The Relationship between Sex Hormones, SHBG and Endothelial Function in Older Men: Data from the PIVUS Study

Background. Epidemiological studies have shown that low sex hormone binding globulin (SHBG) and testosterone (T) and high estradiol (E2) are independent predictors of metabolic syndrome and its components including hypertension in adult-older men (1).Preliminary data support a potential role of low SHBG in the risk of PAD in older men (2). However, despite the emerging link between SHBG, sex hormones and cardiovascular disease, the direct relationship with endothelium function has never been addressed. This is of importance since endothelium-dependent vasodilation assessed by the invasive forearm technique with acetylcholine given in the brachial artery (EDV) has been recently shown to be an independent risk factor of cardiovascular events in the older population (3).Aim of the study. To test the association between SHBG, sex hormones and endothelial function in older men.Methods and Results. Participants 70 years of age of the Prospective Study of the Vasculature in Uppsala Seniors (PIVUS) study (52% women), with complete data on EDV, and the brachial artery ultrasound technique with flow-mediated dilatation (FMD) were evaluated. From the entire sample of 1016 subjects, we used data of 430 men having information on SHBG, estradiol, T and endothelial function. SHBG, E2 and T were assessed by chemiluminescence (Beckman Coulter). The MDC for E2, T and SHBG were 73 pmol/L, 0.35 nmol/L and 2 nmol/L. The interassay coefficients of variation (CV) was <20% for E2, the intra and inter CVs <4 and <7% for T, the interassay CV 5.3% for SHBG. Generalized linear models adjusted for multiple confounders including BMI, C Reactive Protein, fasting insulin, were used to test the relationship between sex hormones, SHBG and endothelial function. In a crude analysis we found a positive significant relationship between SHBG and EDV (β± SE 4235.2 ± 908.5,p <.0001) but not with FMD. The relationship was maintained after adjustment for T (β± SE 3734.6± 1364.7, p=0.0065) and further adjustment for additional confounders including E2(β± SE 2898.9± 1413. 1,p=0.04). No significant relationship was found between T, E2 and EDV or FMD.Conclusions. In older men SHBG, but not testosterone and estradiol, is positively and independently associated with endothelium-dependent vasodilation