Self-rated ill-health strengthens the effect of biomedical risk factors in predicting stroke, especially for men - an incident case referent study

Objectives To examine how self-rated ill-health interacts with biomedical stroke risk factors in predicting stroke and to explore differences between men and women and educational groups. Design An incident case-referent study where the study subjects had participated in a prior health survey. Setting Nested within the Vasterbotten Intervention Program (VIP) and the Northern Sweden MONICA cohorts. Subjects The 473 stroke cases had two referents per case, matched for age, sex and residence, from the same study cohorts. Results Self-rated ill-health independently increased the risk of stroke, specifically for men. The interaction effect between self-rated health and biomedical risk factor load was greater for men than for women. The attributable proportion due to interaction between having a risk factor load of 2+ and self-rated ill-health was 42% for men and 15% for women. Better-educated individuals with self-rated ill-health and two or more of the biomedical risk factors had a higher risk of stroke than the less educated. Calculations of the respective contribution to the stroke cases of self-rated health, hypertension and smoking showed that self-rated ill-health had a role in 20% of the cases and could alone explain more than one-third of the cases among those who rated their health as bad, more so for men than for women. Conclusions The results underscore the importance of including both a gender and a social perspective in discussing the role of self-rated health as a predictor of disease outcome. Physicians must be more gender sensitive when discussing their patient's own evaluation of health in relation to biomedical risk factors. (C) 2003 Lippincott Williams Wilkins.