- Email: [email protected]
Nutrition Education Worksite Intervention for University Staff: Application of the Health Belief Model
RESEARCH BRIEF
Nutrition Education Worksite Intervention for University Staff: Application of the Health Belief Model D O R I S A. A B O O D, E D D, CHES; 1 D AV I D R. B L AC K , P H D, HSPP, CHES, MPH, CPPE, FASHA, FSBM, FAAHB; 2 D I A N E F E R A L , MS, RD 1 1
Department of Nutrition, Food, and Exercise Sciences, Florida State University,Tallahassee, Florida; 2 Department of Health and Kinesiology, Purdue University,West Lafayette, Indiana
and decreased energy, fat, saturated fat, and cholesterol intake to levels consistent with national recommendations.
ABSTRACT
Objective: To evaluate the efficacy of an 8-week worksite nutrition education intervention for university staff using the Health Belief Model (HBM) to promote healthful dietary behaviors that reduce risks for cardiovascular disease and cancer.
KEY WORDS: nutrition education, worksite, Health Belief Model, university staff (J Nutr Educ Behav. 2003;35:260-267.)
Design: 2 2 repeated measures baseline/posttest ex post facto research design.
INTRODUCTION
Participants: Staff employees were randomly assigned to treatment (n = 28) and control groups (n = 25).
Of the 10 leading causes of death in the American adult population, 4 are directly related to nutrition: coronary heart disease, cancer, stroke, and diabetes mellitus.1 In response to these chronic diseases, the US government and other groups have introduced prevention programs. Dietary changes are one means of disease prevention, and the Dietary Guidelines for Americans have been incorporated into nutrition programs to improve dietary intake.2,3 Dietary behavioral changes have become a focal point in the past decade, and significant dietary changes have been reported in clinical trials using selected men and women at high risk for cardiovascular disease (CVD) and cancer.4 Generalizing from these trials, however, is difficult because of rigorous inclusion criteria and protocols, and the programs implemented are often inappropriate for application to the general population. Community-based interventions are needed to determine which approaches might be effectively used to prevent CVD and cancer among free-living populations.5 Worksites, such as university campuses, can be considered communities.Worksite interventions are important because employees are currently facing large increases in health insurance costs. Nutrition education at worksites may be effective in improving health-related behaviors so that insurance costs, sick days, and turnover can be reduced, and employee productivity and morale can be increased.6,7 Worksite interventions may be advantageous too because they are convenient and accessible to workers, allow employees to support one another in their efforts to make behavior changes, and are less expensive, especially compared with programs offered in clinical settings.8,9
Intervention: The intervention focused on specific health beliefs, nutrition knowledge, and dietary practices to demonstrate treatment effect. Main Outcome Measures: Dependent variables were specific health beliefs, nutrition knowledge, and dietary behaviors. Independent variables were demographic characteristics and group assignment. Analyses: Tests of parametric assumptions, power analyses, analysis of variance, and Kuder-Richardson and Pearson product-moment coefficients were computed and specificity of treatment effects was assessed. Results: Perceived benefits of healthy nutrition practices and nutrition knowledge related to cardiovascular disease and cancer significantly improved among the treatment participants, P < .001. Treatment group participants also significantly reduced total calories, fat, saturated fat, and cholesterol intake (each P < .001). Conclusions: The intervention appears to be related to treatment effects and significantly increased nutrition knowledge
Address for correspondence: Doris A.Abood, EdD, CHES, Department of Nutrition, Food and Exercise Sciences, Florida State University, 408 Sandels Building, Tallahassee, FL 32306-1493;Tel: (850) 644-4796; Fax: (850) 645-5000; E-mail: dabood@ mailer.fsu.edu. ©2003 SOCIETY FOR NUTRITION EDUCATION
260
Journal of Nutrition Education and Behavior Volume 35 Number 5
Behavioral theory has increasingly been used to guide nutrition research to improve intervention efficacy.4 One such theoretical framework that has been applied is the Health Belief Model (HBM). The HBM was developed in the 1950s to explain health behavior associated with the failure of people to participate in programs that would reduce disease risk.10-13 The HBM implies that health behaviors are determined by health beliefs and readiness to take action.10 Constructs central to the HBM consist of perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and other mediating variables. The construct of selfefficacy is frequently included in applications of the HBM.14,15 The inclusion of self-efficacy in the HBM assumes that the likelihood of taking action is not only a function of beliefs related to outcomes but also a function of a person’s belief that he or she is behaviorally capable of achieving the desired outcome.14,15 Interventions that also emphasize selfefficacy may increase the likelihood that positive behavioral changes will be executed. A review of the literature suggests that there is a void in research that focuses on university campuses as worksites and theory-based, tailored nutrition interventions for employees. An electronic search using several databases (MEDLINE, PubMed, and ERIC using the dates 1999 through April 2003) did not produce any nutrition education interventions that were theory based and focused on university worksite employees. Broadening the search to include only single-worksite interventions produced only one study of a rural manufacturing site that used community organization strategies to produce changes in fruit and vegetable consumption.16 No studies were found of single worksites that tailored the nutrition education intervention to participants’ baseline knowledge, dietary behaviors, or health beliefs. The purpose of this study was to evaluate the efficacy of a worksite-tailored nutrition education program for university staff employees.The program used the HBM as a theoretical framework to increase nutrition knowledge and improve dietary intake via instruction and dietary exercises. The nutrition education program was designed as an initial prototype to aid in the prevention of CVD and cancer among university worksite employees.
DESCRIPTION OF THE INTERVENTION AND EVALUATION Intervention The treatment group participated in 8 1-hour weekly educational sessions. A registered dietitian who was a member of the research team provided all instruction. Three invariant education sessions were held each week to provide employees with maximum opportunity to attend sessions convenient to their work schedules. Subjects were prohibited from attending more than 1 session per week. The focus of the intervention was to promote knowledge and beliefs con-
September • October 2003
261
ducive to improving or maintaining positive dietary practices for prevention of CVD and cancer.The sessions were didactic interspersed by questions and answers, and information was presented by means of computerized overhead projection, displays, and paper materials.A detailed outline also was distributed at the beginning of each session to provide an overview and for note taking.An overview of the nutritional intervention is presented in Table 1, which shows the topics and content and how constructs of the HBM were integrated into each session of the program.All dietary information presented and recommendations provided were consistent with the US Department of Agriculture’s (USDA) Food Guide Pyramid and the US Department of Health and Human Services’ National Cholesterol Program Guidelines.3,17 The posttest was administered at the end of the intervention. Tailoring the intervention specifically to the target population occurred in two ways: baseline test assessments and prior research findings. Tailoring was based on baseline assessment of nutrition knowledge and dietary behavior status.This assessment was used to guide content selection and determined the information to emphasize in each session. For example, in terms of nutrition knowledge, it was noted that more than 50% of participants incorrectly answered questions pertaining to risk factors associated with CVD and cancer, sources of protein and cholesterol, macronutrient recommendations (percentage of energy), facts related to margarine versus butter, fruit and vegetable intake recommendations, benefits of exercise, hidden sources of fat, the difference between 2% and whole milk, and that cholesterol comes only from animal products. Nutrition education instruction focused on increasing knowledge in these areas. Mean values of initial dietary behavior also revealed that the priority of focus should be on fat and cholesterol intake and secondarily on fruit and vegetable consumption (Table 2). Tailoring of the intervention also was based on HBM constructs. Even though all HBM constructs were addressed during the intervention, perceived benefits and perceived barriers were given priority despite the mean values of perceived benefits falling into the high category and the mean values of perceived barriers falling into the moderate category in comparison with normative values (see Table 2).The reason perceived benefits and perceived barriers were given primary consideration is because prior research has shown perceived susceptibility to be a strong predictor of behavior when a person believes that behavior change would be accompanied by perceived benefits that outweigh perceived barriers.18 In other words, theory would suggest that for a change in nutrition behavior to occur, the perceived benefits of the behavior must outweigh the perceived costs (barriers). An extensive review of the literature has shown that the strongest predictor of nutrition-related behavior change is the benefit-cost ratio.18 Barriers reported in other studies specific to healthy eating, which were the focus of this intervention, included perceived taste, lack of convenience/time, high cost, confusing advertising and diet recommendations, and lack of knowledge of actual food intake.18-21
