Emerging observational evidence suggests that higher levels of testosterone after menopause may increase women’s risk of cardiovascular disease (CVD). The latest data—from the Multi-Ethnic Study of Atherosclerosis (MESA)—appear to firm up this potential association and to underscore the importance of accurate and reliable testosterone measurements in women to help understand the hormone’s role in health and disease.
The latest data from MESA, an ongoing study involving a diverse group of more than 6,000 men and women from across the United States, indicates that postmenopausal women with higher blood levels of testosterone-to-estrogen have a greater risk of developing CVD (19%), coronary heart disease (45%), and heart failure (31%), according to the study’s senior author Erin Michos, MD, MHS, associate director of preventive cardiology at the Johns Hopkins Ciccarone Center for the Prevention of Heart Disease in Baltimore (J Am Coll Cardiol 2018;71:2555–66).
She noted that higher testosterone and higher estrogen were linked to greater and lower heart disease risks, respectively, but the testosterone-to-estrogen ratio seemed to be the best risk indicator. “We found that [this ratio] was a little bit stronger marker of risk,” she said.
Not all observational studies have connected higher testosterone levels with greater risk of heart disease in women, leading to some debate in the field about the hormone’s role in CVD. But results over time from MESA have consistently linked higher testosterone in older women to heart disease risk, according to Michos.
“We have a large accumulating body of evidence that at least in an observational study, women with higher testosterone levels have increased cardiovascular risks,” she said. “They have a worse metabolic profile, more atherosclerosis progression, more endothelial dysfunction, more hypertrophy in their heart, and they’re at higher risk of events.”
Understanding what role testosterone and other hormones might play in women’s heart disease will be key going forward, as will confirming the MESA results in other investigations.
Studies have long shown that men have higher heart disease risks, which led some to suspect testosterone as the culprit. Men’s levels of testosterone gradually decrease over time, but studies of testosterone supplementation in men with low testosterone have suggested an increased risk of heart attacks and stroke, which led the Food and Drug Administration to issue a warning about the potential downsides of this approach.
At the same time, studies have shown that women’s heart disease risks rise after menopause, supporting the hypothesis that estrogen might be protective. However, results in 2002 from the landmark Women’s Health Initiative study linked postmenopausal estrogen supplementation to increased heart disease risks, casting doubt on that theory.
In addition to the MESA results, some research suggests that extremely low concentrations of estrogen or extremely high levels of testosterone may increase women’s heart risks. But other studies haven’t found a connection between testosterone and heart events or subclinical heart disease.
“In men, we have clear associations between sex hormones and cardiovascular risk factors, including blood pressure, lipids, diabetes, or metabolic syndrome,” said Robin Haring, PhD, an epidemiologist and adjunct professor at Monash University in Melbourne, Australia. “In women, I think the picture’s much more diluted because the whole hormonal profile is much more diverse, and there are many more players involved. It’s much less clear than in men where testosterone is the top-notch anabolic hormone.”
Many of the studies to date have measured individual hormones in isolation despite the complexity of hormone interactions and hormone metabolism in women, said Virginia Miller, PhD, director of the Women’s Health Research Center at the Mayo Clinic in Rochester, Minnesota. That, along with studies that have lumped many forms of heart disease into one composite endpoint, may also be contributing to the lack of consistent results in observational studies, said Miller, who also authored an editorial about the most recent MESA results (J Am Coll Cardiol 2018;71:2567–9).
“The way we think about hormonal effects on disease has to become more complex and more precise,” she emphasized. The latest MESA data add to the evidence so far by separately analyzing different forms of heart disease and by looking at multiple hormones simultaneously, Miller noted.
If the association between higher testosterone-to-estrogen ratios and CVD is confirmed in other studies, Michos said, measuring these hormones might be a useful marker for assessing risk. In the meantime, she and her colleagues are trying to determine how a higher testosterone-to-estrogen ratio might increase heart disease risk.
“Our main goal was first to … try to establish a mechanism,” Michos said. Her research team also would like to know if there are any ways to reduce women’s CVD risks beyond traditional interventions, such as healthy lifestyle changes, blood pressure control, lipid treatment, or glucose control.
Doing that will likely require a pharmacogenomics approach that takes into account differences in the way individual women metabolize estrogen and other hormones and how this interacts with specific diseases of the cardiovascular system, Miller said.
“Understanding what is causing the variability in estrogen response among women will really go a long way to help us understand how that affects many diseases in women, not just cardiovascular disease,” she elaborated.
While debate continues about the meaning of the conflicting findings from MESA and some other studies, there’s broad agreement that more accurate measurements of testosterone in women are needed, both for research and for clinical testing.
Accurately measuring the low levels of testosterone in women is very difficult for all assays. Consequently, different labs often produce inconsistent results on the same sample, explained Hubert Vesper, PhD, director of the Clinical Standardization Programs at the Centers for Disease Control and Prevention in Atlanta.
This may happen because of lack of specificity with some assays: With the concentration of testosterone in women being small relative to other interfering compounds in a sample, the interfering compounds can have a high impact on the measurement results. The impact of these interferences can be less pronounced in samples from men where testosterone concentrations are much higher, he said.
“The levels of testosterone are much lower than what we normally see in men, roughly about 20 times lower,” Vesper explained. “At these low concentrations, research studies found that the measurements are not very accurate.”
Most of the studies on testosterone’s role in women’s heart health, including the MESA studies, have used radioimmunoassays to measure testosterone. However, since then mass spectrometry increasingly has been embraced as a more reliable method of measuring women’s testosterone, according to the Endocrine Society’s 2014 guidelines on androgen therapy for women. In fact, that guideline notes that although immunoassays are inexpensive and simple, they “cannot be recommended.”
“The field is moving toward more precise measurements with mass spectrometry, rather than radioimmunoassays,” Michos said. She and her colleagues are applying for grants to repeat her study now that MESA has collected 15 years of follow up data on participants to look at whether changes in hormone levels contribute to heart risk. If this initiative gets funded, the investigators will likely use mass spectrometry this time around.
Both Michos and Haring agreed it’s too soon to start using testosterone measurements to assess heart disease risk in women. The relationship and its mechanism must first be confirmed and relevant clinical cutoffs need to be determined, Michos noted.
As Haring put it, “All the associations between sex hormones and subclinical and clinical cardiovascular disease in women are in the exploratory state of producing evidence.”
Bridget M. Kuehn is a science writer in Chicago, Illinois. +Email: firstname.lastname@example.org