Free androgen index as a determinant of arterial stiffness in menopause: a mediation analysis

OBJECTIVE: Associations of endogenous androgens in menopause with blood pressure (BP) and indices of arterial stiffness are reported, but directional relationships are not clear. Structural equation modeling is a contemporary statistical method, which allows assessment of such relationships and improves pathway understanding. METHODS: We recruited 411 consecutive apparently healthy postmenopausal women who underwent noninvasive vascular evaluation. This included pulse wave analysis (aortic pressures and arterial wave reflections [augmentation index]), measurement of aortic stiffness by pulse wave velocity (PWV), stiffness index (SI), and flow-mediated dilatation. A cumulative marker combining PWV and SI (combined local and aortic arterial stiffness [CAS]) was also assessed. Free androgen index (FAI) was calculated from circulating total testosterone and sex hormone-binding globulin. RESULTS: FAI was an independent determinant of systolic BP (SBP) (P = 0.032), SI (P = 0.042), and PWV (P = 0.027). Under structural equation modeling analysis, FAI was a direct predictor for PWV (beta = 0.149, P = 0.014), SI (beta = 0.154, P = 0.022), and CAS (beta = 0.193, P = 0.02), whereas SBP was a parallel mediator of androgen's vascular effects on PWV (beta = 0.280, P < 0.001) and CAS (beta = 0.248, P = 0.004), but not SI (beta = 0.024, P = 0.404). FAI-induced increase in arterial stiffness via flow-mediated dilatation was not established. FAI was not a determinant of augmentation index. CONCLUSIONS: In healthy postmenopausal women, FAI was directly associated with PWV, SI, and CAS. FAI also directly correlated with SBP, which in turn concurrently increased PWV and CAS. The directional correlations found herein, imply that endogenous androgens may be causally associated with indices of arterial stiffness both directly and indirectly. This hypothesis should be confirmed in further studies with causal design.