reveal that

persons who modify their behavior after myocardial infarction to include regular exercise

have improved rates of survival (O'Connor et al., 1989; Oldridge, Guyatt, Fischer, &

Rimm, 1988). Recent studies also have revealed that intensive multiple interventions

such as smoking cessation, blood lipid reduction, weight control, and physical activity

significantly decreased rate of progression and, in some cases, led to regression in the

severity of atherosclerotic lesions in persons with coronary disease (Gould et al., 1992;

Haskell et al., 1994; Omish et al., 1990; Schuler, Hambrecht, Schlierf, Grunze, et al.,

1992; Schuler, Hambrecht, Schlierf, Niebauer, et al., 1992;). In addition, limited data









indicate that higher-intensity exercise compared with lower-intensity exercise improves

left ventricular ejection fraction in persons with coronary artery disease (Oberman et al.,

1995). Current activity status (ie, persons remaining physically active or having been

sedentary and becoming physically active) revealed the greatest decline in coronary

artery disease risk (Blair, Kohl, Barlow, et al., 1995; Lee, Hein, Suadicani, & Gyntelberg,

1992; Hsieh & Paffenbarger, 1995). Persons who remain sedentary have the highest risk

for all-cause and cardiovascular disease mortality.

In summary, the problem of inactivity in this country has implications that go beyond

what is currently seen to be plaguing our nation's health. We must address this health

risk from many aspects of intervention to begin to make an impact on exercise or

physical activity adoption and adherence.

Benefits of Exercise

Healthy persons as well as many persons with cardiovascular disease can improve

exercise performance with training (Adamopoulos et al., 1993; Coats et al., 1992;

Hambrecht et al., 1995; Kavanagh et al., 1988; Kobashigawa et al., 1994; Sullivan,

Higginbotham, & Cobb, 1989). This improvement is the result of increased ability to use

oxygen to derive energy for work. Exercise training increases maximum ventilatory

oxygen uptake by increasing both maximum cardiac output (the volume of blood ejected

by the heart per minute, which determines the amount of blood delivered to the

exercising muscles) and the ability of muscles to extract and use oxygen from blood.

Beneficial changes in hemodynamic, hormonal, metabolic, neurological, and respiratory

function also occur with increased exercise capacity. These changes also can benefit

persons with impaired left ventricular function, in whom most adaptations to exercise









training appear to be peripheral and may occur with low-intensity exercise (Adamopoulos

et al., 1993; Belardinelli, Georgiou, Scocco, Barstow, & Purcaro, 1995; Coats et al.,

1992; Hambrecht et al., 1995; Kavanagh et al., 1988; Kobashigawa et al., 1994; Sullivan,

Higginbotham, & Cobb, 1989).

Exercise training results in decreased myocardial oxygen demands for the same level

of external work performed, as demonstrated by a decrease in the product of heart rate x

systolic arterial blood pressure (an index of myocardial oxygen demand). These changes

also are beneficial in persons with coronary artery disease, who after exercise training

may attain a higher level of physical work before reaching the level of myocardial

oxygen requirement that results in myocardial ischemia (Trap-Jensen & Clausen, 1971).

Exercise training favorably alters lipid and carbohydrate metabolism. The exercise-

induced increase in high-density lipoproteins is strongly associated with changes in body

weight (King, Haskell, Young, Oka, & Stefanick, 1995; Tran & Weltman, 1985;

Williams, 1996). Regular exercise in overweight women and men enhances the beneficial

effect of a low-saturated fat and low-cholesterol diet on blood lipoprotein levels (Wood,

Stefanick, Williams, & Haskell, 1991). Endurance training has effects on adipose tissue

distribution (Schwartz et al.,1991), and the effect on adipose tissue distribution is likely

to be important in reducing cardiovascular risk (Despres et al., 1990; Donahue, Abbott,

Bloom, Reed, & Yano, 1987; Larsson et al., 1984; Krotkiewski, Bjorntorp, Sjostrom, &

Smith, 1983). Exercise training also has an important effect on insulin sensitivity, (King

et al., 1988; Rosenthal, Haskell, Solomon, Widstrom, & Reaven, 1983) and intense

endurance training has a highly significant salutary effect on fibrinogen levels of healthy

older men (Stratton et al., 1991). In addition, recent data support the role of physical









activity in the prevention and treatment of osteoporosis and certain neoplastic diseases,

notably colon cancer (Lee, 1994).

Developing and maintaining aerobic endurance, joint flexibility, and muscle strength

and endurance are important in a comprehensive exercise program, especially as people

age (Ades, Hanson, Gunther, & Tonino, 1987; Ades and Grunvald, 1990; Ades,

Waldmann, & Gillespie, 1995). Elderly women and men show comparable improvement

in exercise training, and adherence to training in the elderly is high (Ades, Waldmann, &

Gillespie, 1995). Resistance training exercise alone has only a modest effect on risk

factors compared with aerobic endurance training, but it does aid carbohydrate

metabolism through development or maintenance of muscle mass and effects on basal

metabolism (Evans, 1995; Kohrt & Holloszy, 1995). Furthermore, resistance training is

currently recommended by most health promotion organizations for its effects on

maintenance of strength, muscle mass, bone mineral density, functional capacity, and

prevention and/or rehabilitation of musculoskeletal problems (e.g., low back pain)

(ACSM position stand on osteoporosis and exercise: American College of Sports

Medicine, 1995). In the elderly, resistance training is both safe and beneficial in

improving flexibility and quality of life (Ghilarducci, Holly, & Amsterdam, 1989;

Sparling, Cantwell, Dolan, & Niederman, 1990; Stewart, Mason, & Kelemen, 1988).

Persons with cardiovascular disease are usually asked to refrain from heavy lifting and

forceful isometric exercises, but moderate-intensity dynamic strength training is safe and

beneficial in persons at low risk.

Many activities of daily living require more arm work than leg work. Therefore,

persons with coronary artery disease are advised to use their arms as well as their legs in









exercise training. The arms respond like the legs to exercise training both quantitatively

and qualitatively, although ventilatory oxygen uptake is less with arm ergometry.

Although peak heart rates are similar with arm and leg exercise, heart rate and blood

pressure response during arm exercise is higher than leg exercise at any submaximal

work rate. Therefore, target heart rates are designated 10 beats per minute lower for arm

training than for leg training (Clausen, 1976; Franklin, Hellerstein, Gordon, & Timmis,

1989; Franklin, Vander, Wrisley, & Rubenfire, 1983).

Maximum ventilatory oxygen uptake drops 5% to 15% per decade between the ages

of 20 and 80 years (Fletcher, et al., 1995; Pollock, Foster, Knapp, Rod & Schmidt, 1987;

Trappe, Costill, Vukovich, Jones, & Melham, 1996); a lifetime of dynamic exercise

maintains an individual's ventilatory oxygen uptake at a level higher than that expected

for any given age. The rate of decline in oxygen uptake is directly related to maintenance

of physical activity level, emphasizing the importance of physical activity (Jackson et al.,

1995).

Middle-aged men and women who work in physically demanding jobs or perform

moderate to strenuous recreational activities have fewer manifestations of coronary artery

disease than their less active peers (Morris, Clayton, Everitt, Semmence, & Burgess,

1990; Powell, Thompson, Caspersen, & Kendrick, 1987). Meta-analysis studies of

clinical trials reveal that medically prescribed and supervised exercise can reduce

mortality rates of persons with coronary artery disease (Hillsdon, Thorogood, Anstiss, &

Morris, 1995; O'Connor et al., 1989; Oldridge, Guyatt, Fischer, & Rimm, 1988).

In addition to the physical benefits of exercise, both short-term exercise and long-term

aerobic exercise training are associated with improvements in various indices of









psychological functioning. Cross-sectional studies reveal that, compared with sedentary

individuals, active persons are more likely to be better adjusted (Eysenck, Nias, & Cox,

1982), to perform better on tests of cognitive functioning (Spirduso, 1980), to exhibit

reduced cardiovascular responses to stress (Crews & Landers, 1987), and to report fewer

symptoms of anxiety and depression (Lobstein, Mosbacher, & Ismail, 1983). In one

report (Camacho, Roberts, Lazarus, Kaplan, & Cohen, 1991), persons who increased

their activity levels between the mid 1960s and mid 1970s were at no greater risk for

depression than those individuals who were active all along; however, persons who were

active and became inactive were 1.5 times as likely to become depressed by the mid

1980s compared with those who maintained an active lifestyle.

Longitudinal studies also have documented significant improvement in psychological

functioning. Exercise training reduces depression in healthy older men (Blumenthal,

Emery, Madden, et al., 1989), and in persons with cardiac disease (Kavanagh, Shephard,

Tuck, & Qureshi, 1977) or major depression (Martinsen, Medhus, & Sandvik, 1985).

Exercise also improves self-confidence and self-esteem (Folkins and Sime, 1981),

attenuates cardiovascular and neurohumoral responses to mental stress (Blumenthal,

Fredrikson et al., 1990), and reduces some type A behaviors (Blumenthal, Emery, Walsh,

et al., 1988). Although exercise training generally has not been found to improve

cognitive performance (Emery & Blumenthal, 1991), short bouts of exercise may have

short-term facilitative effects (Tomporowski & Ellis, 1986).

It is apparent that the benefits of physical activity go beyond affecting overall physical

health improved physical activity also makes an impact on an individual's social and

psychological well being. Physical activity has been shown to serve as a gateway









behavior or motivator for encouraging changes in diet or smoking behavior (Emmons et

al., 1994). Physical activity can therefore play an important role in comprehensive

disease-prevention programs and health education interventions.

Exercise Adherence

Despite the evidence supporting the health benefits of regular exercise, the overall

participation rates in regular physical activity are low. Advances in technology and

increased mechanization have resulted in a progressive decline in occupational activity

levels in the United States since World War II (Stephens, 1987). Furthermore, despite

the fact that leisure-time activity has probably increased over the last 20 years (Blair,

Mulder, & Kohl, 1987; Stephens, 1987), the prevalence of obesity, heart disease, and

exercise related cancers is increasing. Early research on exercise recruitment and

adherence emphasized the physical characteristics of exercise program participants.

Relative to dropouts, exercise adherers apparently had a higher standard of

cardiorespiratory fitness, less excess weight, and less subcutaneous fat (Massie &

Shepard, 1971; Sidney & Shepard, 1977).

It is estimated that only 10% of the North American population exercise regularly

(Stephens & Casperson, 1993; Stephens & Craig, 1990). Moreover, those who indicated

an exercise program have trouble maintaining it. Dishman (1988) has estimated that

approximately 50% of individuals who begin a structured exercise program will drop out

within the first six months. This statistic seems to hold regardless of the demographic

profile of the sample or the purpose of the exercise (Courneya & McAuley, 1995).

Results have been similar for children, college students, and middle-aged and elderly

persons and in primary prevention, secondary prevention, and worksite settings (Robison









& Rogers, 1994). The issue of non-adherence is particularly important because exercise is

only beneficial if it is maintained for extended periods of time. Thus, it is important to

develop strategies to improve exercise initiation and adherence, especially for persons

who are among the least active--some African-American women, the less educated, the

obese, and the elderly (King et al., 1992). As it is presented in Chapter 1, the problem of

exercise adherence is impacting the potential benefits reaped by any implemented

exercise program and must be addressed.

General Exercise Programs

Persons of all ages should include physical activity in a comprehensive program of

health promotion and should increase their habitual physical activity to a level

appropriate to their capacities, needs, and interest. Activities such as walking, hiking,

stair-climbing, aerobic exercise, calisthenics, resistance training, jogging, running,

bicycling, rowing, swimming, and sports such as tennis, racquetball, soccer, basketball,

and "touch" football are especially beneficial when performed regularly. Brisk walking

also is an excellent choice (Duncan, Gordon, & Scott, 1991; Rippe, Ward, Porcari, &

Freedson, 1988).

The training effect of such activities is most apparent at exercise intensities exceeding

40% to 50% of exercise capacity. (Exercise capacity is defined as the point of maximum

ventilatory oxygen uptake or the highest work intensity that can be achieved.) Evidence

also supports that even low- to moderate-intensity activities performed daily can have

some long-term health benefits and lower the risk of cardiovascular disease (Leon,

Connett, Jacobs, & Rauramaa, 1987; Rippe, Ward, Porcari, & Freedson, 1988; Slattery,

Jacobs, & Nichaman, 1989). Low-intensity activities generally range from 40% to 60%









of maximum capacity. The 40% to 60% of maximum capacity range is similar for young,

middle-aged, and elderly persons. Such activities include walking for pleasure, gardening

and yard work, dancing, and prescribed home exercise. For health promotion, dynamic

exercise of the large muscles for extended periods of time (30 to 60 minutes, three to six

times weekly) is recommended. This may include short periods of moderate intensity

(60% to 75% of maximal capacity activity, approximately 5 to 10 minutes) that total 30

minutes on most days (DeBusk, Stenestrand, Sheehan, & Haskell, 1990).

