Obesity and diet-related diseases are among the most pressing public health challenges of our time. In the last 30 years, the prevalence of obesity has doubled among American adults and tripled among American children. Currently, obesity affects 4 of every 10 adults and 1 of every 5 children in the United States. Although obesity rates appear to be leveling off in the general population and even declining among very young children (2 to 5 years of age), they continue to increase in low-income and minority populations, which are disproportionally impacted. According to the most recent estimates, the prevalence of obesity is 47% among Black and Latino adults, compared with 38% among White adults. Similarly, the prevalence of obesity is 22% among Black children and 26% among Latino children, compared with 14% among White children.
Obesity is associated with a host of diet-related diseases such as diabetes (which affects 10% of adults) and hypertension (which affects 29% of adults). As is the case with obesity, racial and ethnic minorities are disproportionately impacted by diet-related diseases.
Excess body weight and its associated conditions are expensive. The direct costs for obesity are estimated to account for 54% to 59% of total medical costs, and the medical costs for individuals affected by obesity are approximately 30% greater than those for their normal-weight counterparts. The amount of spending on severe obesity, which is defined as a body mass index (BMI) of ≥40 and affects 1 of every 13 adults, is estimated to be $69 billion per year, $8 billion of which is paid for by state Medicaid programs.
The complications of obesity and diet-related diseases challenge the traditional model of health care delivery, in which care primarily rests with physicians, nurses, and other health professionals. These conditions instead require health systems to restructure care delivery to consider underlying drivers, which have traditionally fallen to professionals outside of the health care delivery system. Key among these underlying drivers are the social determinants of health.
What Are the Social Determinants of Health?
The social determinants of health refer to the circumstances in which people are born, grow up, live, work, and age that affect a wide range of health, functioning, and quality-of-life outcomes and risks. More concretely, these determinants include the neighborhood where a patient lives, income, education level, job status, housing situation, and source of transportation, among a host of other factors. Together, the social and economic conditions of people’s lives account for 50% of modifiable contributors to health outcomes, whereas medical care accounts for approximately 20% and behaviors account for the remaining 30%.
A number of health-related social needs are associated with the risk of excess body weight, including poverty, lack of education, food insecurity, and unstable housing. Longitudinal data indicate that societal changes have a strong effect on obesity inequalities according to income. Empirical evidence also suggests that exposure to negative social and environmental influences may exacerbate inequalities in BMI trajectories throughout the life course.
Role of Health Systems in Addressing the Social Determinants of Health
In the U.S., more than 95% of the trillion dollars spent annually on health care is for direct medical services, despite the fact that medical care prevents just 10% of avoidable deaths and more than half of preventable deaths are due to health-related social needs.
The strong influence of nonmedical factors on obesity and diet-related health outcomes mutes the ability of health systems to maximize the impact on patients affected by these conditions, even if quality and access to care are optimal. Therefore, the assumption underlying the present article is that nonmedical factors are relevant and important issues for health systems to address.
A number of empirical examples illustrate the impact of nonmedical factors on health outcomes. For example, children born into lower-income families are more likely to be born small and then to experience quick catch-up growth that leads to overweight and obesity. These same children also have higher rates of diet-related conditions such as cardiovascular disease. In addition, children who are born to parents who have not completed high school are more likely to live in a neighborhood with less access to sidewalks, parks or playgrounds, and recreation centers. Recent evidence suggests that exposure to violence during adolescence is associated with the risk of obesity. It is well documented that low-income neighborhoods do not have the same quality of food options as wealthier ones, with millions of Americans living in poor neighborhoods characterized by a high density of restaurants and stores with inexpensive and low-nutrition options that have a negative impact on health. The disproportionate impact of social, economic, and environmental circumstances on lower-income and minority populations suggests that attention to this issue also has a significant potential to reduce longstanding health disparities in obesity and diet-related diseases.
The good news is that there is widespread recognition among health systems regarding the need to address social and economic circumstances and their influence on health. There is a strong business case for health systems to prioritize these factors as they are important drivers of health care costs. For example, Advocate Health Care, an accountable care organization (ACO) in Chicago, started a malnutrition program that involved patient screening and the implementation of an enhanced nutrition care program offering nutrition education, post-discharge instructions, follow-up calls, and coupons for retail oral nutritional supplements for high-risk patients. As a result of these measures, the organization reported $4.8 million in total savings in 6 months. Another example is the University of Illinois Hospital System, which partnered with the community to launch a housing initiative targeting chronically homeless patients in the emergency room. Participating patients were moved to transitional housing and worked with case managers to devise a long-term solution, an effort that was reportedly associated with a 42% decline in health care costs.
