75
Economic Impact of Physical Activity: Implications for Healthcare Costs, Health Gains and Labor Market Outcomes
Data/Methods: Secondary data on health care costs, health and employment outcomes from the 5 waves (1996-2002) of the Health and Retirement Study were analyzed. Total medical expenditure collected from 1996-2002 were utilized in this study, which included 10965 unique individuals at baseline. Total health care costs, change in self-rated health, depression and labor market outcomes were the main outcome variables of interest. Persistence was measured based on multiple assessments from each individual over time. The average annual health care costs were adjusted with the 2002 Gross Domestic Product. The total health care cost was used since out-of-pocket expenditure would not reflect the complete picture of total expenditures related to health care. The overall perceived health gain was the change in self-reported health in current wave minus the previous year survey during the study period. This variable measures the change in self-reports of health categories as excellent, very good, good, fair and poor. The mental health of a respondent was measured using a score on the Center for Epidemiologic Studies Depression Scale (CESD) with a range from 0-8, higher score meaning more depression. Statistical analyses were conducted using one-part generalized linear model for estimating changes in mean expenditures, health status and employment outcomes.
Results: Average health care costs increased from $4184 in 1996 to $9313 in 2002 in the study sample. About 24% of respondents reported positive changes in self-rated health at baseline, while 20% indicated gains in self-rated health in 2002 and average CESD score increased by20% from baseline. Adjusting for important baseline health measures and other individual level characteristics, the mean estimated cost was 47% lower per each point increase in persistence score. Higher persistence was also independently associated with gains in self-reported health and depression score. The exponentiated coefficient on physical activity indicated that for each unit increase in persistence score, the CESD score decreased by 40%. Physical activity was a significant predictor of decreased absenteeism and likelihood of work limitations due to ill health. For each unit increase in persistence score, the probability of working increased by 28%. Persistence in physical activity was less likely to be associated with work-limitation (odds ratio (OR): 0.37 [95% CI: 0.29-0.46]) and probability of any missed work days (OR: 0.66 [95% CI: 0.53-0.83]), suggesting that higher persistence in physical may reduce productivity losses due to absenteeism and work limitation
Conclusion: Benefits of persistence in regular physical activity may extend beyond direct health care costs and lower indirect costs associated with productivity losses due to absenteeism and work limitation, while improve health outcomes.