262 Table 1.
Abood et al/NUTRITION EDUCATION WORKSITE INTERVENTION BASED ON THE HBM Overview of Nutrition Intervention
Session
Topic
Content
1
Development and Medical, Clinical, and Social Consequences of CVD and Cancer; Risk Reduction
Risk factors and prevalence rates; health consequences; health assessment of risk for CVD and cancer; USDA Food Guide Pyramid and proper nutrition to reduce risks; dietary analysis
2
Macronutrients
Macronutrients (fats, carbohydrates, and protein); Food Guide Pyramid and sources and benefits of recommended intakes; importance and benefits of proper nutrition; reducing barriers to increase probability of dietary changes; interpretation of health assessment results and dietary analysis
3
Macronutrients
Macronutrients and types of fat; hidden sources of fat, low-fat foods, meals, and fat alternatives; benefits of saturated fat reduction and reduction of barriers to taking such actions
4
Fruits and Vegetables
Health-protective role of fruits and vegetables; frequency and portion sizes; fiber, vitamins, minerals, antioxidants, and phytochemicals; benefits of increased fruit and vegetable intake and barrier reduction to taking action
5
Physical Activity and Weight Control
Health benefits of weight control and exercise, including types of exercise; barrier reduction; macronutrients and energy metabolism
6
Meal Patterns and Healthy Eating
Benefits of eating meals regularly; distribution and preparation of simple/convenient low-calorie/highnutrient recipes; ideas for removing barriers to healthful eating patterns
7
Meal Planning and Food Label Reading
Scheduling meals, shopping lists, label reading, cost cutting, food storage, and food preparation
8
Integration of All Previous Topics for a Prescription of Healthy Living
HBM constructs to change nutrition behaviors to reduce risks and for behavior maintenance; supplements, salt, caffeine intake, soft drinks
CVD indicates cardiovascular disease; HBM, Health Belief Model; USDA, US Department of Agriculture.
Control Group Subjects in the treatment and control groups were “blinded” to each other to reduce “contamination effects.”The control group, which was physically housed in a separate building, completed questionnaires only at baseline and postintervention. Subjects in the control group continued to work during the time in which the treatment group received the intervention.The control group received an abbreviated version of the nutrition education intervention 1-month postintervention.
trators granted approval for and agreed to provide 1-hour release time each week for all full-time employees who wished to participate in the intervention.A self-selected sample of subjects employed by the admissions and personnel department or the dean of students office participated, and the group that received treatment or served as the control group was selected randomly as a means to enhance the internal validity of the study. The institutional committee on the use of human research subjects approved the study, and all participants signed a consent form to participate.
Recruitment of Subjects
Data Collection and Instrumentation
Heads of administrative departments at a southern university were approached by the researchers and were asked to endorse a nutrition education program for their employees. Adminis-
The baseline/posttest questionnaire that took between 30 and 45 minutes to complete included 171 items and was divided into 4 sections. Section I was assessment of 5 HBM
Journal of Nutrition Education and Behavior Volume 35 Number 5
September • October 2003
263
Table 2. Means/Standard Deviations to Assess Intervention Effects Baseline Constructs/Variables
Treatment
Posttreatment Control
Treatment
Control
Health Beliefs† Health concern
2.03/1.48
1.88/1.50
1.85/1.40
1.84/1.50
Perceived susceptibility
6.10/1.37
5.96/1.45
5.92/1.42
6.16/1.49
Perceived severity
3.28/1.27
2.76/1.37
2.75/1.44
2.84/1.42
Perceived benefits
4.25/1.03
4.40/1.50
3.17/1.12***
4.80/1.75
Perceived barriers
5.53/1.01
5.40/1.35
4.53/1.00
5.28/1.30
Self-efficacy
4.07/2.03
4.60/2.43
3.57/2.03
4.24/2.37
Nutrition Knowledge Nutrition knowledge assessment
12.80/3.92
13.36/2.95
17.36/4.97***
13.30/2.96
Dietary Behavior Total energy (kcal)
2264/1077
1891/968
1427/503***
1919/884
Total fat (g)
91.99/56.02
81.78/46.48
47.43/30.41***
90.80/51.99
Total protein (g)
84.63/41.89
74.72/42.31
64.89/23.96
70.68/25.65
Fat energy (%)
35.36/9.18
36.25/7.37
21.18/7.63***
40.12/10.84
Total saturated fat (g)
31.38/20.38
28.40/15.43
13.50/5.48***
32.20/18.31
Total cholesterol (mg)
332.23/216.29
227.01/142.19
174.43/106.01***
266.64/154.20
Total fiber (g)
16.93/9.48
14.40/6.65
14.43/9.81
15.68/11.10
Vegetable (servings)
2.36/1.80
2.09/1.22
2.54/1.64
1.80/1.78
Fruit (servings)
2.45/2.19
1.63/1.17
2.79/2.39
1.34/1.17
Treatment group, n = 28; control group, n = 25. ***p < .001. †Health Belief Model score interpretation: health concern: ≤ 2 = high, 3 = moderate, ≥ 4 = low; perceived susceptibility, severity, benefits, barriers, and self-efficacy: ≤ 4 = high, 5-7 = moderate, ≥ 8 = low.
constructs as they relate to nutrition, CVD, and cancer.These 6 constructs were health concerns, perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and self-efficacy. Subjects rated items related to each of these constructs on a 5-point Likert-type scale from 1 (strongly agree) to 5 (strongly disagree). Negatively worded questions were reverse-scored. Lower scores reflect a higher degree of health beliefs. Section II of the questionnaire was the personal health assessment of risks for CVD and cancer.This portion of the questionnaire was developed using a combination of modified questions from the American Cancer Society,22 the American Heart Association,23 and Fahey and colleagues24 and included questions about family history of CVD and cancer and the risk factors for these diseases, such as smoking status, blood pressure, cholesterol, exercise, and diabetes (body weight was not assessed). Questions assessing cancer risk were related to breast, lung, and colorectal cancer because of their high prevalence rates in the United States and the established links to diet. Risk scores for CVD and cancer were individually computed by summing all responses related to each disease. Higher scores represented a higher risk of CVD, cancer, or both. The personal health assessment was used only for the initial assessment to be respectful of participants’ time and because dramatic changes in health status were unexpected in the short term.
Section III was the nutrition knowledge assessment, which was developed specifically for this sample and intervention to assess nutrition knowledge related to CVD and cancer risk. Some questions were adapted from a nutrition quiz developed by Hurley. 25 Question formats were true/false and multiple choice. Questions addressed aspects of protein, fat, carbohydrates, food sources of cholesterol, recommended intakes of fruits and vegetables, and healthful ways of reducing saturated fat, among other topics. One point was given for each correct answer. Section IV assessed dietary behaviors based on a modified food frequency instrument used by Boeckner and colleagues26 and originated by Caudill.27 Modifications of the instrument included the omission, addition, and clarification of certain food items to make the questionnaire appropriate to the cultural customs of the region and participant population. Subjects were asked to indicate the number of times they consumed specific portions of various food items over the previous week.Various portion sizes were listed next to each food item (portions sizes were expressed in this assessment in a manner identical to those used in the Food Guide Pyramid, ie, cups, ounces, slices, tablespoons, etc), and measuring utensils were present during the administration of the questionnaire to illustrate actual portion sizes to aid in recording accuracy. Food analyses were performed using The Food Processor.28 The intake of fruits and vegetables was calculated using serv-
264
Abood et al/NUTRITION EDUCATION WORKSITE INTERVENTION BASED ON THE HBM
ing sizes adopted from the USDA’s Food Guide Pyramid.3 Standard program outputs are total energy, grams of fat, percentage of fat, grams of saturated fat,milligrams of cholesterol, servings of fruits and vegetables, and grams of dietary fiber; all were shown to be associated with CVD and various cancers. Psychometrics A group of 18 graduate students in dietetics provided written and verbal feedback about the questionnaire regarding wording, clarity of items, and response bias. Items were revised based on students’ suggestions. Two nutrition professors also reviewed the instruments after revision for content validity, which resulted in rewording 5 items. No items were discarded based on the feedback of the students or the faculty members. The reading level of the final version of the questionnaire, which was determined by the Flesch-Kincaid Grade Level Index, was the 8th month of the sophomore year of high school.Test–retest reliability was established by administering the questionnaire twice to 20 full-time employees, 1 week apart, who were not participants in the study. The KuderRichardson coefficients used to assess test–retest reliability were .84, .86, .96, and .79, respectively, for health belief, health assessment, nutrition knowledge, and dietary behaviors. Cronbach α and test–retest reliability, respectively, for each of the HBM constructs were as follows: (a) health concern, .80 and .82; (b) perceived susceptibility, .82 and .84; (c) perceived severity, .74 and .79; (d) perceived benefits, .81 and .81; (e) perceived barriers, .82 and .80; and (f) self-efficacy, .75 and .78. Research Design and Statistical Analyses A 2 2 repeated measures baseline/posttest ex post facto research design was selected.29 Data from subjects who participated in > 5 treatment sessions were used for analyses to avoid prospects of a type III error and issues related to an insufficient dose-response relationship.30-32 After selecting data for subjects who met the treatment session criterion, the following initial analyses were computed: (a) normality and homogeneity to test parametric assumptions and (b) multiple regression analyses to assess whether demographic variables, HBM constructs, health assessment, nutrition knowledge, and dietary behaviors were significantly related to group membership. Second, repeated measures analyses of variance (ANOVAs) were used to identify significant differences between the treatment and control groups in HBM constructs, health assessment, nutrition knowledge, and dietary behaviors. Third, the efficacy of intervention was assessed based on “specificity.”30 Specificity or specific intervention effect means that only those constructs or variables of primary intervention emphasis should change significantly and the effect size for those constructs/variables only should be large enough to produce adequate power, in contrast to those variables of secondary focus that should not change significantly, and power should be minimal to marginal. The criteria for specificity were significant (< .05) ANOVA results and a power based on Cohen’s33 power tests
of > 80% (when α was set at .05) for each variable of emphasis. Last, Pearson product-moment correlations were computed to note relationships between salient variables and primary intervention constructs/variables.
