Responding to Reward: Initiation and Maintenance of Regular Exercise in an Incentive-Based Employer Wellness Program
On January 1, 2008, the University of Minnesota initiated the Fitness Rewards Program (FRP) for enrollees of the university’s health insurance program (UPlan). The FRP provides a monthly $20 credit if an individual exercises at a participating facility at least 8 times per month.
We analyze factors influencing individuals’ propensity to initiate and maintain fitness center exercise, including whether exercise behavior varies by demographic, health, and economic attributes as well as beliefs about exercise and weight loss. We also consider whether the program design leads individuals to regulate behavior towards the credit threshold.
Data and Methods. Using UPlan administrative data from 2008-2010, we identified employees eligible to participate in the FRP. For those who participated, we have objective measures of exercise behavior based on ‘swipe card technology’ which measures the number of visits each participant has to a fitness facility in a given month. Our explanatory variables include demographic and employment information (age, sex, job class, location) as well as health status (indicators for chronic and acute conditions as well as a composite risk score). Finally, for a subset of participants we incorporate data from the UPlan’s health risk assessment, pertaining to preferences for making lifestyle behavior changes including exercise and weight loss.
We predict the initiation of fitness center exercise using a longitudinal complementary log-log survival model. We model the maintenance and regulation of exercise behavior by estimating the quantity of monthly fitness center exercise using discrete time-series regression models.
Key Findings: Participants visited a gym in 49% of all FRP-eligible months and earned a credit 23% of the time. Women were more likely to initiate exercise than men (HR: 1.17, 95%CI: 1.09-1.25). We find no significant gender difference in earning at least one credit. Eligible enrollees who indicated plans to lose weight were more likely to initiate exercise (HR: 1.22, 95%CI: 1.08-1.36) and as likely to earn credit (HR: 0.97, 95%CI: 0.84-1.10) compared to employees with no plans for weight loss. In contrast, enrollees with plans to get more exercise were both less likely to initiate exercise (HR: 0.76, 95% CI 0.66-0.85) or earn credit (HR: 0.92, 95%CI: 0.79-1.05). Preliminary findings of exercise maintenance show that as program duration increased, eligible enrollees were more likely to exercise or earn credit, but once a participant initiated exercise they became less likely to continue exercising or earn credit in each subsequent month. Finally, we find that in any given month participants who did exercise were most likely to have exactly 8 gym visits, the amount needed to earn credit, compared to any other visit amount. Descriptively, we find that participants with exactly 8 visits per month had lower income, but were relatively healthy compared to all participants.