Diabetes Care 1 CETFCREVIEW SCIENTIFIC

Social Determinants of Health and Felicia Hill-Briggs,1,2 Nancy E. Adler,3 Seth A. Berkowitz,4 Marshall H. Chin,5 Diabetes: A Scientific Review Tiffany L. Gary-Webb,6 Ana Navas-Acien,7 Pamela L. Thornton,8 and 9 https://doi.org/10.2337/dci20-0053 Debra Haire-Joshu

Decades of research have demonstrated that diabetes affects racial and ethnic minor- ity and low-income adult populations in the U.S. disproportionately, with relatively intractable patterns seen in these populations’ higher risk of diabetes and rates of diabetes complications and mortality (1). With a health care shift toward greater emphasis on population health outcomes and value-based care, social determinants of health (SDOH) have risen to the forefront as essential intervention targets to achieve health equity (2–4). Most recently, the COVID-19 pandemic has highlighted unequal vulnerabilities borne by racial and ethnic minority groups and by disadvantaged communities. In the wake of concurrent pandemic and racial injustice events in the U.S., the American College of Physicians, American Academy of Pediatrics, Society of General Internal Medicine, National Academy of Medicine, and other professional organizations have published statements on SDOH (5–8), and calls to action focus on amelioration of these determinants at individual, organizational, and policy levels (9–11). In diabetes, understanding and mitigating the impact of SDOH are priorities due to disease prevalence, economic costs, and disproportionate population burden (12–14). In 2013, the American Diabetes Association (ADA) published a scientific statement on socioecological determinants of prediabetes and type 2 diabetes (15). Toward the goal of understanding and advancing opportunities for improvement among the population with diabetes through addressing SDOH, ADA convened the current SDOH and diabetes 1Department of Medicine, Johns Hopkins Univer- writing committee, prepandemic, to review the literature on 1) associations of SDOH sity, Baltimore, MD 2 with diabetes risk and outcomes and 2) impact of interventions targeting amelioration Welch Center for Prevention, Epidemiology and of SDOH on diabetes outcomes. This article begins with an overview of key definitions Clinical Research, Johns Hopkins Medical Insti- tutions, Baltimore, MD and SDOH frameworks. The literature review focuses primarily on U.S.-based studies of 3Department of Psychiatry and Behavioral Sci- adults with diabetes and on five SDOH: socioeconomic status (education, income, ences, University of California San Francisco, San occupation); neighborhood and physical environment (housing, built environment, Francisco, CA 4 toxic environmental exposures); food environment (food insecurity, food access); Division of General Medicine and Clinical Epi- demiology, Universityof NorthCarolinaatChapel health care (access, affordability, quality); and social context (social cohesion, social Hill, Chapel Hill, NC capital,socialsupport).Thisreviewconcludeswithrecommendationsforlinkagesacross 5Department of Medicine, University of Chicago, health care and community sectors from national advisory committees, recommen- Chicago, IL dations for diabetes research, and recommendations for research to inform practice. 6Departments of Epidemiology and Behavioral and Community Health Sciences, University of Pittsburgh, Pittsburgh, PA DEFINITIONS OF HEALTH DISPARITIES, HEALTH EQUITY, AND SDOH 7Department of Environmental Health Sciences, Table1displays definitions of keyterms. Differences in diabetes riskand outcomescan Columbia University, New York, NY 8 result from multiple contributors, including biological, clinical, and nonclinical factors National Institute of Diabetes and Digestive and fi Kidney Diseases, National Institutes of Health, (1). A substantial body of scienti c literature demonstrates the adverse impact of a Bethesda, MD particular type of difference, health disparities (16) in diabetes (1,17,18). A pre- 9The Brown School and The School of Medicine, ponderance of health disparities research in the U.S. has examined disparities by race Washington University in St. Louis, St. Louis, MO and ethnicity (3,19). Internationally, the term health equity has traditionally been Corresponding author: Felicia Hill-Briggs, fbriggs3@ used to encompass the range of population inequalities resulting from demographic jhmi.edu and economic characteristics, and this term is used increasingly in the U.S. (20–24). Received 25 September 2020 and accepted 25 Addressing healthy equity necessitates an understanding of social and environmental September 2020 factors that combined account for 50% to 60% of health outcomes (22,25). These © 2020 by the American Diabetes Association. social and environmental factors collectively are known as SDOH (21,26,27). Readers may use this article as long as the work is properly cited, the use is educational and not for profit, and the work is not altered. More infor- SDOH NOMENCLATURES AND CONTEXTUAL FACTORS mation is available at https://www.diabetesjournals The writing committee reviewed the following commonly referenced SDOH frame- .org/content/license. works for classifications and terminology: the World Health Organization (WHO) See accompanying articles, pp. XX, XX, Commission on Social Determinants of Health (28), Healthy People 2020 (29,30), the XX, and XX. Diabetes Care Publish Ahead of Print, published online November 2, 2020 2 Social Determinants of Health and Diabetes Diabetes Care

Table 1—Definitions Term Definition Health disparities A particular type of health difference that is closely linked with social, economic, and/or environmental disadvantage. Health disparities adversely affect groups of people who have systematically experienced greater obstacles to healthbased on their racial or ethnic group;religion; socioeconomic status; sex; age; mental health; cognitive, sensory, or physical disability; sexual orientation or gender identity;geographiclocation; orothercharacteristicshistoricallylinked todiscrimination or exclusion (16). Health equity Equity is the absence of avoidable, unfair, or remediable differences among groups of people, whether those groups are defined socially, economically, demographically, or geographically or by other means of stratification. “Health equity” or “equity in health” implies that ideally everyone should have a fair opportunity to attain their full health potential and that no one should be disadvantaged from achieving this potential (24). Health equity is attainment of the highest level of health for all people. Achieving health equity requires valuing everyone equally with focused and ongoing societal efforts to address avoidable inequalities, historical and contemporary injustices, and the elimination of health and health care disparities (23). Social determinants of health (SDOH) The social determinants of health are the conditions in which people are born, grow, live, work, and age. These circumstances are shaped by the distribution of money, power, and resources at global, national, and local levels. The social determinants of health are mostly responsible for health inequitiesdthe unfair and avoidable differences in health status seen within and between countries (26).

County Health Rankings Model (31,32), inequities that work through interme- housing, transportation, and nutritious food and Kaiser Family Foundation Social De- diary sets of determinants to cause health and social resources such as political power, terminants of Health factors (33). inequities. WHO positions the health social engagement, and control. No single consensus set of factors de- system as an SDOH that plays a role in The three components of SES are in- fine SDOH. Nomenclatures used by each mediating impact of intermediary deter- tercorrelated (38), but each aspect has framework, for shared SDOH factors, are minants on health outcomes (28). unique implications for health. Each com- depicted in Fig. 1. Common among the ponent can be assessed at the individual frameworks is placement of economic REVIEW OF SDOH AND DIABETES or population level (39). For example, and socioeconomic determinants as fore- SDOH inclusion in this review was deter- economic status is often measured by most. Food SDOH factors (e.g., insecurity mined by representation within one or determining a person’s own income. and access) are classified as economic more existing SDOH frameworks and However, it is also assessed by the in- stability (29), neighborhood and built presence of a sufficient body of literature come of the household in which the environment (29), material circumstan- to demonstrate influence of the deter- person resides and by the income level ces (28), or as an independent category minant on diabetes. The reviewed SDOH of the community (e.g., mean household (33). Housing is classified as material are shown in Table 2. Studies of primar- income of the census track in which a circumstances (28), economic stability ily adult populations are described, and person resides) as a proxy for the indi- (29), or neighborhood and built envi- the terms type 2 diabetes mellitus vidual’s household income. Census-level ronment (33). Environmental exposures (T2DM), type 1 diabetes mellitus (T1DM), household income also operates as a (i.e., toxins, air pollution, and water and “diabetes” (indicating either a mixed contextual variable, reflecting the com- quality) are classified as neighborhood T2DM and T1DM sample or diabetes case position and available resources in a de- and physical environment (31) or built ascertainment methods that do not en- fined area. environment (29). Social environment is able specifying clinical diagnostic type) Educational status can be quantified represented as community and societal are used in accordance with the termi- either in years of schooling or highest context (29,33), social cohesion and so- nology used in the respective studies. degree earned. It may be assessed at the cial capital (28), or social factors (31). level of the individual (e.g., the person’s Health care is a SDOH in the WHO, Socioeconomic Status and Diabetes own educational attainment), the house- Healthy People 2020, and County Health Socioeconomic status (SES) is a multidi- hold (the highest grade completed by Rankings models, with access factors as mensional construct that includes edu- anyone in the household), or the com- primary, with or without quality of care cational, economic, and occupational status munity (e.g., percent of high school or factors. (34–36). SES is a consistently strong pre- college graduates in a census track). Each framework posits complex inter- dictor ofdisease onset and progressionat Quantity of education does not capture actions among SDOH factors. WHO ad- all levels of SES for many diseases, in- differences in quality of education that ditionally maps causal priority among cluding diabetes (37). SES is linked to may be relevant to SES measurement the determinants. For example, the up- virtually all of the established SDOH. It is (38). Literacy has emerged as a measure stream sociopolitical context and result- associated with the extent to which in- of educational quality and as potentially ing socioeconomic position are root dividuals and communities can access more reflective of SES than years of cause, structural determinants of health material resources including health care, schooling among African Americans and care.diabetesjournals.org Hill-Briggs and Associates 3

Figure 1—Nomenclatures for shared determinants among four social determinants of health frameworks, the World Health Organization Commission on the Social Determinants of Health, the U.S. Department of Health and Human Services Healthy People 2020, the County Health Rankings Model, and the Kaiser Family Foundation Social Determinants of Health framework. low-income Whites (40,41). Health liter- the SES ladder are more likely to develop middle income, near poor, and poor, was acy, which is directly associated with T2DM, experience more complications, 40.0%, 74.1%, and 100.4%, respectively. literacy and is context specific (42–44), and die sooner than those higher up The difference in diabetes prevalence by and literacy are included as SDOH in on the SES ladder (46,47). The higher a income was greater during this time pe- Healthy People 2020 (29). person’s income, the greater their edu- riod than it had been in a prior period Occupation is itself multidimensional. cational attainment, and the higher their (1999–2002), pointing to widening dispar- It has been measured as employment occupational grade, the less likely they ities in diabetes prevalence associated status (e.g., employed vs. unemployed), are to develop T2DM or to experience its with income. stability (e.g., job insecurity), job type complications. The gradient is steeper at At the neighborhood level, differences (e.g., manual vs. nonmanual, prestige of the bottom, however, and research has in diabetes prevalence by census track the occupation), and working conditions focusedprimarilyonthosewiththelowest are attributable to SES (51,52). For ex- (e.g., shift work, number of hours worked, levels of income and education. ample, in a recent study by Kolak et al. job demands, and control) (39,45). For Income. Prevalence of diabetes increases (52), rate of T2DM was found to be members of large organizations, occupa- on a gradient from highest to lowest significantly higher and concentrated in tional hierarchies of job titles capture income (48,49). In data from the National census tracts characterized by factors in- work conditions as well as qualifications Health Interview Survey (NHIS) covering cluding lower incomes, lower high school and pay (e.g., civil service grades). 2011–2014, Beckles and Chou (50) found graduationrates,moresingle-parenthouse- Associations of SES With Diabetes Incidence, increasing diabetes prevalence at lower holds, and crowded housing. Living in Prevalence, and Outcomes levels of income as reflected in the levels neighborhood census tracts with lower Income, education, and occupation show of ratio of income to poverty level. Com- educational attainment, lower annual a graded association with diabetes prev- pared with those with high income, the income, and larger percentage of house- alence and complications across all levels relative percentage difference in preva- holds receiving Supplemental Nutrition of SES, up to the very top. Those lower on lence of diabetes for those classified as Assistance Program benefits has been

Table 2—SDOH and component factors included in the diabetes review Socioeconomic status Neighborhood and physical environment Food environment Health care Social context Education Housing Food security Access Social cohesion Income Built environment Food access Affordability Social capital Occupation Toxic environmental exposures Food availability Quality Social support 4 Social Determinants of Health and Diabetes Diabetes Care