DESCRIPTION OF THE FINDINGS Initial Analyses None of the parametric assumptions for any of the parametric analyses were violated.The results revealed no significant initial associations of demographic variables, health beliefs, health assessment results, nutrition knowledge, and dietary behaviors and group membership. Subject Characteristics Thirty-eight treatment subjects completed the baseline/ posttest questionnaires. Of the 38, 28 (73.7%) actually participated in the intervention by completing > 5 sessions. Midterm ratings showed that all participants, including those who dropped out, were “very” to “extremely satisfied” with the intervention. All 25 control subjects completed the posttest questionnaire.The average age for subjects in the treatment and control groups was 34.3 and 37.9 years old, respectively. Women comprised the majority of subjects in each group, 96% and 92%, respectively. Education levels for both groups were nearly equally distributed, with 28% and 29%, respectively, reporting the highest education level being a high school degree, 25% and 20% with a vocational or 2-year degree, 21% and 20% with a 4-year college degree, and 25% and 32% with a graduate degree. Approximately 81% and 82% of the treatment and control groups, respectively, earned $34 000 or less. The majority of the participants in both groups were white (57% and 60%, respectively), whereas African American and Hispanic participants accounted for 36% and 32% and 7% and 8%, respectively. In comparison with the population from which it was drawn, not surprisingly, the sample was slightly younger and predominantly female. Specificity Health beliefs. Table 2 shows that only 1 of the 2 HBM constructs of primary emphasis during the intervention changed significantly. Perceived benefits concerning the adoption of positive dietary behaviors increased significantly after treatment, but there was no significant change in perceived barriers, although there was an apparent reduction in the expected direction. Nutrition knowledge. There was a significant increase in nutrition knowledge related to CVD and cancer for the treatment group (Table 2).There was also an association between nutrition knowledge and higher fiber intake, r (26) = .49, P < .005, and between nutrition knowledge and consuming a
Journal of Nutrition Education and Behavior Volume 35 Number 5
lower percentage of total fat and saturated fat energy, r (26) = .55, .52; P < .005, respectively. Dietary behavior. Inspection of Table 2 and comparison of these data with national databases reveal that, at baseline, the treatment group’s mean energy (2264 kcal), fat (92 g/day), cholesterol (332 mg/day), saturated fat (31 g/day), and percentage of energy from fat (35/day) intake were higher than the recommended levels. 34 Following the intervention, energy intake decreased by approximately 840 kcal/day, total fat intake decreased by 45 g/day, and saturated fat and cholesterol dropped by 18 mg/day and 158 mg/day, respectively. These significant reductions are similar to those interventions conducted by others, who reported a 17% (versus 14% in this study) reduction in the percentage of calories from fat,35 similar percentages of energy from fat at baseline, as in this study (34% versus 35% in this study), decreases in the percentage of energy from fat at posttest (12% versus 14% in this study),36 similar reductions in fat grams (36 g versus 44 g in this study),37 equivalent decreases in energy from fat (13% versus 14% in this study),38 and a greater reduction in total fat than that reported in this study (from 79.3 g to 34.8 versus 35 g to 21 g reported in this study).39 Power analyses. The power of detecting a significant difference when the α was set at .05 was > 92% for every one of the constructs/variables of intervention emphasis. All other constructs/variables not of intervention focus were low or marginal power (~10% to ~36%), as expected.
LESSONS LEARNED The goal was to evaluate a theory-based, tailored nutrition intervention for staff employees at a university campus worksite.The intervention was designed to change dietary behavior in a positive manner by systematically modifying specific health beliefs, in particular the benefits and barriers related to making certain dietary changes, while increasing nutrition knowledge and improving dietary behaviors related to CVD and cancer. Perceived benefits of changing dietary behaviors significantly increased. Perceived barriers was the only 1 of the 8 constructs/variables of intervention focus expected to change that did not. The intervention was successful and appears to be associated with producing significant increases in nutrition knowledge and significantly decreasing energy, fat, saturated fat, and cholesterol intake to levels consistent with national recommendations.The intervention was successful in attracting and retaining both African-American and Hispanic participants, who together comprised 43% and 40% of the treatment and control groups, respectively. This program, therefore, may be appropriate for future samples with similar demographic characteristics. An educational feature of the program and an implication for practice was providing participants with an analysis of
September • October 2003
265
their nutrition knowledge, dietary behaviors, and a risk factor score for CVD and cancer prior to the 8-week nutrition education intervention. A priori appraisals of participants’ nutrition knowledge, dietary intake status, and health risks have been conducted in other studies.40,41 Focusing on perceived benefits and perceived barriers was based on the research literature and baseline analyses, which revealed that participants reported more barriers for decreasing fat intake than increasing fruit and vegetable intake. Consequently, the focus of the intervention was on lowering fat, with the anticipation that behavior maintenance or self-change would spontaneously occur in regard to fruits, vegetables, and fiber. The positive change in decreasing fat is consistent with the focus and emphasis of the intervention and instruction. However, only about half of the treatment group reported an increase in fruit or vegetable intake after the intervention. Anticipated spontaneous changes did not occur, which suggests that fruit and vegetable intake needs to be priority independent of initial baseline assessment scores. There may be other reasons fruit and vegetable consumption needs to be emphasized that are related to the negative influences of personal, social, and environmental barriers. One reason is that dieters are probably more familiar and comfortable with restriction of food intake than with adding foods to a diet.Thus, lowering total fat intake may have been easier than adding fruits or vegetables.Adding new behaviors is considered more difficult than avoiding or limiting a behavior.18 A second reason that lowering fat consumption might have been easier is due to the amount of publicity about the health benefits of a lower-fat diet. In contrast, the National Cancer Institute and the National Institutes of Health42 5 A Day program is fairly new, and there has been far less publicity regarding the benefits of adding fruits and vegetables. A third reason is that participants may have believed that a decrease in fat and calorie intake was the most efficient way to produce significant weight loss.The last reason is caloric restriction.A significant decrease in calories may have yielded a diet containing less healthful nutrients, especially if fruit and vegetable intake failed to increase.This problem, however, was not manifested in this study.The findings of the current study are congruent with Herbert and colleagues43 and Peterson and colleagues,44 who found that participants who significantly decreased their energy and fat intake did not necessarily sacrifice micronutrients. The present study may have limitations. Because diet was not observed, self-report bias may have occurred. Nevertheless, self-report is the norm in nutrition “field studies” because of practicality and “realism” in terms of time demands on voluntary study participants and the negative effects on recruitment and retention when time demands initially appear excessive. Local norms were not developed for the food frequency questionnaire. Norms developed elsewhere may not be representative of local customs and habits. Food frequency questionnaire results should be viewed as estimates of change, but, most likely, conclusions would remain unchanged even if it were presumed that absolute
266
Abood et al/NUTRITION EDUCATION WORKSITE INTERVENTION BASED ON THE HBM
values would vary only slightly. Also, it should be noted that the durability of the intervention effect is unknown, and after demonstrating an intervention effect, as occurred in this study, collection of follow-up data would be the next priority in future investigations. In addition, it should be noted that retention was somewhat of a problem, despite making every effort to be accommodating and sensitive to work demands by offering sessions at multiple times during the week. Some participants found it difficult to attend sessions. The reason they gave was having too much work to do and the lure of being able to “catch up” during the hour of the intervention. Future studies also should address the challenges of effectively promoting the adoption of healthful food behaviors, especially those that require the addition of a new food behavior, such as increasing fruits and vegetables. Interventions promoting such behaviors might further concentrate on HBM variables. The employment of psychosocial models might enhance the effectiveness of nutrition interventions that promote the addition of specific foods. In conclusion, this study addresses an area in which there is an exiguity of research by focusing on the application of an 8-week worksite-tailored nutrition education intervention for staff employees at a university. The HBM was successfully used as a theoretical framework to encourage healthful dietary behavior change by increasing nutrition knowledge and promoting the benefits of making prudent dietary changes.This program may serve as a framework for future applications of such programs among this population of worksite employees and as an integral part of the total national effort to reduce the risk of CVD and cancer.