Physical activity may have risks as well as benefits, although risks are relatively

infrequent. Estimates of sudden cardiac death rates per 100,000 hours of exercise range

from 0 to 2 per 100,000 in general populations and from 0.13 per 100,000 to 0.61 per

100,000 in cardiac rehabilitation programs (Haskell, 1978; Koplan, Siscovick, &

Goldbaum, 1985; Van Camp & Peterson, 1986).

In addition to cardiac risks, resistance or strength training programs may present risk

as well. However, studies have demonstrated the cardiovascular safety of maximum

strength testing and training in healthy adults and low-risk cardiac patients (Gordon et al.,

1995). Falls and joint injuries are additional risks associated with physical activity

(especially in older women), but most of these injuries do not require medical treatment.

The incidence of such complications is less in those participating in low-impact activities

such as walking (Carroll et al., 1992; Pollock et al., 1991).

General Worksite Programs

The worksite is believed by many experts to be an optimal arena for making healthful

lifestyle changes, and there has been a tremendous growth in the number of fitness

programs offered where individuals work (Blair, Piserchia, Wilbur, & Crowder, 1986;









Cox, 1984; Edington, 1986; Gebhardt & Crump, 1990). Worksites are considered to be a

key channel for the delivery of interventions designed to reduce chronic disease among

adult populations (Abrams, 1991; Abrams et al.1994; Heimendinger et al.1990; Glasgow

et al., 1995). Worksites provide researchers with access to over 60% of adults in the

United States (US Department of Labor, 1992), as well as to diverse populations in terms

of race/ethnicity, gender, age, and health status.

According to a meta-analysis of 26 worksite exercise and health promotion programs,

adequate participation in worksite programs can reduce absenteeism and employee

turnover, and result in increased productivity. In addition, this meta-analysis suggests

that worksite wellness programs can lead to decreases in body fat and increases in aerobic

power, muscle strength, and flexibility, and enhance mood (Shepard, 1999). Baseline

assessment of an employee's health status can be performed at a relatively low cost and

should include an assessment of physical conditioning. Public health interventions in the

workplace have resulted in an increase in vigorous physical activity by participating

employees that is associated with increases in objective measurements of physical

conditioning (Blair, Piserchia, Wilbur, & Crowder, 1986). As healthcare costs continue

to increase, these programs will become more attractive to both small and large

businesses.

Exercise has received considerable attention at the worksite in recent years because of

its association with reduced risk for cardiovascular and musculoskeletal diseases, obesity,

and metal health problems, and its potential for minimizing the negative effects of such

chronic conditions as diabetes, osteoporosis, and back pain (Bouchard, Shepard,

Stephens, Sutton, & McPherson, 1990; Duncan et al., 1985; Harris, Caspersen, DeFries,









& Estes, 1989; Melby and Hyner, 1988; Powell, Caspersen, Koplan, & Ford, 1989;

Taylor, Sallis, & Needle, 1985). As mentioned earlier, one very plausible approach to

increasing adult fitness is to offer worksite fitness programs, as working adults spend a

major part of their day in the work environment (USDHHS, 1991). Support for physical

fitness and wellness programs in the workplace comes from both the private and public

sectors (Gebhardt & Crump, 1990; USDHHS, 1991; U.S. Office of Personnel

Management, 1991).

The proportion of worksites offering physical fitness programs has increased over

time. For instance, the most recent, 1999 National Worksite Health Promotion Survey

(AWHP, USDHHS, & Mercer, 1999) reported that the prevalence of physical activity

programs, with the intention of changing behavior, increased 36% since 1992 (in 1992

42% of worksites had exercise/fitness activities). The number of worksites offering

physical fitness programs is likely to continue increasing, as occupational settings are

identified as a target for employer sponsored physical activity and fitness programs in the

health objectives for the nation (USDHHS, 1991).

Worksite research suggests that improved physical fitness results in fewer worker

injuries, fewer absences from work due to illness, and increased worker productivity

(Bowne, Russell, Morgan, Optenberg, & Clarke, 1984; Cady, Bischoff, O'Connell,

Thomas, & Allen, 1979; Hilyer, Brown, Sirles, & Peoples, 1990; Lynch, Golaszewski,

Clearie, Snow, & Vickery, 1990; Shepard, Cox, & Corey, 1981; Sirles, Brown, & Hilyer,

1991). A number of worksite-based studies have been conducted on several key risk

factors for chronic disease which included physical activity (Blake et al., 1996; Blair,

Piserchia, et al., 1986; King, Carl, Birkel, & Haskell, 1988). Although there have been









varying outcomes from the single risk factor studies, overall any intervention effects that

have occurred have typically been quite small (Emmons, Linnan, Shadel, Marcus, &

Abrams, 1999). To date, however, there have been fewer randomized trials of worksite

health promotion that targeted physical activity (Abrams, 1991; Abrams et al., 1994;

Heimendinger et al., 1990).

Results show that about 30% of employees within an organization will actually start

participating when a program is offered to them (Oldridge, 1984). Futhermore, studies

show that 3 to 6 months after the start, only 50-60% of the original participants are still

participating in the program (Dishman, 1988; Marcus et al., 1992; Oldrige, 1984).

Adherence is important because lasting benefits of a worksite fitness program are more

likely when workers participate in the exercise program over time. However, as in other

settings, adherence to exercise at the worksite has been a major problem. It is clear that

alternative approaches to the adherence problem are needed (Robison et al., 1992).

Programs that incorporate strategies to increase exercise adherence will likely be more

successful in terms of worker benefit and employer benefit than those without strategies

(Blue & Conrad, 1995).

It is important to note that many weaknesses and limitations exist in the above

mentioned and other (not mentioned) worksite studies which prove challenging when

interpreting results. For example, many of these studies had small sample sizes, had a

minority of health conscious employees participating in the programs, and had biased

evaluators. These are important considerations in research design and implementation

when conducting worksite health promotion studies, hence, few worksite health









promotion studies to date provide clear results with implications to worksite health

promotion program delivery.

Worksite Programs in Schools

In the midst of the ongoing development and implementation of the Comprehensive

School Health Program (Allensworth & Kolbe, 1987), schools are ideal settings for

worksite health promotion programs because they have facilities and professional

resources required to develop and implement the program, including pupil service

professionals, the food service staff, and as health education, physical education, and

home economics teachers. Schoolsite wellness programs have been shown to reduce

weight, body fat, systolic and diastolic blood pressure, anxiety, depression, and smoking,

and to increase exercise and consumption of a more balanced diet.

Schoolsite health promotion programs have decreased absenteeism, health care claim

costs, and the need for substitute teachers, and have improved teacher morale and

productivity (Allensworth & Kolbe, 1987). For example, the Institute for Aerobics

Research (Blair, Collingwood, et al., 1984) assessed the impact of a health promotion

program among four schools on health behaviors, general well-being, job satisfaction,

stress management, and self-concept. The study's conclusions indicated that health

promotion is feasible in a school setting, and significant and important changes occur in

teachers exposed to health promotion programs (Blair, Collingwood, et al., 1984).

Other more recent, successful schoolsite wellness programs have included elements

such as personalized counseling, lipid assessment, and on-site exercise facilities, as well

as modification of the school food service (Glasgow, McCaul, & Fisher, 1993; Lovato &

Green, 1990; Masey, Gilmarc, & Kronenfeld, 1988; Wong, Bauman, & Koch, 1996). In









addition, Allegrante and Michela (1990) implemented a comprehensive Health

Enhancement Program (HEP) at ten New York City schools and found that the HEP had

a significant impact on the morale of teachers and that teachers gave more favorable

ratings of school quality and climate following the implementation of the program. The

researchers concluded that the study provided empirical support for the potential that

school-based workplace health promotion programs have the ability to enhance morale,

improve individual and organizational well-being, and help to meet health-related needs

of teachers and others who work in schools (Allegrante & Michela, 1990).

These literature findings support the notion that schools are viable worksites for

employee health promotion programs, including exercise promotion, and can potentiate

multiple positive outcomes, health-related and otherwise.

Self Efficacy

As presented in Chapter 1, whereas it appears that no single variable solely determines

adherence to either prescribed or self-initiated exercise regimens (Sallis & Hovell, 1990),

self-efficacy (Bandura, 1977, 1986) has been consistently identified as playing an

important role in health (O'Leary, 1985) and exercise behaviors (McAuley, 1992;

McAuley & Jacobson, 1991; Sallis et al., 1996; Sallis and Hovell, 1990). There is a

growing consensus that self-efficacy is among the most important and modifiable

predictors of physical activity behavior (Pate et al., 1995; Sallis, Hovell, Hofstetter, &

Barrington, 1992; USDHHS, 1996). Individuals' beliefs regarding perceived capabilities

in particularized domains are theorized to influence choice of activity, effort expenditure,

and persistence in the face of adversity. Furthermore, self-efficacy also is proposed to

influence thought patterns and emotional reactions (Duncan & McAuley, 1993).









Bandura's self-efficacy theory is a social cognitive model of behavioral causation

which posits that behavior, physiological and cognitive factors, and environmental

influences all function as interacting determinants of one another (Bandura, 1986). This

theory of reciprocal determinism views healthy functioning as being determined by

correspondent influence among the individual's physiological states, behavior, cognition,

and the environment. Self-efficacy cognitions have consistently been shown to be

important determinants of physical activity and exercise behavior as well as social,

clinical, and health-related behaviors (Bandura, 1986; McAuley, 1992; O'Leary, 1985).

As stated in Chapter 1, it is important to realize that self-efficacy is not concerned with

individual skills but, rather, with the judgements of what an individual can do with the

skills he or she possesses.

Individuals with high self-efficacy expectations tend to approach more challenging

tasks, put forth more effort, and persist longer in the face of aversive stimuli. When

faced with stressful stimuli, low efficacious individuals tend to give up, attribute failure

internally, and experience greater anxiety or depression (Bandura, 1982). Expectations of

personal self-efficacy are culled from four major sources of information: mastery

accomplishments, social modeling, social persuasion, and physiological states. Mastery

accomplishments are the most dependable and influential sources of efficacy information

with a history of previous successes facilitating efficacy expectations, whereas previous

failures will result in lowered perceptions of personal efficacy. Social modeling is a

source of efficacy information derived through observation or imaging others engaging in

the task to be performed. Social persuasion is a commonly used technique to bolster

personal efficacy, but is less powerful than information based on personal









accomplishments. Finally, physiological states are postulated to affect behavior through

the cognitive evaluation (efficacy expectations) of the information conveyed by the

anxiety arousal, fatigue, and muscular strain and tension. That is, somatic sensations are

often interpreted as inability to successfully carry out a course of action.

Self-efficacy theory (Bandura, 1986) has generated an enormous literature in multiple

domains of behavioral functioning, including physical activity as a health-promoting

behavior. Because exercise is a complex behavior, which appears particularly difficult

for many individuals to change, it is of little surprise that self-regulatory skills have been

consistently implicated in successful adoption and maintenance of this behavior. The

presence of a robust sense of self-efficacy and the development and nurturing of skills

and strategies to enhance such cognitions, continually have been identified as

determinants of physical activity in acute bouts of exercise in laboratory settings (Ewart,

Taylor, Reese, & DeBusk, 1983; Ewart et al., 1986; McAuley & Courneya, 1992), in

long-term exercise participation (Ewart, Stewart, Gillian, Keleman, 1992; Garcia & King,

1991; McAuley, 1992), and in larger survey population studies (Sallis et al., 1986, 1989).

Self-Efficacy and Exercise Adherence

Exercise self-efficacy is the conviction that one can successfully engage in physical

activity (McAuley, Lox, & Duncan, 1993). Self-efficacy theory predicts that highly self-

efficacious individuals are more likely to adopt or engage in a greater number of like

behaviors than are their counterparts whose personal efficacy has been impaired

(Bandura, 1986). Where exercise is concerned, those who perceive themselves to be

more efficacious with respect to their physical capabilities are more likely to adopt and

maintain a lifestyle in which exercise plays an important role.









Several studies exist that document the role-played by perceived efficacy in adherence

to prescribed exercise programs and maintenance of activity post-program termination in

normal populations. Although not without their individual flaws, they do provide some

support for the contention that self-efficacy influences exercise behavior. Desharnais,

Bouillon, and Godin (1986) in predicting adherence to an 11-week adult exercise

program demonstrated self-efficacy to be more capable of discriminating between

subjects classified as adherers and those classified as dropouts. Similarly, Corbin, Lauric,

Gruger, and Smiley (1984) reported that general self-confidence in sport and physical

activity influenced commitment and involvement in physical activity.