The bad news is that health system changes like these have been slow. In the absence of a full pivot, patient care for obesity and diet-related conditions cannot be optimized.
The Reality of Changing Clinical Care
Readily available clinical tools for obesity and diet-related disease (e.g., behavioral counseling, weight loss medication, bariatric surgery) largely focus on proximal causes (e.g., behaviors) and not the fundamental causes (i.e., social, economic, and environmental conditions that may impact health), which matter tremendously. Most of these clinical tools, with the exception of bariatric surgery, have not demonstrated long-term sustainability and are not as effective among vulnerable groups. For example, in trials examining the effect of intensive behavioral therapy for weight loss, Black participants lose less weight than non-Black participants. This disparity is surely a source of significant frustration for patients and health professionals.
Because the demographics of physicians are not nearly as diverse as those of the general U.S. population, the day-to-day realities of their patients’ lives are unlikely to be immediately obvious. For example, 13% of the U.S. population is Black (compared with 4% of physicians) and 18% of the country is Latino (compared with 4% of physicians). It is also the case that most physicians in the U.S. did not grow up in poverty. These dissimilarities matter, as evidence suggests that sociocultural differences between patients and providers influence communication. Given that individuals from racial minority groups typically live in geographically separate communities with different environmental exposures and that food consumption is strongly influenced by cultural backgrounds, it is quite possible that health care providers may be unfamiliar with the resources and food options in communities of patients from different racial backgrounds. Better knowledge among physicians about the local realities of their patients’ environments may greatly facilitate advice that is more context-appropriate. For example, a physician’s advice to buy more fresh fruits and vegetables may fall on deaf ears if a patient lives in an area without ready access to this produce.
Physicians who are interested in learning more about their patients’ real-world circumstances and using that information to improve patient care face an uphill battle. Many physicians are handicapped by a lack of training and knowledge to effectively counsel patients on behavior modification to improve diet as well as insufficient time to deliver this counseling. Experiences with poverty, violence, hunger, or a host of other adverse events that could influence health outcomes also may not be the first thing that comes to mind when a patient is asked to describe his or her reason for seeking care. Also, for some patients, social or economic realities may be a source of embarrassment that they intentionally do not mention to their doctor. The well-documented negative attitudes of physicians toward patients with obesity (lower levels of respect for patients, weight stigma, pessimism about patient’s desire/ability to lose weight, belief that weight-loss counseling is ineffective, a perception that patients with obesity are not adherent to weight-loss activities) may further mute patients’ willingness to share relevant and important information about their lives, which may impact their disease risk and subsequent care.
Better integrating the realities of adverse social and economic conditions — and their impact on health — into the conversations between doctors and their patients is further complicated by the longstanding belief that the problem of obesity is mostly due to individual choices (proximal causes) rather than broader societal forces (fundamental causes). This perspective is shared widely by physicians, other health professionals (e.g., dietitians, nurses, behavioral psychologists, physical therapists, pharmacists), and Americans in general. Because physicians’ beliefs about the dietary causes of obesity translate into how they counsel patients on actionable issues (e.g., increasing activity, reducing portion size), improved education for physicians about the role of social, economic, and environmental influences on obesity and diet-related disease are important for improving clinical care.
There does appear to be a shift in this direction for new doctors entering the profession. For example, in 2015, the Medical College Admission Test (MCAT) began to include questions about how social inequality can affect the health of a patient. It is interesting to note that primary care physicians (PCPs) who completed medical school more recently generally report feeling more successful in helping patients with obesity to lose weight and that — regardless of when they completed medical school — PCPs overwhelmingly support additional training and practice-based changes to help them improve their obesity care.
While it is certainly the case that an improved understanding among physicians about their patients’ social contexts as well as the role of social and economic circumstances in obesity and diet-related disease are critical to improving communication and subsequent care, that alone is not enough. Even with optimal patient-physician communication and effective evidenced-based practice patterns, it is important to remember that physicians are not social workers. They lack the appropriate training to help patients navigate the social support systems and lack the time to do this well. And they cannot reasonably be expected to play this role in addition to all of their clinical responsibilities.
A well-established focus on cultural competence (the ability of health professionals and systems to provide care to diverse patients) and, more recently, a focus on contextualizing care (the adaption of approaches to account for the context in which patients live) are important steps in the right direction. So too is obesity training, which can improve the quality of care and preparedness to deliver obesity care. However, none of these steps are sufficient to meaningfully address the underlying issues of social and economic constraints faced by patients within the health care system. Structural changes at the health system level are needed.
So, what is currently happening in health systems that may impact care?
