Gender, sexual orientation, and marital/intimate partner relationship status comingle with other social-ecological factors to shape health. In a substudy within a broader survey of social determinants of health in Japan (n = 258), I tested potential associations between this cluster of independent variables and self-rated overall health and mental health as dependent variables. I also investigated other relevant constructs. Perceived gender bias, sexual orientation bias, and lifetime experience of sexual violence/abuse were remarkably prevalent but not associated with the dependent variables. Not being in a relationship legally recognized in Japan and not being in a committed relationship (regardless of legal recognition) were associated with worse self-rated mental health at the bivariate level, though this effect did not withstand multivariate testing. Three bivariate results remained significant in multivariate models adjusted for socio-demographic/economic controls: compared with heterosexual/straight participants, those with non-heterosexual orientations had lower odds of reporting good/very good/excellent overall health; and divorced or widowed people (who were also single) had lower odds of rating both their overall health and mental health as good/very good/excellent, compared with people in heterosexual marriages. Gender was not associated with either dependent variable, though it approached significance with mental health. This was contrary to theory but consistent with other studies and has potentially positive implications.
Anthony DiStefano (Sun,) studied this question.