Long (1984; 1985) and Long and Haney (1988) employed various cognitive and

behavioral modalities to influence self-efficacy and reduce stress in females and to

further examine the role played by efficacy in adherence to a jogging program for

community males and females. In this study, however, both stress inoculation training

and aerobic activity were reported to have increased self-efficacy significantly but not

differentially (Long, 1984, 1985). Dzewaltowski (1989) and Dzewaltowski, Noble, &

Shaw (1990) reported data that compare the relative merits of self-efficacy theory and

attitudinal models of behavior change, specifically, the theory of reasoned action

(Fishbein & Ajzen, 1975) and Ajzen's (1985) theory of planned behavior, in explaining

physical activity behaviors in college undergraduates. In both studies, self-efficacy was a

significant unique predictor of exercise behavior. The comparison of the differing

theoretical approaches revealed efficacy rather than intention to be implicated in the

prediction of physical activity when each variable was statistically controlled.









Clearly, the self-efficacy approach to the prediction of physical activity as a health-

promoting behavior is governed by situationally dependent information processing.

Dishman, Ickes, and Morgan (1980) espoused a more traitlike perspective, in which the

more general dispositional characteristic of self-motivation is proposed as an important

component of activity participation. Supporting Heiby, Onorato, and Sate (1987) and

refuting Weber and Wertheim (1989) evidence for this latter proposition exists in the

literature, and two studies have specifically contrasted the relative utility of the self-

motivation and self-efficacy approaches (Garcia & King, 1991; McAuley & Jacobson,

1991). In a clinical trial involving sedentary, healthy, middle-aged males and females,

Garcia and King (1991) reported efficacy cognitions but not self-motivation to be

positively related to exercise adherence at 6 and 12 months. Moreover, more proximate

aspects of the exercise experience (e.g. exercise, enjoyment, and convenience) did not

account for significant variance beyond that accounted for by self-efficacy. McAuley

and Jacobson (1991) reported similar findings in a study of formerly sedentary, middle-

aged females engaged in aerobic activity. Self-efficacy rather than self-motivation

predicted both in-class exercise participation and activity outside of class.

A large prospective study (McAuley, 1992, 1993), of sedentary males and females,

tested the hypothesis that one would expect efficacy expectations to be more influential

in those stages of exercise participation where demands of continued adherence are

greater (e.g. early stages of adoption, resumption of activity following dropout, injury).

In the first report of this study, path analysis revealed self-efficacy and body fat to be

predictive of exercise frequency and intensity at the midpoint of a 5-month program, but

exercise participation up to that point was the only predictor of adherence over the









remaining period of the program. Such findings are consistent with the perspective that

cognitive control systems play their most important role in the acquisition of behavioral

proficiencies (Bandura, 1989). When behaviors are less demanding and more easily

engaged in (in this case, beyond adoption and adaptation), cognitive control systems give

way to lower control systems (Bandura & Wood, 1989). Clearly, different mechanisms

take on differing degrees of importance at various stages of the exercise process.

In a follow-up study (McAuley, 1993), the self-efficacy-exercise participation

relationship was examined in a more demanding context continued exercise

maintenance following termination of the program. Participants were contacted 4 months

after program completion and interviewed by telephone and surveyed by mail as to their

exercise participation patterns since program termination. Hierarchical regression

analyses indicated that self-efficacy predicted 12.5% of the unique variance in continued

exercise participation and shared a further 14% of the variance with physiological

(VO2max) and behavioral (past exercise frequency and intensity) parameters. However,

only efficacy was a significant individual predictor of behavior. Thus, it appears that

when situations or behaviors become more demanding, efficacy cognitions assume a

more important role. Once again, it is clear that diverse parameters take on varying

degrees of influence at different stages of the exercise process (Dishman, 1990;

McAuley, 1992; Sallis & Hovell, 1990).

Numerous studies by Ewart and colleagues (1983; 1986) as well as McAuley and

others (1991; 1992) showed efficacy to be related to acute bouts of physical activity and

the responses associated with such activity. Cardiovascular and muscular efficacy were

related to aerobic capacity (especially VO2max), muscular strength and endurance, and









cardiac responses in middle-aged males and females (McAuley et al., 1991). In addition,

males reported significantly greater perceptions of efficacy than females at the onset of

an exercise program. However, females following a 20-week exercise program became

as efficacious, and in some domains, more efficacious than their male counterparts.

McAuley and Courneya (1992) detailed the effects of pre-existing exercise efficacy on

perceived exertion and affective responses during graded exercise testing. More

efficacious subjects perceived the exercise bout as less effortful and reported more

positive affective responses. In summary, the studies reported in this section provide

consistent support for the mediational role played by the perceptions of personal efficacy

in predicting the adoption of, and adherence to, exercise regimens.

Social Support

Social support appears to be an important determinant of success in changing health

habits. Social support has been linked to a number of health outcomes, including

adherence to medical regimens (Wallston, Alagna, DeVellis, & DeVellis, 1983) and

success in smoking cessation (Mermelstein, McIntyre, & Lichtenstein, 1983), although

the findings have not always been consistent (Malott, Glasgow, O'Neill, & Klesges,

1984). Social support is one social cognitive mechanism, which has been implied in the

maintenance of various health-promoting regimens. Although definitions of support vary

considerably, the assumption underlying the various models and empirical investigations

of this phenomenon is that supported individuals are physically and emotionally healthier

than non-supported individuals. Social support also has been shown to have both direct

and indirect beneficial effects on measures of psychological well being and self-reported

symptoms (Cohen & Willis, 1985; Cohen, 1988). Prospective studies also link a lack of









social support to premature mortality (Cohen, 1988). Numerous studies have identified

social support in involvement in regular physical activity (Dunbar, Marshall, & Hovell,

1979; Heinzelman, 1973; King & Frederiken, 1984; Wankel, 1984, 1985).

Social support is defined as those activities performed by one individual that assist

another individual in moving toward a desired goal (Caplan, Robinson, French, Caldwell,

& Shinn, 1976). It is a dynamic process in which sources of social support considered

important by the individual adopts a specific health behavior and moves toward the

desired goal. There are various dimensions of general social support, including (a) social

network or the existence or quantity and structure of social relationships and (b) the

functional content of relationships, or behaviors that one person performs in support of

another (Caplan et al., 1976). Cultrona and Russell (1987) have argued that Weiss'

(1974) model of social provisions incorporates all of the major components proposed by

other theorists plus one additional component. The six social support provisions

proposed by Weiss are guidance (advice or information), reliable alliance (material

assistance), reassurance of worth (recognition of competence and value), attachment

(emotional closeness), social integration ( a sense of belonging to a group that shares

similar interests and concerns), and opportunity for nurturance (a sense of belonging to a

group that shares similar interests and concerns), and opportunity for nuturance (the sense

that others rely upon one for personal well-being).

Although these dimensions of general social support may enhance a person's ability to

attain individual goals, they may be very different from the support necessary to adopt

and maintain specific health behaviors such as exercise. Social network encompasses

structural (e.g., type, number, density, proximity) and interactional (frequency, durability,









and intensity) aspects of social relations whereas social support focuses on behavioral or

functional aspects of social relations (i.e., the provision of supportive behavior). Thus,

social support is distinguished from social network in that the latter refers only to the

linkages between people that may or may not provide social support (Israel & Schurman,

1990).

Social Support and Exercise Adherence

There is evidence to suggest that exercisers with supportive spouses are more likely to

continue their exercise programs (Dishman, Sallis, & Orenstein, 1985; Martin & Dubbert,

1982). Several weight-loss studies indicate that spousal support and participation in

treatment enhances weight loss (Brownell & Stunkard, 1981; Dubbert & Wilson, 1984;

Murphy et al., 1982; Rosenthal, Allen, & Winter, 1980). Wallace, Raglin, & Jastremski

(1995) found that the higher monthly attendance to exercise sessions of married couples

may have been enhanced by social support or camaraderie provided by an exercising

spouse rather than the individual factor of self-motivation. Other investigations have

suggested that social support specific to exercise may be a better indicator of exercise

adherence than general perceptions of support among both women and men (Oka, King,

& Young, 1995).

A number of studies in the extant literature suggest that gender is an important

influence on several aspects of interpersonal relationships relevant to the support process

(Antonucci & House, 1983; Lowenthal & Haven, 1968; Stokes & Wilson, 1984).

Antonucci and House (1983) and Fisher and Bishop (1986), suggest that different types

of support influence men and women. Henderson, Byrne, Duncan-Jones, Scott, and

Adcock (1980) demonstrated that social integration produced a stress-buffering effect for









males, whereas emotional support analogous to reassurance of worth served as a

buffering effect for females.

Results of some studies suggest that although the social environment is important in

influencing exercise participation in both men and women, it may play a more significant

role in determining exercise participation in women (Wankel, 1984). For instance, in a

study of adolescent males and females, social influence was an important predictor of

physical activity in females, but not males (Reynolds et al., 1990). Sallis et al. (1989)

found family influence was an important determinant of exercise participation in young

men and older women only. In a later study by Sallis, Hovell, Hofstetter, and Barrington

(1992), support of family and friends was found to be consistently associated with

exercise participation.

Since social support has been identified to be a major determinant of compliance and

adherence (Dishman, Sallis, & Ornstein, 1985), the opportunity for social support from

coworkers is a critical feature of many health promotion programs at the worksite

(Sorensen & Pechacek, 1987; Jose & Anderson, 1991). If social support at work is

associated with employee health risk and predictive of subsequent behavior change

toward healthier lifestyles, worksite health promotion interventions could be improved by

including social support at work into the design and delivery of programs. Hence, both

social support and self-efficacy need to be manipulated into employee exercise promotion

programs for maximum impact on adherence.

The Self-Efficacy/Social Support Relationship to Exercise Adherence

Whereas social support infers the influence of others on an individual's coping

responses, self-efficacy is concerned with the role played by personal or self-referent









resources in adaptive behavior. In many ways, the reception and effects of social support

may well be mediated by the degree of efficacy one is perceived to possess. In essence,

one might argue that social support is delivered based upon the beliefs about what the

recipient is capable of accomplishing. Evidence to support this interdependence of social

support and self-efficacy is provided by studies in a few strikingly different

environments. Cutrona and Troutman (1986) demonstrated that perceived social support

led to reduced postpartum depression in adolescent mothers through the mediation of

parenting self-efficacy. Holahan and Holahan (1987) found that self-efficacy beliefs

functioned indirectly through social support, in the form of social provisions, in the

alleviation of depression in elderly adults. Taylor, Bandura, Ewart, Miller, and DeBusk

(1985) reported that spouses of early myocardial infarction patients who believed their

husbands to have greater cardiac and coping efficacy were more likely to encourage them

to actively pursue rehabilitation and a normal life.

It appears, therefore, that social support and perceived self-efficacy are intimately

related rather than orthogonal influences of health behavior (Ducan & McAuley, 1993).

That is, the examination of the total resources available to individuals in health-

promoting activities such as exercise and physical activity may be more informative in

understanding these behaviors than mere reliance on a single construct (Duncan, 1989).

Specifically, high levels of social support for a given endeavor are thought to boost a

person' level of self-efficacy for that endeavor (Courneya & McAuley, 1995).

Bandura (1986) predicted that self-efficacy operates as a cognitive mediator linking

psychosocial influences to various health promoting behaviors. Duncan and McAuley's

(1993) finding that social support failed to influence exercise behaviors directly, but did









so indirectly, supports the contention that self-efficacy may be an important mediating

variable, when explaining the effects of various provisions of social relationships on such

health promoting behaviors as regular exercise. These researchers found that the

relationship between social support and exercise adherence was mediated by self-

efficacy. Consistent with the theoretical arguments from Bandura (1986) and empirical

data from Duncan and McAuley (1993), social support influenced perceived behavioral

control in a study conducted by Courneya and McAuley (1995). It was found that the

higher the perceived social support, the higher the perceived control over attending a

regular exercise class. Consequently, these researchers concluded that any intervention

aimed at developing social support in the exercise domain is likely to enhance

perceptions of control over exercise.

By providing continued support and enhancing the individual's beliefs in personal

capabilities, social network members may help the individual avoid the downward spiral

which ultimately leads to a relapse of health-impairing behaviors (Willis, 1985).

Therefore, provisions of social relationships such as guidance, attachment, and social

integration may act in concert to produce what is in effect a motivationally based support,

which, in turn, bolsters the individual's self-percepts of efficacy.

As presented in Chapter 1, O'Donnell's Model of Health Promotion Behavior

includes both social support and self-efficacy within the later portion of the model that

includes components from the Theory of Planned Behavior (see figure 1, Chapter 1).

Within O'Donnell's model, self-efficacy beliefs play a central role in predicting both

intentions and behavior. The model also incorporates the construct of social support as an

important determinant of whether one's intentions get translated into action as well as









social barriers and prior experience. This model depicts the self-efficacy/social support

relationship in health education interventions designed to promote health behaviors such

as exercise adherence and has not been tested prior to this study.

Use of Behavioral Techniques to Enhance Social Support and Self Efficacy
in Health Education Interventions

As mentioned in Chapter I, Bandura (1986) poses several methods in which to

increase self-efficacy, mainly, authentic mastery, social modeling, social support and

modifying physiological reactions. In terms of health education application, O'Donnell

(1992) incorporates self-efficacy and social support into a model for health behavior

change. These theoretical concepts increasing self-efficacy and social support within a

health education intervention model to promote health behavior change sets a direction

for behavior techniques to be implemented within health education interventions

promoting health behavior change.

