High, Low Estradiol Levels Linked to Mortality in Men With Chronic Systolic Heart Failure

May 19, 2009

May 18, 2009 (Chicago, Illinois) — Both low and high serum levels of estradiol, compared with mid-range levels, are significantly and independently associated with increased all-cause mortality in men with chronic systolic heart failure, suggests a prospective observational study in the May 13, 2009 issue of the Journal of the American Medical Association [1].

Why serum estradiol levels would show such a U-shaped relationship with mortality in such patients isn't actually known, but the study also showed low and high levels to be associated with different clinical characteristics, "suggesting that the underlying pathophysiological mechanisms are not the same," write the authors, led by Dr Ewa A Jankowska (Military Hospital, Wroclaw, Poland).

Of course, the study does not suggest that low and high estradiol concentrations themselves were the cause of increased mortality, observed Dr Vasan Ramachandran (Boston University School of Medicine, MA) for heartwire . It also doesn't show that low and high levels will make useful prognostic markers in men with heart failure. For that, he said, it would be helpful to know causes of death, something that isn't much discussed in the report.

"Was it from progressive pump failure, or arrhythmias, or thromboembolic events?" he asked. "That would be the next thing to look at." Ramachandran, who isn't connected with the analysis, is codirector of his center's echocardiography/vascular laboratory and one of 14 primary investigators with the Framingham Heart Study.

Jankowska et al prospectively followed 501 men with stable HF, mean age 58, and echocardiographically documented LVEF<45% (mean 28%). Most by far were in NYHA class 2 or 3. Exclusion criteria included recent ACS, coronary revascularization, or unplanned cardiovascular hospitalization; any history of acute or chronic illness known to affect hormones (including infections and autoimmune diseases); and use of hormone therapy, including anabolic steroids, corticosteroids, and thyroid hormones.

Mortality over three years was 34%. With the men grouped according to estradiol-concentration quintiles, the middle quintile prospectively designated as the reference group, adjusted mortality hazard ratios for those in quintiles 1, 2, and 5 were significantly increased.

Serum Estradiol Concentration Quintiles as Predictors of All-Cause Mortality in Men (With Quintile 3 as the Reference Group)

Parameter Quintile 1 Quintile 2 Quintile 3 Quintile 4 Quintile 5
Estradiol concentration (pg/mL) <12.903 12.90–21.79 21.80–30.11 30.12–37.39 >37.40
3-y survival* (%) 44.6 65.8 82.4 79.0 63.6
Quintile as mortality predictor,* HR (95% CI) vs quintile 3 4.17 (2.33–7.45) 2.15 (1.16–3.99) -- 1.22 (0.64–2.31) 2.33 (1.30–4.18)
HR= hazard ratio

*Adjusted for clinical variables (including age, body-mass index, NYHA class, LVEF, biomarkers, renal function, diuretic and aldosterone-blocker use, hypertension, and diabetes) and androgens (total testosterone and dehydroepiandrosterone sulfate)

Men in the highest and lowest quintiles showed different arrays of distinguishing clinical features at baseline that offered clues, perhaps, to pathophysiologies behind their poorer outcomes. Those in quintile 5 had significantly increased levels of bilirubin and liver enzymes and significantly lower serum sodium levels than men in quintile 3 (p<0.01 vs quintile 3 for all differences). Men in quintile 1 had increased serum total testosterone (p<0.001 vs quintile 3) and decreased serum dehydroepiandrosterone (p<0.05), total fat tissue mass (p<0.01), and glomerular filtration rate (p<0.05) and were less likely to be in NYHA class 1–2 rather than class 3–4 (p<0.01 vs quintile 3).

Poorer outcomes with low estradiol levels "are something that we have seen in Framingham," according to Ramachandran. "We've seen it in the context of overt heart failure, and we've seen it in people without overt cardiovascular disease. The outcome we looked at was cardiovascular disease occurrence, here they're looking at people with established heart failure and their outcome is mortality, but the concepts are parallel."

Speculating on the mechanisms behind the risk increase at low estradiol levels, he observed that "estrogen has some vascular protective and cardioprotective influences. So people who have lower levels of estradiol would have less of that. It's probably true even in people with heart failure."

Increased mortality for men in the highest estradiol quintile "I think is a little bit more puzzling." But Ramachandran said it could follow from longstanding evidence that estrogen therapy in men with prostate cancer, for example, increased their risk of thromboembolic events and diminished survival.

He also speculated that men in quintile 5 may have had more severe congestion of the liver, a common part of the heart-failure syndrome. "So despite all the adjustments that they did, the upper end of the estradiol distribution may just be marking out those who are sicker at some level."