ACKNOWLEDGMENT This article is dedicated to Mr. George Abood in respectful memory and appreciation for his contributions and inspirations to improve the health and welfare of humankind whenever and however possible.
REFERENCES 1. McGinnis JM, Foege WH. Actual causes of death in the United States. JAMA. 1993;270:2207-2212. 2. US Dept of Agriculture and the US Dept of Health and Human Services. Dietary Guidelines for Americans. Washington, DC: US Government Printing Office; 1995. 3. US Dept of Agriculture, Human Nutrition Information Service. The Food Pyramid. Washington, DC: US Government Printing Office; 1992. Home and Garden Bulletin No. 249. 4. Glanz K, Patterson RE, Kristal AR, et al. Stages of change in adopting healthy diets: fat, fiber, and correlates of nutrient intake. Health Educ Q. 1994;21:499-519. 5. Glanz K. Behavioral research contributions and needs in cancer prevention and control: dietary change. Prev Med. 1997;26(suppl 1):43-55.
6. Goetzel RZ, Jahr TY, Aldana SG, Kenny GM. An evaluation of Duke University’s Live for Life health promotion program and its impact on employee health. Am J Health Promotion. 1996;10:340-342. 7. Baer JT. Improved plasma cholesterol levels in men after a nutrition education program at the worksite. J Am Diet Assoc. 1991;91: 1391-1397. 8. Glanz K, Seewald-Klein T. Nutrition at the worksite: an overview. J Nutr Educ. 1986;18(suppl 1): S1-S12. 9. Sorenson G, Himmelstein J. Worksite nutrition intervention. In: Ockene I, Ockene J, editors. Prevention of Coronary Heart Disease. Boston: Little, Brown; 1992:517-539. 10. Hochbaum GM. Public participation in medical screening programs: a sociopsychological study. Public Health Service. no. 572, 1958. 11. Rosenstock IM.What research in motivation suggests for public health. Am J Public Health. 1960;50:295-301. 12. Rosenstock IM.Why people use health services. Milbank Mem Fund Q. 1996;44:94-124. 13. Rosenstock IM. Historical origins of the health belief model. Health Educ Monogr. 1974;2:328-335. 14. Bandura A. Social Learning Theory. Englewood Cliffs, NJ: Prentice-Hall; 1977. 15. Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the health belief model Health Educ Q. 1988;15:175-188. 16. Fries EA, Ripley JS, Jigueiredo MI, et al. Can community organization strategies be used to implement smoking and dietary changes in a rural manufacturing work site? J Rural Health. 1999;15:413-420. 17. US Dept of Health and Human Services. National Cholesterol Program: Second Report of the Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults [Adult Treatment Panel III]. Washington, DC: National Institutes of Health; 1993. No. 93-3093. 18. Damrosch S. General strategies for motivating people to change their behavior. Nurs Clin North Am. 1991;26:833-843. 19. Anderson AS, Cox DN, McKellar S, et al. Five, a nutrition education intervention to increase fruit and vegetable intakes: impact on attitudes towards dietary change. Br J Nutr. 1998;90:133-140. 20. McCullock-Melnyk KA. Barriers: a critical review of recent literature. J Nurs Res. 1988;37:196-201. 21. Kayman S. Applying theory from social psychology and cognitive behavioral psychology to dietary behavior change and assessment. J Am Diet Assoc. 1989;89:191-193. 22. American Cancer Society. Cancer:Assessing Your Risk [Brochure].Atlanta, Ga, ACS, 1993. 23. American Heart Association. Test Your Cholesterol IQ [Brochure]. Dallas, Tex, AHA, 1991. 24. Fahey TD, Insel PM, Roth WT.Wellness worksheet: are you at risk for cardiovascular disease? In: Core Concepts in Health. Mountain View, Calif: Mayfield Publishing Company; 1994. 25. Hurley J. Fat or fiction? Nutrition Action Health Letter. 1994;10:5. 26. Boeckner LS, Kohn H, Rockwell K.A risk-reduction course for adults. J Am Diet Assoc. 1990;90:260-263. 27. Caudill E. Food shopping behavior and food item frequency of work mothers [master’s thesis]. Lincoln: University of Nebraska; 1979. 28. The Food Processor,Version 6.03 [nutrition analysis software]. Salem, Ore: ESHA Research; 1995. 29. Rubinson L, Neutens JJ. Research Techniques for the Health Sciences. New York: Macmillan; 1987.
Journal of Nutrition Education and Behavior Volume 35 Number 5 30. Black DR, Tobler N, Sciacca JP. Peer helping/involvement: an efficacious way of meeting the challenge of reducing illicit drug use? J School Health 1998;68:87-93. 31. Windsor R, Baranowski T, Clark N, et al. Evaluation of Health Promotion, Health Education, and Disease Prevention Programs. 2nd ed. Mountain View, Calif: Mayfield; 1994. 32. Timmreck TC. An Introduction to Epidemiology. 3rd ed. Boston: Jones and Bartlett; 2002. 33. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 1988. 34. Tippett KS, Cypel YS, eds. Continuing Survey of Food Intakes by Individuals and the Diet and Health Knowledge Survey, 1994-1996.Washington, DC: US Dept of Agriculture, Agricultural Research Service; 1998. 35. Simon MS, Heilbrun LK, Boomer A, et al.A randomized trial of a lowfat dietary intervention in women at high risk for breast cancer. Nutr Cancer. 1997;27:136-142. 36. Boyd NF, Lockwood GA, Greenberg LJ, et al. Effects of a low-fat high carbohydrate diet on plasma sex hormones in premenopausal women: results from a randomized controlled trial. Br J Cancer. 1997;76:127-135. 37. Coates RJ, Bowen DJ, Kristal AR, et al.The Women’s Health Trial Feasibility Study in Minority Populations: changes in dietary intakes. Am J Epidemiol. 1999;149:1104-1112.
September • October 2003
38. Gorbach SL, Morrill-LaBrode A,Woods MN, et al. Changes in food patterns during a low-fat dietary intervention in women. J Am Diet Assoc. 1990;90:802-809. 39. Buzzard MI,Asp EH, Chlebowski RT, et al. Diet intervention methods to reduce fat intake: nutrient and food group composition of selfselected low-fat diets. J Am Diet Assoc. 1990;90:42-50, 53. 40. Contento IR, Murphy BM. Psycho-social factors differentiating people who reported making desirable changes in their diets from those who did not. J Nutr Educ. 1990;22:6-13. 41. Strychar IM, Potvin L, Pineault R, et al. Change in knowledge and food behaviour following a screening program held in a supermarket. Can J Public Health. 1993;84:382-388. 42. National Cancer Institute and National Institutes of Health. 5 a Day for Better Health: A Baseline Study of American’s Fruit and Vegetable Consumption. Rockville, Md: National Institutes of Health; 1992. 43. Herbert JR, Harris DR, Sorensen G, et al. A work-site nutrition intervention: its effects on the consumption of cancer-related nutrients. Am J Public Health. 1993;83:391-394. 44. Peterson S, Sigman-Grant M, Eissenstat B, et al. Impact of adopting lower-fat food choices on energy and nutrient intakes of American adults. J Am Diet Assoc. 1999;99:177-183.
VISION, MISSION, AND GUIDING PRINCIPLES OF THE SOCIETY FOR NUTRITION EDUCATION Vision Healthy people in healthy communities. Mission To enhance nutrition educators’ ability to promote healthful sustainable food choices and nutrition behaviors. Guiding Principles • • • • •
267
Fiscal responsibility Respect for diversity of opinions and perspectives Trust and willingness to communicate openly and respectfully Knowledge-based decisions Excellence and lifelong learning
• • • •
Professionalism and integrity Inclusiveness in membership Equality among members Rewarding and enjoyable experiences for volunteers and supporters
Nutrition Education Worksite Intervention for University Staff: Application of the Health Belief Model D O R I S A. A B O O D, E D D, CHES; 1 D AV I D R. B L AC K , P H D, HSPP, CHES, MPH, CPPE, FASHA, FSBM, FAAHB; 2 D I A N E F E R A L , MS, RD 1 1
Department of Nutrition, Food, and Exercise Sciences, Florida State University,Tallahassee, Florida; 2 Department of Health and Kinesiology, Purdue University,West Lafayette, Indiana
and decreased energy, fat, saturated fat, and cholesterol intake to levels consistent with national recommendations.
