This September’s issue of the journal Alzheimer’s and Dementia focuses on sex and gender differences, anchored by the review article "Understanding the impact of sex and gender in Alzheimer's disease: A call to action” (Nebel et al., 2018). The article, authored by an Eli Lilly-funded task force of the Society for Women’s Health Research (SWHR), a Washington, DC, advocacy and research organization, energetically lays out the case for the plausibility of sex and gender differences driving patterns of disease distribution in the case of Alzheimer’s disease (AD).
The task force’s accessible article provides a nice opportunity to examine a central conceptual debate in feminist science studies. In the article, the SWHR task force operationalizes ‘sex’ and ‘gender’ as research variables with sweeping explanatory potential in biomedicine. The manner in which they do so reveals a theoretical gulf between a widely practiced genre of bioscience knowledge production in women’s health research and critical feminist approaches to science, medicine, and the body.
This gulf concerns two interrelated conceptual matters: First, what constitutes a sex or gender difference? Second, what is sex, and what is gender? As we hope to show, the concerns we raise are not merely a matter of precision in terminology. They have implications for how research on sex, gender and Alzheimer’s is framed, and, consequently, for the kinds of interventions that researchers, funding organizations, and pharmaceutical companies will pursue for millions living with the condition.
What is a sex or gender difference?
The task force begins with a quandary. Their goal is to motivate the study of sex differences in Alzheimer’s, but while it’s true that AD is more prevalent in women – two-thirds of the people living with AD are women – this is largely, if not entirely, accounted for by the fact that women live longer than men. That is, the epidemiology does not clearly indicate a sex difference in the incidence of (or risk of contracting) Alzheimer’s. So, unlike classically sex-biased disorders such as lupus, AD epidemiology does not suggest a greater biological vulnerability in women compared to men.
Perhaps the brains of females and males decline in quite similar ways, and instead other variables – social-structural, for example – are more critical for driving interventions into dementia. But similarity is not of interest; the null hypothesis is not on the table. Undeterred by the apparent equality in incidence of the disease, Nebel et al. propose another route: sex differences in risk factors, symptoms, and response to treatments. For instance, depression and heart disease have different incidences and courses in men and women, and depression and heart disease are risk factors for AD: this reasoning drives the task force’s call for more research on sex differences and AD.
Certainty that differences between men and women must be a central axis of importance for AD research pervades the expository strategy of the piece. The reader learns, for instance, that in sleep “slow-wave activity is greater in women than men,” that “women exercise less than men,” and that there is a “lifelong female advantage in verbal memory” (pp. 1174-5). These differences are noted without saying how large the differences are, whether they are replicated across populations, and how much overlap between men and women and variation among women and among men is found.,
In their focus on sex differences, the question of biological mechanism moves to the background. For example, the authors note that microvascular disease is more common in women than men, and therefore suggest that researchers investigate whether microvascular disease contributes to the progression of AD in women. But both men and women can have microvascular disease. Does everyone with a history of this disease follow a similar course to AD, or is the association unique to women? Researchers will be more likely to miss these significant questions if they focus only on the presence of sex difference.
The authors place a large explanatory burden on sex differences without careful examination of effect sizes. A good example is the claim that sex differences in verbal recall might underlie the faster decline observed in women after diagnosis compared to men, an intriguing hypothesis that garnered health news headlines this summer. Researchers have suggested that a key test for early stages of dementia involves a skill (verbal recall) that women are better at than men. Perhaps, the hypothesis goes, this test is thus not as sensitive for women. As a result, women end up being diagnosed later, closer to the point of severe disability. The problem with this hypothesis is that everything we know about differences in verbal and related cognitive abilities between males and females suggest that these differences, when they are found, show very small effect sizes. Although this is well known in the literature, we only hear that there is a difference; no criticality or curiosity is raised about the ability of such small differences in a specific cognitive capacity to mask the advance of Alzheimer’s in women and thus explain their much more rapid decline after diagnosis.
