Marital status and mortality: the national longitudinal mortality study

To examine the effect of marital status (married, widowed, divorced/separated, and never-married) on mortality in a cohort of 281,460 men and women, ages 45 years and older, of black and white races, who were part of the National Longitudinal Mortality Study (NLMS).

Major findings are based on assessments of estimated relative risk (RR) from Cox proportional hazards models. Duration of bereavement for the widowed is also estimated using the Cox model.

For persons aged 45-64, each of the non-married groups generally showed statistically significant increased risk compared to their married counterparts (RR for white males, 1.24-1.39; white females, 1.46-1.49; black males, 1.27-1.57; and black females, 1. 10-1.36). Older age groups tended to have smaller RRs than their younger counterparts. Elevated risk for non-married females was comparable to that of non-married males. For cardiovascular disease mortality, widowed and never-married white males ages 45-64 showed statistically significant increased RRs of 1.25 and 1.32, respectively, whereas each non-married group of white females showed statistically significant increased RRs from 1.50 to 1.60. RRs for causes other than cardiovascular diseases or cancers were high (for white males ages 45-64: widowed, 1.85; divorced/separated, 2.15; and never-married, 1.48). The importance of labor force status in determining the elevated risk of non-married males compared to non-married females by race is shown.

Each of the non-married categories show elevated RR of death compared to married persons, and these effects continue to be strong after adjustment for other socioeconomic factors.