The Relationship Between Modifiable Health Risks and Health Care Expenditures
An Analysis of the Multi-Employer Hero Health Risk and Cost Database
Published in: Journal of Occupational Environmental Medicine, v. 40, No. 10, Oct 1998, p. 843-854
Posted on RAND.org on January 01, 1998
This investigation estimates the impact of ten modifiable health risk behaviors and measures and their impact on health care expenditures, controlling for other measured risk and demographic factors. Retrospective two-stage multivariate analyses, including logistic and linear regression models, were used to follow up 46,026 employees from six large health care purchasers for up to 3 years after they completed an initial health risk appraisal. These participants contributed 113,963 person-years of experience. Results show that employees at high risk for poor health outcomes had significantly higher expenditures than did subjects at lower risk in seven of ten risk categories: those who reported themselves as depressed (70% higher expenditures), at high stress (46%), with high blood glucose levels (35%), at extremely high or low body weight (21%), former (20%) and current (14%) tobacco users, with high blood pressure (12%), and with sedentary lifestyle (10%). These same risk factors were found to be associated with a higher likelihood of having extremely high (outlier) expenditures. Employees with multiple risk profiles for specific disease outcomes had higher expenditures than did those without these profiles for the following diseases: heart disease (228% higher expenditures), psychosocial problems (147%), and stroke (85%). Compared with prior studies, the results provide more precise estimates of the incremental medical expenditures associated with common modifiable risk factors after we controlled for multiple risk conditions and demographic confounders. The authors conclude that common modifiable health risks are associated with short-term increases in the likelihood of incurring health expenditures and in the magnitude of those expenditures.