associated with higher risk of progres- education, 9.5% for those with a high with standard work hours (35–40 h per sion to T2DM among adults with pre- school education, and 7.2% for those week) with higher incident diabetes in diabetes (53). with more than a high school education adults with low SES but not in adults with Gaskin et al. (49) examined the in- (61). Having a college education or more high SES. A U.S. population-based survey teraction of individual poverty with neigh- is associated with the lowest odds of on diabetes and occupation found high- borhood poverty and found that, compared diabetes (62). Mirroring findings on in- est prevalence of diabetes among trans- with nonpoor adults living in nonpoor come, temporal trends in diabetes prev- portation workers and lowest prevalence neighborhoods, poor adults living in alence at different levels of education of diabetes among physicians (69,70). nonpoor neighborhoods have increased show increasing disparities in prevalence odds of having diabetes, and poor adults associated with educational attainment SES Interventions and Diabetes Outcomes To date, there is no body of literature living in poor neighborhoods have two- (50). describing impact of change in income, fold higher odds of having diabetes. In The risk of diabetes-related mortality change to higher educational status, or addition, a race-poverty-place gradient demonstrates a gradient from lowest to different employment/occupational sta- was observed. Compared with nonpoor highest education level. Compared with tus on diabetes outcomes, although in- Whites in nonpoor neighborhoods, odds adults with a college degree or higher, come and wage changes, and job changes of diabetes were highest for poor Whites having less than high school education is and loss, do occur naturalistically. Sim- in poor neighborhoods (odds ratio [OR] associated with a twofold higher mor- ilarly, no diabetes outcomes have been 2.51, 95% CI 5 1.31–4.81), followed by tality from diabetes (relative hazard 2.05, reported from interventions directly tar- poor Blacks in poor neighborhoods and 95% CI 1.78–2.35) (54). In adults with geting living wages, early childhood ed- nonpoor Blacks in poor neighborhoods T1DM, not having a college degree is ucation, educational quality, or educational (OR 2.45, 95% CI 1.50–4.01, and OR 2.49, associated with a threefold higher mor- access for poor children and families. 95% CI 1.48–4.19), and finally poorWhites talityfromdiabetes compared withcoun- Studies have examined diabetes self- in nonpoor neighborhoods (OR 1.73, 95% terpartswithacollege degree (63).Lower management interventions in the setting CI 1.16–2.57) (49). educational level is associated with higher of low literacy/health literacy, particu- Adults with T2DM who have a family HbA , with a meta-analysis (56) reporting 1c larly among racial/ethnic minority adults income below the federal poverty level apooledmeandifferenceinHbA of 1c with T2DM and have demonstrated ef- have a twofold higher risk of diabetes- 0.26% (95% CI, 0.09–0.43) between peo- fectiveness of low-literacy adaptions (71) related mortality compared with their ple with low and high educational levels. and health literacy and numeracy tools in counterparts in the highest family in- Regarding literacy/health literacy as a improving diabetes knowledge and self- come levels (54). This pattern of diabe- SDOH, Marciano et al. (64) conducted a care (72–74). A meta-analysis of nine tes-related mortality has been observed meta-analysis of 61 studies of 18,905 intervention trials with 1,874 adults specifically in adults with T1DM as well adults with T1DM or T2DM to determine with T2DM found that literacy-sensitive (55). A meta-analysis by Bijlsma-Rutte associations of health literacy with sev- interventions wereassociated witha small etal.(56)observed aninverseassociation eral diabetes outcomes and found that but statistically significant decrease in between income and HbA levels in higher levels of health literacy were sig- 1c HbA (–0.18%; 95% CI –0.36 to –0.004) people with T2DM, with a pooled mean nificantly associated with lower HbA 1c 1c in comparison with usual clinical care (75) difference in HbA of 0.20% (95% CI 20.05 levels and better diabetes knowledge, 1c in patients regardless of health literacy to 0.46) between people with low and high but not with more frequent self-man- status. Literacy-adapted education and income. Low income is associated with a agement activities. tools may need to be combined with higher risk of experiencing diabetic ketoa- Occupation.Systematic reviews and meta- more comprehensive evidence-based be- cidosis among youth and adults with T1DM analyses have examined several aspects havioral self-management intervention (57) and with higher HbA levels, partic- of occupation in relation to diabetes risk, 1c approaches to achieve substantive clinical ularly among racial/ethnic minority youth although most of this research has been improvements in racial/ethnic minority with lower SES (58,59). conducted outside of the U.S. Ferrie et al. populations with T2DM and low liter- Education. Age-adjusted incidence of di- (65) conducted a meta-analysis of asso- agnosed diabetes in adults is associated ciations of job insecurity with incident acy/health literacy (76,77). In conclusion, also with educational level in a stepwise diabetes and found an association of high despitethelong-standingevidenceforSES pattern. Diabetes incidence is highest job insecurity with higher risk of incident as a key determinant both of diabetes risk (10.4 per 1,000 persons) for adults diabetes (OR 1.19, 95% CI 1.09–1.30). A and outcomes, systematic investigation of with less than a high school education, meta-analysis by Varanka-Ruuska et al. impact on diabetes of change in SES 7.8 per 1,000 persons for those with a (66) found that unemployment was as- remains a gap in the literature. terminal high school education, and 5.3 sociated with increased odds of both per 1,000 persons for those with more prediabetes (OR 1.58, 95% CI 1.07– Neighborhood and Physical than a high school education (60). Di- 2.35) and T2DM (OR 1.72, 95% CI Environment and Diabetes abetes prevalence in the adult U.S. pop- 1.14–2.58). Exposure to shift work is The neighborhood environment in which ulation is similarly inversely associated associated with higher risk of diabetes one lives has been of major interest as a with educational level in a stepwise pat- than working normal daytime schedules setting in which to understand contex- tern. In the U.S., the age-adjusted prev- (67). A meta-analysis by Kivimaki¨ et al. tual and multilevel influences on health alence of diagnosed diabetes is 12.6% (68) reported an association of long work (78).DiezRouxandMair(78)havedescribed for those with less than a high school hours ($55 h per week) as compared the role of historical and contemporary care.diabetesjournals.org Hill-Briggs and Associates 5

residential segregation by race, ethnicity, instability or a commonly used assess- Housing Instability Interventions and Diabetes and SES as the socioeconomic and political ment instrument. Further, because hous- Outcomes. Given its expense, housing is context that produced the patterns of un- ing instability is more likely to occur one of the most difficult health-related equal resource distribution resulting in among individualswith lower SESdwhich social needs to intervene upon. Housing neighborhood environments that maintain is independently associated with higher intervention studies reporting diabetes health inequities. Tung et al. (79) also diabetes prevalencedit is unclear whether outcomes are few; however, there is discuss the multiple intricacies associated housing instability is causally related to some high-quality evidence for housing with how race, place, and poverty con- developing diabetes. One systematic re- interventions. The Moving To Opportu- verge in a dynamic way across various view did not find higher diabetes preva- nityforFairHousingDemonstrationProject spatial contexts and circumstances to lence than in the general population (MTO), a randomized social experiment influence health and propose that un- among persons experiencing homeless- conducted through the Department of derstanding the intersection of these ness, estimating approximately 8% prev- Housing and Urban Development, in contextual influences is needed to pre- alence in adults who do and do not partnership with behavioral scientists vent diabetes inequities. Neighborhood experience homelessness (94). A recent and other federal agencies, was designed and physical environment factors of study using nationally representative to determine what impact moving from a housing, built environment, and envi- data from individuals seen in community high-poverty to a low-poverty census ronmental exposures are reviewed. health centers found that approximately tract would have on multiple outcomes 37% of individuals with diabetes re- (102,103). In 1994–1998, MTO random- Housing ported housing instability. This study ized 4,498 women with children living in Stable housing is a key indicator of eco- also found that individuals with diabetes public housing within high-poverty cen- nomic stability (80) and a core SDOH (80). and housing instability were more likely to sustractsinfivecitiestooneofthreestudy Housing instability refers to a spectrum self-report having an emergency depart- arms. The 1,788 women in the experi- of situations that can range from living ment visit or hospitalization for their di- mental arm received Section 8 vouchers ’ in one s car, staying with relatives or abetes (adjusted OR 5.17, 95% CI 2.08– usable only in low-poverty areas (census friends, having trouble paying rent, suf- 12.87) (82). A cross-sectional study in a tracts with ,10% of the population below fering evictions or frequent moves, pay- single health care system found that the poverty line) along with counseling ing more than 50% of income in rent, housing instability among individuals and assistance in finding a private rental and living in crowded conditions (histor- with diabetes was associated with higher unit. The 1,312 women in the Section fi ically de ned as having more than one outpatient utilization (incident rate ra- 8 arm received traditional unrestricted d person per room) to homelessness the tio 1.31, 95% CI 1.14–1.51) (95). Though vouchers and the usual briefing the local most extreme form of unstable housing not specific to diabetes, additional work Section 8 program provided. The 1,398 – fi (81 85). Homelessness is de ned as has linked housing instability to poor women in the control arm received no “ lacking a regular nighttime residence health outcomes and reduced health vouchers but continued to receive MTO or having a primary nighttime residence care access (91,96–100). A longitudinal project-based assistance. Those who re- that is a temporary shelter or other study in the Department of Veterans ceivedvouchers could choose whether to ” place not designed for sleeping (86). Affairs (VA) health care system found use the vouchers or not. Findings from Asof 2020,theU.S. governmentreported that experiencing homelessness was as- the follow-up survey in 2008 through 567,715 or 17 of 10,000 people in the sociated with higher adjusted odds of 2010 found a 21.6% relative reduction in country are homeless; African Americans . having an HbA1c .8.0% and .9.0%. prevalence of an elevated HbA1c ( 6.5%) accounted for 40% of people experienc- Vijayaraghavan et al. (84) identified un- in the group that moved to low-poverty ing homelessness, while those identify- stable housing as a key barrier to diabetes census tracts compared with the control ing as Hispanic or Latino comprised 22% careamonglow-incomeindividuals.There group, with an absolute difference of of the homeless population (87). A com- was an observed linear decrease in di- 4.31 percentage points (95% CI 27.82 mon theme in conceptual models linking abetes self-efficacy as housing instability to 20.80). The low-poverty group also housing instability to poor health is that increased (b-coefficient 20.94, 95% CI had relative reductions of 13.0% in prev- the instability inherent to the situation 21.88 to 20.01, P , 0.01), which was alence of BMI $35 and relative reduction fi makes it dif cult to attend to preventive partially mediated by food insecurity. of 19.1% in BMI $40 kg/m2, with abso- – services and self-care (83,88 90), leading Qualitative work has found that unsta- lute differences of 4.61 percentage points to worse control of chronic conditions, blehousingmakesitmoredifficulttoen- (95% CI 28.54 to 20.69) and 3.38 per- higher use of acute-care services like gage in self-care, follow self-management centage points (95% CI 26.39 to 20.36), emergency departments, and higher like- routines, afford diabetes medications and respectively(102).Theusualvouchersand – lihood of complications (91 93). supplies, and eat healthy foods (91,92). control arms did not differ. Other MTO Associations of Housing Instability With Diabe- Choice of medication is important, and outcomes among the group randomized tes Incidence, Prevalence, and Outcomes. The considerations should include medication to low-poverty census tracts included prevalence of diabetes among those with cost and the ability to store medication higher housing quality, education, em- housing instability in the U.S., and whether and diabetes care supplies safely. Brooks ployment, andearnings aswell as multiple it differs from that among those without et al. provide a narrative review of con- additional improvements to child and housing instability, is not known. A key siderations for diabetes treatment among adult health (103). A 10–15 year follow-up limitation for the field is that there is individuals experiencing homelessness study found substantial and sustained no single, accepted definition of housing (101). reductions in diabetes prevalence, rates 6 Social Determinants of Health and Diabetes Diabetes Care

of extreme obesity, and improvement in obesity-related outcomes (110–113). and tree-lined areas). The meta-analysis, mentalhealthoutcomesamongtheadults However, a smaller body of research representing 462,220 participants, showed who received vouchers to move to low- has examined associations of the built an association of high exposure with poverty neighborhoods and reduction in environment with diabetes specifically. reductions in the incidence of T2DM (OR extreme obesity among the adults who Smalls et al. (114) reported significant 0.71, 95% CI 0.61–0.85) (117). After dec- received Section 8 vouchers (104). While associations of both walking environ- ades of research, many built environ- not specific to diabetes, a meta-analysisof ment (b 520.040) and neighborhood ment factors related to PA and obesity randomized trials that provided low-barrier activities (b 520.104) with exercise in a risk have been identified for consider- housing support for individuals experienc- southeastern U.S. population with dia- ation in urban planning (118). ing homelessness found significant reduc- betes. A recent U.S. review and meta- Neighborhood-Level Interventions on the tions in health care utilization (105). analysis by Chandrabose et al. (113) Built Environment and Diabetes Outcomes. Housing interventions may facilitate examined longitudinal studies of the Because it is often not feasible or ethical access to diabetes care. The Collabora- built environment and cardiometabolic to randomize neighborhoods to receive tive Initiative to End Chronic Homeless- health. Results showed strong evidence certain structural interventions, natural nessprovidedadultswhowerechronically for impact of walkability on T2DM out- experiment designs are used in which homeless with permanent housing and comes, with four of seven studies (57%) the researcher does not control or with- supportive primary health care and men- showing significant findings in the ag- hold intervention allocation to particular tal health services (106). Placed persons gregated analyses using objective and areas; rather, natural or predetermined were morelikely toreceiveevaluationand perceived measures of walkability. Al- variation in allocation occurs, often as a management services (relative risk [RR] though the methods to determine me- result of policy intervention (119). Several 1.03, 95% CI 1.01–1.04) than unplaced diation by physical activity in most review articles of natural experiments persons (107). Placed persons were more studieswereineffectivetomakecon- summarize the benefits of policy and built likely to receive HbA1c tests (RR 1.10, 95% clusions, one study tested the indirect environment changes on obesity-related CI 1.02–1.19) and lipid tests (RR 1.09, 95% effect of walkability on 10-year change outcomes(112)anddietandPAoutcomes CI 1.02–1.17), while for those without in HbA1c and found a partial mediation (120,121).Thestrongestdiet-relatedstudies baseline diabetes placement was associ- effect for self-reported physical activity were those that evaluated regulations to ated with lower risk of new diabetes using structural equation modeling. For the food environment, and the strongest diagnoses (RR 0.87, 95% CI 0.76–0.99). other measures of built environment, PA-related studies were those that im- Keene et al. (91) suggest the relationship such as neighborhood recreational fa- proved infrastructure for active transport. of stable housing to diabetes manage- cilities or destinations/routes, there was Although this literature does not directly ment is due to its role as a foundation for insufficient data to examine the rela- address diabetes outcomes, improve- prioritizing care and allowing for the tionship with T2DM outcomes. A larger ments in obesity and diet and PA behav- routinization of diabetes management, body of research on built environment iors are relevant to populations with critical to disease control. This suggests and diabetes has been conducted in diabetes and warrant rigorous evaluation the benefits of supportive and stable countries outside of the U.S (110–112). (122). housing may be extended to diabetes In these studies, neighborhood physical care and prevention. A naturalistic qual- activity (PA) environments, specifically Toxic Environmental Exposures itative study of the impact of transitioning better walkability of neighborhoods and Toxic environmental exposures can be to rental-assisted housing among low- access to greenspace, have been consis- naturally occurring (e.g., arsenic in pri- income, housing-insecure adults with T2DM tently associated with lower risk of T2DM vate wells, radon) or introduced into the reported that rental assistance afforded and better outcomes (115,116). Numer- environment through human activity individuals more environmental and fi- ous studies have been conducted on (e.g., pollution, synthetic pesticides) (123). nancial control over life circumstances, walkability measured by macroscale as- Marginalized communities in the U.S. are thereby enabling diabetes routines and pects of the neighborhood, including higher disproportionately exposed to environ- allocation of financial resources to diabe- population density, land use mix, and mental agents that have evidence of an tes care (108). aesthetics, to microscale aspects, includ- association with diabetes, including air ing sidewalks, street connectivity, and pollution, environmental toxicants, and Built Environment street safety. A review by Bilal et al. (115) ambient noise (124–129), and subgroups fi The built environment, as de ned by the on walkability and diabetes incidence that generate the least pollution have U.S. Centers for Disease Control and and prevalence found that more walk- highest exposures (130). Prevention (CDC), includes the physical able neighborhoods are associated with Factors contributing to inequities in parts of where people live and work, such a lower incidence and prevalence of toxic environmental exposures include as infrastructure, buildings, streets, and T2DM. Similarly, Twohig-Bennett and residential segregation and inequity in open spaces (109). Here, built environ- Jones (117) conducted a systematic re- goods and services, due in part to sys- ment factors of walkability and green- view and meta-analysis examining the temic racism in environmental regulation space are reviewed. relationship to diabetes outcomes of and opportunities (128,130–133). Ex- Associations of Built Environment with Diabe- “high” and “low” exposure to green- planatory factors are closer proximity tes Incidence, Prevalence, and Outcomes. A space in neighborhoods (defined as of underserved neighborhoods to nearby robust literature has demonstrated as- open, undeveloped land with natural pollution sources, poor enforcement of sociations of the built environment with vegetation and/or spaces such as parks regulations, and inadequate response to care.diabetesjournals.org

