The Association between Health Risks and Medical Expenditures in a Japanese Corporation

Abstract

Purpose. To examine the relationship between health risks and medical care expenditures in an employer setting in Japan. Design. A cross-sectional, correlational study. Setting. A large Japanese corporation. Subjects. A total of 6543 employees of a large Japanese electronics company, for whom medical expenditures, lifestyle risks and biometric data were available, were included in the analysis. Seventy-six percent were male, and subjects were primarily white-collar workers. Measures. Medical expenditure data were available for fiscal year 2000, including inpatient, outpatient, and total expenditures, measured in Japanese yen. Binary expenditure indicators for those having no claims and those having high claims (90th percentile) were also created. Risk measures included biometric assessment of high blood pressure and high body mass index (BMI, body weight and height) and self-reported stress, lack of exercise, excess alcohol consumption, poor nutrition, current smoking, and recent quitting. High cholesterol and high blood glucose measures were also available for some subjects from company physicals. Results. Average total expenditures were ¥48,017 (US$445). The 90th percentile of the expenditure distribution was approximately ¥111,750 (US$1037). The most commonly reported risk factors were lack of exercise (52.9%), current smoking (35%), stress (33%), and poor nutritional habits (23.6%). Least common were recently quitting smoking (2%), high blood pressure (4.1%), and high blood glucose (9.4%). The prevalence of overweight or obesity was 15.9%. High blood pressure and recent quitting were consistently related to high expenditures, after adjusting for the influence of other predictors. Adjusted expenditures were 76% higher for recent quitters and 22.6% higher for employees with high blood pressure. Males and younger employees had consistently lower expenditures. Current smoking, poor nutrition, and alcohol risk were also associated with lower expenditures. Those with multiple cardiovascular risk factors had adjusted medical expenditures that were 128% higher than those with no cardiovascular risks. Those who had multiple risk factors for stroke had expenditures that were 13% lower than those without stroke risk factors. Conclusions. This paper represents a first step in examining the association between health risks and medical expenditures in Japanese employees. The investigation uncovered some significant levels of risk for lack of exercise, smoking, and stress. Although results indicate some significant associations between health risks and medical expenditures, several unexpected associations were noted that require further study. Such information provides a solid foundation for health promotion efforts in Japan and direction for subsequent investigations of health risks and medical expenditures. Future studies should address important issues of health risk measurement, data collection, and research design.