Excess mortality for the unmarried in rural Bangladesh.

Prospective data from the Matlab surveillance area in rural Bangladesh, 1974-1982, were used in this study to show that divorced and never-married adults (aged 15-44 years) had significantly higher mortality than their currently married peers with differences in disability status accounting for some of this excess risk. Widowed individuals on the other hand had no excess mortality relative to the currently married. A certain proportion of the sharp reduction in mortality associated with remarriage after divorce for men could be attributed to differences in disability between the remarried group and those who remained divorced. Household economic status indicators could account for little of the excess mortality of divorced and never-married individuals relative to their married counterparts. These findings provide support for the hypothesis that a certain proportion of mortality differentials by marital status in this population can be attributed to selection into marriage and remarriage on the basis of underlying disability status. The evidence for the protective effect of increased economic status associated with the currently married state is less convincing.