Behavioral management techniques such as self-monitoring of exercise progress,

improvement charts, personal exercise diaries, and goal-setting strategies may provide

evidence of personal mastery and consequently increase perceptions of personal efficacy

(Duncan & McAuley, 1993). Also, some researchers feel that the key to good exercise

programs, specifically, is to offer a variety of opportunities from which to choose; this

variety enhances not only participant health but also other positive emotions (Baun &

Bernacki, 1988). Thompson and Wankel (1980) have shown that participants who

perceive they have a choice in fitness activities will adhere better to exercise programs.

Hence, some of the key behavioral practices to increasing self-efficacy and social

support for health behavior change, are behavioral contracting and goal setting. Once a

program participant has tentatively committed to a course of action, the participant









requires additional guidance, structure, and support to make and carry through on an

action plan for health behavior change. One effective means for doing this through a

written contract specifying which behavior(s) the participant is going to perform during a

specified time-period. The contract also specifies the rewards that the client will receive

if the behaviors are successfully performed.

The most successful contracts are those whose details are openly negotiated between a

program participant and a counselor (or health educator). This is preferable to foisting a

ready-made contract on the participant (Boehm, 1989; Steckel, 1982). Having the

participant participate actively in the contracting process fosters a high sense of

behavioral control and commitment to the health behavior goals. Also it helps ensure that

the reward given (contingent upon the performance of the behavior) is indeed meaningful

to the client and will act as a reinforcement for future behavior. Built into this approach

is some means of the participant and health educator monitoring of the participant's

behavior. The acts of simply monitoring, recording, and reporting one's behavior

constitute a very powerful set of behavioral strategies for bringing about desired health

behavior change (Wallston, 1994).

Through behavioral contracting, the other key tenet to increasing self-efficacy and

social support to change health behavior is behavioral goal setting. Two main

considerations in behavioral goal setting is incorporating non-health-related goals and

allowing goals to change over time. Wankel (1985) found that although health related

goals (e.g., lose weight, relieve tension and anxiety, prevent cardiovascular disease,

improve physical fitness) were reported to be the most important by all participants, they

did not differentiate between continuing program participants and dropouts. Non-health









related goals such as developing recreational skills, developing social relationships, going

out with friends, releasing competitive drive and satisfying one's curiosity, on the other

hand, although rated less important, did significantly distinguish between adherence and

drop-outs. This suggests that secondary, non-health related goals to be more readily

attainable and more useful for facilitating continued involvement. Bandura (1987) has

recognized the general importance of short-range goals to motivational behavior change

in his social cognitive theory.

A second consideration with respect to behavioral goals is that they may change over

time. The goals for maintaining participation may not be seen as those for initially

undertaking a program (Oldridge, 1982). In an early study of factors affecting exercise

involvement, Heinzelmann and Bagley (1970) found that the most important reasons for

joining an adult fitness program were a desire to feel better and healthier and a concern

about reducing the probability of having a heart attack. The major reasons given for

staying in the program at a later time, however, were the program's organization and

leadership (31%), recreational games (29%), and social aspects or camaraderie (26%).

Calfas, Sallis, Oldenburg, and French (1997) implemented an intervention designed to

increase self-efficacy and social support using a protocol emphasizing specific actions

such as setting goals, recruiting social support, and gradually increasing activity levels,

which are reflected in the behavioral processes. Their results suggested that participants

who received structured counseling were making some of the changes that were targeted

in the intervention. These researchers concluded that their results provided preliminary

support for a conclusion that activity-specific social support and self-efficacy may play a









role in influencing changes in physical activity (Calfas, Sallis, Oldenburg, & French,

1997).

Summary

It is apparent from the research presented in the introduction, the benefits of exercise

and exercise adherence sections of this chapter, that the promotion of exercise or physical

activity is a priority health behavior that must be taken seriously in order to make an

impact on our nation's overall health. In addition, from the research presented in the

sections dealing with self-efficacy, social support, and the relationship of self-efficacy

and social support, these factors can have a dramatic impact on the adherence to exercise

regimens if integrated effectively into exercise promotion programs. As seen in the

behavioral techniques section, the most promising methods to integrating these affective

factors into exercise programs is through behavioral processes such as goal setting,

behavioral contracting and one-on-one consulting. Health education can provide this link

from exercise benefit knowledge to actually engaging in regular exercise as adoption of a

way of life.

Health educators are trained to use behavioral techniques such as those mentioned to

assist others to adopt and adhere to health behaviors. As the research suggests,

behavioral techniques designed to increase self-efficacy and social support can make a

significant impact on exercise adherence. However, no research thus far has tested these

behavioral techniques in the improvement of exercise adherence in worksite exercise

promotion programs. Many worksites, more than ever, are offering exercise programs

however, few are designed to support participation and adherence.









It is therefore the intent of this study, to test a health education intervention designed to

increase self-efficacy and social support through behavioral techniques on adherence to a

worksite exercise promotion program. Specifically, differences in exercise adherence,

self-efficacy and types of social support between employee groups receiving and not

receiving a health education exercise promotion intervention will be assessed to test the

intervention. It is the hope that this health education intervention can demonstrate

effectiveness in increasing exercise adherence in worksite exercise promotion programs

and serve as a model for organizations to use in the future to impact participation and

adherence.
















CHAPTER 3
PROCEDURES AND METHODS OF ANALYSIS

This chapter will cover the population and sample, setting, research design and data

sources, treatment, instruments, and statistical analysis for the present study. The

information in these sections will describe all procedures, methods, and analysis for the

study.

Population and Sample


The population under observation is elementary school employees in one North

Central Florida school district. The gender and race demographics for the school district

are 24% male and 76% female and 70% Caucasian and 28% African-American,

respectively. The age demographics for the district are 13% under 30 years, 21% between

31 and 40 years, 32% between 41 and 50 years, 30% between 51 and 60 years, and 5%

over 61 years. Finally, the job classification demographics for the district are 47%

classified as teachers and 53% classified as career service.



Settings


This study was conducted at four participating elementary school worksites in

Gainesville, FL. The study recruitment health fairs, pre and post program assessments,









and individualized health education sessions were conducted in school classrooms and

media rooms after school hours. Each of these are described in more detail in the

following sections.

Form (Design) and Sources of Data


This study will be a Quasi-Experimental One-to-One Matching Repeated Measures

design. Four elementary school worksites from Alachua County have been designated as

treatment and control sites. The treatment sites (worksites A, B & C) received the health

education intervention with 50 total employees participating in the intervention for the

intervention group analysis. Site D served as a control site having only pre and post

assessment measures performed. Approximately 15 employees participated from the

control site. A diagram depicting the stages of methodology and study design is shown

below followed by specific information for each stage (see Figure 2).

Health
Sites Education
A, B &C Intervention
(Treatment] (Self-Efficacy/ -..
Behavioral Mid-Program
Site -- ^ ^, 1Techniques] Self-Efficacy P P
Subject Pre-Program and Social -- Assessment
Re ent Assessment xecise Program SuppoFt Evaluation
Recruitment I Data Evaluation
S traditional Collection
Program
STurn in
Site D exercise /
[Control]) erification
sheets only)

Figure 2
Methodology Design


Site and Subject Recruitment

Telephone calls and follow-up letters were sent to five Alachua county elementary

schools for potential site recruitment. The four sites that volunteered to participate had









no formal exercise program and had available locations for health education

consultations. These schools also had adequate assessment rooms in which to administer

pre and post program assessments. The assessment rooms are deemed to be adequate if

they conform to the most current ACSM guidelines, including appropriate room

temperature, humidity, ventilation, and adequate lighting (ACSM, 1995). After the

treatment and control sites were selected, subjects were recruited for participation via a

mini health fair consisting of blood pressure measurement, completion of an Institutional

Review Board consent form (see Appendix A), Pre-Participation Questionnaire (see

Appendix B), fitness assessment instruction and other activities. All subjects received

individual counseling to discuss their blood pressure screenings and to explain the

purpose and nature of the exercise intervention and/or assessments to follow (depending

on treatment or control site). Program flyers also were posted and distributed in staff

boxes to advertise and recruit participants for the program.

All subjects interested in participating in the exercise program were required to meet

the screening criteria of no history of cardiac disease or orthopedic injuries that would

preclude participation in an exercise program. Prescreening included completion of the

PPQ, resting heart rate and blood pressure measurements. Subjects were classified as

Class A- apparently healthy exercise participants within the new American Heart

Association and American College of Sports Medicine guidelines (AHA/ACSM, 1998).

Subjects were not excluded from participating based on employment classification,

gender, race, age, or socioeconomic status.









Pre-intervention assessment

All eligible subjects participated in an individual preprogram assessment session.

After checking that prior consent and PPQ information had been given, subjects were

asked to complete the CDSII and social support instruments (see Appendixes C & D).

After completion of the questionnaires, a 10 second baseline heart rate was taken

(multiplied by 6) followed by a 3 minute habituation period and recording of the resting

blood pressure (systolic and diastolic) using a sphygmomanometer and stethoscope.

Following baseline recordings, subjects' Anthropometric measurements including,

weight, height and body composition (as measured by a Lange skin caliper) were

obtained followed by the YMCA aerobic step test. These assessments were recorded

using assessment forms (see Appendix E). However, an exception was made concerning

those participants who had special conditions) with regards to the step test. Those

participants with an orthopedic injury, taking certain medications or having certain other

risk factors that may contraindicate performing a submaximal step test deemed by the

Principal Investigator, were excluded from the step test portion of the assessment.

Basic exercise prescription

At the conclusion of the assessment, a generalized aerobic fitness plan, generated from

the step test, was shared with the subject. This plan required all subjects to perform

aerobic activity with a frequency of three times per week for a minimum duration of 20

minutes per session. Exercise intensity was determined by calculating a target heart rate

range (THRR) for each subject. The aerobic fitness plan was explained and an

assessment team member discussed the nature and benefits of aerobic exercise with each

participant. Exercise Science graduate students and seniors with proper fitness









assessment training served as assessment team members for the project. These students

had CPR certification and demonstrated competency in performing fitness assessments to

the Principal Investigator. The scores from the CDSII and Social Support for Exercise

Habits Scale and assessment measurements served as baseline sources of data.

Exercise Program

Before the start of the program, subjects had to attend one of a number of orientation

meetings held for specific instructions regarding the completion and procedures for the

exercise verification sheets. The on-site exercise program consisted of encouraging

subjects to adhere to the basic exercise prescription of performing three, 20-minute

sessions per week of aerobic activity within their THRR. Various types of aerobic

activities such as brisk walking, biking, aerobic dance and use of cardiovascular

equipment were suggested for participant participation. Participants were allowed to

exercise with each other and with others outside the program as desired. They also could

exercise at home, in the community, or on school grounds.

During the exercise program, participants turned in the monthly exercise verification

sheets (indicating frequency and duration of exercise) as an exercise adherence data

source (see Appendix F). Treatment site participants turned in their verification sheets to

their health educators at the monthly sessions as well as verbally reported their

adherence. Health promotion consultants turned in reports to the Program Coordinator

that were added to the participant profiles as an additional source of qualitative data. The

control site participants turned in their verification sheets at a designated location at each

school site.









Mid-intervention self-efficacy and social support assessment

At the third monthly session of the intervention, the health promotion consultants

administered the self-efficacy and social support measures to the treatment participants.

These instruments provide crucial data points, midway through the study. The goal is to

discern differences in social support relationships and levels of self-efficacy. The

program coordinator personally administered and collected the instruments from the

control site participants at their mid-program point.

Post-intervention assessment

At the end of the five-month exercise program, participants at all four sites were

individually reassessed on the self-efficacy, social, anthropometric, and aerobic fitness

measures (with those exceptions for step test).

Treatment

All treatment groups (sites A, B & C) received the health education intervention

which is designed to promote self efficacy towards exercise through mastery experience

and reinterpreting physical states (concepts described by Bandura). Both mastery

experience and monitoring physical states were accomplished through monthly sessions

between the individual subjects and a health promotion consultant during the course of

the exercise program period. These twenty to thirty minute "exercise strategy" sessions

were scheduled before, during, or after work hours at the school sites for employee

convenience and were run according to a "Personal Exercise Plan" or "PEP" (see

Appendix G).

For the purpose of mastery experience, realistic personal exercise-related goals were

set within the PEP (as a behavioral contract). Participants were asked to identify the types









of physical activity in which they liked to participate, specifying their level of enjoyment;

when and where they were active; and who supported their activity program.

