Objective: Whether endogenous androgen levels contribute to the cardiovascular disease (CVD) risk profile in women is controversial. The purpose of this study was to investigate systematically the relationships between serum levels of endogenous androgens and sex hormone-binding globulin (SHBG) and biochemical CVD risk profile, taking other known risk factors into account.
Design: This community-based cross-sectional study included 587 non-healthcare-seeking-women, aged 18 to 75 years, who were randomly recruited from the community via the electoral roll from April 2002 to August 2003. Participants were euthyroid; had no usage of exogenous steroids; had no history of tubal ligation, hysterectomy, or bilateral oophorectomy; and did not have hyperprolactinemia or polycystic ovarian syndrome. The relationships between total testosterone, SHBG, free testosterone, dehydroepiandrosterone sulfate, and androstenedione and high-sensitivity C-reactive protein (CRP) and lipids were explored using linear regression with natural logarithm (ln) -or square root-transformed data as indicated. Issues of nonlinearity and interaction were addressed by the inclusion of extra regression terms where appropriate. We determined the change in the proportion of variation for each marker of the CVD risk profile explained by the addition of each hormone term to the models, having adjusted for age, body mass index, smoking, alcohol, and exercise.
Results: Menopausal status did not influence the statistical models for high-sensitivity CRP and high-density lipoprotein cholesterol, but for both low-density lipoprotein cholesterol and triglycerides, the proportion of variation explained by the models was substantially less in postmenopausal than in premenopausal women. Almost all of the highly statistically significant findings were related to the addition of the SHBG terms to the models. The changes in r2 values were highly statistically significant for the addition of the SHBG terms to the models for ln CRP and ln high-density lipoprotein for both pre- and postmenopausal women (P ≤ 0.01 and < 0.001, respectively) and for ln triglycerides in postmenopausal (P < 0.001) and premenopausal women (P < 0.01).
Conclusions: Endogenous testosterone and the adrenal preandrogens per se are not significant independent determinants of circulating high-sensitivity CRP or lipoprotein lipids. Our analyses provide further support for the independent predictive value of low SHBG levels for CVD risk profile and an independent contribution of the menopausal transition to the determination of low-density lipoprotein cholesterol and triglycerides.