Anthropometric and lifestyle correlates of serum lipoprotein and apolipoprotein levels among normal non-smoking men and women.

The relationship between serologic predictors of coronary risk and anthropometric as well as lifestyle characteristics was investigated in 61 men (37.5 +/- 8.5 yrs) and 33 women (40.1 +/- 9.0 yrs). All subjects were healthy non-smokers, mostly middle-class bank employees. In bivariate analysis, among both genders the ratio of waist-to-hip circumference (WHR) was the single best predictor of levels of serum LDL-cholesterol, apolipoprotein B, VLDL-cholesterol and triglycerides (positive association) as well as HDL-cholesterol and apolipoprotein A-I (inverse association). In men, body fat as estimated from bioelectrical impedance measurement was the second best predictor of lipoprotein and apoprotein concentrations, whereas in women it was the body mass index (BMI). The additional independent predictive power of WHR and body fat for the lipid profile was tested in multivariate analysis by adding WHR and body fat sequentially to regression models containing already BMI, endurance capacity, exercise, alcohol consumption and age. For example, explained variance of triglyceride distribution rose from 26.3 to 35.1% (P = 0.01 for increase) when body fat was entered into the regression equation, or inclusion of WHR into a model already containing age, the behavioral variables, BMI, and body fat increased the explained variance of LDL/HDL-cholesterol ratio from 20.9 to 27.6% (P = 0.04 for increase). In women, the same regression models were even slightly more predictive for the serum lipid profile. Endurance capacity was related to a low atherogenic risk lipid profile in bivariate analysis but lost much of its predictive power in multivariate analysis, which confirms that the effect of fitness on lipid levels is probably mediated in part by a low body fat content. It is concluded from this cross-sectional investigation that studies which focus on associations between lifestyle and serologic predictors of atherogenic risk should possibly include the WHR and a measure of body fat, since the latter two appear to be closer correlates of serum lipoprotein and apolipoprotein levels than BMI or single behavioral factors, at least among male non-smokers.