Worksite health promotion programs are designed to improve employee health and productivity and reduce the cost of healthcare expenditures.1-14 The spectrum of worksite health promotion includes health risk appraisal (HRA) to identify high-risk conditions or behaviors, environmental supports for behavior change (eg, access to healthy food choices and exercise facilities), corporate policies that support healthy lifestyles (eg, no-smoking policies), financial incentives, and individualized follow-up counseling to reduce risk.12 The most effective interventions in worksites are those that provide sustained individual follow-up risk factor education and counseling.12
We conducted a randomized controlled study to test the relative impact of HRA plus individualized risk factor management using a commercial, evidence-based, comprehensive lifestyle management and cardiovascular disease risk reduction program (INTERVENT USA Inc, Savannah, Georgia) compared with HRA plus information about free worksite health promotion programs on global cardiovascular risk in an employee population. We hypothesized that after 12 months, employees randomized to the disease management (DM) program would have a lower cardiovascular risk score compared to the group assigned to HRA plus information about worksite health promotion programs.
Between 2000 and 2003, we recruited Vanderbilt University employees at higher than average risk for cardiovascular disease based upon an HRA administered by Vanderbilt University's employee wellness program. Inclusion criteria are shown in Table 1. We excluded employees who were established patients in the Vanderbilt Heart Disease Prevention Program or the Vanderbilt Lipid Clinic and those who did not plan to be employed by Vanderbilt 1 year after the start of the trial.
This study was a randomized controlled clinical trial. In 1999, the Vanderbilt University Institutional Review Board approved the protocol, and written informed consent was obtained from all participants. Potential participants were screened using a commercial 39-question HRA (Wellsource Inc Concise version, Portland, OR). Those who met the study inclusion criteria by HRA were invited for additional data collection to confirm their current risk status. This included a baseline history, assessment of smoking status (self-report), dietary assessment, physical activity assessment, body mass index (BMI), heart rate, blood pressure, fasting lipids and lipoproteins, fasting glucose, hemoglobin A1c in participants with known diabetes, and quality of life (Short Form 36, Medical Outcomes Trust Inc, Waltham, Massachusetts). Eligible volunteers were randomly assigned to DM or HRA groups for 1 year, and a baseline Framingham risk score was calculated for each participant.15
Participants randomized to the HRA group were provided a summary report of the evaluation, had a consultation with the project nurse to review their risk factors, and were encouraged to use the health promotion facilities and services offered by the university. The university offers free to all employees a comprehensive worksite wellness and fitness program, with a fully equipped fitness center and a wide range of classes and workshops on exercise, nutrition, stress management, cholesterol control, blood pressure control, smoking cessation, and weight management. Participants were asked to return in 1 year to repeat the same risk factor assessment that was performed at baseline.
Participants randomized to DM met with the same project nurse to receive individualized counseling for nutrition, exercise, smoking cessation (if needed), weight loss (if needed), and stress reduction. They were also encouraged to use the health promotion facilities and services offered by the university. Table 2 lists the goals for risk factor intervention for DM participants. Tailored risk factor intervention was delivered using a commercial, evidence-based, comprehensive lifestyle management and cardiovascular disease risk reduction program (INTERVENT USA Inc). The process involved brief counseling sessions with written materials and audiotapes to teach participants how to make lifestyle changes. Visits were face to face and typically lasted 15 minutes. The treatment plan was individualized to the risk factor profile of each participant. This plan included educational vignettes that focus on nutrition, physical activity, stress management, or smoking cessation. With each visit, the nurse reinforced behavior change through food diaries, physical activity logs, or other homework assignment tracking. In general, visits occurred at least twice per week during weeks 1 to 4, once per week during weeks 5 to 8, once every 2 weeks during weeks 9 to 12, and once per month until the end of the 12-month intervention. The project nurse is highly experienced in behavioral health counseling and was trained in the delivery of the risk reduction program before the study launch. When indicated by national treatment guidelines, medications for control of high blood pressure, high cholesterol, high blood sugar, and smoking were recommended to participants and their physicians. All study information was kept strictly confidential and was not reported to the participants' employer or insurance plans.
The primary end point of the trial was the difference between groups in the change in the average Framingham risk score15 from baseline to 1 year. The score is derived from an algorithm using categorical variables to predict multivariate coronary artery disease risk in patients without overt coronary disease. The Framingham coronary prediction algorithm provides estimates of coronary heart disease risk over the course of 10 years. Separate score sheets are used for men and women, and the factors used to estimate risk are age, blood cholesterol (or low-density lipoprotein cholesterol), high-density lipoprotein (HDL) cholesterol, blood pressure, cigarette smoking, and diabetes mellitus. We created a programming code based on these score sheets. Relative risk for coronary heart disease is estimated by comparison to low-risk Framingham participants. Comparisons of changes in individual risk factors were also assessed.