Table 3—SDOH intervention recommendations from international and national (U.S.) committees Committee Recommended actions Description Commission on the Social Determinants of Health, Improve daily living conditions Put major emphasis on early childhood education and development. Improve living and WHO (2008) (27) working conditions. Create social protection policy supportive of all. Tackle the inequitable distribution of power, money, and Create a strong public sector that is committed, capable, and adequately financed. Ensure resources legitimacy, space, and support for civil society, for an accountable private sector, and for the public to agree to reinvestment in collective action. Measure and understand the problem and assess the Acknowledge there is a problem. Ensure that health inequity is measured. Develop national impact of action and global health equity surveillance systems for routine monitoring of health inequity and the social determinants of health. Evaluate the health equity impact of policy and action. Ensure stronger focus on social determinants in public health research. Committee on Recommended Social and Behavioral Standardize data collection and measurement to Office of the National Coordinator for Health Information Technology and the CMS should include Domains and Measures for Electronic Health facilitate the critical use and exchange of information therecommendedstandardizedmeasuresinthecertification and meaningful use regulations: Records, Institute of Medicine, NASEM (2014) (80) on social and behavioral determinants of health Commonly used measures: race and ethnicity,* residential address,* alcohol use, tobacco use Additional recommended measures: census tract-median income,* education,* financial resource strain,* social connections and social isolation,* depression, intimate partner violence, physical activity, stress Committee on Educating Health Professionals to Create a learning environment for health professionals to Health professional educators should create lifelong learners who appreciate the value of Address the Social Determinants of Health, NASEM foster community collaborations relationships and collaborations for understanding and addressing community-identified (2016) (301) needs and for strengthening community assets. Prepare health professionals to take action on SDOH Topreparehealthprofessionalstotakeactiononthesocialdeterminantsofhealthin,with,and across communities, health professional and educational associations and organizations at the global, regional, and national levels should apply [frameworks for] partnering with communitiestoincreasetheinclusivityanddiversityofthehealthprofessionalstudentbody and faculty. Integrate SDOH into organizational mission and values Governments and individual ministries (e.g., signatories of the Rio Declaration), health professional and educational associations and organizations, and community groups should foster an enabling environment that supports and values the integration of the social determinants framework principles into their mission, culture, and work. Build the evidence base for SDOH learning, intervention, Governments, health professional and educational associations and organizations, and and evaluation approaches community organizations should use [a social determinants] framework and model to guide and support evaluation research aimed at identifying and illustrating effective approaches for learning about the social determinants of health in and with communities while improving health outcomes, thereby building the evidence base. Committee on Integrating Social Needs Care Into the Design health care delivery to integrate social care into Establish organizational commitment to addressing disparities and health-related social Delivery of Health Care to Improve the Nation’s health care, guided by the five health care system needs. Incorporate strategies for screening and assessing for social risk factors and needs. Health, NASEM (2019) (5) activitiesdawareness, adjustment, assistance, Incorporate social risk into care decisions using patient-centered care. Establish linkages alignment, and advocacy between health care and social service providers. Include social care workers in team care. Develop infrastructure for care integration, including financing of referral relationships with ilBig n soits7 Associates and Hill-Briggs select social providers. Build a workforce to integrate social care into health care Social workers and other social care workforces should be providers eligible for delivery reimbursement from payers. Integrate SDOH competencies in medical and health professional credentialing.

Continued on p. 8 oilDtriat fHat n Diabetes and Health of Determinants Social 8

Table 3—Continued Committee Recommended actions Description Develop a digital infrastructure that is interoperable Establish ACA-recommended digital infrastructure for social care. The Office of the National between health care and social care organizations Coordinator should support identification of interoperable, secure, platforms for use across health and social care communities. The Federal Health Information Technology Coordinating Committee should facilitate data sharing across domains (e.g., health care, housing, and education). Analytic and technology implementation must have an explicit focus on equity to avoid unintended consequences such as perpetuation or aggravation of discrimination, bias, and marginalization. Finance the integration of health care and social care CMS should define and use flexibility in what social care constitutes Medicaid-covered services. Health systems, payers, and governments should consider collective financing to spread risk and create shared returns on investments in social care. Health systems subject to community benefit regulations should comply with those regulations and should align their hospital licensing requirements and public reporting with community benefits regulations and should link their community benefits providing social care. Fund, conduct, and translate research and evaluation on Federal (e.g., NIH, AHRQ, PCORI) and state agencies, payers, providers, delivery systems, and the effectiveness and implementation of social care foundations should contribute to advancing research and evaluation of social care through practices in health care settings funding opportunities, researcher support (i.e., cultivate health services, social sciences, and cross-disciplinary researchers), and use of experimental trials, rapid learning cycles, and dissemination of learnings. CMS should fully finance independent state waiver evaluations to ensure robust evaluation of social care and health care integration pilot programs and dissemination. AHRQ, Agency for Healthcare Research and Quality; NIH, National Institutes of Health; PCORI, Patient-Centered Outcomes Research Institute. *SDOH measures. ibtsCare Diabetes care.diabetesjournals.org Hill-Briggs and Associates 9

community complaints (134–138). In ru- insulin resistance and T2DM (159–161). comparison of preterm births among ral and suburban communities, including These findings highlight that populations four studies in different countries, be- Native American Indian communities, more exposed to air pollution are also fore and after the implementation of unregulated private wells are a source disproportionately at risk for developing smoke-free legislation, has shown re- of water contaminants including arsenic diabetes. ductions in diabetes risk (pooled risk and other metals/metalloids, pesticides, There is epidemiologic and experimen- change 218.4%, 95% CI 218.8 to 22), and hazardous chemicals, affecting mil- tal evidence that environmental exposures although the long-term benefits have lions of people (139–141). Both food increase susceptibility to cardiovascular not yet been evaluated (171). packaging and fast-food consumption, disease (CVD) in people with diabetes. Because individuals generally have which can be high in low-income neigh- The evidence is extensive for air pollution limited control over environmental agents, borhoods, can expose people to chem- exposures (162,163). For example, in themost effectiveinterventions willbeat icals known to be endocrine disrupters Medicare patients, a daily increase of the population level, through policy and (142–145). Examples include chemicals 10 mg/m3 in particulate matter ,10 regulation, with a particular focus on released from plastic packaging during mm of aerodynamic diameter was asso- protecting marginalized and underserved microwave heating (142), higher urinary ciated with 2.01% increase in CVD hos- populations. There is evidence that de- phthalate levels associated with fast food pitalizations for those with diabetes clines in air pollution levels and metal (145), and higher urinary bisphenol A compared with 0.94% increase among exposures have contributed to improve- levels from canned foods (146). Certain those without diabetes (162). Short-term ments in CVD development (172,173); personal care and cosmetic products, increases in air pollution exposure are benefits for diabetes development are which are a source of phthalates and also related to higher risk of stroke pending. Research is also needed to metals (e.g., skin-lightening products, mortality in patients with diabetes com- test intermediate strategies at the clinical which are high in mercury), are dispro- pared with those without (164). In an level, such as exposure screening (e.g., portionately marketed to marginalized experimental model, mice with diabe- asking about living near highways or using population subgroups (147). tes exposed to diesel exhaust particles private wells for drinking water) and Associations of Environmental Risk Factors showed increased cardiovascular sus- recommendations to test air or water, With Diabetes Incidence, Prevalence, and ceptibility compared with mice without reduce known sources of exposure (e.g., Outcomes. In 2011, the National Toxicol- diabetes (165). In natural experiments in minimize packaged foods, avoid heating ogy Program at the National Institute of human populations, air pollution expo- food in plastic containers, and minimize Environmental Health Sciences convened sure also resulted in increased vascular use of certain cosmetic products), and an international workshop to evaluate reactivity (166) and inflammation in pa- make home interventions (e.g., install the experimental and epidemiologic ev- tients with diabetes compared with those filters for air or water contaminants) idence on the relationship of environ- without (167). In addition to air pollution, (174–176). mental chemicals with obesity, diabetes, some evidence is also available for met- and metabolic syndrome (148–150). Ev- als.In the Strong Heart Study of American Food Environment and Diabetes idencewasdeemed strongestfor arsenic, Indian adults followed since 1989–1991, The food environment can be defined as with relative risks of diabetes found to the risk of CVD associated with higher the physical presence of food that range from 1.11 to 10.05 in different exposure to arsenic and cadmium was affects a person’s diet; a person’s prox- studies (median 2.69) at high arsenic higher among participants with diabetes imity to food store locations; the distri- exposure levels. More recent systematic compared with those without diabetes bution of food stores, food service, and reviews and meta-analyses present the (168,169). In a clinical trial in patients any physical entity by which food may be growing literature examining multiple with a previous myocardial infarction (Tri- obtained; or a connected system that groups of chemicals (148,151) or specific altoAssessChelationTherapy[TACT]),the allows access to food (177). It is the groups of chemicals (152–154). Overall, beneficial effects of repeated chelation “collective physical, economic, policy the evidence supports an increased risk with disodium edetate on cardiovascular and sociocultural surroundings, oppor- of diabetes in populations exposed to outcomes were greater in patients with tunities and conditions that influence environmental chemicals including arse- diabetes (170). people’s food and beverage choices nic, persistent organic pollutants, phtha- Environmental Exposures Interventions and and nutritional status” (178). It is also lates, and possibly bisphenol. Diabetes. Few studies have evaluated the referred to as the community food en- In 11 prospective studies of air pollu- effect of population-based or clinical vironment (e.g., number, type, location, tion exposure and incident diabetes in interventions related to environmental and accessibility of food outlets such as adults, the pooled hazard ratio (HR) (95% exposures and diabetes prevention or food stores, markets, or both) and the CI) per 10 mg/m3 increment particulate control. The increased risk of diabetes in consumer-level environment (e.g., health- matter of ,2.5 mm aerodynamic diam- populations exposed to environmental ful, affordable foods in stores, markets, eter was 1.10 (1.04–1.17) (155). Other chemicals and the increased suscepti- or both), which interact to affect food reviews have reached conclusions con- bility for diabetes complications in in- choices and diet quality (179,180). Key sistent with this increased diabetes risk dividuals with diabetes exposed to air dimensions of the food environment finding (156–158). The epidemiologic pollution potentially provides an oppor- include accessibility, availability, afford- evidence is also supported by animal tunity for prevention and treatment that ability, and quality (181–184). These experiments showing that air pollution can be particularly relevant for the most factors, which define the quality of the exposure can increase susceptibility to vulnerable populations. For example, a food environment, are of particular 10 Social Determinants of Health and Diabetes Diabetes Care

Table 4—Examples of resources on SDOH available for health care organizations and health care professionals Organization Resource Centers for Disease Control and Prevention (CDC) Tools for Putting Social Determinants of Health Into Action (https://www.cdc.gov/ socialdeterminants/tools/index.htm) National Academies of Science, Engineering, and Questionsfor conducting socialand behavioraldeterminantassessmentand frequencies for Medicine assessing Adler NE, Stead WW. Patients in contextdEHR capture of social and behavioral determinants of health. N Engl J Med 2015;372:698–701. DOI: 10.1056/NEJMp1413945 National Institutes of Health (NIH) Division of The Neighborhood AtlasdFree social determinants of health data for all! Extramural Affairs Kind AJH, Buckingham W. Making neighborhood disadvantage metrics accessible: the neighborhood atlas. N Engl J Med 2018;378:2456–2458. PMCID: PMC6051533 American Academy of Family Physicians The EveryONE Project’s Neighborhood Navigator Toolkit (https://www.aafp.org/patient- care/social-determinants-of-health/everyone-project/neighborhood-navigator/ training-videos.html) American College of Physicians Addressing Social Determinants to Improve Patient Care and Promote Health Equity: An American College of Physicians Position Paper. DOI: 10.7326/M17-2441 American Medical Association Podcast: Social determinants of health: What they are and what they aren’t (https://www .ama-assn.org/delivering-care/patient-support-advocacy/social-determinants-health- what-they-are-what-they-arent) Nonprofit services 211: A service of the United Way that continuously identifies links for all “211” health and human services referral services in the U.S. HealthLeads: A nonprofit offering tools, training and resources for integrating SDOH into accountable care Aunt Bertha: A service that provides links to hundreds of programs serving every U.S. zip code. Free basic use.