Along with the encouraged overall goal of adhering to three, 20 minute aerobic

sessions per week, participants also were encouraged to set a more individualized,

personal exercise-related goal for their monthly contracts (i.e. weight or body fat loss,

improvement in aerobic capacity, stress reduction, etc). Participants were asked to be

specific and realistic when setting personal goals. As participants met their goals, they

were able to revise or set a new goal each month. Within the personal goal section of the

PEP, the health promotion consultant and participant could brainstorm barriers and

strategies for achieving personal goals.

Reinterpreting physical states also was accomplished through these monthly sessions

by teaching subjects to modify or reinterpret their impressions of their physiological

states during exercise. After goal setting and strategizing, the health promotion

consultant recorded and responded to the types of physical symptoms and changes

experienced by the subject within the PEP. This process served as a venue to teach

subjects to interpret gradual changes in physiological symptoms such as fatigue, muscle

tension, aches and pains, dry throat, and shortness of breath, as markers of improved

conditioning and not as a negative experience.

At the end of the PEP, spaces for names of people who would support the participant

and sign as witnesses on the PEP for each of the three weekly reported exercise bouts by

the participant through the verification sheets, were included. At least one verifier was

required for all participant PEP contracts; however, multiple verifiers were encouraged as

long as their names were included.









Once the subject and health promotion consultant agreed upon an overall physical

activity strategy, the health promotion consultant summarized the plan at the bottom of

the PEP, and both signed to indicate a commitment of attempting to achieve the goal for

the next month. The health promotion consultants also could refer participants to the

program website, which contained examples of moderate and vigorous activities to try

and suggestions for how to overcome common barriers. The subject was given four

monthly exercise verification sheets for the program and was asked to turn a log in each

month at the designated worksite location.

Instruments

Causal Dimension Scale (CDSII). The CDSII (McAuley, Duncan, & Russell, 1992)

assessed subjects' self-efficacy causal attributions for previous attrition from exercise

programs. This measure allowed respondents to provide their own open-ended

attribution for not engaging in regular exercise and then code their attribution along

causal dimensions. The CDSII is a recently revised version of Russell's (1982) Causal

Dimension Scale, a measure of how individuals perceive causes. This scale differs from

the original in that it comprises four rather than three self-efficacy causal dimensions,

locus of causality, stability, personal control, and external control. Twelve semantic

differential scales, with three items representing each of the dimensions, comprise the

CDSII. Values can range from 1-9 with higher values representing attributions that are

more internal, stable and either personally or externally controllable. Several studies

(McAuley, Poag, Gleason, & Wraith, 1990; Duncan et al., 1993; and Duncan &

McAuley, 1993) have previously reported good internal consistency for all four

dimensions of the CDSII with standardized alpha coefficients of c= .820 for locus of









causality, a= .957 for stability, c= .956 for personal control, and a= .963 for external

control (Cronbach, 1951).

Social support. Separate scales for family and friend social support for exercise have

been found to have adequate reliability and validity (Sallis, Grossman, Pinski, Patterson,

& Nader, 1987). In the Social Support for Exercise Habits Scales, subjects rate the

frequency (1 indicating "none" through 5 indicating "very often") with which family and

friends support their physical activity in 17 different ways (e.g., "gave me encouragement

to stick with my exercise program"). These scales have undergone factor analysis along

with several methods to determine reliability and validity. The first five questions

comprise friend support for exercise (factor 1 or first subscale) with an Eigen value of

6.3. The last 12 questions comprise family support for exercise (factor 2 or second

subscale) with an Eigen value of 7.2. The test-retest reliabilities range from 0.55 to 0.86

and the alpha coefficients range from 0.61 to 0.91.

For purposes of this study, the instrument was revised and piloted to include a

coworkerr" social support section using the same questions for friend as previously used

for friend support. Gainesville area aerobics participants were used as pilot subjects to

determine item total correlations, response patterns and other feedback for the purposes

of making instrument revisions. The instrument was administered to approximately

eighty respondents containing the same five questions in categories of both friend and co-

worker support and the twelve family support items (see Appendix D). All items had

good response distributions with standard deviations ranging from 0.82 to 1.49. The

Cronbach's alpha was 0.92 and the Split half reliability coefficient was 0.93 for the









instrument, indicating excellent internal consistency and reliability for the instrument

overall.

Pre-Participation Questionnaire (PPQ). All participants were required to complete a

pre-participation questionnaire created for this study, which asked questions regarding

basic demographic information; medications; heart conditions; discomfort during

physical activity; self and family history of disease; hypertension; orthopedic problems;

other physical and medical conditions; and current activity level. This questionnaire

served as part of the screening measure for subject eligibility. As stated earlier, subjects

must have met the screening criteria of no history of cardiac disease or orthopedic

injuries that would preclude participation in an exercise program. Subjects' blood

pressure and resting heart rate measurements were taken along with completion of this

form. Subjects taking certain medications such as beta-blockers were not included in the

analysis of the data due to contraindications in estimating a THRR.

Cardiovascular fitness. The YMCA 3-Minute Step Test was used to estimate VO2max.

This submaximal test involved stepping up and down on a 12-inch high step, at a rate of

24-steps-per-minute (96 beats per minute) for 3 minutes, then immediately stopping and

having heart rate recorded by an assessment team member. Within 5 seconds the tester

counts the pulse with the stethoscope for 15 seconds. The subject can take her or his own

pulse at the same time by palpating the radial artery, providing a double check of the

count. Prior to the test, the assessment team member provides instruction in proper

stepping technique and demonstrates on the step. The participant "warms up" by taking

practice steps onto the step and is encouraged to stretch. A metronome was used to set

the 96 beats per minute cadence for the participants to step in time with the beat. The 15-









second count reflects the heart's ability to recover quickly, with a lower versus higher

count reflecting a better fitness level. The total 1-minute post exercise heart rate

(calculated from the 15 second count) is the score for the test and is converted to an

estimated VO2max by using a maximal oxygen uptake equation (McArdle, 1991) and

recorded.

In terms of validity, according to the American College of Sports Medicine (1998),

direct analysis of expired gases yields the most accurate determination of VO2max,

followed by maximal exercise testing. However, direct analysis is costly and time

consuming and maximal exercise testing requires participants to exercise to the point of

volitional fatigue. In lieu of these obstacles, submaximal testing was developed requiring

a steady-state heart rate attainment for each exercise work rate. Submaximal exercise

testing, though not as precise as maximal exercise testing, can still provide a reasonably

accurate reflection of an individual's fitness. If an individual is given repeated

submaximal exercise tests over a period of weeks and the heart rate response to a fixed

work rate decreases over time, it is likely that the individual's cardiorespiratory fitness

has improved, irrespective of the accuracy of the VO2max prediction (ACSM, 1998).

Anthropometric measures. Three anthropometric measures were obtained from each

subject height, weight (as measured by a calibrated mechanical eye-level physician

scale at each school) and a three-site body composition, as measured by a Lange skin

caliper. All fitness assessment procedures followed the ACSM 1995 guidelines and

procedures. The same testers were used for pre and post test assessments and the three

sites measured in the skinfold body composition test were the triceps, iliac crest (hip),

and thigh for women, and the pectoralis major, abdomen, and thigh for men.









In terms of validity, correlation coefficients between skinfolds and hydrostatically

determined body fatness have consistently ranged from .70 to .90 (AAHPERD, 1980). In

general, the inclusion of three skinfold sites in the regression equation produces a better

prediction (lower standard of error of estimate) of body density than fewer sites.

However, neither the feasibility nor accuracy is improved by using more than three sites

(Pollock & Jackson, 1984).

Statistical Analysis

As mentioned earlier, self-efficacy beliefs and social support for exercise data (using

CDSII and social support scale values) were collected before, during (midpoint of

intervention) and after the intervention. Collecting data at these three points determines

if exercise self-efficacy has changed and what types of social support have occurred over

the course of the intervention. Physical assessments (cardiovascular, weight, and

skinfold) also were measured before and after the intervention as additional outcome

measures of exercise adherence (i.e. adherers will have more improved physical attributes

and cardiovascular levels than non-adherers.) In addition, exercise adherence data was

collected throughout the entire intervention period to determine an exercise adherence

average for the treatment and control groups. Demographic data also was collected

through the Pre-Participation Questionnaire.

Data analysis included basic descriptive statistics of all treatment and control groups.

Statistical differences (p<.05) between the groups using mean average exercise adherence

scores were evaluated using group t-test statistics. A Repeated Measures ANOVA was

used to assess change between the three data collection points for self-efficacy and social

support. A dependent samples t-test was used to assess change in physical outcomes






59


within treatment and control groups. Pearson correlations between exercise history

related variables, self-efficacy and social support, and adherence measures were also

performed to investigate construct relationships. Multiple regression models were then

conducted to determine the independent contributions of self-efficacy and social support.















CHAPTER 4
RESULTS


Introduction

The primary purpose of this study was to assess differences in exercise adherence,

self-efficacy and types of social support between employee groups receiving and not

receiving a health education exercise promotion intervention. All results of this study

including descriptive results, frequency data, correlational analyses, analyses of variance,

and regression analyses will be described in this chapter. These results answer all

primary and secondary research questions posed in Chapter 1 as well as follow the

statistical analysis plan described in Chapter 3. All results will be discussed in Chapter 5.

Descriptive Results

Participant Demographics

Over 70 participants were initially assessed for this study, however only 58 actually

began the program and could be considered in the overall analyses. Fifteen participants

comprised the control group and 43 participants comprised the treatment group. The 15

employees who were initially assessed but decided not to participate in the program "the

non-starters" (treatment and control) were similar demographically to the program

participants. Program participant demographics included in the analyses were gender,

race, age, and job title or classification. Table I contains the program participant,

elementary school employee and overall school district employee demographics

(described in Chapter 1), providing a sample to population comparison.










Table I
Demographic Profile for Program Participants and Representative County


Treatment Control Elementary County
N Frequency N Frequency N Frequency N Frequency
Gender
Male 7 16.3% 0 0.00% 232 13.17% 998 23.85%
Female 36 84.7% 15 100.00% 1530 86.83% 3187 76.15%
Unreported 0 0.00%
Race
Caucasian 31 72.1% 13 86.67% 1270 72.08% 2928 69.96%
African-American 10 23.2% 1 6.67% 448 25.43% 1160 27.72%
Hispanic 0 0.00% 0 0.00% 35 1.99% 78 1.86%
Asian 1 2.3% 0 0.00% 8 0.45% 16 0.38%
Other 1 2.3% 0 0.00% 1 0.06% 3 0.07%
Unreported 0 0.00% 1 6.67%
Age
30 and under 3 7.0% 0 0.00% 370 19.08% 187 12.60%
31-40 9 20.9% 2 13.33% 356 18.36% 315 21.23%
41-50 20 46.5% 8 53.33% 671 34.61% 469 31.60%
51-60 9 20.9% 3 20.00% 443 22.85% 441 29.72%
61 and over 2 4.6% 1 6.67% 99 5.11% 72 4.85%
Unreported 0 0.00% 1 6.67%
Job Title
Teacher 31 72.1% 13 86.67% 918 52.10% 1963 46.91%
Career Service 11 25.6% 2 13.33% 844 47.90% 2222 53.09%
(Office & Custodial)
Office Staff only 7 16.3% 2 13.33%
Custodial only 4 9.3% 0 0.00%
Unreported 1 2.3%



PPQ Descriptive Data

The Pre-Participation Questionnaire (PPQ) administered to all program participants

prior to start of the program, included items regarding current exercise and past exercise

habits. Along with these items were other descriptive items regarding participation in

sports or leisure activities and types of activities commonly engaged. The percentages of

program participants in both treatment and control groups responding "yes" or "no" to

these items are shown in Table II. A frequency breakdown of types of activities reported

by participants prior to the program also are included. Of the "non-starters," 53.3%










reported that they were not currently exercising, 93.3% had not exercised in the past 6

months, and 26.7% had never exercised or engaged in sports or leisure activities.

Table II
Reported Participant Exercise and Sports/Leisure Status

Treatment Control
Variable N Frequency (43) N Frequency (15)
Currently exercising
Yes 15 34.8% 14 93.33%
No 26 60.4% 1 6.67%
Unreported 2 4.6% 0 0.00%
Exercised in past 6 months
Yes 20 46.5% 10 66.67%
No 22 51.2% 4 26.67%
Unreported 1 2.3% 1 6.67%
Have ever exercised
Yes 32 74.4% 10 66.67%
No 10 23.2% 5 33.33%
Unreported 1 2.3% 0 0.00%
Have ever participated in sports
or leisure
Yes 30 69.8% 11 73.33%
No 13 30.2% 4 26.67%
Types of activities from past
Walking 10 24.44% 9 60.00%
Weight training 12 26.67% 2 13.33%
Running / Jogging 7 15.56% 2 13.33%
Aerobics 6 13.33% 2 13.33%
Gym or fitness center 5 11.11% 2 13.33%
Swimming 1 2.22% 2 13.33%
Tennis 2 4.44% 1 6.67%
Cycling 1 2.22% 2 13.33%



Personal Exercise Plan data

Within the Personal Exercise Plan (PEP), treatment participants reported activities

they enjoyed doing, personal goals, strategies to achieve personal goals, barriers, and

strategies to overcome barriers with their health promotion consultant. A summation of

this information reported by all treatment group participants over the course of the

program is listed in Table III. Ninety-one percent of participants reported exercising at










home, 53% reported exercising in a gym or recreational center, and 35% reported

exercising at the worksite. In addition, 84% of participants reported exercising in the

evening and 26% reported exercising in the morning.

