Older Women Often Excluded From Clinical Research: Age Bias or Gender Bias?

Jocalyn P. Clark, MSc

A recent article published in a special issue of the Canadian Medical Association Journal on Diversity and Women's Health described poor inclusion and representation of women in clinical drug trials for treatment of myocardial infarction (MI). Despite heart disease being a leading cause of disability and death among North American women, especially older women, less than one-quarter of the patients included in the studies were women and the average age of participants was only 62 years. The work of Rochon and colleagues at the University of Toronto extends earlier findings of Gurwitz et al. at the University of Massachusetts who reviewed the literature for a 30 year period up to 1991 and found that women represented only 20% of MI drug trial participants. Most of these trials excluded patients over the age of 75 years. Traditionally, older people have been poorly represented in clinical trials because they are more difficult to study: they tend to have coexisting illnesses, they use other medications that may interact with study drugs, and the elderly are more vulnerable to adverse drug effects. Additional reasons for explaining women's exclusion include fear of harming a fetus, hormonal fluctuations that may complicate responses to medication, and the use of estrogens which may be protective for some diseases. In spite of these difficulties, many research funding and regulatory policies now mandate the inclusion of women and elderly people in clinical trials evaluating therapies for which these patients are targeted.

But does the exclusion of older women stem from age bias or gender bias? When asked this question several experts in Geriatrics and Women's Health offered some comments in support of increasing older women's inclusion in clinical research. Dr. Barry Goldlist, Director of the Division of Geriatric Medicine at the University of Toronto and Head of the combined Geriatric Service at Toronto Hospital and Toronto Rehabilitation Institute, suggests that many of the underlying "ageist" attitudes also reflect "sexist" attitudes. Dr. Goldlist states that "the cardiologist's description of a patient who is 'not a CCU candidate' is most frequently an older woman. Whatever the underlying prejudice, it is women who get the short end." Dr. Donna Stewart, a psychiatrist at Toronto Hospital and a leader in women's health, agrees that age and gender combine to create a "double whammy" that has implications for both clinical research and practice. In the case of heart disease, where women's cardiovascular risk is so strongly age-dependent, patient exclusions on the basis of age effectively dismiss the female patients who are the most important targets for the therapies. The physiological differences, hormonal changes, and increased comorbidity that may result in women responding differently to drugs than men are the very factors that require us to study them. According to a recent review by Morgan and colleagues, women's smaller body mass, higher percentage of body fat, hormonal fluctuations, and the presence of polypharmacy contribute to gender differences in drug responses and higher rates of adverse drug reactions among women. Due to these gender-related differences, Dr. Paula Rochon of the Baycrest Centre for Geriatric Care suggests that extrapolating research evidence from predominantly male, young, and fit individuals is simply not good enough to meet the goals of providing quality care to older women.

Dr. Michael Gordon, Head of Medicine at the Baycrest Centre for Geriatric Care and Professor of Medicine at the University of Toronto, takes a more pragmatic approach. He acknowledges that studying older people in clinical trials is extremely difficult and that this challenge predominantly affects women because they represent a larger number of the elderly. It is difficult to control for all the confounding variables when older people tend to have other medical conditions, especially those that affect cognition. These factors may also influence their compliance with trial protocols. A strategy that Dr. Gordon suggests would be to test new drugs for efficacy and tolerability in traditionally sampled (exclusive) research populations, then once effectiveness and safety are generally established, conduct suitable trials including the patient groups for which the drug is likely to be targeted in practice.

It is reasonable to expect that researchers should carefully weigh the risks and benefits of including older women in clinical trials. But applying age exclusions a priori in clinical studies disproportionately excludes women and amounts to both an age and gender biased way of conducting research. This practice prevents the very data collection researchers and clinicians need to make informed decisions when treating their older women patients. The same principles of a women's health perspective that promote inclusive and quality health care for women underlie the call for increased inclusion of older people in research studies. Older women's full participation in clinical research will greatly benefit the health and medical care of this very important patient population.

Suggested reading

1. Gurwitz JH, Col NF, and Avorn J. The exclusion of the elderly and women from clinical trials in acute myocardial infarction. Journal of the American Medical Association, 1992;268(11):1417-1422.
2. Morgan NA, Colling CL, and Fye CL. Cardiovascular diseases in women: An equal opportunity killer. Journal of the American Pharmaceutical Association, 1996;NS36(6):360-9.
3. Rochon PA, Clark JP, Binns MA, Patal V, and Gurwitz JH. Reporting of gender-related information in clinical trials of drug therapy for myocardial infarction. Canadian Medical Association Journal, 1998;159(4):321-327.
4. Wenger NK. Coronary heart disease: An older women's major health risk. British Medical Journal, 1997;315:1085-1090.

Jocalyn Clark is a PhD candidate in the Department of Public Health Sciences and The Centre for Research in Women's Health at the University of Toronto.