importance in marginalized communities, diabetes rates (185). Haynes-Maslow and examined neighborhood, social, and which may have poor access to super- Leone (186) similarly found availability physical environments and T2DM in markets and healthy foods but abundant of full-service restaurants to be associ- 3,700 African Americans through the access to fast-food outlets and energy- ated with lower prevalence of diabetes in Jackson Heart Study and found higher dense foods and are often dispropor- adults and availability of fast-food restau- density of unfavorable food stores was tionately impacted by physical hazards rants generally to be associated with associated with a 34% higher T2DM in- (e.g., vacant houses, traffic, and crime) higher diabetes prevalence. However, cidence after adjusting for individual- (78). At their root, differences in the the study reported variability in associa- level risk factors. In a longitudinal food environment can be caused by tions among numerous food environment employee cohort, Herrick et al. (190) government policies and incentives, characteristics based on county composi- found that living in a zip code with higher and the legacy of such policies as red- tion (low poverty/low minority, low poverty/ supermarket density was associated lining and segregation. medium minority, high poverty/low mi- with a reduction in T2DM risk, while Associations of Food Environment With nority), highlighting complexities in un- zip codes with a higher percentage of Diabetes Incidence, Prevalence, and derstanding patterns among variables of poverty and zip codes with higher walk- Outcomes county socioeconomic status, demo- ability scores were both associated with FoodAccessandAvailability.Cross-sectional graphics, food availability, and diabetes higher diabetes risk. Christine et al. (191) studies have shown associations between prevalence (186). reported long-term exposure to residen- food access, availability, geographic char- Several observational, longitudinal tial environments that offer resources to acteristics, and T2DM prevalence. Ahern studies report neighborhood resources support healthy diets and PA was asso- et al. (185) examined 3,128 counties in general, and access and availability of ciated with a lower incidence of T2DM, across the U.S. for food access (assessed the food environment in particular, as although results varied by measurement as percent of households with no car associated with diabetes prevalence and method. living more than one mile from a grocery incidence (187). A systematic review by Studies have also examined both food store) and food availability (assessed as den Braver et al. (188) found availability and PA environments in combination and number of fast-food restaurants, full-service of fast-food outlets and convenience diabetes risk. Meyer et al. (192) com- restaurants, grocery stores, convenience stores to be associated with higher bined measures of neighborhood food stores, and per capita sales in dollars from T2DM risk/prevalence and perceived and PA environments and weight-related local farms made directly to consumers). healthfulness of the food environment outcomes (N 5 14,379) of the Coronary Higher access to food was associated with to be associated with lower diabetes risk/ Artery Risk Development in Young Adults lower T2DM rates in metro and nonmetro prevalence, but no association was found (CARDIA) study, examining population counties, and higher availability of full- between density of grocery stores and density–specific (less than vs. greater servicerestaurants andgrocery storesand T2DM risk/prevalence. Heterogeneity than 1,750 people per square kilometer) lower availability of fast-food and conve- across the studies prevented the conduct clustersof neighborhood indicators: road nience stores were associated with lower of meta-analyses. Gabreab et al. (189) connectivity, parks and PA facilities, care.diabetesjournals.org Hill-Briggs and Associates 11

Table 5—SDOH and diabetes research recommendations Research recommendation 1 ConsensusisneededaroundlanguageandmetricsassociatedwithSDOHand Establish consensus core SDOH definitions and metrics diabetes care that move beyond health care and capture the impact of social advantage and disadvantage in population settings. Clarity and consistency in measurement, evaluation, and reporting of progress will allow for appropriate planning of interventions, allocation of resources, and analysis of impact in achieving equity goals. Research recommendation 2 Examinations of potential differences in pathways or impacts of SDOH based Examine specificities in SDOH pathways and impacts among on characteristics including diabetes type or diagnostic category (e.g., different populations with diabetes T1DM vs. T2DM, gestational diabetes mellitus, prediabetes), age group (e.g., children and youth, adults, older adults), and different SES (wealthy vs. middle class vs. poor) are needed. In addition, complexities of SDOH pathways and impacts for different racial/ethnic groups, based on historical drivers and policies, warrant elucidation to inform intervention and mitigation strategies. Research recommendation 3 Multisector partnerships, comprising academic institutions, government Prioritize a next generation of research that targets SDOH as sectors (e.g., housing, education, justice), and public health entities are the root cause of diabetes inequities requiredin ordertodesignandtestobservationalandinterventionstudies to better understand and intervene on SDOH as root causes of diabetes disparities. Priorities need to move from compensatory to the next- generation of research that will be larger in scope, addressing foundational causes of disparities (e.g., policy, systems change), and tested over time across sectors. Complex studies, examining the interactive effects of multifaceted systems that influence SDOH, will also transform andmove translational efforts toward large-scale solutionsthat promote equity for all populations and mitigate the influence of SDOH on diabetes outcomes. Research recommendation 4 For clinical research programs, dissemination and implementation methods Use dissemination and implementation science to ensure SDOH will shorten the translation gap from discovery to impact of evidence- considerations are embedded within diabetes research and based interventions by addressing the complexity of integrating and evaluation studies adapting evidence-based practices to real-world community and clinical settings. This will assure all populations benefit from the billions of U.S. tax dollars spent on research to prevent diabetes and to improve diabetes population health. Research studies must also consider the potential influence of either positive or negative SDOH (e.g., wealth or economic security vs. poverty, food security vs. insecurity, stable vs. unstable housing) on intervention appropriateness and outcomes, on study recruitment and participation, and on study outcomes and conclusions. Research recommendation 5 Training on SDOH and their influence on diabetes prevention and treatment Train researchers in methodologies and experimental techniques is needed. Training priorities include interdisciplinary science, multisector for multisector and next generation SDOH intervention studies collaboration research approaches, and methods to advance root cause research on SDOH. Additionally, increasing diversity among research workforces, and fostering educational experiences encompassing multisector partners will develop a workforce that is congruent with promoting diabetes health equity.

and food stores/restaurants. In lower– built environments with glycemic con- unhealthy foods would lead to differ- population density areas, higher food and trol in adults with diabetes ascertained ences in purchasing patterns and result- PA resource diversity relative to other from the New York City A1C Registry ing diets and that those differences clusters was significantly associated from 2007 to 2013. Individuals who lived would be more prominent for individuals with higher diet quality (192). In higher– continuously in the most advantaged of lower SES. In a longitudinal study, they population density areas, a cluster with residential areas, including greater ratio examined food affordability and neigh- relatively more natural food/specialty of healthy food outlets to unhealthy borhood price of healthier food relative stores, fewer convenience stores, and food outlets and residential walkability, to unhealthy food and its association more PA resources was associated with achieved increasedglycemic control and with T2DM and insulin resistance. Higher higher diet quality. Neighborhood clus- took less time to achieve glycemic con- prices of healthy foods relative to un- ters were inconsistently associated trol compared with the individuals who healthy foods were found to be associated with BMI or insulin resistance and lived continuously in the least advan- with lower odds of having a high-quality not associated with fast-food consump- taged residential areas (193). diet; however, there was no association tion, or walking, biking, or running (192). Food Affordability. Kern et al. (194) note with diabetes incidence or prevalence Tabaei et al. (193) examined associations that it is reasonable to expect that large (194). More studies are needed in this of residential socioeconomic, food, and differences in price between healthy and area. 12 Social Determinants of Health and Diabetes Diabetes Care

Food Insecurity. Food insecurity is defined Food Environment Interventions and Results of these neighborhood invest- as not having adequate quantity and Diabetes ments on measured BMI, blood pressure, quality of food at all times for all house- Three studies reported food bank and HbA1c, and HDL cholesterol will be forth- hold members to have an active, healthy pantry interventions with food inse- coming. In sum, food environment fac- life (195,196). Approximately 20% of di- cure clients with T2DM (196,220,221). tors of food unavailability, inaccessibility, abetes patients report household food Seligman et al. (196) conducted a pilot and insecurity each demonstrate asso- insecurity (197), and food insecurity is a program in Texas, California, and Ohio ciations with worse diabetes risk and risk factor for poor diabetes manage- with a pre/post design, encompassing outcomes, and interventions including di- ment (196). Researchers have investi- provision of diabetes-appropriate food, abetes-targeted food and self-management gated several pathways through which blood glucose monitoring, self-management care at food banks and pantries and food insecurity may worsen T2DM out- support, and primary care referrals. The increasing grocery store presence in comes (198–200). First, in the nutritional study resulted in improvements in HbA1c, low-income neighborhoods are few, but pathway, food insecurity is associated fruit and vegetable consumption, self- collectively they demonstrate the potential with lower diet quality (201), which is in efficacy and medication adherence. In a to impact diabetes risk, clinical outcomes, turn associated with higher HbA1c. Food randomized controlled trial of the inter- and psychosocial outcomes. insecurity incentivizes more affordable, vention, Seligman et al. (220) found im- energy-dense foods that can directly provements in nutritional consumption, Health Care and Diabetes raise serum glucose (e.g., refined carbo- food security, and distress but no clinical Health care as a SDOH includes access, hydrates, processed snacks and sweets, changes. Palar et al. (221) found reduction affordability, and quality of care factors. sugar-sweetened beverages, etc.) and in BMI but not HbA and better nutri- 1c In the U.S., these factors are highly cor- may lead to greater insulin resistance tional and psychosocial outcomes. related with race/ethnicity, SES, and place/ (202,203). Conversely, low or inconsis- Studies have examined effect of su- geographic region (19). tent food availability can increase risk of permarket gain or loss on T2DM out- hypoglycemia. Second, via a compensa- comes. A study conducted within the Associations of Health Care With Diabetes tory pathway, behavioral strategies nec- setting of the Kaiser Permanente North- Incidence, Prevalence, and Outcomes essary to cope with the immediate ern California Diabetes Registry linked Access. In population-based studies, hav- problem of food insecurity can inadver- clinical measures to metrics from a geo- ing health insurance is the strongest tently undermine T2DM management. graphic information system based on partic- predictor of whether adults with diabe- For example, financial resources that ipants’ residential addresses (115,222,223). tes have access to diabetes screenings might otherwise have been used for Results over 4 years of tracking super- and care (227). Uninsured adults in the medications or diabetes care supplies market change in low-income neighbor- U.S. population have a higher likelihood are diverted to meet dietary needs hoods showed that relative to no change of having undiagnosed diabetes than (197,204–206). Third, through the psy- in supermarket presence, supermarket adults with insurance (228). Compared chological pathway, the state of food loss was associated with worse HbA1c with insured adults with diabetes, the insecurity, in which meeting basic needs trajectories, especially among those with uninsured have 60% fewer office visits is outside an individual’s control, under- highest HbA1c. Supermarket gain in neigh- with a physician, are prescribed 52% fewer mines self-efficacy and increases depres- borhoods was associated with marginally medications, and have 168% more emer- sive symptoms and diabetes distress better HbA1c outcomes, but only for those gency department visits (229). Liese et al. (207–210). Several studies have reported a with near-normal HbA1c baseline values (230) found that, among adolescents and relationship between food insecurity and (223). In a natural experiment design, the young adults with T1DM or T2DM, com- adverse diabetes outcomes (211,212), Pittsburgh Hill/Homewood Study on Eat- pared with having private insurance, hav- and a review by Barnard et al. (213) has ing, Shopping, and Health (PHRESH) tested ing state or federal health insurance was suggested that food insecurity among the effects of adding a supermarket, associated with higher HbA1c values by patients with and at high risk for along with other neighborhood invest- 0.68%, and having no insurance was as- T2DM may be particularly toxic because, ments, on cardiometabolic risk factors sociated with higher HbA1c by 1.34%. in addition to issues of accessing sufficient among a randomly selected cohort of Having insurance has also been found calories overall, the dietary quality of the residents from two low-income, urban, to attenuate associations of financial foods eaten is even more important than and predominately African American barriers with higher HbA1c (231). for the general population. Several cross- matched neighborhoods (222,224). Re- Geographic access to adult and pedi- sectional studies report a relationship sults for the intervention neighborhood atric endocrinologists varies substantially between food insecurity and T2DM di- (receiving the supermarket) showed im- by state and county in the U.S (232), with abetes outcomes (214–216), including proved perceived access to healthy food disparities in access in many of the geo- poor metabolic control (217,218), expe- (225), and the prevalence of diabetes graphic regions with highest diabetes rience of severe hypoglycemia in low- increased less in the neighborhood with prevalence and socioeconomic disadvan- income and low-education samples (218), the supermarket than in the comparison tage (232,233). Similarly, factors that lower diabetes self-management behav- neighborhood. Since the initiation of the increase odds of having a diabetes self- ioral adherence and worse glycemic con- supermarket, many other investments management educationprogram ina geo- trol (219), and increased outpatient visits including greenspace, housing, and com- graphic area include a higher percentage but not increased emergency department/ mercial spaces have been implemented of the population with at least a high inpatient visits (95,212). in the intervention neighborhood (226). school education, a higher percentage of care.diabetesjournals.org Hill-Briggs and Associates 13