Causal Dimension Scale Reason List

Within the Causal Dimension Scale (CDSII) that all program participants were given

as a measure of self-efficacy towards exercise, participants were asked to list the main

reason as to why they are unable to exercise. A summation of these reasons is listed in

Table IV.

Table IV
Reported Reasons for Not Engaging in Regular Exercise Indicated on the CDSII


**** Poor time management/over scheduling
*** Lack of motivation/initiative
** Tired/lack of energy
Medical reason/illness
No reason
Do not enjoy exercise
Exercise is not a priority
Fear of injury
Lack of a routine
Weather
**** Over 60% of participants
*** Over 20% of Participants
** Over 15% of Participants
Over 10% of Participants



Self Efficacy and Social Support Results

Descriptive Statistics

Descriptive data for the self-efficacy and social support measures can be broken down

into means and standard deviations for scores on each measure. For self-efficacy, a total

self-efficacy score was computed for both treatment and control groups (a possible score

of "9" could be attained for each item) and the items were then summed. For social














Table III
Summation of Personal Exercise Plan (PEP) Information Reported by Participants


Activities Enjoyed Personal Strategies to Achieve Barriers Strategies to Overcome
Goals
Personal Goals Barriers


Aerobics
Basketball
Biking
* Bowling
Canoeing
Circuit training
Dancing
Exercise tape
Gardening
Golf
Fitness center
* Hiking
Jumping-rope
Racquet ball
Rollerblade
Rowing
Running
Stair climbing
Swimming
Tennis
*** Walking
Weights
Yoga/Stretching


Control diabetes
Improve cardiovascular health
Improve overall health
Improve self image
Increase endurance
Increase energy
* Increase lean body mass
Increase strength
Lower cholesterol
Maintain weight loss
* Make exercise a habit
Manage stress better
Try new exercises
** Weight loss


*** Reported by over 60% of participants
** Reported by over 40% of participants
* Reported by over 20% of participants


** Consistency
Do weight bearing exercise
Exercise in THR
* Exercise more often
Exercise with more intensity
Join a fitness center
Keep an exercise log
Maintain a positive attitude
Partner/peer motivation
Self motivation
Use time management skills
Vary exercise routine
Watch diet


Family obligations
Feeling stressed
* Medical reasons/Injury
Over eating
** Poor time-management
Self
** Feeling unmotivated/tired
Unexpected incidents
Unsupportive environment
Weather


Be less critical on self
Bring kids with them
Develop a back-up plan
Display reminders
Do exercise indoors
Do more indoor activity
Do weight-bearing exercise
Improve time-management
** Look in the mirror
Make exercise a priority
* Not procrastinating
Partner/peer motivation
* Personal rewards
Self motivation
* Watch diet










support, three subtotal scores for coworker, those living with participants, and those not

living with participants social support was computed (a possible score of "4" for each

item could be attained) and these items were then summed. A total social support score

also was computed for both groups (adding all three subtotals). The self-efficacy and

social support descriptive information for both treatment and control groups is shown in

Table V.

Table V
Self Efficacy and Social Support Descriptive Statistics

Self-Efficacy Social Support

Baseline BSE Total BSSCW BSSL BSSNL BSS Total

Treat Ctrl Treat Ctrl Treat Ctrl Treat Ctrl Treat Ctrl

67.3 64.5 6.9 4.3 14.8 15.7 5.2 7.4 26.2 27.4
x
SD 18.1 16.1 6.0 4.3 12.1 13.9 5.3 7.3 17.2 21.1

Midpoint MSE Total MSSCW MSSL MSSNL MSS Total

Treat Ctrl Treat Ctrl Treat Ctrl Treat Ctrl Treat Ctrl

73.9 62.9 8.3 6.6 15.9 19.8 6.0 7.4 30.2 33.8
x
SD 17.5 9.7 5.8 3.1 11.4 13.7 5.9 7.2 15.9 19.9

Post PSE Total PSSCW PSSL PSSNL PSS Total

Treat Ctrl Treat Ctrl Treat Ctrl Treat Ctrl Treat Ctrl

69.2 63.8 8.9 4.5 17.0 17.4 6.9 6.1 32.8 28.0
x
SD 13.0 11.6 6.0 4.1 12.0 14.8 6.7 7.0 19.0 18.9

(N) Treatment = 43
(N) Control = 15
BSE, MSE, and PSE = baseline, midpoint, and post program total self-efficacy
BSSCW, MSSCW, and PSSCW = baseline, midpoint, and post program social support from coworker
BSSL, MSSL, and PSSL = baseline, midpoint, and post program social support from those living with
participants
BSSNL, MSSNL, and PSSNL = baseline, midpoint, and post program social support from those not living
with participants
BSS, MSS, and PSS = baseline, midpoint, and post program total social support









Differences in Self-Efficacy and Social Support between Groups

An independent samples or group t-test statistic was used to compare self-efficacy and

social support scores between treatment and control groups using the means and standard

deviations shown in Table V. For total self-efficacy, the treatment groups had no

significant difference on total self-efficacy scores than the control group prior to the

program (baseline). However, at program mid-point, the treatment groups self-efficacy

scores were significantly higher (t = -2.74, DF=41.4, p=0.03) than the control group. At

program end (post), self-efficacy decreased in the treatment groups and increased slightly

in the control group. Thus, the treatment groups ended with non-significantly higher self-

efficacy than the control group.

For social support, three subscales of social support including coworker, those living

with participants and those not living with participants and total social support were

analyzed for differences between treatment and control groups. For pre-program

(baseline) coworker social support, the treatment groups had non-significantly higher

scores than the control group. At program midpoint, both treatment and control groups'

coworker social support increased but did not significantly differ. At program end (post),

the treatment groups coworker social support increased once again, however the control

groups scores decreased, ending with significantly higher (t=-2.9, DF=30.0, p=0.006)

coworker social support scores for the treatment groups versus the control group.

For social support from those living with participants (SSL), the treatment groups had

non-significantly lower support at baseline than the control group at baseline. At

program midpoint, both treatment and control groups had increased SSL, but did not

significantly differ. At program end, the treatment group had increased SSL scores again,









from midpoint however, the control group had decreased SSL. Both treatment and

control groups still did not significantly differ in SSL.

For social support from those not living with participants (SSNL), the treatment

groups had non-significantly lower support than the control group. At program midpoint,

the treatment groups had slightly increased SSNL from baseline however, the control

group support remained the same and did not differ significantly from the treatment

groups' scores. For program end, the treatment group had increased SSNL support again,

from midpoint however the control group had decreased SSNL. Both treatment and

control groups still did not significantly differ in SSNL.

For total social support from all three subscales, the treatment groups had non-

significantly lower scores that the control group at baseline. At program midpoint, both

treatment and control groups had increased total social support from baseline, but did not

significantly differ. At program end, the treatment group had increased total support

again, from midpoint, however the control group had decreased total support. Both

treatment and control groups still did not significantly differ in total social support.

Overall, the treatment groups had a significant increase in self-efficacy compared to

the control group at program midpoint but did not significantly differ from the control

group at program end. For types of social support, both treatment and control groups had

increased social support at program mid-point for co-worker (SSCW), cohabiting support

(SSL), and total social support. However, for SSL and total social support, the treatment

group had even higher support at program end and the control group had non-

significantly lower support than the treatment group. For co-worker support, the

treatment groups had significantly higher support at program end than the control group.









As for non-cohabiting social support (SSNL), the treatment groups had gained support

at both program mid-point and end however, the control group had the same level of

support at program mid-point, had decreased support at program end but did not

significantly differ from the treatment groups.

Differences in Self-Efficacy and Social Support within Groups

A Repeated Measures (GLM) analysis was conducted to determine differences within

treatment and control group between pre, mid and post program on both self-efficacy and

social support measures however, no significant differences were found. Only pre versus

mid program self-efficacy nearly significantly differed at the 0.10 alpha level for the

treatment groups (SE=2.93, t=1.74, p=0.09) in an dependent t-test.

Dependent t-tests were conducted to determine differences between pre and midpoint

and post program social support measures for both treatment and control groups. For the

control group, only the difference in mid program versus post program co-worker social

support was significant (SE=1.19, t=-2.23, p=0.05). For the treatment group however,

pre versus post program co-worker and total social support differed significantly

(SE=0.71, t=2.14. p=0.04 and SE=2.29, t=2.42, p=0.02, respectively).

Assessment Measures Results

Descriptive Statistics

The descriptive data for the pre and post program assessment measures included means

and standard deviations for pulse, blood pressure, weight, body composition, and

estimated VO2max for the cardiovascular step test for both treatment and control groups

before (baseline) and after (post) delivery of the program. This data is shown in Table

VI.












Table VI
Assessment Measures Descriptive Statistics

Baseline Post Program
Treatment Control Treatment Control
(N=43) (N=15) (N=43) (N=15)

x SD X SD SD x SD
Resting Heart 75.5 9.8 77.9 20.5 73.1 13.5 68.6 6.7
Rate
Blood pressure 126.4 13.8 131.1 13.0 119.5 13.0 120.2 6.3
Systolic
Blood pressure 80.7 8.6 80.6 8.4 71.3 8.4 71.8 6.4
Diastolic
Weight 161.1 33.9 144.3 16.4 163.4 33.5 142.1 14.4
(in pounds)
Body 29.9 5.8 28.5 4.9 29.7 5.8 26.6 4.4
Composition (%)
Step Test 44.4 6.0 43.3 4.2 42.7 14.4 20.6 6.8
(Estimated
VO2max)________________


Assessment Measure Differences between Treatment and Control Groups

As with analyzing differences in self-efficacy and social support, an independent

samples or group t-test statistic was used to compare differences in assessment measures

between treatment and control groups using the means and standard deviations shown in

Table VI. No significant differences existed for any of the pre-program assessment

measures between treatment and control groups except for weight (t=-2.43, DF=48.7,

p=0.02). The treatment group had a significantly higher mean weight than the control

group (see Table VI).

For the assessment measures at post program, no significant differences existed for

resting heart rate and systolic and diastolic blood pressure between the treatment and

control groups. However, weight (t=-3.14, DF=46.2, p=0.003), body composition

percentage (t=-2.01, DF=27.7, p=0.05), and VO2max (t=-6.52, DF=28.9, p=0.0001)

differed significantly between treatment and control groups at program end.









Assessment Measures Differences within Treatment and Control Groups

A dependent samples t-test was used to determine significant differences between

each assessment measure for pre and post program within the treatment and control

groups. For the control group, there was no significant difference between pre and post

program measures for resting heart rate, weight, and body composition percentage.

However, the control group did improve significantly in systolic (SE= 3.08, t=-3.6,

p=0.004) and diastolic blood pressure (SE=2.24, t=-3.36, p=0.006), but significantly

worsened in estimated VO2max(SE=1.8,

T=-12.6, p=0.0001).

For the treatment groups, there were no significant differences between pre and post

program measures for weight, body composition percentage, and VO2max. However, the

treatment group did improve significantly in systolic (SE= 2.39, t=-2.97, p=0.005) and

diastolic blood pressure (SE=1.72, t=-5.87, p=0.0001). The treatment groups also

improved significantly in resting heart rate but only at the .10 alpha level (SE=1.83, t=-

5.87, p=0.09).

Adherence Results

Descriptive Statistics

The adherence measures for the 14-week exercise program included cardiovascular

exercise frequency and duration, strength training frequency, and leisure activity

frequency. For cardiovascular exercise frequency each week, participants were given a

"0" for exercising under 3 times per week (under program goal), a "1" for exercising

three times per week (at program goal), and a "2" for exercising over three times per

week (over the program goal). For cardiovascular exercise duration each week,










participants were given a "0" for exercising under 20 minutes per session (under program

goal), a "1" for exercising at 20 minutes per session (at program goal ), and a "2" for

exercising over 20 minutes per session (over the program goal). For strength training and

leisure activity frequency, the number of days of activity were recorded each week.

The overall program weekly means and standard deviations for each type of physical

activity are shown in Table VII. Types of leisure activities reported by participants

included golf, playing with their children, bowling, dancing, and yard work.

Table VII
Adherence Descriptive Statistics for 14-Week Program

Treatment Control
(N=43) (N=15)

xSD SD
Cardiovascular Frequency 0.8 0.7 0.9 0.7
*(times/week)
Cardiovascular Duration 1.3 0.6 1.4 0.7
**(length of session)
Strength Training Frequency 0.7 1.1 0.7 1.0
(Days/Week)
Leisure Activities 0.1 0.2 0.1 0.1
(Days/Week)
* For times/week, 0 = exercising under 3 times per week, 1 = exercising three times per week and 2 =
exercising over three times per week
** For length of session, 0 = exercising under 20 minutes per session, 1 = exercising at 20 minutes per
session, and 2 = exercising over 20 minutes per session.

