Differential contribution of testosterone and estradiol in the determination of cholesterol and lipoprotein profile in healthy middle-aged men

The role of endogenous sex steroids in the association between male gender and cardiovascular risk remains unclear. We performed a cross-sectional analysis of the role of endogenous testosterone (T) and estradiol (E-2) as well as their respective biologically active fractions, in the determination of lipids and lipoproteins in an occupation-based cohort of 715 healthy middle-aged men. Serum T, sex hormone binding globulin (SHBG) and E-2 were measured by immunoassays; free T (FT) and free E-2 (FE2) were calculated using a validated equation. Serum total cholesterol (Chol), HDL-cholesterol (HDL-Chol), apolipoproteins Al (ApoAl), B (ApoB), E (ApoE), ApoE phenotype, lipoprotein a (Lpa), fibrinogen, C-reactive protein (CRP), systolic (SBP) and diastolic blood pressure (DBP) were assessed. Serum levels of T and FT, correlated positively with HDL-Chol and ApoAl with Spearman correlation coefficients, partialised for age and body mass index (BMI), ranging between 0.14 and 0.17 (P < 0.001); FT was associated with total Chol and ApoB levels (r = 0.12 for both T and FT; P < 0.01). After adjustment for age and BMI, both serum E2 and FE2 levels correlated significantly with ApoE (r = 0.25 and r = 0.26 for E-2 and FE2, respectively; P < 0.001). Free and total E-2 were associated with both SBP and DBP with correlation coefficients partialised for age and BMI ranging between 0.11 and 0.13 (P < 0.01). No correlation was found between any of the studied sex steroids, fibrinogen, Lpa or CRP. In multiple linear regression analyses, T was the most important independent hormonal determinant of HDL-Chol levels, when E-2, SHBG and exogenous factors were considered in the model (P < 0.01), whereas E-2 contributed mostly in the determination of ApoE levels (P < 0.001) and SBP (P < 0.01). When FT and FE2 were considered in multivariate analyses as independent hormonal variables, FT was the most significant predictor of HDL-Chol (P < 0.01) and ApoB (P < 0.01) concentrations. Moreover, in the same multivariate model, ApoE (P < 0.001) concentration as well as SBP (P < 0.001) was most affected by FE2 levels in comparison with FT. In conclusion, our findings do suggest a differential role of T and E-2 in the determination of traditional cardiovascular risk factors in healthy middle-aged men. In the determination of both HDL-Chol and ApoB levels endogenous (F)T may be involved, whereas (F)E-2 may contribute to the determination of ApoE levels in this study group of 715 healthy middle-aged men. Regarding the observational design of the study, the physiological relationship of the observed associations between sex steroids and cardiovascular risk factors remains to be unravelled. (C) 2002 Elsevier Science Ireland Ltd. All rights reserved.