insured individuals, and a lower rate of Quality. Having insurance is the stron- populations with diabetes have been unemployment (234). DeVoe et al. (235) gest single predictor of whether adults diabetes-focused in content and goals found that among adults with diabetes, with diabetes are likely to meet indi- and have utilized structured curricula having both insurance and a usual source vidual quality measures of diabetes care (254); however, one series of studies of care, rather than one or the other, (242). Sociodemographic disparities in reported use of a standardized, all-con- conferred the greatest odds of receiving at care quality are well documented in dition CHW intervention and found mod- least minimum diabetes health care. Be- national reports and recommendations est gain in diabetes outcomes along with ing uninsured and without a usual source (2) and appear to remain consistent over additional health benefits (257,258). of care was associated with three to five time (243). In a U.S. population-based Organizational Interventions. Systematic times lower odds of adults receiving an study of achievement of a composite reviews report improvements in quality HbA1c screen, blood pressure check, or diabetes treatment goal from 2005 to of diabetes care among racial/ethnic mi- access to urgent care when needed (235). 2016, data from 2013 to 2016 showed norities resulting from quality improve- Among adolescents and young adults with that non-Hispanic Blacks had lower odds ment employing health information diabetes who had state or federal health of achieving a composite diabetes quality technology (i.e., patient registries in the insurance, not having any usual source of measure than non-Hispanic Whites (ad- electronic health record, computerized provider (primary care or diabetes spe- justed OR 0.57, 95% CI 0.39–0.83), and decisionsupportfor providers, reminders, cialist) was associated with higher HbA1c women had lower odds than men (ad- centralized outreach for diabetes pa- than having a usual source of provider, justed OR 0.60, 95% CI 0.45–0.80), with tients overdue for specific services) and HbA1c was similar whether in primary no improvement in diabetes treatment (245,259,260). There is also evidence care or specialist care (230). gaps from prior time periods (2005–2008 of effectiveness of self-management in- Affordability. On average, health care and 2009–2012), especially for minori- terventions delivered directly to under- costs of people with diabetes are 2.3 ties, women, and younger adults (227). served patients with diabetes when times those of people without diabetes Within insured settings, disparities have interventions are designed to overcome (229). Approximately 14% to 20% of been reported among Blacks as compared barriers. For example, the Centers for adults with diabetes report reducing with Whitesdin measures including di- Medicare & Medicaid Services (CMS)- or delaying medications due to cost lated eye exam taken; LDL test taken; LDL, sponsored National Diabetes Prevention (236–238). Among adults with diabetes blood pressure,orHbA1c control;and statin Program (DPP) Medicaid demonstration who are prescribed insulin, rates may therapy (244–246). A study of 21 VA fa- found CDC-recognized DPP lifestyle change be .25% (236,239). Cost-related or cost- cilities found Blacks with diabetes were programs were effective in achieving per- reducing nonadherence (CRN) is associ- more likely than Whites with diabetes to formance measures among Medicaid re- ated with income, insured status, and receive care at lower-performing facilities cipients in Maryland and Oregon, and type of insurance. Adults with diabetes overall, which explained some racial differ- additional strategies (i.e., transportation with an annual household income of ences in diabetes quality measures (246). assistance and child care) facilitated the ,$50,000 are more likely to engage in highretentionreportedoverthe12months CRN than their counterparts with in- Health Care Interventions and Diabetes of DPP visits (261). In a series of studies, a come $$50,000, and uninsured adults Community Health Workers. Several sys- problem-based self-management training with diabetes are more likely to engage in tematic reviews have concluded that addressing multiple life barriers to care CRN than those with insurance (236). community health worker (CHW) inter- in low-income and minority populations Within a diabetes clinic population of ventions using trained lay workforces was adapted for low literacy and prevalent adults with T1DM or T2DM prescribed are effective for multiple outcomes in diabetes-related functional limitations insulin, odds of CRN were three times underserved African American and His- (e.g., low vision, physical disability, and higher for those with Medicaid or no panic adults with T2DM and comorbid mildcognitiveimpairment)thatimpede insurance compared with those with conditions (247–250). CHWs have been self-management education (73,262). The Medicare (239). Piette et al. (240) found integrated into care delivery (251,252) approach has proven effective in improv- differences based on health system with reimbursement in some states (253). ingclinicaloutcomes(HbA1c,bloodpressure), model. Compared with VA patients with Roles of CHWs include patient naviga- self-care behaviors, and self-management diabetes, risk of CRN was found to be tion, appointment scheduling, visit atten- knowledge and problem-solving skills in almost three times higher for privately dance, patient education, home-based low-income, racial/ethnic minority, and ru- insured patients and four to eight times monitoring, assessment of social needs ral populations (76,263,264). higher for patients with Medicare, Med- and connection with social services, social Policy. Studies have examined the impact icaid, or no health insurance (240). support, and advocacy (252,254). Re- of the Affordable Care Act (ACA) on Higher financial stress, financial insecu- ported outcomes include better diabetes insurance coverage and health care ac- rity, and financial barriers are associ- knowledge and self-care behaviors, in- cess for patients with diabetes (265). ated with likelihood of CRN (231,238). creasedqualityoflife,reducedemergency Analyses of NHIS data from 2009 and People with CRN experience poorer di- visits and hospitalizations, reduced costs, 2016 found an increase nationwide of abetes management, higher HbA1c,and and modest improvements in glycemic 770,000 more adults with diabetes aged decreased functional status (231,240). control (247–250,255), using home-based 18 to 64 years with health insurance Deaths have been reported from in- or integrated health team delivery coverage in 2016, with a significant in- sulin CRN among youth and adults with models (252,256). A majority of the crease in coverage seen among Whites, T1DM (241). CHW interventions designed for adult Blacks, and Hispanics, people with family 14 Social Determinants of Health and Diabetes Diabetes Care

income ,$35,000, and people across relationshipsaswellastheirperceptions exposure, those in the highest quartile of educational attainment strata (less than of those relationships. Categories include exposure to everyday racism had a 31% high school and more than high school) emotional support, tangible support, in- increased risk of diabetes (HR 1.31, 95% (266). Among people with diabetes in the formational support, and companionship CI 1.20–1.42), and women with the high- lowest income strata, the proportion of (282–285). Social support is theorized to est exposure to lifetime racism had a 16% income spent on health costs decreased operate by either buffering the effects increased risk (HR 1.16, 95% CI 1.05– significantly from 6.3% to 4.8% (266). of poor health or by directly impacting 1.27); both associations were mediated Other studies found increased access to health (285,286). by BMI (298,299). Further work is needed care, diabetes management, and health to understand the multiple ways that the status among people with diabetes in Associations of Social Context With Diabetes social environment influences inequities Medicaid expansion states as compared Incidence, Prevalence, and Outcomes in diabetes outcomes. with their counterparts in non–Medicaid A systematic review by Florˆ et al. (287) Social Context Interventions and Diabetes expansion states (267); increased rates concluded that social capital was posi- Outcomes of diabetes detection and diagnosis tively associated with diabetes control To our knowledge, there is no empirical among Medicaid patients with undiag- among different populations, indepen- research on social capital or social co- nosed diabetes in states with Medicaid dent of the quality or quantity of social hesion interventions and impact on di- expansion (268); and reduction in cost- capital. However, the few studies avail- abetes outcomes, but a body of literature related medication nonadherence rates able and variations among populations has examined effects of social support. and uninsured rates among people with and measures limit the ability to draw The systematic review by Strom and diabetes following ACA (269). firm conclusions related to dimensions of Egede (284) of 18 observational studies social capital and whether the associa- of adults with T2DM found that higher tion is the same at the individual or Social Context and Diabetes levels of social support were associated Several multidimensional factors shape neighborhood level (272,288–290). Ge- with outcomes including better glycemic the social environment as a determi- breab et al. (189), using data from the control, knowledge, treatment adher- nant of health (270), including social Jackson Heart Study, examined social ence, quality of life, diagnosis awareness capital, social cohesion, and social sup- cohesion, measured as trust in neigh- and acceptance, and stress reduction port (28,29). Social capital is defined as bors, shared values with neighbors, will- (284). Lack of social support has been the features of social structures that ingness to help neighbors, and extent to linked with increased mortality and di- serve as resources for collective action which neighbors get along. The study abetes-related complications in T2DM (e.g., interpersonal trust, reciprocity revealed higher neighborhood social co- (291,295). Strom and Egede’s review of norms, and mutual aid) (271–273). Bond- hesion was associated with a 22% lower 16 social support intervention studies ing social capital refers to trusting and incidence of T2DM (189). Studies dem- demonstrated improved diabetes-related co-operative relations between mem- onstrating the relationship between so- outcomes (clinical, psychosocial, and/or bers of a network who see themselves cialsupportanddiabetes have associated self-management behavior change) in as being similar in terms of their shared increased social support with better gly- adults with T2DM, and improvements in social identity; by contrast, bridging so- cemic control and improved quality of life clinical outcomes (HbA , blood pressure, cial capital refers to aspects of respect (291–295), while lack of social support 1c lipids) appeared to be unrelated to the and mutuality between people who do not has been associated with increased mor- source or delivery (i.e., peer support, cou- share social identities (e.g., differing by tality and diabetes-related complications ples/spouse, or nurse manager). race/ethnicity, social class, age) (274–276). (291). With regard to preferencesdin a Racism, discrimination, and inclusion ver- A number of studies suggest social study conducted before the coronavirus susexclusionaremacro-levelsocialcapital cohesion, social capital, and social sup- disease 2019 pandemicdSarkar et al. factors that impact health (28). port may influencedor be influenced (300) found that, compared with White Social cohesion refers to the extent of bydracism and discrimination (296). adults with diabetes, Hispanics with di- connectednessand solidarityamong groups Racism interacts with other social enti- abetes preferred telephone-based and in a community (271,277) and has two di- ties, creating a set of dynamic, interde- group support (including promotoras), mensions: reduction of inequalities and pendent components that reinforce each while African Americans demonstrated patterns of social exclusion of population other, sustaining racial inequities and more variability in their preferences (i.e., subgroups from full participation in so- promoting both institutional- and individual- telephone, group, internet). Reliance on ciety (278) and strengthening of social re- level discrimination across various sectors support from family and community lationships and interactions (279–281). Social of society impacting diabetes incidence tended to be higher in minority popula- cohesion actions facilitate the goal of keep- (296,297). For example, Whitaker et al. tions, while Whites relied more on media ing the society united, not only through so- (298) documented associations of major and health care professionals (300). cial relations, community ties, and intergroup and everyday discrimination experien- harmony but also through reducing bias ces with incident diabetes among a di- and discrimination toward economically dis- verse sample of 5,310 middle-aged to LINKAGES ACROSS HEALTH CARE advantaged groups within a society, such as older adults from the Multi-Ethnic Study AND COMMUNITY SECTORS TO women and ethnic minorities (28). of Atherosclerosis. The Black Women’s ADDRESS SDOH Social support describes experiences in Health Study found that, when compared International and U.S. national commit- individuals’ formal and informal personal with women in the lowest quartile of tees have convened to provide guidance care.diabetesjournals.org Hill-Briggs and Associates 15

on SDOH intervention approaches. These evidence of feasibility (305). In addition, effective intervention to compensate for expert committee recommendations are Table 4 displays publicly available resour- centuries of legal racial discrimination in not specific to any disease; rather, they ces and tools to aid providers in address- educational access and quality, a next- are applicable to all conditions and pop- ing individual patients’ social needs. generation intervention might target ulations of health inequity. Table 3 dis- the education sector and implement plays recommendations from the WHO delivery of high-quality early education Commission on Social Determinants of DISCUSSION to all within both the public and private Health (27), the National Academies of There is SDOH evidence supporting as- school systems and with equitable ed- Science, Engineering, and Medicine (In- sociations of SES, neighborhood and ucational funding for sociodemographic stitute of Medicine, NASEM) Committee physical environment, food environ- populations. Similarly, while partner- on the Recommended Social and Be- ment, health care, and social context ships to bring bags of healthy groceries havioral Domains and Measures for Elec- with diabetes-related outcomes. Inequi- to low-income families living in food tronic Health Records (80), the NASEM ties in living and working conditions and deserts are important to compensate Committee on Educating Health Profes- the environments in which people reside for food deserts, a next-generation ap- sionals to Address the Social Determi- have a direct impact on biological and proach might target historical redlining nants of Health (301), and the NASEM behavioral outcomes associated with di- and zoning policies that are the root Committee on Integrating Social Needs abetes prevention and control (12,48). cause of absence of supermarkets and Care into the Delivery of Health Care to Life-course exposure based on the length freshfood markets in minorityand lower- Improve the Nation’s Health (5). of time one spends living in resource- income neighborhoods (312–314). The WHO recommendations are unique deprived environmentsddefined by pov- The review has limitations. First, the in their emphasis on root-cause, multi- erty, lack of quality education, or lack of undertaking was designed to summarize sector interventions designed to remove health caredsignificantly impacts dispar- literature on the range of SDOH identi- the SDOH as a barrier to health equity. ities in diabetes risk, diagnosis, and out- fied as having impact on diabetes out- The NASEM recommendations are based comes (12,48,306). Although the review comes. As such, this article describes in the health care sector and, collectively, reports SDOH intervention studies for findings from systematic reviews and focus on integration of SDOH into the aspects of housing, built and food envi- meta-analyses as well as more recent health care mission, operations, and fi- ronment, and health care, there appears published studies on the named SDOH; it nancial model. Accountable care organ- to be relatively limited U.S.-based re- was not designed as a primary systematic izations, value-based purchasing, and search examining impact on diabetes of review of all published research on the shared savings programs could be in- interventions designed to target educa- topic. Second are limitations of the re- tentionally designed to support and in- tion, income, occupation, toxic environ- search itself, including wide variability in centivize health care systems to address mental exposures, social cohesion, and measures and definitions used in studies patients’ health-related social needs as a social capital. within an SDOH area, making it more strategy to improve health outcomes (5). In the U.S., integrating social context difficult to describe outcomes for an The Accountable Health Communities is into health care delivery has become a pri- SDOH area in a consistent or uniform one current CMS demonstration project ority strategy (5–8). A clinical context manner or to report quantitative out- examining impact on health care costs of alone, however, is too narrow to accom- comes derived from meta-analyses. three models for health care response to modate systemic SDOH influences. Struc- Third, this review was U.S.-focused; con- SDOHthroughlinkageswithcommunity tural and legal interventions are needed clusions from SDOH research in other services: awareness (screening for social to address root causes driving SDOH countries, which in some instances may needs within the health care setting and (27,307). Similarly, additional emphasis utilize more standardly defined SDOH patient referral to services using an in- is neededona nextgenerationofresearch variables (e.g., occupation) are not part ventory ofavailable local communityserv- that prioritizes interventions impacting of this initial review. Finally, the many ices), assistance (screening, referral, plus the root causes of diabetes inequities, complexitiesofSDOHandtheirpotentially navigation to enable access to and use of rather than compensatory interventions different pathways and impacts on pop- communityservices),andalignment(screen- assisting the individual to adapt to in- ulationsarebeyondthescopeofthisinitial ing, referral, community service navigation, equities(18,308). For example, in the U.S., review and require attention to specificity plus partner alignment using a “backbone” proficient literacy and resulting health in designs of future SDOH research in organization for capacity building, data literacy are disproportionately low in mar- diabetes. sharing among community and health ginalized populations and communities Recommendations for SDOH research care partners, and scaling of services) (42), with historical sociopolitical root in diabetes resulting from this SDOH re- (302). Many health care systems are uti- causes. U.S. antiliteracy laws for Blacks, view are described in Table 5 and include lizing electronic medical records and whichprohibitedBlacksfrombeingtaught establishing consensus SDOH definitions health information exchanges to capture to read or write, persisted until the 1930s and metrics, designing studies to exam- SDOH data and commercially available in some states (309,310), and laws pro- ine specificities based on populations, SDOH algorithms to identify patients at hibiting African Americans from attending prioritizing next-generation interven- social risk and trigger service referrals public and private schools Whites at- tions, embedding SDOH context within (303). NASEM provided assessment ques- tended continued until 1954 and 1976, re- dissemination and implementation sci- tions to capture SDOH domains and spectively (311). Although adapting health ence in diabetes, and training research- frequencies for assessment (304) with materials for low-literacy suitability is an ers in methodological techniques for 16 Social Determinants of Health and Diabetes Diabetes Care

