Purpose: To study the predictive ability of each of the eight scales of SF-36 on 13-year all-cause mortality and incident coronary heart disease (CHD) in a general middle-aged population.
Methods: The population-based, longitudinal "Life-conditions, Stress and Health" study, in 2003-2004 enrolled 1007 persons aged 45-69 years (50% female), randomly sampled from the general population in Östergötland, Sweden. Variables at baseline included the SF-36 (health-related quality of life, HRQoL) and self-reported disease. Incident CHD (morbidity and mortality) and all-cause mortality data for the study population during the first 13 years from baseline were obtained from national Swedish registries.
Results: Seven of the eight SF-36 scales predicted CHD (sex- and age-adjusted Hazard Ratios up to 2.15; p ≤ 0.05), while only the Physical Functioning scale significantly predicted all-cause mortality. Further adjustments for presence of (self-reported) disease did not, in most cases, alter these significant predictions.
Conclusion: Low SF-36 scores predict risk of CHD, also after adjustment for present disease, supporting the biopsychosocial model of health and disease. Measures of HRQoL yield important information and can add to the cardiopreventive toolbox, including primary prevention efforts, as it is such a simple and relatively inexpensive tool.
Keywords: Health-related quality of life; Heart disease prevention; Morbidity; Mortality.