Back to Articles

Improving Social Determinants of Health 

September 23, 2019 | Comments

For many of us, a successful medical intervention means we don’t feel sick any longer and whatever illness we are being treated for is either gone or under control. To get to this point, we generally see a doctor or other healthcare provider, give a history of our illness, undergo a physical examination and various tests, and then receive treatment. The treatment could be, among many possibilities, taking a medication, undergoing a surgical procedure, or seeing a psychotherapist.

This sequence of events may lead us to believe the whether we get better or not largely depends on receiving the correct diagnosis and treatment for an illness. Most of us also understand that an important component to whether we get sick in the first place and how we respond to treatment depends on our intrinsic genetic make-up: some people are more prone to particular diseases because of the genes we inherit from our parents.

         It comes as a surprise, then, that physiology and genetics account for only about 20 to 30 percent of the reason people develop health conditions and their chances of recovery. The major factors influencing these things are called the social determinants of health (SDOH). Studies show that as much as 80 percent of the chance an individual will get sick and will get better has to do with their[1] income, housing situation, education level, nutrition, social support, and access to healthcare. Poor, hungry, lonely, and homeless people are the most likely among us to become ill and the least likely to get better, even if they receive the most up-to-date care.

         In fact, the number of deaths in the U.S. attributable to social factors is “comparable” to the number attributable to pathophysiological and behavioral factors according to a 2011 study published in the American Journal of Public Health. According to that study “Approximately 245 000 deaths in the United States in 2000 were attributable to low education, 176 000 to racial segregation, 162 000 to low social support, 133 000 to individual-level poverty, 119 000 to income inequality, and 39 000 to area-level

Among the many things that influence health, social determinants have a greater effect than genetic endowment or pathophysiology (image: Shutterstock)

         A striking example of this phenomenon comes from a recent NIH-funded study that showed that living in racially segregated neighborhoods is associated with high blood pressure and that moving away from segregated areas is associated with blood pressure decreases and reductions in heart attacks and strokes

Even when receiving apparently good–and equal–care, low-income people have worse outcomes than higher-income people, as the NIH-Funded ALLHAT study has shown. We often, correctly, think that reducing salt intake, losing weight, exercising more, and taking antihypertensive medications are ways to treat hypertension, but in this case moving to a different neighborhood also seems to be an effective intervention. This phedmnomenon has sometimes been called the “zip code effect” on health.

Do SDOH Interventions Work?

         While the finding of this huge influence of SDOH is now uncontroversial, there is widespread disagreement about whether interventions to improve the factors that make up SDOH work to improve health and whether they have unintended adverse consequences. Do improving access to healthcare by expanding the number of people eligible for Medicaid, improving nutrition by providing food stamps, and boosting income for those living a poverty levels produce measurable health benefits or do they merely make people dependent on welfare? Moreover, even if SDOH interventions work to improve individual and public health, can we afford them?

         There is, of course, an overriding value-based aspect to questions about providing welfare that cannot be addressed with studies and data. Society has a moral obligation to help people who cannot work and are consequently poor to obtain the basic necessities of life: food, shelter, and healthcare. Almost no one would dispute providing public assistance to a severely disabled person who is physically unable to work and has no source of income. Beyond such an obvious case, however, there is disagreement about how to define who should be eligible for welfare and what mechanisms of providing assistance work best.

         A common refrain is that welfare makes people “lazy.” There is a general consensus that the best outcome for someone who receives public assistance is to get to the point that they no longer need it. We also generally agree that very sick, starving people are unlikely to be able to find or sustain income-producing work. But some argue that various welfare programs actually impede the transition to financial independence by removing the incentive to find work and that they therefore waste public funds.

         While we cannot address the value-based aspects behind welfare using scientific evidence, we at Critica did wonder whether there are data addressing the question about whether welfare has any positive effects on health and health cost outcomes. In the course of looking into this issue, we also learned that there are data addressing the broader question of whether welfare inhibits people from finding independent work. This commentary is the first in a series that will address what we have learned exploring studies that ask the basic question, “does welfare help people?”

