Estimation of cardiovascular risk in French men according to age: the IPC cohort

Abstract

The aim of this study was to evaluate the role of several risk factors on cardiovascular (CVD) mortality according to age. In this analysis, we present equations for predicting CVD mortality in 3 age groups of men. The study population included 94,249 men aged from 30 to 95 years, who had a health check-up at the IPC Center between 1972 and 1988. The follow-up period for mortality extended up to 25 years (mean 14 years), ending in December 1997. The population was divided into 3 age groups: 30–50 years (n=61,440; 510 CVD deaths), 50–65 years (n=29,579; 1149 CVD deaths) and > 65 years (n=3230; 632 CVD deaths). These equations were elaborated using Cox regression models, separately in each of the 3 age groups. Equations were evaluated on 80% of the population and validation was assessed with the remaining 20% of the population. The following equations were estimated in each age group. Risk equation for 30–50 y.o., Y= (0.07*AGE) + (0.210*SBP) + (0.25*TOB) + (0.27*CHOL) + (0.64*LVH)+(1.24*CVH)+(0.21*HR)−(0.29*PhA) Risk equation for 50–65 y.o., Y= (0.08*AGE) +(0.17*SBP) +(0.14*TOB) +(0.13*CHOL) +(1.13*CVH)+(0.21*HR) −(0.13*FH)+(0.23*DIAB)+(0.14*SR) Risk equation for >=65 y.o., Y= (0.07*AGE)+(0.07*PP)+ (0.57*CVH)+(0.14*HR) 1 year; SBP: 10 mmHg; TOB: 0=non smoker, 1=ex-smokers, 2=current smokers; CHOL: 1 mol/l of cholesterol; LVH: 0=absence, 1= presence of left ventricular hypertrophy; CVH: 0=absence, 1=presence of cardiovascular disease; HR: 0=Heart rate <60 bpm, 1=60–80 bpm, 2=80–100 bpm, 3= >100 bpm; FH: 0=absence, 1=presence of family history of CVD; SR =1 cm/Hour of sedimentation rate; DIAB: 0=absence, 1=presence of diabetes, PP= 10 mmHg of pulse pressure, PhA: 0=no regular physical activity, 1= regular physical activity. These equations allow us to evaluate the survival probability (S) as an exponential function of the risk equation: S= S0exp (Y) (S0 the baseline survival probability). In conclusion, the role of several risk factors is modified with age. Age, SBP (PP in older subjects), HR and the presence of cardiovascular disease were significant determinants of cardiovascular mortality in all age groups, whereas the role of other risk factors, especially lipid disorders and lifestyle, are predominant in younger subjects. These results, which are drawn from a large French cohort, show that using different risk equations in younger and older subjects may improve the evaluation of CVD risk.