Addressing Health-Related Social Needs Through Delivery System Reform
There is quite a bit of encouraging activity at the federal level that aims to integrate health care delivery with social services to begin addressing health-related social needs. This progress generally focuses on delivery or payment reform as well as on identifying at-risk patients. Three promising activities include accountable care models, Delivery System Reform Incentive Payment (DSRIP) projects, and Accountable Health Communities (AHCs).
Rise of Accountable Care Models
Included in the 2010 Affordable Care Act delivery reforms were ACOs, which are responsible for achieving the “triple aim” of better health, improved patient experiences, and lower costs. ACOs are one example of many newer value-based payment models that are moving toward payment for outcomes rather than process measures. Generally, ACOs create clinical-community linkages by bridging primary and behavioral health care with community-based services that help to address the health-related social needs for vulnerable populations (e.g., food insecurity, inadequate housing).
The State Innovation Models (SIM) initiative, which supports state-based multi-payer health care delivery and payment system reform and includes a recognition of the role of social and economic circumstances on health, is a common funding mechanism for establishing Medicaid ACO programs. Through the SIM program, the Centers for Medicare and Medicaid Services (CMS) has awarded $950 million in grants since 2013 to 34 states, 3 territories, and the District of Columbia that together represent >60% of the U.S. population. Starting in the second round of grants, states were required to develop a statewide plan to improve population health, defined as health outcomes, patterns of health determinants, and the policies or interventions that link the two. State plans to address population health included health improvement initiatives targeting chronic conditions and focusing on decreasing health disparities among vulnerable low-income and underserved populations (e.g., the homeless, American Indians). These projects have included activities such as establishing links between primary care and community-based organizations and incorporating community health workers into care teams.
In 2012, Oregon used a Section 1115 waiver (which authorizes the spending of federal dollars on delivery system reforms that otherwise would not be available under current law) to implement a statewide accountable care model involving coordinated care organizations (CCOs), which focus on delivering coordinated care that bridges the community with the health system. These CCOs appear to be connecting with community partners and are beginning to address social and economic circumstances that impact health; for example, one CCO worked with providers and the Meals on Wheels program to deliver meals to Medicaid enrollees who needed food assistance as part of their recovery following discharge from the hospital. Importantly, CCOs are required to have a community advisory council in which at least half of the members are Medicaid beneficiaries. Evidence suggests that CCOs are associated with reductions in spending growth and improvement in some quality domains.
In 2017, Partners HealthCare Choice, one of the six organizations in the Massachusetts’ Medicaid ACO, launched a pilot initiative that includes incentives to address health-related social needs through partnerships with community-based organizations. This program requires comprehensive screening to identify patients who are adversely impacted by social and economic conditions and integrates community partners to provide more tailored care along with additional targeted resources and support services as needed. In some cases, this tailored care involves in-clinic encounters with community health workers who are able to assess a patient’s social or economic situation and act as a bridge to community agencies or resources. More intensive services (e.g., assistance in completing applications for housing assistance, home visits by community health workers who can more specifically assess a patient’s unmet social and economic needs, etc.) also may be provided. Across ACOs, the nonmedical needs that are most commonly addressed are those related to transportation, housing, and food insecurity.
Delivery System Reform Incentive Payment Projects
DSRIP projects are also authorized under a Section 1115 waiver and provide significant funding to states to support hospitals and eligible providers in changing how they provide care to Medicaid beneficiaries. This waiver mechanism was specifically designed to provide supplemental funding for safety-net providers (mainly acute-care hospitals) caring for low-income, vulnerable populations, although the focus has evolved since the first state (California) was funded in 2010. Now, eight states have DSRIP programs with a total budget of $33 billion. The DSRIP waivers generally focus on process improvements (e.g., infrastructure development or system redesign) and outcomes improvements (e.g., clinical health or population-based improvements), but they also can focus on improvements designed to foster stronger connections between health care providers and social services agencies.
In 2014, New York — which has the country’s largest Medicaid budget — began using a DSRIP waiver that includes transitional housing services as an optional transformation project for participating provider groups. Hospitals in these participating provider groups are required to partner with supportive housing and home care providers. Also included in New York’s Medicaid transformation (of which the DSRIP waiver is one component) is the requirement for certain providers in value-based purchasing agreements to implement at least one intervention focused on a social determinant of health (e.g., housing, food security) from a comprehensive menu provided by the state. The more recently approved DSRIP waiver in Washington State also includes a focus on supportive housing. Programs supporting housing access may be particularly relevant to addressing obesity and obesity-related disease as it has been estimated that over half of chronically homeless urban adults are affected by overweight or obesity, with chronically homeless women and Latino adults at highest risk.