future SDOH intervention studies. By ad- Society scientific statement. J Clin Endocrinol Committee on National Health Promotion and dressing these critical elements, there is Metab 2012;97:E1579–E1639 Disease Prevention Objectives for 2020. Phase I potential for progress to be realized in 2. Centers for Medicare & Medicaid Services, report,2008. Accessed 25 October 2020. Avail- Office of Minority Health. The CMS Equity Plan able from http://www.healthypeople.gov/sites/ achieving greater health equity in diabe- for Improving Quality in Medicare, September default/files/PhaseI_0.pdf tes and across health outcomes that are 2015. Accessed 25 October 2020. Available from 17. Dankwa-Mullan I, Rhee KB, Williams K, et al. socially determined. https://www.cms.gov/About-CMS/Agency-Information/ The science of eliminating health disparities: OMH/OMH_Dwnld-CMS_EquityPlanforMedicare_ summary and analysis of the NIH summit recom- 090615.pdf mendations. Am J Public Health 2010;100(Suppl. Acknowledgments. The authors express appre- 3. U.S. Department of Health and Human Serv- 1):S12–S18 ciation to Malaika I. Hill and Mindy Saraco of the ices, Office of the Secretary, Office of the As- 18. Thornton PL, Kumanyika SK, Gregg EW, et al. American Diabetes Association; Elizabeth A. Vrany, sistantSecretaryforPlanningandEvaluation,and New research directions on disparities in obesity and Johns Hopkins University School of Medicine; and Office of Minority Health. HHS Action Plan to type 2 diabetes. Ann N Y Acad Sci 2020;1461:5–24 Shelly Johnson, Washington University in St. Louis, Reduce Racial and Ethnic Health Disparities: 19. Institute of Medicine Committee on Under- for providing technical assistance for this review. Implementation Progress Report 2011-2014. standing Eliminating Racial Ethnic Disparities in Funding. F.H.-B. is supported in part by the Johns Washington, DC, Office of the Assistant Secretary Health Care. Unequal Treatment: Confronting Hopkins Institute for Clinical and Translational for Planning and Evaluation, 2015 Racial and Ethnic Disparities in Health Care. Research (ICTR), which is funded in part by grant 4. Chin MH. Creating the business case for achieving Smedley BD, Stith AY, Nelson AR, Eds. Wash- UL1TR003098 from the National Center for Ad- health equity. J Gen Intern Med 2016;31:792–796 ington, DC, National Academies Press, 2003 vancing Translational Sciences (NCATS), a compo- 5. National Academies of Sciences, Engineering, 20. Braveman P, Arkin E, Orleans T, Proctor D, nent of the National Institutes of Health (NIH) and and Medicine. Integrating Social Care Into the Plough A. What Is Health Equity? And What NIH Roadmap for Medical Research. F.H.-B. is also Delivery of Health Care: Moving Upstream to Difference Does a Definition Make? Princeton, supported inpart by NIH National Heart, Lung, and Improve the Nation’s Health. Washington, DC, NJ, Robert Wood Johnson Foundation, 2017 Blood Institute (NHLBI) grant T32HL07180. D.H.-J. National Academies Press, 2019 21. U.S. Department of Health and Human Serv- is supported in part by NIH National Institute 6. Byhoff E, Kangovi S, Berkowitz SA, et al.; ices. Healthy People 2020: An Opportunity to of Diabetes and Digestive and Kidney Diseases Society of General Internal Medicine. A Society Address the Societal Determinants of Health in (NIDDK) grant P30DK092950. M.H.C. is supported of General Internal Medicine position state- the United States. Secretary’s Advisory Commit- in part by NIH and NIDDK grant P30DK092949. ment on the internists’ role in social determi- tee on National Health Promotion and Disease T.L.G.-W. is supported in part by NHLBI grant nants of health. J Gen Intern Med 2020;35: Prevention Objectives for 2020, 26 July 2010. R01HL131531. A.N.-A. is supported in part by Na- 2721–2727 Accessed 25 October 2020. Available from tional Institute of Environmental Health Sciences 7. Council on Community Pediatrics; Duffee JH, http://www.healthypeople.gov/2010/hp2020/ grants P42ES010349 and P30ES009089. S.A.B. is Kuo A, Gitterman BA. Poverty and child health in advisory/SocietalDeterminantsHealth.htm supported in part by NIDDK grant K23DK109200. the United States. Pediatrics 2016;137:e20160339 22. Marmot M, Allen JJ. Social determinants of The findings and conclusion in this report are 8. Daniel H, Bornstein SS, Kane GC; Health and health equity. Am J Public Health 2014; those of the authors and do not necessarily rep- Public Policy Committee of the American College 104(Suppl. 4):S517–S519 resent the official position of the Johns Hopkins of Physicians. Addressing social determinants to 23. U.S. Department of Health and Human Serv- ICTR, NCATS, NIH, NIDDK, or any other institution improve patient care and promote health equity: ices, Office of Disease Prevention and Health mentioned in the article. an American College of Physicians position pa- Promotion. Foundation health measures: dispar- Duality of Interests. M.H.C. reports being co- per. Ann Intern Med 2018;168:577–578 ities. Accessed 25 October 2020. Available from coordinator of the “Bridging the Gap: Reducing 9. Peek ME, Vela MB, Chin MH. Practical lessons https://www.healthypeople.gov/2020/about/ Disparities in Diabetes Care National Program for teaching about race and racism: successfully foundation-health-measures/Disparities Office,” supported by the Merck Foundation, leading free, frank, and fearless discussions. Acad 24. World Health Organization. Health equity. a consultant to the Patient-Centered Outcomes Med. 1 September 2020 [Epub ahead of print]. Accessed 25 October 2020. Available from Group, and a member of the Bristol-Myers Squibb DOI: 10.1097/ACM.0000000000003710 https://www.who.int/topics/health_equity/en/ Company Health Equity Advisory Board. S.A.B. 10. Vela M, Blackman D, Burnet D, et al. Ra- 25. County Health Rankings & Roadmaps. Mea- received personal fees for service on an advisory cialized violence and health care’s call to action, 6 sures & data sources, 2019. Accessed 2 Februrary 2020. board about prioritizing food insecurity research June 2020. KevinMD.com. Accessed 7 June 2020. Available from https://www.countyhealthrankings topics for the Aspen Institute. T.L.G.-W. received Availablefromhttps://www.kevinmd.com/blog/ .org/explore-health-rankings/measures-data-sources personal fees for service on an advisory board 2020/06/racialized-violence-and-health-cares- 26. World Health Organization. About social about prioritizing food insecurity research topics call-to-action.html determinants of health, 2020. Accessed 7 Febru- for the Aspen Institute. No other potential conflicts 11. Chin MH, King PT, Jones RG, et al. Lessons for ary 2020. Available from https://www.who.int/ of interests relevant to this article were reported. achieving health equity comparing Aotearoa/ social_determinants/sdh_definition/en/ Author Contributions. F.H.-B. researched data New Zealand and the United States. Health Policy 27. Commission on the Social Determinants of and wrote the manuscript. N.E.A. contributed to 2018;122:837–853 Health. Closing the gap in a generation: health writing and reviewing/editing the manuscript. 12. Haire-Joshu D, Hill-Briggs F. The next gen- equity through action on the social determinants S.A.B.researcheddataandcontributedtowriting eration of diabetes translation: a path to health of health. Final report of the Commission on Social and reviewing/editing the manuscript. M.H.C. equity. Annu Rev Public Health 2019;40:391–410 Determinants of Health. Geneva, World Health Orga- contributed to writing and reviewing/editing the 13. Golden SH, Maruthur N, Mathioudakis N, et al. nization, 2008. Accessed 25 October 2020. Available manuscript. T.L.G.-W. researched data and con- The case for diabetes population health improve- from https://www.who.int/social_determinants/ tributed to writing and reviewing/editing the ment: evidence-based programming for population final_report/csdh_finalreport_2008.pdf manuscript. A.N.-A. researched data and con- outcomes in diabetes. Curr Diab Rep 2017;17:51 28. Solar O, Irwin A. A conceptual framework for tributed to writing the manuscript. P.L.T. re- 14. Hill-Briggs F. 2018 Health Care & Education action on the social determinants of health. Social searched data and contributed to writing and Presidential Address: the American Diabetes Determinants of Health Discussion Paper 2 (Policy reviewing/editing the manuscript. D.H.-J. re- Association in the era of health care transfor- and Practice). Geneva, World Health Organiza- searched data and contributed to writing and mation. Diabetes Care 2019;42:352–358 tion, 2010. Accessed 25 October 2020. Available reviewing/editing the manuscript. 15. Hill JO, Galloway JM, Goley A, et al. Scientific from https://www.who.int/sdhconference/ statement: socioecological determinants of pre- resources/ConceptualframeworkforactiononSDH_ diabetes and type 2 diabetes. Diabetes Care eng.pdf References 2013;36:2430–2439 29. U.S. Department of Health and Human Serv- 1. Golden SH, Brown A, Cauley JA, et al. Health 16. U.S. Department of Health and Human Serv- ices, Office of Disease Prevention and Health Pro- disparities in endocrine disorders: biological, ices. Section IV: Advisory Committee findings and motion. Healthy people 2020: social determinants clinical, and nonclinical factorsdan Endocrine recommendations. In The Secretary’s Advisory of health. Accessed 25 October 2020. Available care.diabetesjournals.org Hill-Briggs and Associates 17

from https://www.healthypeople.gov/2020/topics- 46. Agardh E, Allebeck P, Hallqvist J, Moradi T, for Disease Control and Prevention, US Depart- objectives/topic/social-determinants-of-health Sidorchuk A. Type 2 diabetes incidence and ment of Health and Human Services, 2018. Accessed 30. Centers for Disease Control and Prevention. socio-economic position: a systematic review 25 October 2020. Available from https://www.cdc About social determinants of health (SDOH). and meta-analysis. Int J Epidemiol 2011;40:804– .gov/diabetes/pdfs/library/diabetesreportcard2017- Accessed 14 August 2020. Available from https:// 818 508.pdf www.cdc.gov/socialdeterminants/about.html. 47. Brown AF, Ettner SL, Piette J, et al. Socio- 62. Borrell LN, Dallo FJ, White K. Education and 31. Remington PL, Catlin BB, Gennuso KP. The economic position and health among persons diabetes in a racially and ethnically diverse pop- County Health Rankings: rationale and methods. with diabetes mellitus: a conceptual framework ulation. Am J Public Health 2006;96:1637–1642 Popul Health Metr 2015;13:11 and review of the literature. Epidemiol Rev 2004; 63. Secrest AM, Costacou T, Gutelius B, Miller 32. County Health Rankings & Roadmaps. 26:63–77 RG, Songer TJ, Orchard TJ. Association of socio- County Health Rankings Model, 2014. Accessed 48. Braveman PA, Cubbin C, Egerter S, Williams economic status with mortality in type 1 di- 25 October 2020. Available from https:// DR, Pamuk E. Socioeconomic disparities in health abetes: the Pittsburgh Epidemiology of Diabetes www.countyhealthrankings.org/resources/county- inthe UnitedStates:whatthe patternstellus. Am Complications Study. Ann Epidemiol 2011;21: health-rankings-model J Public Health 2010;100(Suppl. 1):S186–S196 367–373 33. Artiga S, Hinton E. Beyond health care: the 49. Gaskin DJ, Thorpe RJ Jr, McGinty EE, et al. 64. Marciano L, Camerini AL, Schulz PJ. The role role of social determinants in promoting health Disparities in diabetes: the nexus of race, pov- of health literacy in diabetes knowledge, self- and health equity. Issue Brief, May 2018. Kaiser erty, and place. Am J Public Health 2014;104: care, and glycemic control: a meta-analysis. J Gen Family Foundation. Accessed 25 October 2020. 2147–2155 Intern Med 2019;34:1007–1017 Available from http://files.kff.org/attachment/ 50. Beckles GL, Chou CF. Disparities in the prev- 65. Ferrie JE, Virtanen M, Jokela M, et al.; IPD- – issue-brief-beyond-health-care alence of diagnosed diabetes United States, Work Consortium. Job insecurity and risk of 34. Saegert SC, Adler NA, Bullock HE, Cauce AM, 1999-2002 and 2011-2014. MMWR Morb Mortal diabetes: a meta-analysis of individual partici- – Liu WM, Wyche KF. Report of the American Wkly Rep 2016;65:1265 1269 pant data. CMAJ 2016;188:E447–E455 Psychological Association Task Force on Socio- 51. Drewnowski A, Rehm CD, Moudon AV, 66. Varanka-Ruuska T, Rautio N, Lehtiniemi H, economic Status. Washington, DC, American Arterburn D. The geography of diabetes by census et al. The association of unemployment with Psychological Association, 2007 tract in a large sample of insured adults in King glucose metabolism: a systematic review and 35. DuttonDB, Levine S.Socioeconomicstatus County, Washington, 2005-2006. Prev Chronic Dis meta-analysis. Int J Public Health 2018;63:435– and health: overview, methodological cri- 2014;11:E125 446 tique, and reformulation. In Pathways to 52. Kolak M, Abraham G, Talen MR. Mapping 67. Gan Y, Yang C, Tong X, et al. Shift work and Health: The Role of Social Factors.Bunker census tract clusters of type 2 diabetes in a diabetes mellitus: a meta-analysis of observa- primary care population. Prev Chronic Dis 2019; P, Gomby DS, Kehrer BH, Eds. Menlo Park, tional studies. Occup Environ Med 2015;72:72– 16:E59 CA, The Henry J. Kaiser Family Foundation, 78 53. Schmittdiel JA, Dyer WT, Marshall CJ, Bivins 1989, pp. 29–69 68. KivimakiM,VirtanenM, KawachiI,et al.Long R. Using neighborhood-level census data to pre- 36. Adler NE, Boyce T, Chesney MA, et al. working hours, socioeconomic status, and the dict diabetes progression in patients with Socioeconomic status and health. The chal- riskofincidenttype2diabetes:ameta-analysisof laboratory-defined prediabetes. Perm J 2018; lenge of the gradient. Am Psychol 1994;49: published and unpublished data from 222 120 22:18–096 15–24 individuals. Lancet Diabetes Endocrinol 2015;3: 54. Saydah S, Lochner K. Socioeconomic status 37. Adler NE, Newman K. Socioeconomic dis- 27–34 and risk of diabetes-related mortality in the U.S. parities in health: pathways and policies. Health 69. Gallup-Sharecare. The Face of Diabetes in Public Health Rep 2010;125:377–388 Aff (Millwood) 2002;21:60–76 the United States, State of American Well-being, 55. Scott A, Chambers D, Goyder E, O’Cathain A. 38. Braveman PA, Cubbin C, Egerter S, et al. 2017. Accessed 25 October 2020. Available from Socioeconomicinequalitiesinmortality,morbidity Socioeconomicstatusin healthresearch:onesize https://wellbeingindex.sharecare.com/wp-content/ and diabetes management for adults with type 1 fi – does not t all. JAMA 2005;294:2879 2888 diabetes: a systematic review. PLoS One 2017;12: uploads/2017/12/The-Face-of-Diabetes-in-the-United- 39. Shavers VL. Measurement of socioeconomic e0177210 States-2017.pdf status in health disparities research. J Natl Med 56. Bijlsma-Rutte A, Rutters F, Elders PJM, Bot 70. Witters D, Liu D. Diabetes rate greatest Assoc 2007;99:1013–1023 among transportation workers. Gallup, Inc., SDM, Nijpels G. Socio-economic status and HbA1c 40. Dotson VM, Kitner-Triolo MH, Evans MK, in type 2 diabetes: a systematic review and meta- 2017. Accessed 25 October 2020. Available Zonderman AB. Effects of race and socioeco- analysis. Diabetes Metab Res Rev 2018;34:e3008 from https://news.gallup.com/poll/214097/ fl nomic status on the relative in uence of edu- 57. Lindner LME, Rathmann W, Rosenbauer J. diabetes-rate-greatest-among-transportation- cation and literacy on cognitive functioning. J Int Inequalities in glycaemic control, hypoglycaemia workers.aspx – Neuropsychol Soc 2009;15:580 589 and diabetic ketoacidosis according to socio- 71. Hill-Briggs F, Schumann KP, Dike O. Five-step 41. Sisco S, Gross AL, Shih RA, et al. The role of economic status and area-level deprivation in methodology for evaluation and adaptation of early-life educational quality and literacy in ex- type 1 diabetes mellitus: a systematic review. print patient health information to meet the , plaining racial disparities in cognition in late life. J Diabet Med 2018;35:12–32 5th grade readability criterion. Med Care 2012; Gerontol B Psychol Sci Soc Sci 2015;70:557–567 58. Borschuk AP, Everhart RS. Health disparities 50:294–301 42. Rudd RE. Health literacy skills of U.S. adults. among youth with type 1 diabetes: a systematic 72. Cavanaugh KL. Health literacy in diabetes Am J Health Behav 2007;31(Suppl. 1):S8–S18 review of the current literature. Fam Syst Health care: explanation, evidence and equipment. Di- 43. Kutner M, Greenberg E, Jin Y, Boyle B, Hsu Y, 2015;33:297–313 abetes Manag (Lond) 2011;1:191–199 Dunleavy E (Eds.). Literacy in Everyday Life: 59. Walker RJ, Garacci E, Palatnik A, Ozieh MN, 73. Hill-Briggs F, Renosky R, Lazo M, et al. De- Results From the 2003 National Assessment of Egede LE. The longitudinal influence of social velopment and pilot evaluation of literacy-adapted Adult Literacy (NCES 2007–480). Washington, determinants of health on glycemic control in diabetes and CVD education in urban, diabetic DC, NationalCenter for EducationStatistics,2007 elderly adults with diabetes. Diabetes Care 2020; African Americans. J Gen Intern Med 2008;23: 44. Kutner M, Greenburg E, Jin Y, Paulsen C. The 43:759–766 1491–1494 Health Literacy of America’s Adults: Results From 60. Centers for Disease Control and Prevention. 74. White RO, Eden S, Wallston KA, et al. Health the 2003 National Assessment of Adult Literacy National Diabetes Statistics Report, 2017. At- communication, self-care, and treatment satis- (NCES 2006-483). Washington, DC, National Cen- lanta, GA, Centers for Disease Control and Pre- faction among low-income diabetes patients in a ter for Education Statistics, 2006 vention, US Department of Health and Human public health setting. Patient Educ Couns 2015; 45. Gornick ME. 2. Measuring the effects of Services, 2017. Accessed 25 October 2020. Avail- 98:144–149 socioeconomicstatusonhealthcare. InGuidance able from https://dev.diabetes.org/sites/default/ 75. Kim SH, Lee A. Health-literacy-sensitive di- for the National Healthcare Disparities Report. files/2019-06/cdc-statistics-report-2017.pdf abetes self-management interventions: a sys- Swift EK, Ed. Washington, DC, National Acade- 61. Centers for Disease Contol and Prevention. tematic review and meta-analysis. Worldviews mies Press, 2002 Diabetes Report Card 2017. Atlanta, GA, Centers Evid Based Nurs 2016;13:324–333 18 Social Determinants of Health and Diabetes Diabetes Care

