The typical pharmaceutical user is elderly and female. The typical clinical research participant? Middle aged and male. However, elderly women often metabolize drugs differently and are more likely to experience side effects than younger men.
The importance of research including a diverse range of participants across gender, age, socioeconomic status, ethnicity, and co-existing medical problems is increasingly recognized. And it’s not just clinical research: it includes everything from making sure that the cells and animals used in pre-clinical research are no longer more likely to be male, to setting research priorities and looking at the impact of interventions.
The problems with diversity in medical research
Despite a decades-long push to include more diversity in research, many groups are still underrepresented. For example, Joel Ray, a researcher at St. Michael’s Hospital, studied whether ethnicity affects birth weight. He found that our standard weight charts – which were developed in the 1960s based on predominantly white, northern European babies – weren’t very good at judging which newborns were underweight for other ethnicities. Twelve percent of babies with South Asian parents who were of normal weight for their ethnic group would be classified as significantly underweight under the standard charts, and seven percent of babies with parents from Africa or the Caribbean would be. That puts them at the risk of unnecessary interventions, like staying in hospital longer or having testing.
Even women – who should be easy to find and recruit into clinical trials for research studies – tend to be underrepresented. This has led to issues around increased side effects in women, which have only been noticed in post-market surveillance. For example, Zolpidem (Ambien) is a sedative medication that the FDA recently recommended have a lower dose for women, since women eliminate Zolpidem from their bodies more slowly than men, and taking it might impact their ability to drive in the morning. “A number of drugs have been withdrawn from the market as a result of adverse events that disproportionately affected women,” says Paula Rochon, vice-president of research at Women’s College Hospital.
The elderly are also routinely excluded. That can be explicit, with studies excluding participants over a certain age or it can be the result of excluding people with other diseases in an attempt to have cleaner data. “The problem with older people, in short form, is multiple morbidities,” says Michael Gordon, professor of geriatrics at the University of Toronto. But that can leave out the key users of a drug since most diseases are disproportionately represented in the elderly, such as dementia, heart issues and cancer.
“People are realizing that it’s not just age [that needs to be included], it’s looking at the comorbidities,” says Sharon Straus, a geriatrician and director of the Knowledge Translation Program at St. Michael’s Hospital. She adds that it’s important to test medications in people who have similar comorbidities to those most likely to take the drug. A cholesterol lowering medication, for example, should be tested in people with diabetes. “We should try to look at the common clusters of comorbidities, and target those,” she explains.
One of the pressures against doing that, says Peter Juni, director of the Applied Health Research Centre at St. Michael’s Hospital, is that the pharmaceutical industry “doesn’t want to gamble.” They want to test their drugs in the people most likely to benefit, and least likely to be harmed – and older people with multiple issues are decidedly not that group.
On the other end of the spectrum, youth are often excluded from research, says Joanna Henderson, a Clinician Scientist in the Child, Youth and Emerging Adult Program at the Centre for Addiction and Mental Health in Toronto. There are often concerns about the potentially harmful effects of drugs on growing bodies, and it can be difficult to distinguish between the effects of the drugs and normal developmental progress.
Then there are logistical challenges of including both youth and their parents in longer term clinical trials, and of following such a rapidly changing group over time. But not including them as participants in drug testing means that the drug’s effects on puberty, growth, and weight management are unknown. Even treatments that aren’t drug-based can have unintended consequences for children and teenagers. “Cognitive Behavioural Therapy (CBT) was widely tested in adult populations, and shown to be effective across a few diagnostic areas,” says Henderson. “But with CBT, you leave each session with tasks to do in the coming week – homework. Often the parents have made a really big effort to get their child into treatment, and if the child doesn’t complete the homework, it can play out in ways that can be unhelpful.” That issue also results in much higher drop-out rates from trials than you see in adults.
Similar unknowns come up when investigating rural and remote groups, and Indigenous people and the Francophone population. “We need research to identify innovative, effective interventions” in these groups, says Alain Gauthier, a faculty investigator at Laurentian University’s Centre for Rural and Northern Health Research (CRaNHR). Even if patients from Northern Ontario are included in studies, he says, small sample sizes mean that they often can’t be reported on separately.
Ideally, groups participating in drug trials and studies would reflect the people who will use the therapy. But that can be difficult, says Juni, and it’s not always required. “In most situations, the average treatment effect that you observe will actually hold [across multiple subgroups],” he explains. “You could safely assume that if you don’t have clinical evidence or a strong biological argument that there’s a difference, and you have, for example, 20 percent women, that’s underrepresented to a certain extent, but it’s still probably safe to assume that [your treatment] works as well in females as males.”
And sometimes, he says, “you need to compromise between what is feasible and what is desirable.” Pragmatic trials – randomized trials that look at patients in regular practice, rather than in controlled clinical settings – could be an answer to this, he argues.
Other solutions include groups like CRaNHR, which both does its own research and partners with larger urban centres, offering access to those often-underrepresented groups. Having those groups act as partners in research also improves the chances that they’ll be participants in trials.
Changing the way research funding is provided has helped push this forward. Including more women, children and the elderly has been explicitly supported by the Canadian Institutes of Health Research, including an expectation that applicants will incorporate sex and gender into their research. Meanwhile, the U.S.’s National Institutes of Health also requires that women and minorities be included in research.
That’s changing the conversation – but it still hasn’t gone far enough, says Straus. “The bulk of my elderly patients would not be eligible for most of the trials that we see published,” she says. “If that’s the best available evidence, then that’s what we have to use, but we should also push for research that would include my patients.”