Accountable Health Communities
In 2016, the CMS Innovation Center launched the AHCs model, which will test whether systematically identifying and addressing the health-related social needs of Medicare and Medicaid beneficiaries through screening, referral, and community navigation services will impact health care costs (including costs associated with obesity and obesity-related diseases) and reduce health care utilization. Over a 5-year period, the model will provide support to community bridge organizations to test promising service-delivery approaches aimed at linking beneficiaries with community services that may address their health-related social needs (e.g., housing instability, food insecurity, utility needs, interpersonal violence, and transportation needs), many of which have been linked to increased obesity risk. Currently, 31 participating organizations are working with a variety of community partners to assist high-risk beneficiaries in accessing community services that are responsive to their needs.
More Work Is Needed
For many health systems, the rapid growth of ACOs and other value-based payment models has made addressing social and economic needs related to health a priority. But most health systems are in the early stages of these activities. For this reason, evidence about impact is limited and distinctions between population-level progress versus headline-grabbing stories about short-term success are not entirely clear. Nevertheless, there are a number of important steps that health systems can take that should allow them to meaningfully address health-related social needs and, by doing so, better care for patients with obesity and diet-related disease.
To move the needle, considerable empirical evidence is needed about (1) who will benefit most from health systems reforms to address health-related social needs; (2) what are the most effective social and economic interventions that reduce barriers to healthy choices; (3) where in the community are partnerships with the health system most impactful; (4) when in the life course are interventions related to social determinants of health most critical; and (5) why health systems should prioritize a focus on addressing health-related social needs (Fig. 1). Clearly laying out the business case for necessary changes will likely require the collection of new data, which sometimes requires significant investment. Answers to these questions and many others are critical for scalability and are important for maximizing the value of these activities for health care systems.
Focus on Those in Highest Need
Going forward, it will be particularly important to address health-related social needs within the growing number of Medicaid ACOs because these organizations tend to serve higher-need and higher-cost patients who have complex behavioral, economic, and social needs that are often not addressed by the current health services. To date, there has been considerable federal investment in this area, as outlined above. However, the SIM initiative (also described above), which is funded by the CMS Innovation Center, may not be sustainable given that its funding is time-limited. And because the SIM programs are not specifically targeted to the Medicaid population, the ability to address health-related social needs among those most at risk (i.e., lower-income Americans) may be limited.
Leverage Electronic Medical Records
Electronic medical records have an important role to play as health systems explore ways to integrate data related to the social determinants of health into patients’ clinical records. Electronic medical record vendors have begun to develop new tools for the purpose of capturing these determinants (e.g., Cerner’s HealtheIntent and Epic’s Healthy Planet) and using them for population health management (e.g., resolving gaps in care, assessing social and economic conditions to identify care needs, etc.). However, these tools are not standardized and are not designed to be interoperable across multiple systems.
Once standards are established and the data are collected, a tricky issue will be determining what actions health systems should take to help address health-related social needs. For example, should patients be referred to community services, and do such referrals lead to better clinical outcomes by helping patients address important social needs? Stronger evidence in this area is needed both to encourage scalability and to realize benefits in an era in which rewards are mostly based on improving clinical outcomes or reducing expenditures (as indicated by the rise of ACOs, for example).
When using data obtained from electronic health records, care should be taken not to aggregate groups in such a way that important heterogeneity is masked, as there is considerable diversity within and between vulnerable populations. For example, it has been found that chronically homeless women are more likely to face complications of overweight and obesity than men and the relationship between education and obesity may be moderated by gender. It is possible that efforts to address health-related social needs may differentially impact some groups. Therefore, it will be important to avoid spurious results resulting from big data. Leveraging data to better understand what works also may be an effective way to avoid negative unintended consequences such as increased stigma among already vulnerable populations.
Broader Societal Change Is Essential
Health systems do not alone bear the responsibility for addressing health-related social needs. Significant policy, systems, and environmental changes outside of the health system are equally (if not more) critical in order to have a meaningful impact on the prevalence of obesity and diet-related disease. A number of promising activities have the potential to improve population health and health equity. Key among these activities are taxes on sugary beverages, changes to federal nutrition assistance programs, and federal menu labeling.
Beverage taxes, ranging from 1 to 2 cents per ounce, have been successfully passed in seven localities in the U.S. and are expected to significantly reduce purchases of sugary beverages. Results from Berkeley California, the first U.S. city to pass a beverage tax, indicate that the intake of sugary beverages declined by 25% in low-income neighborhoods after tax implementation. Because sugary beverage consumption is highest in lower-income communities, beverage taxes may reduce persistent disparities in sugary beverage consumption and, as a result, help to reduce the prevalence of obesity and related diseases among groups impacted by poverty.