ABSTRACT
Objective: To evaluate the efficacy of an 8-week worksite nutrition education intervention for university staff using the Health Belief Model (HBM) to promote healthful dietary behaviors that reduce risks for cardiovascular disease and cancer.
KEY WORDS: nutrition education, worksite, Health Belief Model, university staff (J Nutr Educ Behav. 2003;35:260-267.)
Design: 2 2 repeated measures baseline/posttest ex post facto research design.
INTRODUCTION
Participants: Staff employees were randomly assigned to treatment (n = 28) and control groups (n = 25).
Of the 10 leading causes of death in the American adult population, 4 are directly related to nutrition: coronary heart disease, cancer, stroke, and diabetes mellitus.1 In response to these chronic diseases, the US government and other groups have introduced prevention programs. Dietary changes are one means of disease prevention, and the Dietary Guidelines for Americans have been incorporated into nutrition programs to improve dietary intake.2,3 Dietary behavioral changes have become a focal point in the past decade, and significant dietary changes have been reported in clinical trials using selected men and women at high risk for cardiovascular disease (CVD) and cancer.4 Generalizing from these trials, however, is difficult because of rigorous inclusion criteria and protocols, and the programs implemented are often inappropriate for application to the general population. Community-based interventions are needed to determine which approaches might be effectively used to prevent CVD and cancer among free-living populations.5 Worksites, such as university campuses, can be considered communities.Worksite interventions are important because employees are currently facing large increases in health insurance costs. Nutrition education at worksites may be effective in improving health-related behaviors so that insurance costs, sick days, and turnover can be reduced, and employee productivity and morale can be increased.6,7 Worksite interventions may be advantageous too because they are convenient and accessible to workers, allow employees to support one another in their efforts to make behavior changes, and are less expensive, especially compared with programs offered in clinical settings.8,9
Intervention: The intervention focused on specific health beliefs, nutrition knowledge, and dietary practices to demonstrate treatment effect. Main Outcome Measures: Dependent variables were specific health beliefs, nutrition knowledge, and dietary behaviors. Independent variables were demographic characteristics and group assignment. Analyses: Tests of parametric assumptions, power analyses, analysis of variance, and Kuder-Richardson and Pearson product-moment coefficients were computed and specificity of treatment effects was assessed. Results: Perceived benefits of healthy nutrition practices and nutrition knowledge related to cardiovascular disease and cancer significantly improved among the treatment participants, P < .001. Treatment group participants also significantly reduced total calories, fat, saturated fat, and cholesterol intake (each P < .001). Conclusions: The intervention appears to be related to treatment effects and significantly increased nutrition knowledge
Address for correspondence: Doris A.Abood, EdD, CHES, Department of Nutrition, Food and Exercise Sciences, Florida State University, 408 Sandels Building, Tallahassee, FL 32306-1493;Tel: (850) 644-4796; Fax: (850) 645-5000; E-mail: dabood@ mailer.fsu.edu. ©2003 SOCIETY FOR NUTRITION EDUCATION
260
Journal of Nutrition Education and Behavior Volume 35 Number 5
Behavioral theory has increasingly been used to guide nutrition research to improve intervention efficacy.4 One such theoretical framework that has been applied is the Health Belief Model (HBM). The HBM was developed in the 1950s to explain health behavior associated with the failure of people to participate in programs that would reduce disease risk.10-13 The HBM implies that health behaviors are determined by health beliefs and readiness to take action.10 Constructs central to the HBM consist of perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and other mediating variables. The construct of selfefficacy is frequently included in applications of the HBM.14,15 The inclusion of self-efficacy in the HBM assumes that the likelihood of taking action is not only a function of beliefs related to outcomes but also a function of a person’s belief that he or she is behaviorally capable of achieving the desired outcome.14,15 Interventions that also emphasize selfefficacy may increase the likelihood that positive behavioral changes will be executed. A review of the literature suggests that there is a void in research that focuses on university campuses as worksites and theory-based, tailored nutrition interventions for employees. An electronic search using several databases (MEDLINE, PubMed, and ERIC using the dates 1999 through April 2003) did not produce any nutrition education interventions that were theory based and focused on university worksite employees. Broadening the search to include only single-worksite interventions produced only one study of a rural manufacturing site that used community organization strategies to produce changes in fruit and vegetable consumption.16 No studies were found of single worksites that tailored the nutrition education intervention to participants’ baseline knowledge, dietary behaviors, or health beliefs. The purpose of this study was to evaluate the efficacy of a worksite-tailored nutrition education program for university staff employees.The program used the HBM as a theoretical framework to increase nutrition knowledge and improve dietary intake via instruction and dietary exercises. The nutrition education program was designed as an initial prototype to aid in the prevention of CVD and cancer among university worksite employees.
DESCRIPTION OF THE INTERVENTION AND EVALUATION Intervention The treatment group participated in 8 1-hour weekly educational sessions. A registered dietitian who was a member of the research team provided all instruction. Three invariant education sessions were held each week to provide employees with maximum opportunity to attend sessions convenient to their work schedules. Subjects were prohibited from attending more than 1 session per week. The focus of the intervention was to promote knowledge and beliefs con-
September • October 2003
261
ducive to improving or maintaining positive dietary practices for prevention of CVD and cancer.The sessions were didactic interspersed by questions and answers, and information was presented by means of computerized overhead projection, displays, and paper materials.A detailed outline also was distributed at the beginning of each session to provide an overview and for note taking.An overview of the nutritional intervention is presented in Table 1, which shows the topics and content and how constructs of the HBM were integrated into each session of the program.All dietary information presented and recommendations provided were consistent with the US Department of Agriculture’s (USDA) Food Guide Pyramid and the US Department of Health and Human Services’ National Cholesterol Program Guidelines.3,17 The posttest was administered at the end of the intervention. Tailoring the intervention specifically to the target population occurred in two ways: baseline test assessments and prior research findings. Tailoring was based on baseline assessment of nutrition knowledge and dietary behavior status.This assessment was used to guide content selection and determined the information to emphasize in each session. For example, in terms of nutrition knowledge, it was noted that more than 50% of participants incorrectly answered questions pertaining to risk factors associated with CVD and cancer, sources of protein and cholesterol, macronutrient recommendations (percentage of energy), facts related to margarine versus butter, fruit and vegetable intake recommendations, benefits of exercise, hidden sources of fat, the difference between 2% and whole milk, and that cholesterol comes only from animal products. Nutrition education instruction focused on increasing knowledge in these areas. Mean values of initial dietary behavior also revealed that the priority of focus should be on fat and cholesterol intake and secondarily on fruit and vegetable consumption (Table 2). Tailoring of the intervention also was based on HBM constructs. Even though all HBM constructs were addressed during the intervention, perceived benefits and perceived barriers were given priority despite the mean values of perceived benefits falling into the high category and the mean values of perceived barriers falling into the moderate category in comparison with normative values (see Table 2).The reason perceived benefits and perceived barriers were given primary consideration is because prior research has shown perceived susceptibility to be a strong predictor of behavior when a person believes that behavior change would be accompanied by perceived benefits that outweigh perceived barriers.18 In other words, theory would suggest that for a change in nutrition behavior to occur, the perceived benefits of the behavior must outweigh the perceived costs (barriers). An extensive review of the literature has shown that the strongest predictor of nutrition-related behavior change is the benefit-cost ratio.18 Barriers reported in other studies specific to healthy eating, which were the focus of this intervention, included perceived taste, lack of convenience/time, high cost, confusing advertising and diet recommendations, and lack of knowledge of actual food intake.18-21
