The impact of spousal behavior changes on smoking, drinking and physical activity: The longitudinal survey of middle-aged and elderly persons in Japan.

The purpose of this study is to assess how one spouse's behavior change can influence their partner's successful behavior changes in smoking, drinking and physical activity. We used data from 10-wave prospective annual surveys of 9417 married couples (discrete-time person-years = 118,876) aged 50-59 years in the Longitudinal Survey of Middle-aged and Elderly Persons in Japan. A logistic generalized estimating equation model with discrete-time design was used among individuals who smoked at baseline to examine the impact of their spouse's health behaviors (i.e. quit smoking, stable non-smoker, or started smoking in reference to stable smoker) on changes in their own behavior (quitting smoking) which lasted one year or more. Similarly, reducing alcohol intake and starting physical activity were individually analyzed. Partners of spouses who had quit smoking had higher odds of quitting smoking themselves than partners of spouses who were stable smokers. The multivariable odds ratios[95%CI] in men and women were 1.94[1.23-3.07] and 2.89[1.81-4.52]. An association was found in partners of spouses who had been stable non-smokers (OR:1.64[1.33-2.03] and 2.20[1.66-2.94]), but not after spouses had started smoking (OR:1.29[0.71-2.36] and 1.27[0.54-2.99]). Similar associations were found for reducing alcohol intake and starting physical activity although for physical activity, the association was still found after the spouse had become physically inactive. Couples affect each other's health behaviors. Both male and female participants had higher odds of adopting positive health behavior changes if these changes had previously been made by their spouse.