76. Hill-Briggs F, Lazo M, Peyrot M, et al. Effect of 93. Gelberg L, Andersen RM, Leake BD. The relationship with diabetes diagnosis and man- problem-solving-based diabetes self-management Behavioral Model for Vulnerable Populations: agement among homeless persons in New York training on diabetes control in a low income patient application to medical care use and outcomes for City. Am J Epidemiol 2019;188:1120–1129 sample. J Gen Intern Med 2011;26:972–978 homelesspeople.HealthServ Res2000;34:1273– 108. Keene DE, Henry M, Gormley C, Ndumele C. 77. Cavanaugh K, Wallston KA, Gebretsadik T, 1302 ‘Then I found housing and everything changed’: et al. Addressing literacy and numeracy to im- 94. Bernstein RS, Meurer LN, Plumb EJ, Jackson transitions to rent-assisted housing and diabe- prove diabetes care: two randomized controlled JL. Diabetes and hypertension prevalence in tes self-management. Cityscape 2018;20:107– trials. Diabetes Care 2009;32:2149–2155 homeless adults in the United States: a system- 118 78. Diez Roux AV, Mair C. Neighborhoods and atic review and meta-analysis. Am J Public Health 109. Centers for Disease Control and Preven- health. Ann N Y Acad Sci 2010;1186:125–145 2015;105:e46–e60 tion. Built environment assessment tool manual, 79. Tung EL, Cagney KA, Peek ME, Chin MH. 95. Berkowitz SA, Meigs JB, DeWalt D, et al. 2019. Accessed 9 March 2020. Available from Spatial context and health inequity: reconfigur- Material need insecurities, control of diabetes https://www.cdc.gov/nccdphp/dnpao/state- ing race, place, and poverty. J Urban Health 2017; mellitus, and use of health care resources: local-programs/built-environment-assessment/ 94:757–763 results of the Measuring Economic Insecurity index.htm 80. Institute of Medicine. Capturing Social and in Diabetes study. JAMA Intern Med 2015;175: 110. Drewnowski A, Buszkiewicz J, Aggarwal A, Behavioral Domains and Measures in Electronic 257–265 Rose C, Gupta S, Bradshaw A. Obesity and the Health Records: Phase 2. Washington, DC, The 96. Burgard SA, Seefeldt KS, Zelner S. Housing built environment: a reappraisal. Obesity (Silver National Academies Press, 2014 instability and health: findings from the Michigan Spring) 2020;28:22–30 81. Baggett TP, Berkowitz SA, Fung V, Gaeta JM. Recession and RecoveryStudy. Soc Sci Med 2012; 111. Martin A, Ogilvie D, Suhrcke M. Evaluating Prevalence of housing problems among commu- 75:2215–2224 causal relationships between urban built envi- nity health center patients. JAMA 2018;319:717– 97. Charkhchi P, Fazeli Dehkordy S, Carlos RC. ronment characteristics and obesity: a method- 719 Housing and food insecurity, care access, and ological review of observational studies. Int J 82. Berkowitz SA, Kalkhoran S, Edwards ST, health status among the chronically ill: an anal- Behav Nutr Phys Act 2014;11:142 Essien UR, Baggett TP. Unstable housing and ysis of the Behavioral Risk Factor Surveillance 112. Mayne SL, Auchincloss AH, Michael YL. diabetes-related emergency department visits System. J Gen Intern Med 2018;33:644–650 Impact of policy and built environment changes and hospitalization: a nationally representative 98. Stahre M, VanEenwyk J, Siegel P, Njai R. on obesity-related outcomes: a systematic re- study of safety-net clinic patients. Diabetes Care Housing insecurity and the association with view of naturally occurring experiments. Obes 2018;41:933–939 health outcomes and unhealthy behaviors, Rev 2015;16:362–375 83. Kushel MB, Gupta R, Gee L, Haas JS. Housing Washington State, 2011. Prev Chronic Dis 113. Chandrabose M, Rachele JN, Gunn L, instability and food insecurity as barriers to 2015;12:140511 et al. Built environmentandcardio-metabolic health care among low-income Americans. J 99. Gibson M, Petticrew M, Bambra C, Sowden health: systematic review and meta-analysis Gen Intern Med 2006;21:71–77 AJ, Wright KE, Whitehead M. Housing and health of longitudinal studies. Obes Rev 2019;20: 84. Vijayaraghavan M, Jacobs EA, Seligman H, inequalities: a synthesis of systematic reviews of 41–54 Fernandez A. The association between housing interventions aimed at different pathways link- 114. Smalls BL, Gregory CM, Zoller JS, Egede LE. instability, food insecurity, and diabetes self- ing housing and health. Health Place 2011;17: Assessing the relationship between neighbor- efficacy in low-income adults. J Health Care 175–184 hood factors and diabetes related health out- Poor Underserved 2011;22:1279–1291 100. Shaw M. Housing and public health. Annu comes and self-care behaviors. BMC Health Serv 85. Children’s Health Watch. Our survey. Ac- Rev Public Health 2004;25:397–418 Res 2015;15:445 cessed 25 October 2020. Available from https:// 101. Brooks LK, Kalyanaraman N, Malek R. Diabetes 115. Bilal U, Auchincloss AH, Diez-Roux AV. childrenshealthwatch.org/methods/our-survey/ care for patients experiencing homelessness: Neighborhood environments and diabetes risk 86. U.S. Congress. H.R.558 - Stewart B. McKinney beyond metformin and sulfonylureas. Am J Med and control. Curr Diab Rep 2018;18:62 homeless assistance act, 1987. Accessed 25 2019;132:408–412 116. Leal C, Chaix B. The influence of geographic October 2020. Available from https://www 102. Ludwig J, Sanbonmatsu L, Gennetian L, life environments on cardiometabolic risk fac- .congress.gov/bill/100th-congress/house-bill/558 et al. Neighborhoods, obesity, and diabetesda tors: a systematic review, a methodological 87. Henry M, Watt R, Mahathey A, Ouellette J, randomized social experiment. N Engl J Med assessment and a research agenda. Obes Rev Sitler A; Abt Associates. The 2019 Annual Home- 2011;365:1509–1519 2011;12:217–230 less Assessment Report (AHAR) to Congress; Part 103. Orr L, Feins JD, Jacob R, et al. Moving to 117. Twohig-Bennett C, Jones A. The health 1: Point-in-Time Estimates of Homelessness, Opportunity for Fair Housing Demonstration benefits of the great outdoors: a systematic January 2020. U.S. Department of Housing and Program: Interim Impacts Evaluation. Washing- review and meta-analysis of greenspace expo- Urban Development. Accessed 25 October 2020. ton, DC, U.S. Department of Housing and Urban sureandhealthoutcomes.EnvironRes2018;166: Available from https://www.hudexchange.info/ Development Office of Policy Development and 628–637 resource/5948/2019-ahar-part-1-pit-estimates- Research, 2003. Accessed 25 October 2020. 118. Durand CP, Andalib M, Dunton GF, Wolch J, of-homelessness-in-the-us/ Available from https://www.huduser.gov/Publications/ Pentz MA. A systematic review of built environ- 88. Cutts DB, Meyers AF, Black MM, et al. US pdf/MTOFullReport.pdf ment factors related to physical activity and Housing insecurity and the health of very young 104. Sanbonmatsu L, Marvokov J, Porter N, et al. obesity risk: implications for smart growth urban children. Am J Public Health 2011;101:1508–1514 The long-term effects of Moving to Opportunity planning. Obes Rev 2011;12:e173–e182 89. Frederick TJ, Chwalek M, Hughes J, on adult health and economic self-sufficiency. 119. Petticrew M, Cummins S, Ferrell C, et al. Karabanow J, Kidd S. How stable is stable? De- Cityscape2012;14:109–136 Natural experiments: an underused tool for fining and measuring housing stability. J Com- 105. Baxter AJ, Tweed EJ, Katikireddi SV, public health? Public Health 2005;119:751–757 munity Psychol 2014;42:964–979 Thomson H. Effects of Housing First approaches 120. MacMillan F, George ES, Feng X, et al. Do 90. Meltzer R, Schwartz A. Housing affordability on health and well-being of adults who are natural experiments of changes in neighborhood and health: evidence from New York City. Hous homeless or at risk of homelessness: systematic built environment impact physical activity and Policy Debate 2016;26:80–104 review and meta-analysis of randomised con- diet? a systematic review. Int J Environ Res Public 91. Keene DE, Guo M, Murillo S. “That wasn’t trolled trials. J Epidemiol Community Health Health 2018;15:217 really a place to worry about diabetes”: housing 2019;73:379–387 121. Benton JS, Anderson J, Hunter RF, French access and diabetes self-management among 106. Tsai J, Gelberg L, Rosenheck RA. Changes in DP. The effect of changing the built environment low-income adults. Soc Sci Med 2018;197:71–77 physical health after supported housing: results on physical activity: a quantitative review of the 92. Quensell ML, Taira DA, Seto TB, Braun KL, from the Collaborative Initiative to End Chronic risk of bias in natural experiments. Int J Behav Sentell TL. “I need my own place to get better”: Homelessness. J Gen Intern Med 2019;34:1703– Nutr Phys Act 2016;13:107 patient perspectives on the role of housing in 1708 122. Amuda AT, Berkowitz SA. Diabetes and the potentially preventable hospitalizations. J Health 107. Lim S, Miller-Archie SA, Singh TP, Wu WY, built environment: evidence and policies. Curr Care Poor Underserved 2017;28:784–797 Walters SC, Gould LH. Supportive housing and its Diab Rep 2019;19:35 care.diabetesjournals.org Hill-Briggs and Associates 19

