Analysis of the association between exercise habits and socio-economic status using a double hurdle model

The present study was conducted to re-examine the socio-economic factors that determine exercise habits. Previous attempts to identify such factors using quantitative analysis of social survey data have focused on 3 main socio-economic factors related to health and exercise habits: occupational condition, educational attainment, and economic status. However, the methodology of that research prompted criticism, as it overlooked the possibility that the decision of an individual to adopt exercise habits is a 2-step process. The present study assumed that when individuals decide to take up exercise in leisure time, they first decide whether or not to do so, and then decide the time they intend to devote to it. For this study, an analytical model was developed and used to examine the association between leisure time exercise habits and socio-economic status. A quantitative analysis of Sport Life Survey data collected in 2020 was conducted. The double hurdle model was employed to estimate the association between exercise habits and socioeconomicstatus. The double hurdle model combines a selection model with a model for a continuous variable, and allows different independent variables for each model. This makes it possible to estimate the influence of independent variables on the decision to take up exercise habits, while at the same time allowing analysis of the influence of independent variables on the amount of exercise if an individual decides to adopt it. The results showed that all variables related to socio-economic status had a statistically significant influence on whether an individual decided to take up exercise, but not on the length of time devoted to it. Individuals who were more highly skilled, those with higher educational attainment, and those with higher household incomes were more likely to take up exercise habits. However, the influence of socio-economic status on exercise habits was not significant. This suggested that, in Japan, there is no marked socio-economic inequality of exercise habits. The limitations of this research and directions for future research were also discussed.