The null hypothesis was that the change in Framingham global risk scores between baseline and end-of-trial assessments would be similar for the DM and HRA groups. The interaction effect of group membership and time of assessment was assessed via a 2 (DM, HRA) × 2 (baseline, end-of-trial) mixed model analysis of variance of the global risk scores using generalized estimating equations to account for the lack of independence in the repeated assessments. Power analysis revealed that assuming an α level of.05 and power of.80, a sample size of 60 per group was adequate to detect a statistically significant group difference of approximately one global risk score point. Statistical analysis was performed in 2005 and 2007.
A total of 133 volunteers consented to participate in the study. Of these, 126 completed the HRA and were randomized to DM (n = 68 [54.0%]) and HRA (n = 58 [46.0%]) groups. The baseline characteristics are reported in Table 3. Changes in the Framingham risk score from changes in the score's component risk factors from baseline to 1 year are presented in Figure 1. The baseline and follow-up lipid and blood pressure measures are summarized in Table 4. All risk factors except low-density lipoprotein cholesterol improved in the DM group, with statistically significant changes in HDL cholesterol, total cholesterol-to-HDL ratio, and diastolic blood pressure. Among those who completed the study, there were 3 smokers in the DM group and 1 smoker in the HRA group at baseline. At 1 year, 2 of the smokers in the DM group and the 1 smoker in the HRA group had stopped smoking. The baseline and follow-up BMI values for participants who completed the study were 33.3 ± 5.9 and 33.0 ± 5.6 for the DM group and 32.2 ± 6.3 and 32.7 ± 6.4 for the HRA group, respectively. These changes in BMI within groups were not statistically significantly different (P = .140).
Change in Framingham Risk Score
Figure 2 shows the change in Framingham risk score within the DM and HRA groups. The difference noted between groups at baseline was not statistically significant (P = .279). At 1 year, a statistically significant change from baseline was observed in the mean risk score between DM and HRA groups (P = .013). The DM group had a decrease in the mean score (−1.33, decrease of 22.6%), whereas the HRA group had a slight rise in the average Framingham risk score from baseline to 1 year (+0.20, increase of 4.3%).
There were no significant differences in the use of selected medication classes at baseline (Table 3). During the study period, 4 patients in each group received a new prescription for lipid lowering therapy, and 1 patient in the HRA group discontinued lipid drug therapy. Seven patients in the DM group and 5 patients in the HRA group received new prescriptions for antihypertensive medication. One patient in each group discontinued antihypertensive therapy. There was no change in aspirin therapy in the DM group. In the HRA group, 2 patients began aspirin therapy and 1 patient discontinued aspirin. There were 2 new prescriptions for diabetes medication in the DM group and 1 in the HRA group. There was no statistically significant difference between groups for these changes in medical therapy during the trial.
One year after entry into the study, only 77 of the initial 126 participants completed the assessments necessary to calculate the follow-up Framingham risk score. A higher percentage of the HRA participants (70.7%, n = 41) were available for follow-up assessments compared to those in the DM program (52.9%, n = 36), P = .042. Participants dropped out of the study for the following reasons: 23 had job constraints, 6 moved away from the area, 1 was diagnosed with esophageal cancer, and 31 were lost to follow-up. Those lost to follow-up tended to be slightly younger (mean [SD] age, 45.10 [9.33] years) than those who completed 1-year assessments (mean [SD] age, 48.27 [8.15] years), P = .047. There were no significant differences in gender (P = .904) or in baseline Framingham risk score (P = .336) for those retained in the study and those not retained.
We found that among employees at increased cardiovascular risk identified through worksite HRA screening, those randomized to a disease management program had a statistically significant reduction in the mean Framingham risk score after 1 year compared with a slight rise in those randomized to HRA with environmental supports. We observed an almost 2 mg/dL rise in the median HDL cholesterol in the DM group. In observational studies, every 1 mg/dL increment in HDL cholesterol is associated with a 2% decreased risk of coronary artery disease in men and 3% decreased risk in women.16 There was a 7-mm Hg reduction in median systolic blood pressure in the DM group. A 5-mm Hg reduction in systolic pressure is associated with a 9% reduction in death from coronary artery disease and 14% reduction in death from stroke.17 The increase in HDL cholesterol and decrease in systolic blood pressure were not attributable to changes in medication, which were not significantly different between groups. A more likely explanation was lifestyle change, particularly an increase in physical activity, that could result in these physiologic changes.