So far, we’ve looked at the expository strategy of Nebel et al. as they build the case for greater attention to differences between men and women with Alzheimer’s. Now, we turn to how the article utilizes the terms “sex” and “gender.”
What is sex, and what is gender?
Much to their credit, the SWHR task force that authored this article explicitly considers both “sex”-related and “gender”-related factors in Alzheimer’s disease. The SWHR—architect of the controversial 2016 NIH policy requiring study of sex as a biological variable in pre-clinical research—has been criticized in the past for deemphasizing the role of gender variables in producing health inequities among women and men. The task force’s attention to potential gender-related factors represents a positive advance over an often-singular focus on sex-related biological variables in many recent reviews of this sort. But, how do the article’s authors understand “sex” and “gender”?
Nebel et al. begin by defining sex as “biological and physiological differences between men and women” and gender as “environmental, social, and cultural influences on the biological factors in women and men.” On this definition, sex is biological and gender is sociocultural. The problem is that the biological doesn’t fit neatly into “sex,” nor the sociocultural into “gender,” as the paper’s own examples soon reveal.
Under “examples of sex differences in risk factors for AD,” the paper lists cardiometabolic risk factors, depression, and sleep, whereas under “examples of gender and sociocultural risk factors,” we find education, exercise, marital status, and caregiving. But are these sex, or gender?
Depression is listed as a biological “sex difference,” but diagnosis and expression of depression in women as compared to men has well-documented cultural dimensions. Sleep quality – another proposed “sex difference” – is entangled with culturally- and socially-dependent factors like depression, stress, caregiving, economic security, and diet. On the “gender” side, education – a social and cultural influence – alters the biological structure of the brain, exercise impacts hormone levels, and marital status and caregiving affect stress profiles. The line between sex as biology and gender as culture turns out to be a very difficult one to draw.
Under the heading “sex-specific risk factors for women,” the authors include hormone therapy (HT) and gynecological surgeries such as oophorectomy. These interventions affect biology, but they are also human medical interventions. Access to HT and frequency of oophorectomy varies across time, space, and socioeconomic class, and may itself be inflected by norms and expectations surrounding aging and reproductive health. In fact, in two instances, the authors themselves present direct and compelling support for this very point, noting that in many studies those who use HT are “healthier and better educated than other women,” and that bilateral oophorectomies before menopause are more common among women exposed to "adverse childhood experiences or adult abuse.” HT and gynecological surgeries may be risk factors for AD, but if they are, their effects cannot be understood separately from socio-cultural factors.
We can only speculate about the presuppositions undergirding the authors’ intuitions about what counts as sex or gender. What seems to separate what is called “sex” from that dubbed “gender” is some implicit notion of what is “biological.” Perhaps they think that we can biologically measure sleep quality and we can quantify cardiovascular risk factors and disease progress using biometrics, but that education, exercise (understood here as frequency of and access to exercise), and caregiving are not quite as measurable using conventional bioscientific probes. Or, perhaps they see the “sex” factors as impacting the body “directly,” whereas the biological implications of gender-related variables are more “indirect”. In either case, these approaches sap the concept of “gender” of its critical power and indeed render the gender/sex distinction incoherent, as they would have us move putatively gender-related factors into the “sex” column as soon as we figure out how to directly measure or observe their biological consequences.
As we encounter these multiple interrelated risk factors posited as either “sex” or “gender”, the grounding premises of the article as a whole begin to fall apart. The authors present AD as a sexed biological syndrome with some interesting social and cultural variations or influences. Instead, the problematic of the interaction between categories of “sex” and “gender” in the article forces us to see Alzheimer’s disease as a diversely expressed condition underpinned by causes at once material and sociocultural, structural and interpersonal, which are accreted in the body and expressed as forms of progressive dementia, generally in old age. This way of thinking about the broader phenomenon of AD in relation to gender and sex is quite different from that which implicitly frames the article’s assumptions about high priority avenues of intervention. If we want to understand AD, and sex differences in AD, we need to go beyond an approach treating sex and gender as discrete, additive variables.