262 Table 1.
Abood et al/NUTRITION EDUCATION WORKSITE INTERVENTION BASED ON THE HBM Overview of Nutrition Intervention
Session
Topic
Content
1
Development and Medical, Clinical, and Social Consequences of CVD and Cancer; Risk Reduction
Risk factors and prevalence rates; health consequences; health assessment of risk for CVD and cancer; USDA Food Guide Pyramid and proper nutrition to reduce risks; dietary analysis
2
Macronutrients
Macronutrients (fats, carbohydrates, and protein); Food Guide Pyramid and sources and benefits of recommended intakes; importance and benefits of proper nutrition; reducing barriers to increase probability of dietary changes; interpretation of health assessment results and dietary analysis
3
Macronutrients
Macronutrients and types of fat; hidden sources of fat, low-fat foods, meals, and fat alternatives; benefits of saturated fat reduction and reduction of barriers to taking such actions
4
Fruits and Vegetables
Health-protective role of fruits and vegetables; frequency and portion sizes; fiber, vitamins, minerals, antioxidants, and phytochemicals; benefits of increased fruit and vegetable intake and barrier reduction to taking action
5
Physical Activity and Weight Control
Health benefits of weight control and exercise, including types of exercise; barrier reduction; macronutrients and energy metabolism
6
Meal Patterns and Healthy Eating
Benefits of eating meals regularly; distribution and preparation of simple/convenient low-calorie/highnutrient recipes; ideas for removing barriers to healthful eating patterns
7
Meal Planning and Food Label Reading
Scheduling meals, shopping lists, label reading, cost cutting, food storage, and food preparation
8
Integration of All Previous Topics for a Prescription of Healthy Living
HBM constructs to change nutrition behaviors to reduce risks and for behavior maintenance; supplements, salt, caffeine intake, soft drinks
CVD indicates cardiovascular disease; HBM, Health Belief Model; USDA, US Department of Agriculture.
Control Group Subjects in the treatment and control groups were “blinded” to each other to reduce “contamination effects.”The control group, which was physically housed in a separate building, completed questionnaires only at baseline and postintervention. Subjects in the control group continued to work during the time in which the treatment group received the intervention.The control group received an abbreviated version of the nutrition education intervention 1-month postintervention.
trators granted approval for and agreed to provide 1-hour release time each week for all full-time employees who wished to participate in the intervention.A self-selected sample of subjects employed by the admissions and personnel department or the dean of students office participated, and the group that received treatment or served as the control group was selected randomly as a means to enhance the internal validity of the study. The institutional committee on the use of human research subjects approved the study, and all participants signed a consent form to participate.
Recruitment of Subjects
Data Collection and Instrumentation
Heads of administrative departments at a southern university were approached by the researchers and were asked to endorse a nutrition education program for their employees. Adminis-
The baseline/posttest questionnaire that took between 30 and 45 minutes to complete included 171 items and was divided into 4 sections. Section I was assessment of 5 HBM
Journal of Nutrition Education and Behavior Volume 35 Number 5
September • October 2003
263
Table 2. Means/Standard Deviations to Assess Intervention Effects Baseline Constructs/Variables
Treatment
Posttreatment Control
Treatment
Control
Health Beliefs† Health concern
2.03/1.48
1.88/1.50
1.85/1.40
1.84/1.50
Perceived susceptibility
6.10/1.37
5.96/1.45
5.92/1.42
6.16/1.49
Perceived severity
3.28/1.27
2.76/1.37
2.75/1.44
2.84/1.42
Perceived benefits
4.25/1.03
4.40/1.50
3.17/1.12***
4.80/1.75
Perceived barriers
5.53/1.01
5.40/1.35
4.53/1.00
5.28/1.30
Self-efficacy
4.07/2.03
4.60/2.43
3.57/2.03
4.24/2.37
Nutrition Knowledge Nutrition knowledge assessment
12.80/3.92
13.36/2.95
17.36/4.97***
13.30/2.96
Dietary Behavior Total energy (kcal)
2264/1077
1891/968
1427/503***
1919/884
Total fat (g)
91.99/56.02
81.78/46.48
47.43/30.41***
90.80/51.99
Total protein (g)
84.63/41.89
74.72/42.31
64.89/23.96
70.68/25.65
Fat energy (%)
35.36/9.18
36.25/7.37
21.18/7.63***
40.12/10.84
Total saturated fat (g)
31.38/20.38
28.40/15.43
13.50/5.48***
32.20/18.31
Total cholesterol (mg)
332.23/216.29
227.01/142.19
174.43/106.01***
266.64/154.20
Total fiber (g)
16.93/9.48
14.40/6.65
14.43/9.81
15.68/11.10
Vegetable (servings)
2.36/1.80
2.09/1.22
2.54/1.64
1.80/1.78
Fruit (servings)
2.45/2.19
1.63/1.17
2.79/2.39
1.34/1.17
Treatment group, n = 28; control group, n = 25. ***p < .001. †Health Belief Model score interpretation: health concern: ≤ 2 = high, 3 = moderate, ≥ 4 = low; perceived susceptibility, severity, benefits, barriers, and self-efficacy: ≤ 4 = high, 5-7 = moderate, ≥ 8 = low.
constructs as they relate to nutrition, CVD, and cancer.These 6 constructs were health concerns, perceived susceptibility, perceived severity, perceived benefits, perceived barriers, and self-efficacy. Subjects rated items related to each of these constructs on a 5-point Likert-type scale from 1 (strongly agree) to 5 (strongly disagree). Negatively worded questions were reverse-scored. Lower scores reflect a higher degree of health beliefs. Section II of the questionnaire was the personal health assessment of risks for CVD and cancer.This portion of the questionnaire was developed using a combination of modified questions from the American Cancer Society,22 the American Heart Association,23 and Fahey and colleagues24 and included questions about family history of CVD and cancer and the risk factors for these diseases, such as smoking status, blood pressure, cholesterol, exercise, and diabetes (body weight was not assessed). Questions assessing cancer risk were related to breast, lung, and colorectal cancer because of their high prevalence rates in the United States and the established links to diet. Risk scores for CVD and cancer were individually computed by summing all responses related to each disease. Higher scores represented a higher risk of CVD, cancer, or both. The personal health assessment was used only for the initial assessment to be respectful of participants’ time and because dramatic changes in health status were unexpected in the short term.
Section III was the nutrition knowledge assessment, which was developed specifically for this sample and intervention to assess nutrition knowledge related to CVD and cancer risk. Some questions were adapted from a nutrition quiz developed by Hurley. 25 Question formats were true/false and multiple choice. Questions addressed aspects of protein, fat, carbohydrates, food sources of cholesterol, recommended intakes of fruits and vegetables, and healthful ways of reducing saturated fat, among other topics. One point was given for each correct answer. Section IV assessed dietary behaviors based on a modified food frequency instrument used by Boeckner and colleagues26 and originated by Caudill.27 Modifications of the instrument included the omission, addition, and clarification of certain food items to make the questionnaire appropriate to the cultural customs of the region and participant population. Subjects were asked to indicate the number of times they consumed specific portions of various food items over the previous week.Various portion sizes were listed next to each food item (portions sizes were expressed in this assessment in a manner identical to those used in the Food Guide Pyramid, ie, cups, ounces, slices, tablespoons, etc), and measuring utensils were present during the administration of the questionnaire to illustrate actual portion sizes to aid in recording accuracy. Food analyses were performed using The Food Processor.28 The intake of fruits and vegetables was calculated using serv-
264
Abood et al/NUTRITION EDUCATION WORKSITE INTERVENTION BASED ON THE HBM
ing sizes adopted from the USDA’s Food Guide Pyramid.3 Standard program outputs are total energy, grams of fat, percentage of fat, grams of saturated fat,milligrams of cholesterol, servings of fruits and vegetables, and grams of dietary fiber; all were shown to be associated with CVD and various cancers. Psychometrics A group of 18 graduate students in dietetics provided written and verbal feedback about the questionnaire regarding wording, clarity of items, and response bias. Items were revised based on students’ suggestions. Two nutrition professors also reviewed the instruments after revision for content validity, which resulted in rewording 5 items. No items were discarded based on the feedback of the students or the faculty members. The reading level of the final version of the questionnaire, which was determined by the Flesch-Kincaid Grade Level Index, was the 8th month of the sophomore year of high school.Test–retest reliability was established by administering the questionnaire twice to 20 full-time employees, 1 week apart, who were not participants in the study. The KuderRichardson coefficients used to assess test–retest reliability were .84, .86, .96, and .79, respectively, for health belief, health assessment, nutrition knowledge, and dietary behaviors. Cronbach α and test–retest reliability, respectively, for each of the HBM constructs were as follows: (a) health concern, .80 and .82; (b) perceived susceptibility, .82 and .84; (c) perceived severity, .74 and .79; (d) perceived benefits, .81 and .81; (e) perceived barriers, .82 and .80; and (f) self-efficacy, .75 and .78. Research Design and Statistical Analyses A 2 2 repeated measures baseline/posttest ex post facto research design was selected.29 Data from subjects who participated in > 5 treatment sessions were used for analyses to avoid prospects of a type III error and issues related to an insufficient dose-response relationship.30-32 After selecting data for subjects who met the treatment session criterion, the following initial analyses were computed: (a) normality and homogeneity to test parametric assumptions and (b) multiple regression analyses to assess whether demographic variables, HBM constructs, health assessment, nutrition knowledge, and dietary behaviors were significantly related to group membership. Second, repeated measures analyses of variance (ANOVAs) were used to identify significant differences between the treatment and control groups in HBM constructs, health assessment, nutrition knowledge, and dietary behaviors. Third, the efficacy of intervention was assessed based on “specificity.”30 Specificity or specific intervention effect means that only those constructs or variables of primary intervention emphasis should change significantly and the effect size for those constructs/variables only should be large enough to produce adequate power, in contrast to those variables of secondary focus that should not change significantly, and power should be minimal to marginal. The criteria for specificity were significant (< .05) ANOVA results and a power based on Cohen’s33 power tests
of > 80% (when α was set at .05) for each variable of emphasis. Last, Pearson product-moment correlations were computed to note relationships between salient variables and primary intervention constructs/variables.