Several limitations should be noted. There were a large number of dropouts from both study groups. One contributing factor might have been the high employee turnover rate observed commonly at major medical centers.18 The higher dropout rate in the DM group may indicate the resistance of individuals to make behavioral changes, which may have been due to well-documented barriers to lifestyle intervention.19 The sample was mostly women, and generalizability of the results to men is uncertain. Bias may have been introduced by having the same project nurse provide feedback to both treatment groups. Another limitation is that we did not measure the extent to which participants in either group participated in wellness activities offered by the university. Furthermore, we did not measure the frequency of physician visits for participants in either group. We do not know if the difference we observed between groups is durable, although evidence suggests that over a 5-year period, nearly half of transition from medium or high-risk status to low risk among employees participating in a worksite health promotion program occurs during the first year of the program.20 The potential impact of our intervention on hard clinical outcomes, healthcare claims, employee satisfaction, absenteeism, and presenteeism was not measured. Finally, the true cost of program delivery was not measured.
In summary, we confirmed the feasibility of using a commercially available cardiovascular risk reduction program and demonstrated its superiority to HRA in a work environment enriched with health promotion programs as measured by the Framingham risk score.
1. Breslow L. A health promotion primer for the 1990s. Health Aff (Millwood VA)
2. Aldana SG. Financial impact of health promotion programs: a comprehensive review of the literature. Am J Health Promot
3. Stokols D, Pelletier KR, Fielding JE. Integration of medical care and worksite health promotion
4. Pelletier KR. A review and analysis of the clinical and cost-effectiveness studies of comprehensive health promotion and disease management programs at the worksite: update VI 2000-2004. J Occup Environ Med
5. Ozminkowski RJ, Dunn RL, Goetzel RZ, Cantor RI, Murnane J, Harrison M. A return on investment evaluation of the Citibank, N.A., health management program. Am J Health Promot
6. Ozminkowski RJ, Ling D, Goetzel RZ, et al. Long-term impact of Johnson & Johnson's Health & Wellness Program on health care utilization and expenditures. JOccup Environ Med
7. Goetzel RZ, Ozminkowski RJ, Bruno JA, Rutter KR, Isaac F, Wang S. The long-term impact of Johnson & Johnson's Health & Wellness Program on employee health risks. J Occup Environ Med
8. Mills PR, Kessler RC, Cooper J, Sullivan S. Impact of a health promotion program on employee health risks and work productivity. Am J Health Promot
9. Novak B, Bullen C, Howden-Chapman P, Thornley S. Blue-collar workplaces: a setting for reducing heart health inequalities in New Zealand? N Z Med J
10. Wilson JB, Osterhaus MC, Farris KB, et al. Financial analysis of cardiovascular wellness program provided to self-insured company from pharmaceutical care provider's perspective. J Am Pharm Assoc (2003)
11. John EJ, Vavra T, Farris K, et al. Workplace-based cardiovascular risk management by community pharmacists: impact on blood pressure, lipid levels, and weight. Pharmacotherapy
12. Matson Koffman DM, Goetzel RZ, Anwuri VV, Shore KK, Orenstein D, LaPier T. Heart healthy and stroke free: successful business strategies to prevent cardiovascular disease. Am J Prev Med
13. Loeppke R, Taitel M, Richling D, et al. Health and productivity as a business strategy. J Occup Environ Med
14. Okie S. The employer as health coach. N Engl J Med
15. Wilson PW, D'Agostino RB, Levy D, Belanger AM, Silbershatz H, Kannel WB. Prediction of coronary heart disease using risk factor categories. Circulation
16. Maron DJ. The epidemiology of low levels of high-density lipoprotein cholesterol in patients with and without coronary artery disease. Am J Cardiol
17. Whelton PK, He J, Appel LJ, et al. Primary prevention
of hypertension: clinical and public health advisory from The National High Blood Pressure Education Program. JAMA
18. Waldman JD, Kelly F, Arora S, Smith HL. The shocking cost of turnover in health care. Health Care Manage Rev
19. King AC, Blair SN, Bild DE, et al. Determinants of physical activity and interventions in adults. Med Sci Sports Exerc
20. Musich S, McDonald T, Hirschland D, Edington DW. Examination of risk status transitions among active employees in a comprehensive worksite health promotion
program. J Occup Environ Med