In the field of medicine, patients are treated on the basis of scientific knowledge as much as possible. However, many of the scientific proofs are based on studies carried out on a select group comprising mainly white male subjects. There is potential for this approach to have dangerous consequences for other patients, such as a heart attack going unnoticed, medication not working or side effects. Cardiologist Harriette Verwey, vascular medicine internist Jeanine Roeters van Lennep and neuroepidemiologist Arfan Ikram believe that diversity deserves to be granted a more significant role in medical science. But exactly what that role ought to be is still up for debate. Should study populations reflect the diversity found in society or would that muddy the waters in scientific research?
What applied to men ought to apply to women
Diversity plays a significant role in medicine, even if that was not always the prevailing view. “It used to be assumed, simplistically, that what applied to men ought to apply to women too”, explains Harriette Verwey. The cardiologist (now retired) is one of the first doctors in the Netherlands to have raised awareness for the differences between men and women in medical science.
When a major study of cardiovascular diseases was carried out in women at the end of the 1990s, a whole world opened up for Verwey. “Women with symptoms like chest pressure and pain between the shoulder blades were studied, leading us to understand that cardiovascular diseases present differently in women than they do in men, often with less clear-cut symptoms.”
Until recently, there was an unwritten rule that research into the effects of medicines or the clinical presentation of diseases was always carried out on the same demographic group: white males. The upshot of this is that very little is known about pretty much all the other types of patient, which is evident in clinical practice. For example, it is commonly the case that cardiovascular diseases in women do not follow the ‘classical’ presentation that doctors are familiar with in men, resulting in these diseases being underdiagnosed in women. Thus putting women at greater risk of poor outcomes including an untimely death.
“Women also prove to have more specific side effects on multiple medications, despite these medications having been approved for use”, explains Jeanine Roeters van Lennep, vascular internist at Erasmus MC. She says that it is still all too common for major studies to use a ‘standard study population’ of white male subjects. Women and non-white patients are largely regarded as exceptions and are studied in ‘subanalyses’. “It isn’t representative of the wider population if a study encompasses 200 men and a subanalysis is carried out on twenty women”, says Roeters van Lennep.
Verwey and Roeters van Lennep would like studies to henceforth be carried out with study populations that are representative of society. For example, in the Randstad region of the Netherlands this could mean half of the subjects in a study being women, but ethnic diversity being factored in too. “52 percent of people in society have a migrant background. If we’re to be able to develop better treatments for all our patients, this needs to be reflected in our study populations. Quite simply, the trials needs to be randomised”, says Verwey.
Heterogeneity muddies the waters
Arfan Ikram, epidemiologist and chair of research funding organisation ZonMW, takes a slightly different stance. He shares the view that there is a lack of diversity in medical science, but does not regard ensuring that studies represent a precise reflection of society as an ideal solution, not least because the study populations are often limited in size.
“No doubt the hope is that it would put us in a better position to estimate whether a drug would definitely work for everyone”, Ikram explains. “But there’s a fallacy in operation there. Heterogeneity actually muddies the waters. It would result in a decrease in statistical strength, which will detract from the reliability.” Moreover, there would be too many factors (so-called ‘confounders’) to take into consideration if scientists were to put together genuinely representative research populations. “Age, sex, ethnicity, lifestyle and place of residence are all important in medicine.”
Draw the line pragmatically
Ikram uses an example to explain the solution he envisages: “Imagine two vaccines need to be trialled and the funding is only enough to include a limited number of participants. You might choose to test vaccine A on men and vaccine B on women. That would give you enough statistical strength for that study to be able to reach a clear conclusion.” It will not then be the case from the outset that you will need to repeat both studies if you want to know whether vaccine A will also work in women and vaccine B in men. “You see, as a scientist you have infer whether you know enough from previous, similar studies on the effects of vaccine A in women’s bodies and vaccine B in men’s bodies to conclude that you can expect to see the same effect. If you don’t have enough information on this, then you will indeed have to repeat both studies, but that then raises the matter of funding for two additional studies.”
If researchers want to study all medicines or risk factors in all groups from the outset, then it becomes never-ending, thinks Ikram. “We need to draw the line pragmatically.” He believes the solution lies in making strategic decisions on which homogeneous study populations are to be used in which research. He calls that ‘diversity between studies’. You conduct one study on population A, the next on population B and then in population C. At some point it is population A’s turn again. Effectively that could mean studying the relationship between smoking and lung cancer in Europeans, cardiovascular diseases in Asians or salt intake and high blood pressure only in Africans. Thus building an increasing body of knowledge in which all populations are represented to a sufficient, similar degree whilst preserving the methodological strength of an individual study.
Medications without any effect
Verwey does not regard this as a logical solution. “No single country has a uniform or homogeneous population. Why would you only include black or white people in studies if the diseases aren’t limited to those groups? This needs to be taken seriously. It needs to reflect society.” Verwey offers the example of studies in the USA which show that a lot of high blood pressure medications originally trialled in white patients and subsequently approved by the FDA are not having any effect in black patients. Consequently, one sizeable group of patients was being prescribed medications that had no effect whatsoever for a protracted period, resulting in ineffective treatment and putting them at considerable risk.
Mixed or heterogeneous study populations are a priority for Roeters van Lennep too. Which is why she is intending to investigate the hurdles that are ensuring that study populations are still commonly homogeneous. “Women, ethnic minorities and patients from lower socioeconomic classes are not being included enough. We want to analyse how that could be different, to enable future research to always be as inclusive as possible.”