DESCRIPTION OF THE FINDINGS Initial Analyses None of the parametric assumptions for any of the parametric analyses were violated.The results revealed no significant initial associations of demographic variables, health beliefs, health assessment results, nutrition knowledge, and dietary behaviors and group membership. Subject Characteristics Thirty-eight treatment subjects completed the baseline/ posttest questionnaires. Of the 38, 28 (73.7%) actually participated in the intervention by completing > 5 sessions. Midterm ratings showed that all participants, including those who dropped out, were “very” to “extremely satisfied” with the intervention. All 25 control subjects completed the posttest questionnaire.The average age for subjects in the treatment and control groups was 34.3 and 37.9 years old, respectively. Women comprised the majority of subjects in each group, 96% and 92%, respectively. Education levels for both groups were nearly equally distributed, with 28% and 29%, respectively, reporting the highest education level being a high school degree, 25% and 20% with a vocational or 2-year degree, 21% and 20% with a 4-year college degree, and 25% and 32% with a graduate degree. Approximately 81% and 82% of the treatment and control groups, respectively, earned $34 000 or less. The majority of the participants in both groups were white (57% and 60%, respectively), whereas African American and Hispanic participants accounted for 36% and 32% and 7% and 8%, respectively. In comparison with the population from which it was drawn, not surprisingly, the sample was slightly younger and predominantly female. Specificity Health beliefs. Table 2 shows that only 1 of the 2 HBM constructs of primary emphasis during the intervention changed significantly. Perceived benefits concerning the adoption of positive dietary behaviors increased significantly after treatment, but there was no significant change in perceived barriers, although there was an apparent reduction in the expected direction. Nutrition knowledge. There was a significant increase in nutrition knowledge related to CVD and cancer for the treatment group (Table 2).There was also an association between nutrition knowledge and higher fiber intake, r (26) = .49, P < .005, and between nutrition knowledge and consuming a
Journal of Nutrition Education and Behavior Volume 35 Number 5
lower percentage of total fat and saturated fat energy, r (26) = .55, .52; P < .005, respectively. Dietary behavior. Inspection of Table 2 and comparison of these data with national databases reveal that, at baseline, the treatment group’s mean energy (2264 kcal), fat (92 g/day), cholesterol (332 mg/day), saturated fat (31 g/day), and percentage of energy from fat (35/day) intake were higher than the recommended levels. 34 Following the intervention, energy intake decreased by approximately 840 kcal/day, total fat intake decreased by 45 g/day, and saturated fat and cholesterol dropped by 18 mg/day and 158 mg/day, respectively. These significant reductions are similar to those interventions conducted by others, who reported a 17% (versus 14% in this study) reduction in the percentage of calories from fat,35 similar percentages of energy from fat at baseline, as in this study (34% versus 35% in this study), decreases in the percentage of energy from fat at posttest (12% versus 14% in this study),36 similar reductions in fat grams (36 g versus 44 g in this study),37 equivalent decreases in energy from fat (13% versus 14% in this study),38 and a greater reduction in total fat than that reported in this study (from 79.3 g to 34.8 versus 35 g to 21 g reported in this study).39 Power analyses. The power of detecting a significant difference when the α was set at .05 was > 92% for every one of the constructs/variables of intervention emphasis. All other constructs/variables not of intervention focus were low or marginal power (~10% to ~36%), as expected.
LESSONS LEARNED The goal was to evaluate a theory-based, tailored nutrition intervention for staff employees at a university campus worksite.The intervention was designed to change dietary behavior in a positive manner by systematically modifying specific health beliefs, in particular the benefits and barriers related to making certain dietary changes, while increasing nutrition knowledge and improving dietary behaviors related to CVD and cancer. Perceived benefits of changing dietary behaviors significantly increased. Perceived barriers was the only 1 of the 8 constructs/variables of intervention focus expected to change that did not. The intervention was successful and appears to be associated with producing significant increases in nutrition knowledge and significantly decreasing energy, fat, saturated fat, and cholesterol intake to levels consistent with national recommendations.The intervention was successful in attracting and retaining both African-American and Hispanic participants, who together comprised 43% and 40% of the treatment and control groups, respectively. This program, therefore, may be appropriate for future samples with similar demographic characteristics. An educational feature of the program and an implication for practice was providing participants with an analysis of
September • October 2003
265
their nutrition knowledge, dietary behaviors, and a risk factor score for CVD and cancer prior to the 8-week nutrition education intervention. A priori appraisals of participants’ nutrition knowledge, dietary intake status, and health risks have been conducted in other studies.40,41 Focusing on perceived benefits and perceived barriers was based on the research literature and baseline analyses, which revealed that participants reported more barriers for decreasing fat intake than increasing fruit and vegetable intake. Consequently, the focus of the intervention was on lowering fat, with the anticipation that behavior maintenance or self-change would spontaneously occur in regard to fruits, vegetables, and fiber. The positive change in decreasing fat is consistent with the focus and emphasis of the intervention and instruction. However, only about half of the treatment group reported an increase in fruit or vegetable intake after the intervention. Anticipated spontaneous changes did not occur, which suggests that fruit and vegetable intake needs to be priority independent of initial baseline assessment scores. There may be other reasons fruit and vegetable consumption needs to be emphasized that are related to the negative influences of personal, social, and environmental barriers. One reason is that dieters are probably more familiar and comfortable with restriction of food intake than with adding foods to a diet.Thus, lowering total fat intake may have been easier than adding fruits or vegetables.Adding new behaviors is considered more difficult than avoiding or limiting a behavior.18 A second reason that lowering fat consumption might have been easier is due to the amount of publicity about the health benefits of a lower-fat diet. In contrast, the National Cancer Institute and the National Institutes of Health42 5 A Day program is fairly new, and there has been far less publicity regarding the benefits of adding fruits and vegetables. A third reason is that participants may have believed that a decrease in fat and calorie intake was the most efficient way to produce significant weight loss.The last reason is caloric restriction.A significant decrease in calories may have yielded a diet containing less healthful nutrients, especially if fruit and vegetable intake failed to increase.This problem, however, was not manifested in this study.The findings of the current study are congruent with Herbert and colleagues43 and Peterson and colleagues,44 who found that participants who significantly decreased their energy and fat intake did not necessarily sacrifice micronutrients. The present study may have limitations. Because diet was not observed, self-report bias may have occurred. Nevertheless, self-report is the norm in nutrition “field studies” because of practicality and “realism” in terms of time demands on voluntary study participants and the negative effects on recruitment and retention when time demands initially appear excessive. Local norms were not developed for the food frequency questionnaire. Norms developed elsewhere may not be representative of local customs and habits. Food frequency questionnaire results should be viewed as estimates of change, but, most likely, conclusions would remain unchanged even if it were presumed that absolute
266
Abood et al/NUTRITION EDUCATION WORKSITE INTERVENTION BASED ON THE HBM
values would vary only slightly. Also, it should be noted that the durability of the intervention effect is unknown, and after demonstrating an intervention effect, as occurred in this study, collection of follow-up data would be the next priority in future investigations. In addition, it should be noted that retention was somewhat of a problem, despite making every effort to be accommodating and sensitive to work demands by offering sessions at multiple times during the week. Some participants found it difficult to attend sessions. The reason they gave was having too much work to do and the lure of being able to “catch up” during the hour of the intervention. Future studies also should address the challenges of effectively promoting the adoption of healthful food behaviors, especially those that require the addition of a new food behavior, such as increasing fruits and vegetables. Interventions promoting such behaviors might further concentrate on HBM variables. The employment of psychosocial models might enhance the effectiveness of nutrition interventions that promote the addition of specific foods. In conclusion, this study addresses an area in which there is an exiguity of research by focusing on the application of an 8-week worksite-tailored nutrition education intervention for staff employees at a university. The HBM was successfully used as a theoretical framework to encourage healthful dietary behavior change by increasing nutrition knowledge and promoting the benefits of making prudent dietary changes.This program may serve as a framework for future applications of such programs among this population of worksite employees and as an integral part of the total national effort to reduce the risk of CVD and cancer.
