Using health-related quality of life to predict cardiovascular disease events.

ABSTRACT

PURPOSE: Although strong associations between self-reported health and mortality exist, quality of life is not conceptualized as a cardiovascular disease (CVD) risk factor. Our objective was to assess the independent association between health-related quality of life (HRQOL) and incident CVD. METHODS: This study used the REasons for Geographic And Racial Differences in Stroke data, which enrolled 30,239 adults from 2003 to 2007 and followed them over 10 years. We included 22,229 adults with no CVD history at baseline. HRQOL was measured using the SF-12 Physical Component Summary (PCS) and Mental Component Summary (MCS) scores, which range from 0 to 100, with higher scores indicating better HRQOL. Scores were normed to the general US population with mean 50 and standard deviation 10. We constructed a four-level HRQOL variable (1) individuals with PCS & MCS < 50, (2) PCS < 50 & MCS ≥ 50, (3) MCS < 50 & PCS ≥ 50, and (4) PCS & MCS ≥ 50, which was the reference. The primary outcome was incident CVD (non-fatal myocardial infarction (MI), fatal MI or coronary heart disease (CHD) death, fatal and non-fatal stroke). Cox proportional hazards models examined associations between HRQOL and CVD. RESULTS: Median follow-up was 8.4 (IQR 5.9-10.0) years. We observed 1766 CVD events. Compared to having PCS & MCS ≥ 50, having MCS & PCS < 50 was associated with increased CVD risk (aHR 1.46; 95% 1.24-1.70), adjusting for demographics, comorbidities, and CVD risk factors. Associations between MCS & PCS < 50 and CVD were consistent for CHD (aHR 1.54 [1.26-1.89]) and stroke (aHR 1.35 [1.05-1.72]) endpoints. CONCLUSIONS: Given strong, adjusted associations between poor HRQOL and incident CVD, self-reported health may be an excellent complement to current approaches to CVD risk identification.