The Flawed Logic Behind the ‘Social Determinants of Health’ Theory of Medicine

Over the past few years, a movement has emerged to expand the scope of healthcare research and practice to “social determinants of health” (SDOH) – social, economic, and environmental conditions that can be construed as influencing individuals’ health.

Some commonly-cited examples of SDOH include income, employment, educational quality, housing quality, transportation, and so on.

Now, the authors of a commentary published in JAMA Surgery, titled “Quantifying Social Vulnerability and Its Impact on Health Care Delivery, Payment, and Performance,” argue that SDOH data should be integrated into patients’ electronic health records.

This means that SDOH data will likely inform individual care decisions by healthcare professionals.

The authors justify this position by arguing that SDOH “should dictate how we deliver the most appropriate care to each patient.”

They then go a step further, arguing that “[s]urgical research directed at health care delivery should consider these data in a similar way to the contributions of medical comorbidities.”

Yet the fundamental assumption behind the authors’ argument – that SDOH are the cause of disparities in health outcomes between populations, with the authors arguing that SDOH are “drivers of inequities” – is seriously flawed.

Although SDOH may be correlated with disparities in health outcomes, the evidence that SDOH cause poor health outcomes is shoddy and weak, at best.

Much of the scholarship on the topic confuses social and economic conditions that correlate with poor health outcomes with the actual causes of those outcomes, ignoring other factors such as individual agency and health decisions that contribute to health outcomes.

As Manhattan Institute Senior Fellow Chris Pope lays out, “SDOH literature is dominated by sloppily designed observational studies that do not seriously attempt to disentangle causation from correlation or to control for obvious sources of bias.”

Citing Yale University economist Jonathan S. Feinstein, Pope notes that smoking, for example, is more common among poorer people and explains “much of the disparity in health outcomes.”

In other words, although income levels may be correlated with worse health outcomes, the cause of that disparity (in this example) is individual choice. Whereas the SDOH theory may attribute that disparity to poverty, the real culprit is not so easily pinned down.

As another example, a study published in The Quarterly Journal of Economics, examined the question of nutritional inequality: why do higher-income groups of people eat healthier than lower-income groups of people?

The study found that 90% of this “nutritional inequality” is driven by differences in demand; in other words, it was the choices of the lower-income people, not the environmental factors, that caused this disparity.

The “food desert” theory of nutritional inequality, which attributes the fact that lower-income people eat unhealthy foods to lack of supply of healthy food, had the causality backward.

To further illustrate the point that adjusting SDOH does not necessarily produce positive health outcomes, we can look at the results of an experiment with Universal Basic Income (UBI).

Under the SDOH theory, in which income levels determine people’s health outcomes, increasing income should improve health outcomes.

But when nineteen counties in Texas and Illinois tested out the UBI program, recipients “reported no increase in access to or utilization of health care,” UBI did not lead to lasting “physical or mental health improvements,” and “recipients were four percentage points more likely to report a disability or health problem that limits the work they can do.”

Advocates for the SDOH theory of health disparities are thus making the elementary mistake of confusing correlation with causation, or more pointedly, ignoring other causal and confounding factors that undermine the SDOH theory’s explanatory value.

This is not to say that SDOH cannot ever contribute to health inequities, but rather that evidence for SDOH as “drivers of inequities” is wanting, to say the least.

In a nutshell, “social determinants of health” don’t actually “determine” health.

In addition, many studies purporting to show the effects of SDOH on health outcomes fail to consider the effects of personal agency, such as patient’s health choices or adherence to health programs, when identifying health disparities that they attribute to SDOH.

For instance, one of the studies cited by the JAMA Surgery article identifies disparities in emergency gall bladder removals for individuals with higher social vulnerability scores compared to lower scores.

However, the study fails to make any mention of personal choices that may have impacted the need for the emergency procedure, such as decisions to seek care at earlier junctures. The social vulnerability index used in the study does include factors that could potentially impact these decisions, such as distance to the hospital, but does not disaggregate the actual healthcare choices made by the patients from its analysis.

This omission implicitly assumes the causal effect of SDOH without considering the confounding variables that may be correlated with SDOH.

And finally, there is the authors’ position that “[s]urgical research directed at health care delivery should consider these data in a similar way to the contributions of medical comorbidities.”

One of the problems with this position, beyond the causal issues identified earlier, is the lack of a limiting principle; should surgical research consider everything?

SDOH, by their very nature, describe just about every conceivable factor in an individual’s environment that could be construed as affecting an individual’s health. The universe of possibilities is endless.

The resulting policy implications are likewise too broad and removed from the focus of healthcare. Put simply, it is not the role of medical professionals to become transit activists or education reformers.

Such activities are beyond the purview of medicine, and this mission creep comes at the detriment of medicine’s core focus.