The revenue generated by beverage taxes also has the potential to help address other health-related social needs. For example, the beverage tax in Philadelphia has generated $79 million in the first year, and those funds will be used to institute universal pre-kindergarten, establish community schools (improving access to education), and rebuild public parks (enhancing environmental features related to health). Lessons learned from tobacco taxes suggest that this policy mechanism may be very effective for reducing the consumption of sugary beverages.
Federal Nutrition Assistance Programs
A suite of 15 nutrition assistance programs, administered by the U.S. Department of Agriculture (USDA), makes up the federal nutrition safety net. Together, these programs have a budget of about $100 billion and touch the lives of one in four Americans daily. The 15 programs are intentionally varied in size, target, and scope, with nearly all of the budget spent on five programs: Supplemental Nutrition Assistance Program (SNAP), National School Lunch Program (NSLP), Special Supplemental Nutrition Program for Women, Infants, and Children (WIC), School Breakfast Program (SBP), and Child and Adult Care Food Program (CACFP) (Table I).
Given the reach of these programs, it is important to think critically about changes that can improve the nutrition of participants. SNAP is by far the largest program, with a cost of more than $70 billion each year. In a typical month, SNAP helps more than 40 million low-income Americans, half of whom are children. Currently, SNAP benefits can be used to purchase almost any food or beverage, including sugary beverages; therefore, one important policy change would be to restrict the ability to use SNAP benefits to purchase such drinks. Moreover, among federal nutrition assistance programs in which sugary beverages are currently available without reimbursement (e.g., CACFP), another important policy change would be to prohibit such beverages from being served at all.
Even small changes have the potential to impact millions of low-income Americans because the reach of these programs is so broad. Arguments against these proposed changes include the potential of difficulty structuring the elimination of sugar-sweetened beverage purchases because of challenges classifying these products, but much of that heavy lifting has already been done in areas in which beverage taxation has gone into effect. Health systems can play a role by helping to triage eligible patients into appropriate nutrition assistance programs and transition between those programs as needed.
The 2010 Affordable Care Act included the federal menu labeling rule, which mandates that calorie information be posted on menus and menu boards in restaurants or similar retail food establishments with more than 20 outlets. This rule was implemented in May 2018 after multiple delays. The potential impact of this policy is large because about one-third of adults and children eat at fast-food restaurants on a typical day. A number of studies have examined the potential impact of the federal menu labeling rule on consumer and restaurant behavior. Generally, the literature on this topic suggests that the largest impact from menu labeling results not from individuals changing their behavior in response to calorie information, but rather from reformulation by restaurants. Specifically, evidence suggests that menu labeling has little to no impact on calories purchased at fast-food restaurants and limited impact on calories purchased in coffee shops and cafeterias. Furthermore, other evidence suggests that large chain restaurants have decreased the calorie content of newly introduced menu items by about 60 calories (representing a 12% decline), that large chain restaurants that post calorie information have items that are lower in calories than chains that do not post such information, and that large chain restaurants are dropping high-calorie items from their menus, potentially in response to public pressure to offer healthier alternatives. The reasons for these changes in restaurant behavior are not well understood. Some possibilities include the anticipation of public backlash following the widespread posting of calorie information and consumer demand for healthier items.
Making Forward Progress
Unmet health-related social and economic needs increase the risks of developing obesity and diet-related disease, reduce the ability to manage these conditions, increase health care costs, and lead to avoidable utilization of health care resources. Because the prevalence of obesity and diet-related conditions is not determined by health services received, but rather by the vast number of influences outside of the health care system, the growing focus on acknowledging and addressing the social determinants of health within the health system represents a critical shift. Ultimately, health is determined not by isolated factors, but rather by a patient’s interconnected experiences with social, economic, environmental, and clinical influences. The acknowledgement of these factors by health care professionals and health care systems — while recognizing that they cannot fix this problem — is necessary to address how these components may impact clinical outcomes. However, there is limited information on how health systems can best intervene to improve patient outcomes.
A number of promising health system delivery reforms are underway. Going forward, it will be important to keep an eye out for the evidence emerging from these initiatives and, once there is clear evidence of an impact, rapid scaling will be needed to make a real difference. Any efforts by health systems will need to be complemented by coordinated and purposeful policy, systems, and environmental changes designed to reduce the underlying prevalence of obesity and diet-related conditions. There also needs to be wide recognition that no single change will solve this problem and that the social and environmental structures underlying obesity and diet-related disease will not be solved by health systems alone. But health systems can and should play an important role.