Hello all –
Sorry for my delayed response to the homework assignments. I appreciated all your forum submissions for Week 4 on biological mechanisms. I think for most of us, this content is the least familiar of all the topics in this course, so I appreciate you diving in to the readings and the assignment.
A lot of you had interesting comments, which I will weave in to my response below, rather than calling out 1 or 2 people’s responses overall.
For the first question about gene by environment interactions, I think it is important to separate out the broad point that the Hertzmann paper was making – about the fact that gene by environment interactions exist and contribute to differences in health and developmental trajectories between individuals – from the question of whether these interactions have an impact on health disparities on a population level. Thinking about this more specific question about health disparities, one way to answer this is to fall back on assumptions about differential distributions of genes by social categories such as race/ethnicity, which then can be unlocked/exacerbated by differential social environments, leading to disparities. While this is slightly more nuanced than simplistic beliefs about genetic differences alone leading to health disparities, it is still dependent on the same problematic assumptions about genetic definitions of race. However, we can still consider gene/environment interactions to be relevant even if we do not make claims about genetic differences by social categories – e.g. even if groups have exactly the same distribution of genetic predispositions, those exposed to environmental stimuli will be at greater risk. Of course, not all health disparities will be produced by gene/environment interactions, as some environmental risks will not be dependent on genetic predispositions (e.g. some environmental exposures are equally toxic to all people).
One interesting point made by Rebecca Kim was that even in the rare cases in which there are differential distributions of genetic risk factors by social categories, a focus on these genetic differences can contribute to a lack of attention to similarly patterned, and more influential, social risks:
“In my specific area of interest, differences in outcomes of chronic liver disease for patients with different social determinants of health, there has been a gene identified with studies showing an association with worse outcomes for patients with non alcoholic fatty liver disease if the gene is present. This gene also happens to be more prevalent in Latino populations according to the literature. I worry that this gene is now being use to explain why patients who identify as Latino are having poorer outcomes, when in reality, it may be more often due to social determinants of health. As expected, not all liver patients undergo genetic testing, so many assumptions have to be made based on these findings in limited studies. Additionally, I can imagine that if genetic testing does become more common in hepatology, a Latino patient that tests negative for this gene may then be thought of as lower risk for poorer outcomes - the test would provide inappropriate reassurance. When in reality, despite not have the genetic risk factor, the patient may be at risk for poorer outcomes due to social determinants of health, which then may be ignored. “
Chris Albach made a related point about the risk of focusing on mechanisms (below). I think this an excellent point, but also appreciate Dr. LeWinn’s points about the value of understanding biological mechanisms I think both of these perspectives have value, and it is important to carefully consider how to have research on mechanisms contribute positively to the goal of improving health equity, as opposed to contributing to problematic narratives.
Chris’s answer to Question 1: “Although their specific examples are compelling, I am personally always wary of incorporating genetic explanations for health disparities, even if they are used in context of an interaction with the environment, due to the risk of perpetuating dangerous and racist stereotypes based on genes. Particularly with behavior, which they use in their examples, there are possibly an infinite number of factors that affect one’s behavior, many of which are nearly impossible to quantify or measure. It was previously presented in this class that the majority of illness stems from social conditions and that genes, in fact determine a relatively small proportion of disease and disease course. I worry that the focus on the mechanism of how SDH lead to poor health outcomes centers the intervention on the outcome rather than the exposure (SDH), which are in large part due to this country’s violent white supremacist history and current practices.”
There are similar issues with Question 2 about epigenetics. Understanding how epigenetics can produce intergenerational effects has implications for interventions to address health disparities, and many of you commented on the importance of thinking across generations and across the life course as a result of this epigenetic mechanism of biological embedding. For example, Michelle Lee stated:
“Biological consequences of environmental stressors can be passed on to the next generation, who then bear the consequences of these embeddings and are subsequently predisposed to more prejudices, which then cause even further biological embeddings. Thus, a vicious cycle begins where biological consequences and societal prejudices can compound with every generation... For instance, in the hypothesized transgenerational flow of chronic disease, the environment in which a grandmother was raised can have biological implications on her grandchild’s oocytes. If the grandmother suffered from poor nutrition due to systemic racism, her subsequent offspring could be at higher risk for chronic diseases/cancers and also be subjected to systemic racism. Now the grandchild bears “double penalty” – poor health secondary to her grandmother’s environment and current environmental stressors. However, identifying these epigenetic mechanisms can allow for interventions specifically designed to stop this domino effect”
However, a few of you also expressed concern about a focus on epigenetic changes. Two examples are below.
Stephanie Frazin: “Focusing on epigenetic sources of disparities creates channels for medical communities to place blame on patients for their outcomes, while the true source may be in disparate care.”
Chris Albach: “Similar to my response above, epigenetic explanations for how traumatic experiences or SDH can become encoded in the genome and cause intergenerational disease is compelling and exciting for the broader medical community. However, the push to develop scientific explanations for how structural violence leads to poor health outcomes through methods only actionable by the most educated people on the planet further concentrates knowledge and power to an already privileged few. I think a reasonable argument can be made that pandering to a need for a biological proof of the relationship between racism and health encourages the silencing of the voices of many, many people of color who have been exclaiming this for centuries. We know with certainty, from the work of mostly non-scientists, that white supremacy, discrimination, and other political, social, and historical forces of domination dictate nearly every aspect of our lives, including health. The solution then, is to dismantle these systems of domination, which has also been known and worked on by non-scientists for centuries.”
Finally, for question 3, I want to highlight Jonathan Amatruda’s and Griffin Collins’ responses – Jonathan’s as a good overall summary and Griffin’s as pointing out issues specific to measurement and analysis of SES:
Jonathan’s: “Robinette et al. (2016) laid out the connections between environmental stresses and physiology, using allostatic load as a measure of various health outcomes and neighborhood income. The authors use “low SES neighborhood” as a proxy for accumulated psychological, social, and physical stressors that put inhabitants at risk for poor health outcomes. This approach mirrors the socioecological model in that it weaves through the layers that comprise the lived experience and shape health. Low-resource neighborhoods are disproportionately affected by crime and violence. Fears over personal safety lead to stress and may affect neighborhood social cohesion, sowing the seeds of isolation. Neighborhood crime deters outdoor physical activity, depriving residents of options for exercise and failing to support healthy behaviors. These neighborhoods are also farther from economic opportunity, cementing income disparities. Sources of affordable, healthy food are rare, raising risks of obesity and dyslipidemia. Poor living conditions in low SES neighborhoods may contribute to stress and poor sleep. Each of these “hits” can affect health, though in many different ways. Robinette and co-authors manage this by using allostatic load as a composite measure of the ways that stresses accumulating at each level of the socioecological model can shape health”
Griffin’s: “It is clear from their study that simply evaluating one measure of socioeconomic status as a surrogate for socioeconomic status as a whole is inadequate and would miss significant factors that mediate relationships between SES and outcome. Further, it demonstrates that broadening the measures included in SES assessment would provide information about specific factors contributing to a particular outcome that could be informative when identifying possible interventions to improve health outcomes.”
Thanks all! Looking forward to this week, when we will talk more about measurement.
Christine