Background: In clinical practice, individuals at increased risk of cardiovascular diseases (CVD) are identified on the basis of age, sex, smoking, blood pressure, and serum total and high-density lipoprotein cholesterol. We examined whether CVD risk prediction improved when obesity (body mass index ≥30 kg/m(2)) and premature (<70 years) parental myocardial infarction (MI) were added to the classical risk factor model.
Methods: Risk factors were measured in 1993-97 in 12,818 participants (53% female) aged 35-65 in the Dutch MORGEN project. Cases of fatal and nonfatal CVD during 10 years of follow up were identified through record linkage. Classical risk factor equations, obtained by Cox proportional hazard analysis, were extended with obesity, paternal MI, and maternal MI. We calculated the net reclassification index (NRI), a measure for correct reclassification of subjects, to check improvement in risk prediction using 5 and 10% increments in absolute CVD risk.
Results: A CVD event occurred in 280 men and 140 women. Obesity and maternal MI were positively and significantly related to total CVD after adjustment for classical risk factors (both hazard ratios ∼1.5). Adding obesity and parental MI to CVD risk prediction yielded a significant NRI of 4.5% in men and a non-significant NRI of 2.6% in women when 5% risk categories were used. For 10% categories, the NRIs were slightly larger (5.5% and 3.3%, respectively). The improvements in risk prediction were mainly due to obesity.
Conclusion: Modest improvements in CVD risk prediction can be obtained when obesity and, to a lesser extent, parental MI are added to the risk function.
Keywords: Cardiovascular disease; myocardial infarction; net reclassification index; obesity; parental history; risk factors; risk prediction.