         On the federal level, there are six welfare programs: Medicaid, Temporary Assistance for Needy Families (TANF), Supplemental Nutrition Assistance Program (SNAP, commonly knowns as food stamps), the Earned Income Tax Credit, Housing Assistance, and Supplemental Security Income.

         The amount and quality of research involving each of these six programs varies considerably. It is also difficult to find quality research involving state and local welfare programs. How much help a person can receive depends to a large extent on which jurisdiction they live, so studies of federal programs must account for this local variability. Prospective studies involving random assignment and adequate control conditions are rare in this field; most studies are observational and therefore can only establish associations rather than cause and effect relationships.

         Nevertheless, clever research designs and analytic methods have been employed that do allow some broad if somewhat tenuous observations. We will review some of them here, with others to follow in future commentaries.

Attending to SDOH Improves Health and Reduces Healthcare Costs

         Although there is no question that providing food, income, and housing to people costs money, there is mounting evidence that addressing SDOH not only improve health outcomes, it can reduce some healthcare costs. For example, one healthcare system recently initiated a program of screening its Medicaid patients for SDOH and then having care management teams help them with issues like food insecurity, transportation, and health literacy. The effort resulted in a 26.3 percent drop in inpatient admissions and 9.7 percent drop in emergency department (ED) visits . Numerous cities like Houston have also found that similar programs that help link patients to social services actually decrease healthcare utilization and improve outcomes like overall patient functioning.

         A program in Massachusetts delivers food to people eligible for both Medicare and Medicaid, often categorized in the insurance industry as “dual eligibles.” Compared to a matched group of people who were not given food, the group that received meals had fewer ED visits, decreased ambulance usage, and lower medical spending. Studies show that Medicaid eligibility during childhood reduces adult ED use and hospitalizations.

         Similar findings have been reported for SNAP, better known as the food stamp program. Although it is sometimes alleged that people who receive food stamps are really able-bodied adults who are just gaming the system, the fact is that more than 80% of all households that receive food stamps include children, elderly people, or people with disabilities. Furthermore, studies show that while food insecurity is associated with higher health care costs, SNAP is associated with both improved current and long-term health and with reduced healthcare costs.

         In 2009 revisions were made to SNAP that increased access to whole grains, fruits, vegetables, and low-fat milk for women, infants and children. An evaluation of these changes showed that they improved pregnancy and infant health outcomes.

         The city of Los Angeles has a new “Housing for Health” initiative in which it uses healthcare dollars to pay for housing for homeless people with complex medical and behavioral health problems. A RAND study found that the program has decreased ED visits and hospital admissions, improved mental health functioning, decreased the number of individuals arrested and jailed, and decreased public services’ costs by 60% in the year after receiving housing. Some healthcare systems have even made the leap to providing housing for their members. CVS has announced plans to invest $50 million in affordable housing and Kaiser Permanente announced last year it would invest $200 million for programs to remedy housing insecurity and homelessness.

         There are now many such studies from organizations including the Robert Wood Johnson Foundation, Moody’s Analytics, and the Geisinger Health System that show a clear association between improving the various elements that make up SDOH and both improved health outcomes and reduced healthcare costs. Large health insurers have apparently noticed: Large health insurers have apparently noticed: Aetna and CVS recently launched a joint program that uses a new tool to track the impact of SDOH on healthcare costs.

Medicaid Expansion Improves Health

         A provision of the Affordable Health Care Act (ACA) allows states to expand Medicaid coverage to adults with household incomes at or below 138 percent of the federal poverty level. Because some states have chosen to expand Medicaid under the ACA, while others have not, researchers are presented with a unique opportunity to evaluate the program’s outcomes. In some instances, for example, a state decides to expand its Medicaid program, while another state with very similar demographics does not. Comparing outcomes in the two states is, of course, not a prospectively randomized approach and there are always important differences between even very similarly appearing states that can only be controlled for statistically. Nevertheless, such studies give us an opportunity to address questions like whether getting Medicaid stifles employment and if receiving Medicaid leads to improved health outcomes and/or reduced spending in other areas.

