Self-rated health in relation to age and gender:: Influence on mortality risk in the Malmo Preventive Project

Aims: A study was undertaken to examine whether poor self-rated health (SRH) can independently predict all-cause mortality during 22-year follow-up in middle-aged men and women. Subjects and methods: Data are derived from a population-based study in Malmo, Sweden. This included baseline laboratory testing and a self-administered questionnaire. The question on global SRH was answered by 15,590 men ( mean age 46.4 years) and 10,089 women (49.4 years). Social background characteristics ( occupation, marital status) were based on data from national censuses. Mortality was retrieved from national registers. Results: At screening 4,261 (27.3%) men and 3,085 (30.6%) women reported poor SRH. Among subjects rating their SRH as low, 1,022 ( 24.0%) men and 228 (7.4%) women died during follow-up. Corresponding figures for subjects rating their SRH as high were 1801 (15.9%) men and 376 (5.4%) women. An analysis of survival in subjects reporting poor SRH revealed an age-adjusted hazard risk ratio (HR, 95% CI) for men HR 1.5 (1.4 - 1.7), and for women HR 1.4 ( 1.2 - 1.6). The corresponding HR after adjusting for possible social confounders was for men HR 1.3 (1.1 - 1.4), and women HR 1.1 (0.9 - 1.4). When additional adjustment was made for biological risk factors the association for men was still significant, HR 1.2 ( 1.1 - 1.3). Conclusion: Poor SRH predicts increased long-term mortality in healthy, middle-aged subjects. For men the association is independent of both social background and selected biological variables. The adjustment for biological variables can be questioned as they might represent mediating mechanisms in a possible causal chain of events.