ACKNOWLEDGMENT This article is dedicated to Mr. George Abood in respectful memory and appreciation for his contributions and inspirations to improve the health and welfare of humankind whenever and however possible.
REFERENCES 1. McGinnis JM, Foege WH. Actual causes of death in the United States. JAMA. 1993;270:2207-2212. 2. US Dept of Agriculture and the US Dept of Health and Human Services. Dietary Guidelines for Americans. Washington, DC: US Government Printing Office; 1995. 3. US Dept of Agriculture, Human Nutrition Information Service. The Food Pyramid. Washington, DC: US Government Printing Office; 1992. Home and Garden Bulletin No. 249. 4. Glanz K, Patterson RE, Kristal AR, et al. Stages of change in adopting healthy diets: fat, fiber, and correlates of nutrient intake. Health Educ Q. 1994;21:499-519. 5. Glanz K. Behavioral research contributions and needs in cancer prevention and control: dietary change. Prev Med. 1997;26(suppl 1):43-55.
6. Goetzel RZ, Jahr TY, Aldana SG, Kenny GM. An evaluation of Duke University’s Live for Life health promotion program and its impact on employee health. Am J Health Promotion. 1996;10:340-342. 7. Baer JT. Improved plasma cholesterol levels in men after a nutrition education program at the worksite. J Am Diet Assoc. 1991;91: 1391-1397. 8. Glanz K, Seewald-Klein T. Nutrition at the worksite: an overview. J Nutr Educ. 1986;18(suppl 1): S1-S12. 9. Sorenson G, Himmelstein J. Worksite nutrition intervention. In: Ockene I, Ockene J, editors. Prevention of Coronary Heart Disease. Boston: Little, Brown; 1992:517-539. 10. Hochbaum GM. Public participation in medical screening programs: a sociopsychological study. Public Health Service. no. 572, 1958. 11. Rosenstock IM.What research in motivation suggests for public health. Am J Public Health. 1960;50:295-301. 12. Rosenstock IM.Why people use health services. Milbank Mem Fund Q. 1996;44:94-124. 13. Rosenstock IM. Historical origins of the health belief model. Health Educ Monogr. 1974;2:328-335. 14. Bandura A. Social Learning Theory. Englewood Cliffs, NJ: Prentice-Hall; 1977. 15. Rosenstock IM, Strecher VJ, Becker MH. Social learning theory and the health belief model Health Educ Q. 1988;15:175-188. 16. Fries EA, Ripley JS, Jigueiredo MI, et al. Can community organization strategies be used to implement smoking and dietary changes in a rural manufacturing work site? J Rural Health. 1999;15:413-420. 17. US Dept of Health and Human Services. National Cholesterol Program: Second Report of the Expert Panel on Detection, Evaluation, and Treatment of High Blood Cholesterol in Adults [Adult Treatment Panel III]. Washington, DC: National Institutes of Health; 1993. No. 93-3093. 18. Damrosch S. General strategies for motivating people to change their behavior. Nurs Clin North Am. 1991;26:833-843. 19. Anderson AS, Cox DN, McKellar S, et al. Five, a nutrition education intervention to increase fruit and vegetable intakes: impact on attitudes towards dietary change. Br J Nutr. 1998;90:133-140. 20. McCullock-Melnyk KA. Barriers: a critical review of recent literature. J Nurs Res. 1988;37:196-201. 21. Kayman S. Applying theory from social psychology and cognitive behavioral psychology to dietary behavior change and assessment. J Am Diet Assoc. 1989;89:191-193. 22. American Cancer Society. Cancer:Assessing Your Risk [Brochure].Atlanta, Ga, ACS, 1993. 23. American Heart Association. Test Your Cholesterol IQ [Brochure]. Dallas, Tex, AHA, 1991. 24. Fahey TD, Insel PM, Roth WT.Wellness worksheet: are you at risk for cardiovascular disease? In: Core Concepts in Health. Mountain View, Calif: Mayfield Publishing Company; 1994. 25. Hurley J. Fat or fiction? Nutrition Action Health Letter. 1994;10:5. 26. Boeckner LS, Kohn H, Rockwell K.A risk-reduction course for adults. J Am Diet Assoc. 1990;90:260-263. 27. Caudill E. Food shopping behavior and food item frequency of work mothers [master’s thesis]. Lincoln: University of Nebraska; 1979. 28. The Food Processor,Version 6.03 [nutrition analysis software]. Salem, Ore: ESHA Research; 1995. 29. Rubinson L, Neutens JJ. Research Techniques for the Health Sciences. New York: Macmillan; 1987.
Journal of Nutrition Education and Behavior Volume 35 Number 5 30. Black DR, Tobler N, Sciacca JP. Peer helping/involvement: an efficacious way of meeting the challenge of reducing illicit drug use? J School Health 1998;68:87-93. 31. Windsor R, Baranowski T, Clark N, et al. Evaluation of Health Promotion, Health Education, and Disease Prevention Programs. 2nd ed. Mountain View, Calif: Mayfield; 1994. 32. Timmreck TC. An Introduction to Epidemiology. 3rd ed. Boston: Jones and Bartlett; 2002. 33. Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: Lawrence Erlbaum; 1988. 34. Tippett KS, Cypel YS, eds. Continuing Survey of Food Intakes by Individuals and the Diet and Health Knowledge Survey, 1994-1996.Washington, DC: US Dept of Agriculture, Agricultural Research Service; 1998. 35. Simon MS, Heilbrun LK, Boomer A, et al.A randomized trial of a lowfat dietary intervention in women at high risk for breast cancer. Nutr Cancer. 1997;27:136-142. 36. Boyd NF, Lockwood GA, Greenberg LJ, et al. Effects of a low-fat high carbohydrate diet on plasma sex hormones in premenopausal women: results from a randomized controlled trial. Br J Cancer. 1997;76:127-135. 37. Coates RJ, Bowen DJ, Kristal AR, et al.The Women’s Health Trial Feasibility Study in Minority Populations: changes in dietary intakes. Am J Epidemiol. 1999;149:1104-1112.
September • October 2003
38. Gorbach SL, Morrill-LaBrode A,Woods MN, et al. Changes in food patterns during a low-fat dietary intervention in women. J Am Diet Assoc. 1990;90:802-809. 39. Buzzard MI,Asp EH, Chlebowski RT, et al. Diet intervention methods to reduce fat intake: nutrient and food group composition of selfselected low-fat diets. J Am Diet Assoc. 1990;90:42-50, 53. 40. Contento IR, Murphy BM. Psycho-social factors differentiating people who reported making desirable changes in their diets from those who did not. J Nutr Educ. 1990;22:6-13. 41. Strychar IM, Potvin L, Pineault R, et al. Change in knowledge and food behaviour following a screening program held in a supermarket. Can J Public Health. 1993;84:382-388. 42. National Cancer Institute and National Institutes of Health. 5 a Day for Better Health: A Baseline Study of American’s Fruit and Vegetable Consumption. Rockville, Md: National Institutes of Health; 1992. 43. Herbert JR, Harris DR, Sorensen G, et al. A work-site nutrition intervention: its effects on the consumption of cancer-related nutrients. Am J Public Health. 1993;83:391-394. 44. Peterson S, Sigman-Grant M, Eissenstat B, et al. Impact of adopting lower-fat food choices on energy and nutrient intakes of American adults. J Am Diet Assoc. 1999;99:177-183.
VISION, MISSION, AND GUIDING PRINCIPLES OF THE SOCIETY FOR NUTRITION EDUCATION Vision Healthy people in healthy communities. Mission To enhance nutrition educators’ ability to promote healthful sustainable food choices and nutrition behaviors. Guiding Principles • • • • •
267
Fiscal responsibility Respect for diversity of opinions and perspectives Trust and willingness to communicate openly and respectfully Knowledge-based decisions Excellence and lifelong learning
• • • •
Professionalism and integrity Inclusiveness in membership Equality among members Rewarding and enjoyable experiences for volunteers and supporters