Theory matters: A call for curiosity
Among the heterogeneous world of women’s and gender diversity health advocacy, the SWHR is distinguished by its close ties to the pharmaceutical industry and the National Institutes of Health (NIH). The authors of "Understanding the impact of sex and gender in Alzheimer's disease: A call to action" disclose their individual and collective connections to the pharmaceutical industry, and explicitly connect their interest in sex differences to the NIH’s “precision medicine” initiative. Needless to say, pharmaceutical companies and the precision medicine initiative are not in the business of developing cures for gender inequalities in care work or the distribution of wealth. They use molecular biological approaches to mine individual or group-based differences to target therapies and interventions. This context is important because it helps explain elements of the article’s exposition: the focus on discrete, clinically-assessable variables that might be operationalized in the context of precision medical tests or therapies, and the lack of accountable engagement with a more diverse and critical community of women’s and gender health theory, scholarship, and advocacy, including feminist perspectives.
We, too, want to see individuals with AD receive better diagnosis, access earlier interventions, and live dignified lives. Gender and sex must be attended to in that endeavor. But the case for sex differences as a route to more “precision” approaches to AD, as presented by the SWHR task force, is, from the perspective of gender theory, oversimplified and overdrawn. We are not persuaded that comparing men and women, within a frame appealing to sex as biology and gender as culture, is a promising approach to understanding the causes, mechanisms, and phenomenology of Alzheimer’s disease.
The SWHR is ready to “call for action.” We instead call for epistemic humility. Without good theory, an appeal for attending to male-female differences makes sex salient without providing the interpretational resources to understand the meaning of these results. From our perspective, sex differences should invite curiosity—about theories, concepts, and categories, and the ways that they structure research and medical practice. Heeding this call, we suggest, will have implications both for our basic scientific knowledge of the complex condition we understand as “Alzheimer’s disease,” and for the lives and experiences of those affected by it.
How to cite this blog
Reiches, Meredith, Boulicault, Marion, Bruch, Joseph, Noll, Nicole, Shattuck-Heidorn, Heather, Weir, Brianna, and Richardson, Sarah. “Theory Matters: Sex, Gender, and Alzheimer’s Disease,” GenderSci Blog, October 10, 2018, https://projects.iq.harvard.edu/gendersci/blog/theory-matters-sex-gender-and-alzheimer’s-disease.
Statement about intellectual labor
The GenderSci Lab is committed to an equity- and justice-based approach to scholarly research, which includes the practice of citations and intellectual attribution. As such, we’ve followed the CLEAR Lab’s (an innovative feminist, anti-colonial marine laboratory) guidance in determining author order. Reiches drew up synthesizing notes from our lab discussion that were used, along with notes from others, to compile the initial draft. Richardson drafted the blog post and integrated lab member comments and edits during the revision process. All authors provided substantive contributions to the ideas expressed in this post and participated in the final revision and preparation of the blog post.
 Notably, this lack of criticality about claims of group differences in AD also appears when the article’s authors briefly touch on race, writing that “African-Americans are two to three times more likely to be diagnosed with AD or related dementias” (Nebel et al., 2018, pp. 1179). However, the citation that they provide does not support this claim. The cited study regards the genetic make-up of residents from a small area in northern Manhattan; it does not find that AD is more prevalent among African Americans, but is rather an analysis of whether APOE carries a genetic risk for AD.
 As the authors list potentially AD-relevant sex differences, the major confounder of age is inconsistently noted, leaving it unclear whether the seeming sex difference is driven by sex or by other factors. In the one instance in which age is explicitly acknowledged, the case of different risk profiles in men and women with a particular variant of the APOE genotype, the authors concede that the difference may be entirely explained by age.