Adherence Levels

Based on the adherence analyses and the overall program goal, the participants can be

classified into four groups of adherers dropouts, those exercising below program goal,

at program goal, and above program goal for both cardiovascular exercise frequency

and duration. Ranges were set for classification of participants around the above

mentioned values given for cardiovascular frequency and duration adherence for each

week of the program.










Dropouts were those participants with weekly averages below 0.1. Below program

goal exercisers are those participants with averages between 0.1 and 0.6. At program

goal exercisers are those participants with weekly averages between 0.7 and 1.6. Above

program goal exercisers are those participants with weekly averages above 1.7. The

frequency breakdown of the percentage of participants in each of the adherence levels for

cardiovascular frequency and duration is shown in Tables VIII and IX, respectively.

Table VIII
Levels of Adherence for Cardiovascular Frequency

Treatment Control
*(N=25) *(N=14)
n Frequency n Frequency
Dropouts 5 20% 2 14%
Exercising Below Program Goal 7 28% 5 36%
Exercising At Program Goal 9 36% 2 14%
Exercising Above Program Goal 4 16% 4 29%
Due to missing data, only 25 treatment participants and 14 control participants had complete adherence
data for the analysis.

Table IX
Levels of Adherence for Cardiovascular Duration

Treatment Control
*(N=25) *(N=14)
n Frequency n frequency
Dropouts 3 12% 1 7%
Exercising Below Program Goal 1 4% 2 14%
Exercising At Program Goal 13 52% 3 21%
Exercising Above Program Goal 8 32% 8 57%
Due to missing data, only 25 treatment participants and 14 control participants had complete adherence
data for the analysis.


Differences in Adherence between Treatment and Control Groups

For purposes of the overall program goal which was exercising for three, 20-minute

bouts of cardiovascular activity a week, only cardiovascular frequency and duration were

used as adherence measures for comparison between the treatment and control groups.

For adherence differences between treatment and control groups, no significant

differences existed (using independent samples t-test) in either cardiovascular frequency









or duration, or strength training frequency, based on overall program weekly means (see

Table VI). As for levels of adherence between treatment and control groups, the

treatment groups had a higher percentage of dropouts than the control group but a lower

percentage of participants exercising below the program goal. The treatment groups also

had a higher percentage of participants exercising at the program goal but had a lower

percentage of participants exercising above the program goal than the control group.

For strength training and leisure activity frequency, only number of occasions reported

each week by participants was calculated. For the treatment groups, 60% of participants

reported some type of strength training at least one time over the course of the program.

Thirty-two percent of these participants reported engaging in strength training less than

once per week. Only 16% of these participants reported engaging in strength training

between one and two times per week and only 12% reported engaging in strength training

more than twice per week.

Only 24% of treatment participants reported engaging in leisure activities at least one

time over the course of the program. Twelve percent of these participants reported

engaging in leisure activity once per week and the other 12% engaged in leisure activity

more than two times per week.

For the control group, 57% of participants reported some type of strength training or

leisure activity at least one time over the course of the program. Twenty-eight percent of

these participants reported engaging in strength training less than once per week. Only

7% of these participants reported engaging in strength training between one and two

times per week and 21% reported engaging in strength training more than twice per

week.









Thirty-six percent of treatment participants reported engaging in leisure activities at

least one time over the course of the program. All of these participants averaged less

than one time per week of leisure activity.


Correlational Analyses

For interest of construct relationships, several Pearson Correlation Coefficients were

conducted including program end self-efficacy with both frequency and duration

adherence, and program end social support. Also, the relationship between reported

exercise six months prior to the program (exercise history) with both types of adherence

and program end self-efficacy was analyzed. In addition, prior leisure/sport participation

(leisure/sport history) with both types of adherence and program end self-efficacy was

analyzed.

No significant relationship was found for program end self-efficacy with either

frequency or duration adherence. However, program end self-efficacy was positively

correlated with program end social support (p = 0.02). Exercise history was positively

correlated with both types of adherence cardiovascular frequency (p = 0.01) and

cardiovascular duration (p = 0.04). Exercise history was not however, related to program

end self-efficacy. Finally, leisure/sport history was not related to either type of

cardiovascular adherence or program end self-efficacy.

To analyze the independent contributions that program end self-efficacy or social

support may have on adherence a regression model analysis was conducted, which

controlled for both self-efficacy and social support independently. There were no

significant independent contributions from either program end self-efficacy or social

support with respect to either cardiovascular frequency or duration adherence.









Summary

This chapter revealed all analyses needed to answer all the primary and secondary

research questions presented in Chapter 3. The overall participant demographic and

descriptive results were presented as well as the descriptive statistics and within and

between group differences for self-efficacy, social support, assessment measures, and

adherence data. Correlational construct analyses are also presented along with

independent contributions of self-efficacy and social support on adherence. Chapter 5

will discuss these results in detail and use them to answer the proposed research

questions.
















CHAPTER 5
DISCUSSION AND CONCLUSIONS

Introduction

In this chapter, all descriptive and analytic results (including research questions) will

be discussed as well as programming aspects of the intervention and strengths and

weaknesses of the study. Study conclusions will be given as well as recommendations

for future research and the implication for health education.

Discussion of Results

Demographics

For the program participant demographic data it is apparent that the majority of

participants in this study are female, Caucasian, between the age of 31 and 60, and

teachers. However, in the treatment groups in particular, 16% of participants are males,

23% are African-American, and 26% are office staff and custodial employees. Despite

the slightly skewed distribution of demographics, the sample is representative of the

county under study, particularly county elementary school employees (refer to Table I).

As for the non-starter group of individuals who were initially assessed but did not start

the program, this group was demographically similar, but had some distinct differences in

exercise history (discussed in a subsequent section). Overall, in terms of demographics,

most elementary schools across the country are comprised largely of female teachers

within these reported age ranges. With respect to race, this sample was









representative of the county population however, a more even distribution of ethnicity

would have been more consistent with national demographics.

Exercise Habits

It is clear from a comparison of exercise habits between the treatment and control

groups that a much higher percentage of participants in the treatment groups reported that

they were not currently exercising, had not exercised in the past six months, and had

never engaged in sports or leisure activities, prior to the start of the program. Moreover,

93% of participants in the control group reported currently exercising, prior to the

program. This would indicate a highly self-selected group for program participation

within the control group. However, the treatment groups were comprised of participants

with more varied fitness levels and of more participants who were recruited rather than

self-selected. The distribution of those who reported never having exercised was about

equal between treatment and control groups.

However, what is interesting about these findings, is that 53% of the treatment group

and 27% of the control group reported not exercising in the past six months but 60% of

the treatment group and 93% of the control group reported currently exercising. This

observation may be due to participants wanting to make themselves feel better about not

exercising in the past six months and reporting that they are currently exercising. Most

likely, even though participants were not exercising in the past six months, they may have

wanted to make themselves feel better or appear better to program staff (when

completing the Pre-Participation Questionnaire) and thereby reported they were currently

exercising because they would start exercising in the current week.









As for the "non-starters," noticeable differences were seen in the exercise histories

between the treatment and control program participants and the non-starters. Fifty-four

percent of the non-starters were not currently exercising as opposed to 26% in the

treatment group and 7% in the control group. A startling 93% of non-starters had not

exercised in the past six months compared to 22% in the treatment groups and 27% in the

control group. The percentage of non-starters never having exercised or participated in

sports or leisure activity was similar to all groups of program participants. These

findings strongly support the notion that those with poorer exercise history (i.e. little past

exercise) are less likely to start an exercise program.

As an anecdotal note however, these 15 non-starters had understandable personal

reasons and obstacles for not beginning the program. These reasons ranged from

transferring to another worksite, experiencing job changes, needing a more medically

supervised exercise program, and caring for an ill spouse.

Personal Exercise Plan Data

In the Personal Exercise Plan (PEPs), participants had the opportunity to share several

valuable pieces of information with program consultants regarding their exercise

activities, goals, barriers, and strategies. From the information provided in Table III, it is

apparent that most participants chose walking as a preferred exercise activity, wanted to

loose weight for a personal goal, and cited poor time management and lack of motivation

as main barriers to achieving their goals. It is no surprise that walking was the number

one activity among participants for cardiovascular exercise. Virtually anyone, without

major physical disabilities, can walk. Walking is easy to do and can be done practically

anywhere. Moreover, moderate physical activities such as walking, when compared with









higher intensity fitness activities, are associated with a higher rate of adoption and

maintenance of physical activity patterns in a population base; this is particularly true for

women (Sallis, Haskell, Fortmann, Vranizan, Taylor, & Solomon, 1986).

Also interesting is that lack of motivation and feeling tired were cited equally as

important as lack of time management. This barrier to exercise is an important

justification for the role of the program consultants in motivating participants and

convincing participants that moderate, regular exercise will increase energy over time. In

addition, consistency was cited by 46% of participants as a strategy to achieve personal

goals. This reinforces the importance of consistent exercise to achieve exercise-related

benefits and goals. It seems that most individuals who start an exercise program remain

consistent in their exercise routine for perhaps the first month, then become rather

inconsistent in their routine thereafter. Inconsistent or sporadic exercise will not lead to

significant improvements in exercise-related goals.

A final interesting finding in regards to the PEPs was that 91% of participants reported

exercising at home and 84% reported exercising in the evenings. These significant

proportions of participants call for exercise interventions targeting at-home, after work

exercisers. An intervention such as the one tested in this study allows for participants to

exercise where and when it was convenient for them. The literature supports the notion

that if program participants are given the freedom to not only choose their own activities

(Baun & Bernacki, 1988; Thompson & Wankel, 1980) but to do those activities where it

is most convenient for them, those participants will be more successful in adhering to

their exercise program.











Reasons for Not Exercising.

As is shown in Table IV, the common reasons cited on the CDSII by participants for

not engaging in regular exercise, was poor time management/over scheduling, followed

by lack of motivation, lack of energy and medical reasons. This reinforces the barriers

cited in the PEPs. This finding also is consistent with the literature. Shepard (1988)

found that one of the most commonly cited reasons for not participating in a corporate

exercise program were "lack of time."

Even though the number one reason for being unable to exercise or not regularly

engaging in exercise was lack of time, lack of energy and motivation closely followed.

To encourage and promote regular exercise, individuals need assistance with time

management (i.e. scheduling exercise) and motivation. These two catalysts for exercise

were the reoccurring proponents within the program intervention that proved most

successful in forming exercise habits.

Self-Efficacy and Social Support

Primary research question #1: "Does exercise self-efficacy increase during and

after implementation of a health education intervention to promote exercise

adherence?" The self-efficacy and social support descriptive data shown in Table V

depict a greater overall increase in self-efficacy and social support for the treatment

groups as compared to the control group. More specifically, to answer the first primary

research question stated above, comparisons of pre, mid and post program self-efficacy

means for the treatment group show a near significant (significance is measured at the p

0.05 alpha level) increase in self-efficacy from pre-program to mid-program (p = 0.09)









and a slight but insignificant increase in self-efficacy from pre-program to post-program.

The treatment group self-efficacy mean had actually decreased from mid-program to post

program (midpoint x = 73.9 and post x = 69.2).

It is supposed that with the excitement of starting a new program and meeting with the

health promotion consultants for the first and second times, that the treatment participants

experienced the greatest increase in self-efficacy. Towards the end of the program

however participants had more difficulty meeting with consultants due to increased hectic

schedules in the school term and were not exercising as regularly. This decrease in time

and energy for the exercise program may attribute to the decrease in self-efficacy at

program end.

Primary research question #2: "What types of social support for exercise are

received by subjects during and after implementation of a health education

intervention to promote exercise adherence?" To answer the second primary research

question, the social support subscale total means (i.e. the group mean subscale total

score) in Table V were divided by the number of items in each subscale to compute an

average item score for each social support subscale. These computations had to be

performed in order to make equal comparisons across the three subscales, which had

unequal numbers of items (i.e. coworker and cohabiting support had 5 items and non-

cohabiting support has 13 items).

The greatest form of social support for the treatment groups was co-worker support

(MSSCW x= 1.66, PSSCW- = 1.78). The support levels for those living and not living

with participants was virtually the same (MSSL x = 1.22, PSSL x = 1.31 and MSSNL









x = 1.20, PSSNL x = 1.38, respectively). These findings are not surprising since it is

expected that social support from co-workers would be higher than other forms of

support in a worksite program. In fact, one treatment site actually started an exercise

group, comprised of program participants that met after school hours twice a week in the

"assessment" classroom, to exercise to aerobic videotapes. These findings are supported

by a meta-analysis of social influence and exercise (Carron, Hausenblaus, & Mack, 1996)

which revealed that family did not represent the strongest source of social influence for

adherence behavior, rather the influence of important others had a stronger impact. This

study strongly indicates that the "important others" in the exercise program were in fact

coworkers. Perhaps, if spouses, significant others, and friends were allowed to

participate, the other types of social support would increase, however this is unknown.

