Living arrangements as determinants of myocardial infarction incidence and survival: A prospective register study of over 300,000 Finnish men and women.

Living with a spouse is associated with a reduced risk of coronary heart disease mortality in middle age, but it remains unclear whether marriage and other living arrangements are important both for the development of the disease and the survival following incidence. Cohabitation and living alone have also become more common in many Western societies and thus warrant further study. We explored the association between living arrangements and myocardial infarction (MI) incidence and fatality. We used a population-based register sample of adults aged 40-60 in Finland in 1995 (n = 302,885) followed up until the end of 2007. MI incidence and mortality were identified from hospital discharge records and cause of death register (5917 incident cases in men and 1632 in women). Living with a marital partner was contrasted to three alternatives: cohabiting with non-marital partner, co-residence with persons other than a partner and living alone. MI incidence and long-term fatality were analysed with Cox proportional hazards regression with time-varying covariates and first-day fatality with logistic regression. Men who were married had a lower risk of MI incidence even after adjusting for socioeconomic factors - i.e. education, occupation, income, wealth and employment status - with small differences between the other living arrangement groups. For women the effects of living arrangements on incidence were fully explained by the same socioeconomic factors. However, our findings revealed that living arrangements were strong determinants for survival after MI independent of other socio-demographic factors. The results demonstrate greater fatality associated with living alone in men and suggest that cohabitation in midlife may be associated with a greater fatality risk in women. The social support and control offered by a marital relationship may protect from MI fatality in particular.