AS758 - Postmenopausal androgens, sex hormone binding globulin, and cardiovascular risk: a cohort study

Investigator Names and Contact Information

Laura Harrington (laura.b.harrington@kp.org)

Tracy Madsen

Introduction/Intent

Cardiovascular (CV) event risk differs markedly by age between men and women. Among men, CV event risk increases nearly linearly with advancing age; however, among women, CV event risk increases dramatically following postmenopause. It has been speculated that peri- and postmenopausal changes in endogenous sex hormone profiles are major drivers of this increased risk accompanying the menopausal transition. However, most research to-date has focused on changing levels of estrogens, even though postmenopausal estrogen levels drop abruptly in postmenopause and have only been inconsistently associated with CV event risk. In contrast, female adrenal androgen levels decline more gradually, by about 80% between the ages of 20 and 70 years,1 suggesting that androgen levels may become of greater importance to risk predication in postmenopause than estrogens. Among men, higher serum dehydroepiandrosterone (DHEA) and testosterone (T) levels have been associated with less coronary artery calcification2 and a lower risk of major adverse CV events3; however, in women, associations between postmenopausal androgen profiles in relation to CV risk remains largely unknown. This is despite data demonstrating key associations between sex hormone binding globulin (SHBG), a transporter and regulator of bioavailable T, with stroke, diabetes, and coronary heart disease (CHD) risk in prior work by our group.4,5

In the US, women spend more than 1/3 of their lives in postmenopause6, yet postmenopausal changes in androgen profiles across the decades spent in postmenopause are unclear. Furthermore, it is unknown how postmenopausal androgenic profiles and changes relate longitudinally to markers of CV risk (e.g. C-reactive protein [CRP], fibrinogen, high-density lipoprotein [HDL]) and the risk of CV events such as stroke, coronary heart disease (CHD), and venous thromboembolism (VTE). An improved understanding of postmenopausal androgen levels, profiles, and changes in relation to markers of CV risk is critical to elucidating mechanisms underlying the increased CV event risk in postmenopause. This knowledge could inform future genetic and therapeutic targets aimed at reducing postmenopausal CV risk, while improving postmenopausal prediction modeling and risk stratification.

We hypothesize that postmenopausal androgen levels and profiles will be associated longitudinally with markers of CV risk and with CV event risk among older women. We propose an ancillary study to the Women’s Health Initiative (WHI) that will leverage existing stored biospecimens and markers of CV risk previously measured across two to three timepoints spanning 20 to 30 years (1993-1997, 2012, 2023) among postmenopausal participants 50 to 79 years of age at baseline, also followed for up to 30 years for CV events. This cohort study nested in the WHI will include ~774 participants in total which includes ~200 CV events. Using stored biospecimens, we will newly measure postmenopausal levels of adrenal androgens (total T, dihydrotestosterone [DHT], and DHEA sulfate [DHEAS]), SHBG, and estrogens (estrone [E1] and estradiol [E2]) at two to three timepoints per participant across 20 to 30 years and will characterize postmenopausal profiles of androgens and SHBG. Existing markers of CV risk measured at these same timepoints are systolic blood pressure (SBP), CRP, creatinine, Homeostatic Model Assessment for Insulin Resistance (HOMA-IR), total cholesterol, HDL, low-density lipoprotein (LDL) cholesterol, and triglycerides. Factor VII antigen (FVII:Ag), factor VII activity (FVIIC), and fibrinogen were measured at baseline only. CV events of interest (stroke, CHD, and VTE) were adjudicated over the same time period.

Specific Aims

Aim 1: Characterize postmenopausal androgen profiles across 20 to 30 years using latent profile analysis and principal component analysis.

Aim 2: Evaluate the association of time-varying postmenopausal androgen levels and profiles over 20 to 30 years in relation to rates of change in markers of CV risk, above and beyond those associated with chronologic aging.

Aim 3: Evaluate postmenopausal androgen levels and profiles over time in relation to CV event risk.