Estradiol and HDL Function in Women - A Partnership for Life.

High-density lipoprotein (HDL) cholesterol concentration has been widely recognized as a key negative risk factor for cardiovascular and metabolic disease. However, failure of clinical trials aimed at raising levels of circulating HDL cholesterol to reduce cardiovascular risk has shifted the emphasis of research toward understanding HDL’s protective functions in the vasculature. In addition to reverse cholesterol transport, HDL has anti-inflammatory, antioxidant, vasodilatory, and antithrombotic functions protective of the endothelium, which are impaired in cardiovascular diseases (CVD) such as hypertension and coronary artery disease as well as metabolic diseases including obesity and type 2 diabetes mellitus (1). HDL function, not HDL cholesterol (HDL-C), is therefore a better biomarker for cardiovascular risk prediction. CVD risk is higher in men compared with women until women reach the menopause; thereafter, CVD risk increases in women with every year after menopause and eventually exceeds that of men. Menopause is characterized by a decrease in estradiol (E2), implicating the sex hormone in mediating CVD risk in women. In fact, the protective association of HDL-C with brachial artery flow-mediated dilation is only present in menopausal women in the highest tertile of circulating E2, indicating a role for E2 in the modulation of HDL function (2). Understanding the relationship between E2 and HDL’s protective functions in women is key, given that women experience precipitous changes in E2 at 2 major periods in their adult lives—an increase during pregnancy and a decrease at the menopause.

The maternal adaptation to pregnancy exposes the mother to metabolic and inflammatory changes that in nonpregnant individuals would be expected to impair vascular function, such as insulin resistance, hypertriglyceridemia, and increased inflammation. However, vascular function is enhanced in pregnancy, and despite increases in plasma triglycerides, HDL-C increases throughout gestation. It has therefore been hypothesized that HDL synthesized in pregnancy is of increased functionality to overcome the detrimental metabolic profile associated with pregnancy adaptation (1). The increase in plasma HDL-C concentration is concomitant with pregnancy-associated E2 production that has a multitude of direct and indirect effects on lipid and lipoprotein metabolism and the vascular endothelium. These include potent antioxidant effects, reducing LDL particle size and preventing adhesion molecule expression in endothelial cells (reviewed in (3)). Interestingly, many E2 functions overlap with HDL functions. Indeed, E2 has been shown to directly associate with HDL in women, but not in men, thereby enabling increased endothelial nitric oxide (NO) synthase (eNOS) activation and NO production beneficial to vascular function (4). HDL-associated E2 may directly boost HDL function or increased E2 levels may stimulate the production of more functional HDL, either of which could counteract the adverse maternal metabolic adaptations to pregnancy. Of note, HDL from the hypertensive pregnancy disorder preeclampsia has reduced functionality (5) against a background of reduced E2 compared to healthy pregnancies. Despite our increasing knowledge of E2 effects on HDL function in women, studies to date on the association between HDL, E2, menopause, and worsening cardiovascular risk have largely focused on circulating HDL-C concentration.

The recent article by El Khoudary et al (2021), from the Study of Women’s Health Across the Nation (SWAN), has now begun to address this knowledge gap (7). The authors explored the association of E2 with measures of HDL composition and function across the menopause, using time since final menstrual period (FMP) as the independent variable. Commendably, almost 20% of study samples were defined as late or postmenopausal despite difficulties in recruiting sufficient numbers of these women, due to exclusion of those taking hormone replacement therapy (HRT). Women were followed from early menopause in a longitudinal manner. In addition to HDL-C, HDL particle size, phospholipids, triglycerides, and cholesterol efflux capacity (CEC) were measured. Positive associations between E2 and larger HDL size, higher concentrations of HDL phospholipids and triglycerides and HDL-CEC were found after adjusting for factors that may affect hormone levels (cycle day of the blood draw) and HDL metrics (inflammation and body mass index). These characteristics of HDL observed with higher E2 are indicative of HDL with altered composition and function. Larger HDL particle size and increased HDL phospholipids are anti-atherogenic and are associated with improved HDL-CEC in a study of coronary artery disease (6). Notably, HDL-CEC was the only direct measure of HDL function in the study. Conversely, increased HDL triglycerides are a risk factor for atherosclerosis and indicate HDL of pro-atherogenic nature, reflective of the increased CVD risk inherent in the menopause and the duality of HDL function. After more than 2 years post-FMP, the positive association between E2 and HDL triglycerides was markedly strengthened, suggesting an increase in CVD risk and that timing of any intervention to increase E2 levels may be critical (7).

El Khoudary et al (2021) also explored the relationship between follicle-stimulating hormone (FSH) and HDL function across the menopause (7). Across all time-points, higher FSH was associated with higher total HDL particles (linked with lowered CVD risk) but fewer large HDL particles, lower overall particle size and reduced HDL-CEC per particle (all indicative of reduced HDL function). However, FSH was positively associated with apolipoprotein AI, the predominant functional HDL protein, and increased HDL particle number up to 6 months post-FMP, both positive indicators of HDL function. After 6 months post-FMP, FSH correlated with smaller HDL particles, up to 5 years post-FMP. As acknowledged by the authors, these contradictory findings are difficult to put into context given the lack of research into FSH and HDL together. There is perhaps an indirect effect of FSH on the liver in the first 6 months post-FMP, leading to a transient increase in apolipoprotein AI on HDL. Also, in the absence of kinetic turnover studies of HDL metabolism it can be difficult interpret HDL size with respect to function and plasma clearance as a single assessment of size does not indicate whether size is increasing or decreasing at that moment in time. It is clear that more research is required to understand the role of FSH on cardiovascular risk in menopause, although this study’s findings provide a platform on which to build.

Moving forward to the clinical implications of this study, it would be prudent to understand whether there is a threshold effect for the positive association between E2 and HDL-CEC, perhaps providing a target E2 concentration for optimal HDL function. The timing of HRT may be important for maximum reduction of cardiovascular risk postmenopause. The route by which E2 is administered may be significant in maximizing CVD risk reduction, as transdermal E2 prescribed during perimenopause had no effect on HDL-C but did reduce HDL-CEC compared with placebo (8). Experimentally, HDL-CEC was the only functional aspect of HDL considered in this first foray into understanding the relationship between HDL function and endogenous hormones during menopause. Future studies should consider measuring other HDL functional properties, such as endothelial cell anti-inflammatory assays, the ability of HDL paraoxonase-1 to prevent low-density lipoprotein oxidation, and in view of HDL-associated E2 effects on eNOS (4), HDL-mediated NO production across the menopause.

In their well-designed study El Khoudary et al (2021) have demonstrated both indirectly (through particle size and lipid analysis) and directly (through cholesterol efflux capacity) that E2 concentration across the menopause transition is positively associated with HDL function, although E2 can be associated with pro-atherogenic effects further out from FMP. These data raise questions regarding how and when to begin E2 replacement therapy with the aim of preventing CVD. This study adds to the broader knowledge of E2 and HDL vascular protection, a key partnership throughout a woman’s life.