For the control group, support from individuals living with participants was the

greatest form of support (MSSL x = 1.52, PSSL x = 1.34). Support from those not living

with participants was the second greatest form of support (MSSNL x = 1.48. PSSNL x =

1.22) and support from co-workers was the least form of social support (MSSCW x=

1.32. PSSCWx-= 0.9). These comparisons also are not surprising since most of the

control group participants were already regular exercisers prior to the program and did

not receive the worksite intervention. These participants were receiving more support for

their exercise from individuals living with them than the treatment groups at baseline

(treatment BSSL x = 14.8 and control BSSL x = 15.7), which continued to increase at

mid and post program. Also of interest, is the fact that co-worker support remained low

without the treatment intervention but in lieu of assessments being performed at the









worksite. It is supposed that merely conducting fitness assessments onsite is not

sufficient to increase coworker support for exercise.

Adherence

Primary research question #3: "Do the intervention groups have greater

adherence to the exercise prescription than non-intervention group?" To answer the

third primary question, comparisons of means for cardiovascular frequency and duration,

shown in Table VII, indicate that the treatment groups have slightly lower cardiovascular

exercise adherence than the control group (treatment x = 0.8, x = 1.3 and control x = 0.9,

x = 1.4, respectively). In addition, no significant differences existed (using independent

samples t-test) in either exercise cardiovascular frequency or duration and strength

training frequency, based on overall program weekly means (see Table VI). This could

again be explained by the high percentage (93%) of control group participants who were

already exercising. These control group participants seemed to merely continue their

regular exercise throughout the program, whereas many of the treatment group

participants were just exercising again for the first time in many months, even years (i.e.

53% of treatment group reported not exercising in the past 6 months, as stated in previous

section).

As for levels of cardiovascular exercise frequency and duration adherence between

treatment and control groups, shown in Tables VIII and IX, the treatment groups had a

higher percentage of dropouts than the control group but a lower percentage of

participants exercising below the program goal. The treatment groups also had a higher

percentage of participants exercising at the program goal but had a lower percentage of

participants exercising above the program goal than the control group.









In other words, even though the treatment groups had higher dropout rates than the

control group, the treatment group had more participants exercising at the program goal.

The control group seemed to have a split of participants exercising either below or above

the program goal (frequency: 36% below goal, 29% above goal). The treatment group

had most participants exercising at the program goal (frequency at goal = 36%, duration

at goal = 52%). As stated above, the control group exercisers continuing to exercise at

their previous levels may explain this. However, the treatment groups seemed to more

closely follow the program exercise goals due to the treatment intervention.

Moreover, in terms of overall program dropout levels, the treatment groups had only

20% participants drop out in terms of cardiovascular exercise frequency and 12% drop

out in terms of cardiovascular exercise duration. The control group had only 14% drop

out in terms of cardiovascular frequency and 7% drop out in cardiovascular exercise

duration. For both treatment and control groups these percentages are below the dropout

rates of most studies which show that 3 to 6 months after program start, only 50-60% of

the original participants are still participating in the program (Dishman, 1988; Marcus, et

al., 1992; Oldridge, 1984). These percentages show that, in spite of the high percentage

of previous exercisers in the control group, that the intervention served to greatly improve

exercise adherence as compared to most other studies.

As for strength training and leisure activity frequency, 16% of treatment and 7% of

control participants reported engaging in strength training between one and two times per

week. Twelve percent of treatment and 21% of control participants reported engaging in

strength training more than twice per week. Twelve percent of treatment and 36% of

control participants engaged in leisure activity once per week and 12% of treatment









participants engaged in leisure activity more than two times per week. Even though

strength training and leisure activity were not part of the exercise program, it is

interesting to note the percentages of participants in both groups that chose to engage in

these types of activities.

Assessment Measures

Secondary research questions #1-5: Does the intervention group have a greater

improvement in blood pressure, resting heart rate, weight loss, body composition,

and cardiovascular fitness (i.e., improved estimated VO2max) than the non-

intervention group?" To answer secondary research questions 1-5, the within group

comparisons for the assessment measures reveal that the treatment (intervention) groups

had greater improvement in diastolic blood pressure (treatment p = 0.0001 and control p

= 0.006) and resting heart rate (treatment p = 0.09 and control p = 0.11) than the control

group. However, neither treatment nor control groups significantly improved in weight,

body composition or estimated VO2max and both treatment and control groups

significantly improved in systolic blood pressure.

The improvement in resting heart rate and blood pressure in the treatment groups is

assumed to be due to the fact that most participants in the treatment group had started

exercising again, after a great lapse of not exercising. So even after just 14 weeks of

exercise, blood pressure and resting heart rate improved.

Although, the control group also had significantly increased systolic (p = 0.004) and

diastolic blood pressure (p = 0.006), even with the high percentage of previous exercisers.

This may be due to the previous exercisers improving their consistency of exercise during

the program as compared to their poorer consistency of exercise before the program.









As for weight and body composition, the adherence levels of both treatment and

control groups were perhaps, not high enough and program length not long enough, to

reveal significant improvements in these measures.

Correlational Analyses

Before the last two secondary research questions are answered by the correlational

relationship findings, the relationship of self-efficacy and social support deserves

mention. In Chapter 4, program end self-efficacy and social support were positively

correlated (p = 0.02). This finding is consistent with the literature (Courneya &

McAuley, 1995; Cutrona & Troutman, 1986; Duncan & McAuley, 1993; Holahan &

Holahan, 1987; Taylor, et al., 1985) which supports the notion that self-efficacy is

mediated by social support, as described in Chapter 2. This relationship is described in

more detail in the subsequent conclusions section.

Secondary research questions # 6 & 7: "Does a relationship exist between

previous exercise participation and adherence?" and "Does a relationship exist

between previous sports or active leisure participation and adherence?" To answer

the last two secondary research questions, the correlational analyses reveal that previous

exercise participation (exercise history) was positively correlated with both types of

adherence (cardiovascular frequency p = 0.01, duration p = 0.04), however previous

leisure/sport activity did not correlate with adherence.

The relationship between exercise history and program adherence (participation) is

supported by the literature. According to Kendzierski (1990), experienced exercisers

already have a plan of action for starting an exercise program (i.e. decide they'll start, tell

their friends, buy some new walking/running shoes, pick a day to start, etc.). In addition,









once the experienced exerciser takes the first actions in a behavioral sequence for starting

an exercise program, they are more likely to continue through the sequence (Kendzierski,

1990).

Aside from the research questions posed in this study, an additional interesting

correlational finding was that exercise history or exercise experience was not

significantly correlated with pre, mid, or post program self-efficacy. This finding seems

contrary to the theory of self-efficacy in that having confidence in the ability to perform a

behavior leads to higher self-efficacy for that behavior. It seems reasonable that

previously engaging in the behavior (in this case, exercise) would increase confidence in

one's ability to re-engage in the behavior. However, the previous exercisers in the

control group may be an exception to the theory.

Independent Contributions of Self-Efficacy and Social Support on Adherence

Neither program end self-efficacy or total social support had an independent

contribution to cardiovascular adherence. However, this finding should be taken with

caution for two reasons. First, since program end self-efficacy decreased from mid-

program self-efficacy, program end self-efficacy may not truly reflect the self-efficacy

level of the program for most program participants since the largest increase in self-

efficacy was between pre and mid program. Second, program end total social support

reflects all social support subscales which may not provide the best measure of social

support for participants since the subscales means varied greatly between treatment and

control groups.









Discussion of Programming Aspects

Program planning. Some aspects of program planning unique to this study

intervention included scheduling assessments and advertising recruitment health fairs and

assessments. Developing assessment schedules with school employees provided a unique

challenge to program staff in that assessments could only be scheduled during after

school hours (i.e. after students had left, but when employees were still on-site) and

around holiday breaks. These circumstances not only left a two-hour window to conduct

assessments in the afternoons, but also required development of an assessment rotation

through four school worksites.

Another unique program planning aspect of this intervention was developing a

marketing plan. It was first thought to post flyers or posters in the staff/teacher lounges

however, since every employee in most schools had a personal mailbox in the school

office, placing flyers and memos in staff boxes seemed more effective. An extremely

unique marketing tool to school worksites was the use of the public address or "PA"

system throughout the schools. Physical Education teachers/Wellness Coordinators

alerted program staff to the use of the PA system in making "all calls" throughout all

rooms in the school as well as in calling teachers in individual classrooms. In the

marketing plan, announcements were written for office staff to read during regular

morning announcements and immediately after students left in the afternoons as a

reminder to employees participating in the program. These announcements could be used

to advertise recruitment health fairs and assessment schedules.

Program implementation. Several unique aspects of program implementation arose

in this study. These aspects included working with physical education (PE)









teachers/wellness coordinators, teacher in-service days, "using word of mouth," and

organizational/work styles of elementary school employees.

Depending on the planning style, and the motivation and attitude of the PE teacher at a

given school, the course of the program implementation differed slightly. In the schools

that had highly motivated and organized PE teachers, more employees were recruited for

participation and participants were reminded by their PE teachers about upcoming

assessments and other program updates. PE teachers with these characteristics also

assisted program consultants in "tracking down" participants for their scheduled sessions.

The characteristics of these motivated PE teachers are those commonly held by a

"program champion." A program champion is a key employee who has a secure position

in the organization (or school in this case), the respect of peers, and sufficient influence

to help move progressive programming ideas that challenge the norms of the worksite

culture regarding the program being delivered (Wilson & Glaros, 1994, chap. 21). In

many cases, the program champion is critical to the success of implementation of a

worksite health promotion program. It can be concluded that those PE teachers serving

as program champions had a definite impact on the success of this program delivery.

A second unique aspect of program implementation was the use of teacher in-service

days. At one school, in particular, the PE teacher suggested holding pre-program

assessments on their teacher in-service day. Since students were off for the day and

employees could come by at any time during morning hours for assessments, the program

assessment team was able to perform over 30 employee assessments in a morning. This

situation proved most convenient for employees and effective for completing a large

number of assessments.









A third unique aspect of program implementation was using "word of mouth" as a

marketing tool throughout the program's implementation. Since teachers at most schools

were arranged in teams according to grade, "word of mouth" could travel quickly among

teams and other co-workers. By continually updating certain program participants about

the program and having them bring the information back to their team or other co-

workers, program information and reminders could be spread quickly.

A final aspect, in terms of program implementation, which proved challenging at

times, was the organizational/work style of the majority of the teachers participating in

the program. Due to the hectic, energy absorbing nature of some teachers' workdays,

tutoring obligations, and after school meetings, these teachers had difficulty keeping

appointments made with program consultants. Some teachers would simply forget they

had scheduled a session with their consultant or that they had an off-site meeting to

attend. The program consultants, in turn, would have difficulty from time-to-time

"tracking down" teachers for after school program sessions. Despite an observation of

occasional forgetfulness and disorganization of some teachers, all participants made

continued efforts throughout the program to meet with their consultant.

Program evaluation. Both submission of exercise verification sheets and submission

of PEPs deserve discussion as critical components in the program evaluation process.

For submission of exercise verification sheets, employees were to submit their sheets

monthly to either their program consultant or PE teacher's box. However, several

employees submitted all or part of their sheets to program staff at the time of their post

program or final assessment. This proved challenging in terms of data entry, however the

sheets had complete information. This occurrence could again be explained by the









forgetful nature of some employees and inconvenience of having to submit sheets

monthly.

As for submission of Personal Exercise Plans (PEPs), program consultants had to use

carbon paper to generate multiple copies of the PEP. Three copies of the PEP were

generated for each session. One copy was given to the participant, one was submitted to

the PI for the participant files, and one was kept by the consultant to use as reference for

subsequent sessions (in the frequent case when the participant failed to bring their copy to

the session). This process enabled constant feedback to the PI regarding participant

sessions. In addition, consultants submitted a monthly PEP summary with information

summarizing all of their assigned participants' PEP information. Consultants also

provided overall program feedback in a program-end staff meeting.

Overall, several components of the worksite exercise program planning,

implementation, and evaluation were found to attribute to the effectiveness of the

program in this study and can be recommended for future programs. First, considering

employees' schedules and work styles is important when planning assessments and

consultations. At the school sites, health coordinators should take advantage of teacher

in-service days and schedule around after school meetings. Second, thoroughly

investigating all marketing and advertising sources (mailboxes, PA system, "word of

mouth", etc.) within the worksite organization (or schoolsite organization in this case) by

asking involved employees can aid in participant recruitment. Many successful schoolsite

health promotion programs (Glasgow, McCaul, & Fisher, 1993; Masey, Gilmarc, &

Kronenfeld, 1988; Wong, Bauman & Koch, 1996; Lovato & Green, 1990) have used




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs