Adding nontraditional risk factors, age, and race to traditional models modestly improved ischemic stroke prediction (AUC increased from 0.79 to 0.84 in women and 0.76 to 0.80 in men).
Cohort (n=14,685)
Yes
Does the addition of nontraditional risk factors and markers of subclinical disease to a basic traditional risk factor model improve the prediction of ischemic stroke in middle-aged adults?
Adding nontraditional risk factors and subclinical disease markers to traditional risk models provides only modest improvements in predicting ischemic stroke, which are insufficient to influence clinical or public health practices.
The authors assessed the increase in the predictivity of ischemic stroke (IS) resulting from the addition of nontraditional risk factors and markers of subclinical disease to a basic model containing only traditional risk factors (current smoking, diabetes mellitus, systolic blood pressure, antihypertensive therapy, prior coronary disease, and left ventricular hypertrophy) among 14,685 middle-aged persons in the Atherosclerosis Risk in Communities Study. Participants were recruited from four US communities in 1987-1989. Risk prediction scores for IS through 2000 were estimated from Cox models. The ability to predict which persons would develop IS was assessed by means of the area under the receiver operating characteristic curve-the probability that persons with IS had a higher risk score than those without IS. Among 22 nontraditional factors considered, the joint addition of body mass index, waist:hip ratio, high density lipoprotein cholesterol, albumin, von Willebrand factor, alcohol consumption, peripheral arterial disease, and carotid artery wall thickness modestly and statistically significantly improved prediction of future IS over a risk score that included traditional factors. Further improvement was obtained by adding age and race. For women, the area under the receiver operating characteristic curve went from 0.79 to 0.83 to 0.84; for men, it went from 0.76 to 0.78 to 0.80. These modest improvements are not enough to influence clinical and public health efforts to reduce the community burden of IS.
Lloyd E. Chambless (2004) conducted a cohort in Ischemic stroke (n=14,685). Addition of nontraditional risk factors, age, and race to risk prediction model vs. Basic model with traditional risk factors was evaluated on Predictivity of ischemic stroke assessed by area under the receiver operating characteristic curve. Adding nontraditional risk factors, age, and race to traditional models modestly improved ischemic stroke prediction (AUC increased from 0.79 to 0.84 in women and 0.76 to 0.80 in men).