Over Decades, Women Still Missing in NIH-Funded Trials for CVD

Women continue to be underrepresented in cardiovascular clinical trials sponsored by the US National Institutes of Health (NIH), with no improvement seen over the past two decades, a new analysis shows.

This stasis has occurred despite the agency’s own emphasis on inclusion of women and other minority groups in research.

“We know historically that women have been underrepresented in clinical trials, but what's less understood is what we're trying to do to include women at a higher proportion and if we're really moving the needle on appropriate representation of women,” said senior study author Timothy B. Plante, MD, MHS (University of Vermont, Colchester).

He told TCTMD that the focus of this analysis, published recently in the American Journal of Cardiology, was on recruitment efforts in NIH-sponsored clinical trials of interventions for cardiovascular disease.

“Unfortunately, it seems almost that recruitment is an afterthought in a lot of clinical trials,” even without the added goal of enrolling more women, Plante said. Many trials fail as a result of not being able to reach recruitment targets, and even those that do succeed may fail to get “appropriate representation of underrepresented groups,” he noted, pointing out that their group did a similar study looking at Black participants back in 2021.

Unlike with race/ethnicity, where the biological differences are minimal, he continued, the sex-related differences are “tremendous.”

“The exposure to estrogen and testosterone throughout a lifetime have pleomorphic, generalized, very different effects on cardiovascular disease, risk factors, and cardiovascular disease outcomes. We've known that [for years]. At the same time, because men get heart disease younger, there's been sort of a greater emphasis historically on heart disease in men,” said Plante. “But guess what? Heart disease is also the number one killer in women. And we need to understand how our interventions to mitigate cardiovascular disease affect both men and women.”

Harriette Van Spall, MD, MPH (McMaster University, Hamilton, Canada), commenting on the results for TCTMD, said the findings are yet another reminder that women only account for a minority of those enrolled in trials of CVD.

“We want evidence regarding the efficacy and safety of interventions in all the people who live with the disease, including women. And we know that sex matters when it comes to pharmacokinetics, pharmacodynamics, volume of distribution, body size,” and other factors that are relevant when dosing drugs and sizing devices, she stressed. “We want to make sure that everyone living with the disease is represented in our trials.”

Beyond this, however, it’s important to understand why women—and other groups—are currently underrepresented and come up with solutions, said Van Spall. “We need to move from documenting gaps that we know have existed for more than 20 years to closing them.”

Few Set Recruitment Goals

The researchers, led by Yingfei Wu, MD (NYU Langone Health, New York, NY), searched ClinicalTrials.gov for NIH-sponsored studies of interventions for CVD in adults that were completed between 2000 and 2019. To be included, the studies must have had at least one US site and at least 100 participants.

The investigators used the participation-to-prevalence ratio (PPR), a measure that compares the proportion of women in the studies to the American Heart Association (AHA)-estimated prevalence of disease by sex, to gauge representation. If the PPR value fell between 0.80 and 1.2, it was considered adequate, while < 0.80 was underrepresentation and > 1.2 was overrepresentation.

The science of advertisement and recruitment . . . is decades behind where we are with scientific knowledge for diagnosis and prevention of heart disease. Timothy B. Plante

They found 62 trials, with a median enrollment size of 305 participants, that met their criteria and had protocols that were either publicly available or could be requested by the investigators. Women, on average, made up 39% of the study cohorts, but just 13% of the studies reported gender-specific subgroup analyses.

One in five trials specified goals for recruiting female participants, with a median target of 41%. Two-thirds of these actually met their target.

Most of the 62 studies (82%) used active recruitment strategies—such as through electronic medical records—while 13% used passive strategies and 2% used a combination of both. A minority (16%) pursued community-based recruitment, with even fewer getting community input on study design (6%) or listing community members as co-authors (3%). One in five paid participants beyond travel expenses.

Importantly, among the recruitment strategies, there were no differences in the proportion of women enrolled. And there were no gains in representation from 2002 to 2017.

Women made up a larger slice of studies testing behavioral interventions, but a smaller slice of those looking at drugs and procedures. Trials of hypertension and heart failure tended to enroll more women than those on other topics. Overall, most studies had PPR values < 0.8, apart from those devoted to hypertension (PPR = 1.14) and hyperlipidemia (PPR = 0.81).Why so few trial participants are women isn’t entirely clear, but Van Spall said much of it can be traced to sex-specific exclusion criteria. These can include being of childbearing age, not being on contraception, or not having had a hysterectomy, she pointed out. “So basically the presence of any anatomy that yields a child is the exclusion criteria.”

Although trialists might consider these reasonable due to potential safety concerns, they also could deter enrollment of women more broadly. “On the ground level it may be that staff don’t want to ask about those questions. [Either] they don't approach women or many women are approached but excluded by virtue of being within childbearing years,” she said.

Perhaps as a result, guideline recommendations for pregnant women are based on evidence of very low quality, said Van Spall, who recently co-authored a comment for Nature Medicine asserting that pregnant and lactating women should be included in clinical trials of CVD. “We have to wait for decades of observational data to know if something is harmful to the mother, potentially teratogenic, potentially harmful to a breastfeeding child. And that, on balance, puts pregnant women at more risk than including them in a trial in which there is careful monitoring [and] no biological or physiological plausibility of harm.”

Enrollment is less imbalanced, too, when women are at the helm as principal investigators, she observed.

But Van Spall said the lack of progress isn’t an indication that researchers are simply resistant to change.

“Clinical trials are really hard to run. It's easy to comment on what other investigators should or shouldn't be doing. In reality, there's a very tiny proportion of us that lead trials. It's hard work. It takes many years to get a single paper. And when you're running a trial, you just want to enroll patients, any patients, who will consent to the enrollment so that you have enough statistical power to assess your treatment effect,” she commented. That alone is tough, she added, even without going to the next level of asking whether the trial population is representative.

Basically the presence of any anatomy that yields a child is the exclusion criteria. Harriette Van Spall

Plante, too, emphasized that researchers aren’t purposely leaving out women.

A major factor, said Plante, is that the recruitment techniques have never evolved. “This is not a failure of the scientific community to rigorously assess a drug or an intervention for its outcome. We're great at that,” he noted, adding, “I don't think it's at all shocking that we're not moving the needle on engaging underrepresented groups, because the science of advertisement and recruitment . . . is decades behind where we are with scientific knowledge for diagnosis and prevention of heart disease.”

Unlike with a disease like cancer, where potential trial participants are a “captive audience,” Plante explained, in that they’re already being treated for a known condition at a specialized center, cardiovascular trials tend to be done at academic medical centers in patients who have a common disease or are at risk of developing it. The latter scenario runs the risk of missing potential participants who lack insurance or have other barriers.

Researchers continue to rely on newspaper or radio ads, or posters on the subway, to spread their message and haven’t yet adopted modern marketing strategies, he noted. A social media or digital campaign could better target people in specific areas who have certain demographic characteristics, though this approach requires a unique skill set and can be costly. And there are privacy concerns.

How best to pursue these new strategies, and to do so ethically, is the focus of their work as part of an AHA-sponsored Strategically Focused Research Network on the science of diversity and clinical trials, Plante said.

His hope, he said, is that “the cardiology community can come together and emphasize that we need to improve the recruitment of women and other underrepresented groups so that we can have the best knowledge to save lives from heart disease moving forward.”