A new NIH rule won’t be enough to make clinical research more inclusive

A quiet but revolutionary new national health policy goes into effect this week, ushering in changes that could lead to important medical discoveries that benefit most Americans. There’s just one problem. Implementing the change will require that our country’s health researchers make some fundamental changes in how they do business.

Under the National Institutes of Health’s new Inclusion Across the Lifespan policy, federally supported medical research must include patients of all relevant ages or explain their exclusion. Since most studies already include adults, and a mandate to include children has existed since 1998, the novelty in this policy is the stipulation that clinical research include people age 65 and older.

That’s a big group. It currently includes both Donald Trump and Nancy Pelosi, Ruth Bader Ginsberg and Clarence Thomas, as well as the 50 million other older Americans, along with the rest of us who get lucky enough down the road to make it into elderhood.

It’s not breaking news that, on average, older people get sick more often than younger people, have more medical conditions, take more medications, and are more likely to end up in the hospital. Given that, you’d think that the inclusion of older individuals in studies of the drugs and procedures for which they represent a substantial proportion of recipients would be a no-brainer — like including children in educational research. Yet that hasn’t been the standard in U.S. medical science. The consequences of this exclusion are profound.

An article this month in JAMA Network Open on the adverse effects of medical treatment on mortality in the U.S. since 1990 proved what those of us who specialize in the care of older adults have long suspected: People age 70 or older are nearly 20 times more likely to be killed by their medical care than those between the ages of 15 and 49. One of the main reasons for this is that clinicians like me don’t know how most treatments affect old people because they are routinely excluded from the research that informs treatment.

A systematic review of all clinical trials published in the top five medical journals over the course of a single year found both direct and indirect ways that current medical research puts older adults at risk. Almost 66 percent of the studies excluded elders: 20.2 percent used upper age limits and 45.6 percent had age-associated condition exclusions. But even when old people were included, the structure of many studies raised questions about the applicability of the results to their care. A majority lacked age-specific subgroup analyses, which help identify different treatment responses in different age groups; in the minority that included such analyses, half didn’t consider obvious age-related confounders, such as diseases more common in old age — but aren’t due to old age — that affect the outcome. Only 27 percent of these influential trials included outcomes of particular relevance to elders, including health status and quality of life.

This week, a new review of 50 years of NIH-funded clinical trials on the top causes of hospitalization or disease burden in older adults came to similar conclusions. Thirty-three percent of the studies had upper age limits, while 37 percent excluded participants based on medications commonly used by older people and 30 percent excluded those with diseases common among older people. Equally problematic, 67 percent of the trials had average patient ages younger than the actual averages for the diseases being studied.

Certain explanations for excluding older adults have powerful face value, and researchers tend to invoke those justifications. Some argue that studying old people poses unique biological challenges, such as distinguishing the effects of age from those of the intervention or accounting for the influences of participants’ chronic diseases or the medications they are taking. Others lament the difficulties and ethical challenges of recruiting particular types of individuals for studies. They note that getting to a trial center is more onerous for people with functional or health challenges, and that obtaining informed consent is more complicated in people with cognitive impairment.

One problem with such justifications is that they can become self-fulfilling. Compared to the diversity of the U.S., whites are overrepresented in clinical research while people from other racial and ethnic groups are significantly underrepresented. The oft-cited explanation is that those folks are less willing to participate in trials. Yet a meta-analysis of 20 studies that included more than 70,000 people found minimal differences in consent rates based on race or ethnicity. There were, however, significant differences in the numbers of people from minority groups who were invited to participate. In other words, individuals from racial or ethnic minority groups participated less often in research because they were less likely to be asked, possibly because researchers assumed they wouldn’t want to participate.

The second fallacy on which researchers base exclusions from clinical studies is the widespread belief that research on “special populations” helps only those populations. That assumption is inconsistent with many groundbreaking studies that illuminated universal biological processes by studying rare variants. It also leads to missed opportunities for unexpected discoveries. In one recent study, for example, young and middle-aged adults who received more aggressive treatments for colon cancer fared worse than the older adults who received what some clinicians would call less care.

In hospitals and research studies, as in life more generally, when older individuals have bad outcomes, people often say, “Well of course, they were old,” without considering that the very treatments meant to help were actually harmful. It’s impossible to distinguish age from disease, or better from worse, without looking at diseases and treatments in different age groups and various sorts of people. And clinicians can’t safely prescribe medications and other treatments for old people if researchers haven’t studied those interventions in old bodies.

These are just some of the reasons why the NIH’s new policy is so important. The problem is that while the policy is a laudable first step toward including more older individuals in clinical research, the histories of similar policies for other population subgroups tell us that rules are not sufficient to make it happen.

In 1989, the NIH mandated the inclusion of women and minorities in research, a principle that became law four years later. The NIH’s 1998 policy on inclusion of children was developed “because medical treatments applied to children are often based upon testing done only in adults, and scientifically evaluated treatments are less available to children due to barriers to their inclusion in research studies.”

While such policies have increased inclusion, they have not solved the fundamental problem that most studies still teach us more about certain sorts of people than others. “Twenty years and still counting,” lamented one review article on the inclusion of women. The title of another summed up the state of the science this way: “Diversity in clinical and biomedical research: A promise yet to be fulfilled.”

Complexity is universal in human lives and working with it, not imposing artificial simplifications, must become the standard in medical research. We can’t understand the contributors to disease or health or the efficacy of medical treatments if we are not adequately studying the 51 percent of the U.S. population that is female, the more than 40 percent who have brown or black skin, the 24 percent who are children, or the 15 percent who are over age 65. Reductionism may make for cleaner science, but it also results in research results of too little use to too many patients.

The NIH’s new lifespan policy is much needed and long overdue. Today, when old age lasts decades and offers unprecedented options for work, play, and learning, it’s also a significant opportunity to transform medical science and care in ways that will benefit us all.

But none of the NIH’s inclusion policies will succeed if they doesn’t invest in developing best practices for recruiting all types of people into studies. To achieve the inclusivity essential for optimal, generalizable medical research, it will also need to provide researchers with the practical and financial support to make inclusion a priority.

There is an inherent illogic, and a profound injustice, to keeping entire classes of people out of research studies because of their differences from an accepted “norm,” and then using the results of those studies to treat those same people despite their differences. Here’s hoping that this time the world’s largest funder of biomedical research does the right thing — for Americans of all ages and backgrounds, and for medical science.

Louise Aronson, M.D., is professor of geriatrics at the University of California, San Francisco; director of UCSF Medical Humanities; and author of “Elderhood: Redefining Aging, Transforming Medicine, Reimagining Life” (Bloomsbury, June, 2019).