Comparison of risk factors for the competing risks of coronary heart disease, stroke, and venous thromboembolism

Methods for formal comparison of competing risks may clarify uncertainties about the associations of atherosclerotic risk factors with the development of venous thromboembolism (VTE). For a median of 20.1 years, the Physicians' Health Study (1982-2003) followed 18,662 US male physicians with no prior myocardial infarction, stroke, VTE, or cancer and for whom reported risk factor information was available at baseline. The authors used methods of competing risk survival analysis to compare relative hazard rates associated with age, hypertension, elevated cholesterol, diabetes, cigarette smoking, alcohol consumption, exercise frequency, body mass index, and height. During follow-up, coronary heart disease (CHD) occurred first in 1,348 men, stroke in 902 men, and VTE in 358 men. Incidence of all three outcomes increased with age, but the rate of increase was strongest for stroke. Hypertension, elevated cholesterol, diabetes, and smoking were associated with increased rates of CHD and stroke, with comparable magnitudes, but had no association with VTE. Conversely, higher body mass index was more strongly associated with risk of VTE than of either CHD or stroke, and taller men had a significantly increased risk of VTE but a lower risk of CHD. CHD and stroke have broadly comparable risk factor profiles that differ widely from the profile for VTE.Methods for formal comparison of competing risks may clarify uncertainties about the associations of atherosclerotic risk factors with the development of venous thromboembolism (VTE). For a median of 20.1 years, the Physicians' Health Study (1982-2003) followed 18,662 US male physicians with no prior myocardial infarction, stroke, VTE, or cancer and for whom reported risk factor information was available at baseline. The authors used methods of competing risk survival analysis to compare relative hazard rates associated with age, hypertension, elevated cholesterol, diabetes, cigarette smoking, alcohol consumption, exercise frequency, body mass index, and height. During follow-up, coronary heart disease (CHD) occurred first in 1,348 men, stroke in 902 men, and VTE in 358 men. Incidence of all three outcomes increased with age, but the rate of increase was strongest for stroke. Hypertension, elevated cholesterol, diabetes, and smoking were associated with increased rates of CHD and stroke, with comparable magnitudes, but had no association with VTE. Conversely, higher body mass index was more strongly associated with risk of VTE than of either CHD or stroke, and taller men had a significantly increased risk of VTE but a lower risk of CHD. CHD and stroke have broadly comparable risk factor profiles that differ widely from the profile for VTE.