         A study that compared two states that did expand their Medicaid programs, Kentucky and Arkansas, to a state that did not, Texas, concluded that “there were some major improvements in access to primary care and medications, affordability of care, utilization of preventive services, care for chronic conditions, and self-reported quality of care and health”. Another study of Medicaid expansion came to similar conclusions.

         A study published in the Journal of the American Medical Association assessed whether Medicaid expansion resulted in any improvement in rates of low birth weight or preterm births. Although there was no association between overall birth outcomes and Medicaid expansion, the study did find that Medicaid expansion is associated with significant improvements in disparities between black and white infants. Because black infants die of complications from prematurity and low birthweight at almost four times the rate as white infants, the study suggests Medicaid expansion is one tool that may be important in closing well-recognized health disparities in the U.S

 Welfare Does Not Make People “Lazy”

         The data reviewed so far seem to indicate mostly positive effects of SDOH interventions on health and healthcare spending. A concern of many, however, is that federal public assistance programs lock people into permanent lives of unemployment and lassitude. If you can get money, food, housing, and healthcare without working, the trope goes, then why look for a job?

There is actually very little evidence that on a population basis, welfare discourages people from working. For example, a recent study of the ACA’s Medicaid expansion option showed that it “did not result in significant changes in employment, job switching, or full-versus part-time status”. In other words, expanding access to a public assistance program—in this case Medicaid—did not motivate people to quit their jobs or refuse to look for work. At least two other studies confirm the finding that Medicaid expansion does not impact the labor market or employment. An analysis of seven randomized studies evaluating cash transfer programs in poor countries found no evidence that they discourage work.

In fact, Derek Thompson, writing in The Atlantic, observed that there are now “a fleet of studies…converging on the consensus that even radical welfare programs—including basic-income programs and what are called conditional cash transfers-don’t make people any less productive”.

Of particular note are studies Thompson identifies that show that Medicaid and other welfare programs are particularly beneficial for children. When children are covered under these programs, these studies show that they have higher rates of high-school graduation, higher incomes as adults, and less use of welfare assistance as adults. Thompson calls poverty a “slow-motion trauma” and concludes from his review of studies on welfare outcomes that “It’s axiomatic that relieving children of an ambient trauma improves their lives and, indeed, relieved of these burdens, children from poorer households are more likely to follow the path from high-school graduation to college and then full-time employment.”

One proposed way to encourage welfare recipients to transition to income-producing jobs is to add work requirements to eligibility rules. The argument here is that requiring people to do some work in order to quality for receiving Medicaid or food stamps will encourage poor people to get jobs and help them emerge from poverty rather than become “lazy” welfare recipients forever.

         In the early years of the TANF program, proponents of work requirements could cite some rigorously conducted, controlled trials that showed a small increase in work among people subjected to work requirements. However, longer term studies fail to find such an effect: five years after exiting the TANF program people who were subjected to work requirements were as like to have jobs as those who were not. What does seem to work to help people who are receiving public assistance get jobs are education and job training programs.

The data we have reviewed here suggest four conclusions:

  1.     Improving the elements that make up SDOH improves health outcomes
  2.     Improving SDOH can lower healthcare costs
  3.     Receiving public assistance does not discourage employment
  4.     Education and job training rather than work requirements help welfare recipients get jobs

Perhaps the idea that improving the social determinants of health yields benefits to individuals and to society is catching on. In July, a bipartisan coalition introduced a bill into the U.S. House of Representatives that would provide resources to states to address SDOH issues. Seeing Congress act on the evidence in this case is certainly heartening.


[1] Yes, we know that “an individual” is singular and “their” is plural and therefore don’t agree. But we are trying to adopt gender neutral pronouns. So please bear with us. It may sound “off” to the ear and we will undoubtedly forget sometimes and revert to he or she. But we are trying. 


More Like This