123. Landrigan PJ, Fuller R, Acosta NJR, et al. The 139. Powers M, Yracheta J, Harvey D, et al. a systematic review of the human epidemiolog- Lancet Commission on pollution and health. Arsenic in groundwater in private wells in rural ical evidence. Environ Int 2019;132:104768 Lancet 2018;391:462–512 North Dakota and South Dakota: water quality 155. Yang BY, Fan S, Thiering E, et al. Ambient air 124. Casey JA, Morello-Frosch R, Mennitt DJ, assessment for an intervention trial. Environ Res pollution and diabetes: a systematic review and Fristrup K, Ogburn EL, James P. Race/ethnicity, 2019;168:41–47 meta-analysis. Environ Res 2020;180:108817 socioeconomic status, residential segregation, 140. Fox MA, Nachman KE, Anderson B, Lam J, 156. Eze IC, Hemkens LG, Bucher HC, et al. As- and spatial variation in noise exposure in the Resnick B. Meeting the public health challenge of sociation between ambient air pollution and di- contiguous United States. Environ Health Per- protecting private wells: proceedings and rec- abetes mellitus in Europe and North America: spect 2017;125:077017 ommendations from an expert panel workshop. systematic review and meta-analysis. Environ 125. Evans GW, Kantrowitz E. Socioeconomic Sci Total Environ 2016;554-555:113–118 Health Perspect 2015;123:381–389 status and health: the potential role of environ- 141. Lee D, Murphy HM. Private wells and rural 157. Janghorbani M, Momeni F, Mansourian M. mental risk exposure. Annu Rev Public Health health: groundwater contaminants of emerging Systematic review and metaanalysis of air pol- 2002;23:303–331 concern. Curr Environ Health Rep 2020;7:129– lution exposure and risk of diabetes. Eur J 126. Hajat A, Hsia C, O’Neill MS. Socioeconomic 139 Epidemiol 2014;29:231–242 disparities and air pollution exposure: a global 142. Groh KJ, Backhaus T, Carney-Almroth B, 158. Liu F, Chen G, Huo W, et al. Associations review. Curr Environ Health Rep 2015;2:440– et al. Overview of known plastic packaging- between long-term exposure to ambient air 450 associated chemicals and their hazards. Sci Total pollution and risk of type 2 diabetes mellitus: 127. Miao Q, Chen D, Buzzelli M, Aronson KJ. Environ 2019;651:3253–3268 a systematic review and meta-analysis. Environ Environmental equity research: review with fo- 143. Nguyen VK, Kahana A, Heidt J, et al. A Pollut 2019;252(Pt B):1235–1245 cus on outdoor air pollution research methods comprehensive analysis of racial disparities in 159. Liu C, Bai Y, Xu X, et al. Exaggerated effects and analytic tools. Arch Environ Occup Health chemical biomarker concentrations in United of particulate matter air pollution in genetic 2015;70:47–55 States women, 1999-2014. Environ Int 2020; type II diabetes mellitus. Part Fibre Toxicol 128. Mohai P, Lantz PM, Morenoff J, House JS, 137:105496 2014;11:27 Mero RP. Racial and socioeconomic disparities in 144. Varshavsky JR, Morello-Frosch R, Woodruff 160. Liu C, Fonken LK, Wang A, et al. Central IKKb residential proximity to polluting industrial fa- TJ, Zota AR. Dietary sources of cumulative phtha- inhibition prevents air pollution mediated pe- cilities: evidence from the Americans’ Changing lates exposure among the U.S. general popula- ripheral inflammation and exaggeration of type II Lives Study. Am J Public Health 2009;99(Suppl. 3): tion in NHANES2005-2014.EnvironInt 2018;115: diabetes. Part Fibre Toxicol 2014;11:53 S649–S656 417–429 161. Liu C, Ying Z, Harkema J, Sun Q, Rajagopalan 129. Perez AC, Grafton B, Mohai P, Hardin R, 145. Zota AR, Phillips CA, Mitro SD. Recent fast S. Epidemiological and experimental links be- Hintzen K, Orvis S. Evolution of the environ- food consumption and bisphenol a and phtha- tween air pollution and type 2 diabetes. Toxicol mental justice movement: activism, formaliza- lates exposures among the U.S. population in Pathol 2013;41:361–373 tion and differentiation. Environ Res Lett 2015; NHANES, 2003-2010. Environ Health Perspect 162. Zanobetti A, Schwartz J. Are diabetics more 10:105002 2016;124:1521–1528 susceptible to the health effects of airborne 130. Tessum CW, Apte JS, Goodkind AL, et al. 146. Hartle JC, Navas-Acien A, Lawrence RS. The particles? Am J Respir Crit Care Med 2001; Inequity in consumption of goods and services consumption of canned food and beverages and 164:831–833 adds to racial-ethnic disparities in air pollution urinary Bisphenol A concentrations in NHANES 163. Zanobetti A, Schwartz J. Cardiovascular exposure. Proc Natl Acad Sci U S A 2019;116: 2003-2008. Environ Res 2016;150:375–382 damage by airborne particles: are diabetics 6001–6006 147. Zota AR, Shamasunder B. The environmen- more susceptible? Epidemiology 2002;13:588– 131. Apelberg BJ, Buckley TJ, White RH. Socio- tal injustice of beauty: framing chemical ex- 592 economic and racial disparities in cancer risk posures from beauty products as a health 164. Zeka A, Zanobetti A, Schwartz J. Individual- from air toxics in Maryland. Environ Health Per- disparities concern. Am J Obstet Gynecol 2017; level modifiers of the effects of particulate matter spect 2005;113:693–699 217:418.e1–418.e6 on daily mortality. Am J Epidemiol 2006;163: 132. Hilpert M, Johnson M, Kioumourtzoglou 148. Kuo CC, Moon K, Thayer KA, Navas-Acien A. 849–859 MA, et al. A new approach for inferring traffic- Environmental chemicals and type 2 diabetes: an 165. Nemmar A, Subramaniyan D, Yasin J, Ali BH. related air pollution: use of radar-calibrated updated systematic review of the epidemiologic Impact of experimental type 1 diabetes mellitus crowd-sourced traffic data. Environ Int 2019; evidence. Curr Diab Rep 2013;13:831–849 on systemic and coagulation vulnerability in mice 127:142–159 149. Maull EA, Ahsan H, Edwards J, et al. Eval- acutely exposed to diesel exhaust particles. Part 133. Jones MR, Diez-Roux AV, Hajat A, et al. uation of the association between arsenic and Fibre Toxicol 2013;10:14 Race/ethnicity, residential segregation, and ex- diabetes: a National Toxicology Program work- 166. O’Neill MS, Veves A, Zanobetti A, et al. posure to ambient air pollution: the Multi-Ethnic shop review. Environ Health Perspect 2012;120: Diabetesenhances vulnerability to particulate air Study of Atherosclerosis (MESA). Am J Public 1658–1670 pollution-associated impairment in vascular re- Health 2014;104:2130–2137 150. Thayer KA, Heindel JJ, Bucher JR, Gallo MA. activity and endothelial function. Circulation 134. Bellavia A, Zota AR, Valeri L, James-Todd T. Role of environmental chemicals in diabetes and 2005;111:2913–2920 Multiple mediators approach to study environ- obesity: a National Toxicology Program work- 167. O’Neill MS, Veves A, Sarnat JA, et al. Air mental chemicals as determinants of health shop review. Environ Health Perspect 2012;120: pollution and inflammation in type 2 diabetes: disparities. Environ Epidemiol 2018;2:e015 779–789 a mechanism for susceptibility. Occup Environ 135. Dodd-Butera T, Beaman M, Brash M. En- 151. SongY,ChouEL,BaeckerA,etal.Endocrine- Med 2007;64:373–379 vironmental health equity: a concept analysis. disrupting chemicals, risk of type 2 diabetes, and 168. Moon KA, Guallar E, Umans JG, et al. Annu Rev Nurs Res 2019;38:183–202 diabetes-related metabolic traits: a systematic Association between exposure to low to mod- 136. Gee GC, Payne-Sturges DC. Environmental review and meta-analysis. J Diabetes 2016;8: erate arsenic levels and incident cardiovascular health disparities: a framework integrating psy- 516–532 disease. A prospective cohort study. Ann Intern chosocial and environmental concepts. Environ 152. Evangelou E, Ntritsos G, Chondrogiorgi M, Med 2013;159:649–659 Health Perspect 2004;112:1645–1653 et al. Exposure to pesticides and diabetes: a sys- 169. Tellez-Plaza M, Guallar E, Howard BV, et al. 137. Krometis LA, Gohlke J, Kolivras K, Satterwhite tematic review and meta-analysis. Environ Int Cadmium exposure and incident cardiovascular E, Marmagas SW, Marr LC. Environmental health 2016;91:60–68 disease. Epidemiology 2013;24:421–429 disparities in the Central Appalachian region of 153. Jaacks LM, Staimez LR. Association of per- 170. Lamas GA, Goertz C, Boineau R, et al.; TACT the United States. Rev Environ Health 2017;32: sistent organic pollutants and non-persistent Investigators. Effect of disodium EDTA chelation 253–266 pesticides with diabetes and diabetes-related regimen on cardiovascular events in patients 138. Lewis J, Hoover J, MacKenzie D. Mining and health outcomes in Asia: a systematic review. with previous myocardial infarction: the TACT environmental health disparities in Native Amer- Environ Int 2015;76:57–70 randomized trial. JAMA 2013;309:1241–1250 ican communities. Curr Environ Health Rep 2017; 154. Radke EG, Galizia A, Thayer KA, Cooper GS. 171. Been JV, Nurmatov UB, Cox B, Nawrot TS, 4:130–141 Phthalate exposure and metabolic effects: van Schayck CP, Sheikh A. Effect of smoke-free 20 Social Determinants of Health and Diabetes Diabetes Care

legislation on perinatal and child health: a sys- foods and incidence of type 2 diabetes mellitus: 202. Drewnowski A, Eichelsdoerfer P. Can low- tematic review and meta-analysis. Lancet 2014; the Multi-Ethnic Study of Atherosclerosis. Arch income Americans afford a healthy diet? Nutr 383:1549–1560 Intern Med 2009;169:1698–1704 Today 2010;44:246–249 172. Peterson GCL, Hogrefe C, Corrigan AE, Neas 188. den Braver NR, Lakerveld J, Rutters F, 203. Darmon N, Drewnowski A. Contribution of LM, Mathur R, Rappold AG. Impact of reductions Schoonmade LJ, Brug J, Beulens JWJ. Built en- food prices and diet cost to socioeconomic in emissions from major source sectors on fine vironmental characteristics and diabetes: a sys- disparities in diet quality and health: a systematic particulate matter–related cardiovascular mor- tematic review and meta-analysis. BMC Med review and analysis. Nutr Rev 2015;73:643–660 tality. Environ Health Perspect 2020;128:17005 2018;16:12 204. Bomberg EM, Neuhaus J, Hake MM, 173. Ruiz-Hernandez A, Navas-Acien A, Pastor- 189. Gebreab SY, Hickson DA, Sims M, et al. Engelhard EM, Seligman HK. Food preferences Barriuso R, et al. Declining exposures to lead and Neighborhood social and physical environments and coping strategies among diabetic and non- cadmium contribute to explaining the reduction and type 2 diabetes mellitus in African Ameri- diabetic households served by US food pantries. J of cardiovascular mortality in the US population, cans: the Jackson Heart Study. Health Place 2017; Hunger Environ Nutr 2019;14:4–17 1988-2004. Int J Epidemiol 2017;46:1903–1912 43:128–137 205. Billimek J, Sorkin DH. Food insecurity, pro- 174. Flanagan SV, Braman S, Puelle R, et al. 190. Herrick CJ, Yount BW, Eyler AA. Implications cesses of care, and self-reported medication Leveraging health care communication channels of supermarket access, neighbourhood walkabil- underuse in patients with type 2 diabetes: results forenvironmentalhealthoutreachinNewJersey. ity and poverty rates for diabetes risk in an from the California Health Interview Survey. J Public Health Manag Pract 2020;26:E23–E26 employee population. Public Health Nutr 2016; Health Serv Res 2012;47:2159–2168 175. Hadley MB, Vedanthan R, Fuster V. Air 19:2040–2048 206. Herman D, Afulani P, Coleman-Jensen A, pollution and cardiovascular disease: a window 191. Christine PJ, Auchincloss AH, Bertoni AG, Harrison GG. Food insecurity and cost-related ofopportunity.NatRevCardiol2018;15:193–194 et al. Longitudinal associations between neigh- medication underuse among nonelderly adults 176. Wong KH, Durrani TS. Exposures to endo- borhood physical and social environments and in a nationally representative sample. Am J Public crine disrupting chemicals in consumer products-a incident type 2 diabetes mellitus: the Multi- Health 2015;105:e48–e59 guide for pediatricians. Curr Probl Pediatr Adolesc Ethnic Study of Atherosclerosis (MESA). JAMA 207. Silverman J, Krieger J, Kiefer M, Hebert P, Health Care 2017;47:107–118 Intern Med 2015;175:1311–1320 Robinson J, Nelson K. The relationship between 177. Centers for Disease Contol and Prevention. 192. Meyer KA, Boone-Heinonen J, Duffey KJ, food insecurity and depression, diabetes distress General food environment resources, 2010. Ac- et al. Combined measure of neighborhood food and medication adherence among low-income cessed 10 July 2019. Available from https://www and physical activity environments and weight- patients with poorly-controlled diabetes. J Gen .cdc.gov/healthyplaces/healthtopics/healthyfood/ related outcomes: the CARDIA study. Health Intern Med 2015;30:1476–1480 general.htm Place 2015;33:9–18 208. Leung CW, Epel ES, Willett WC, Rimm EB, 178. Swinburn B, Sacks G, Vandevijvere S, et al.; 193. Tabaei BP, Rundle AG, Wu WY, et al. Asso- Laraia BA. Household food insecurity is positively INFORMAS. INFORMAS (International Network ciations of residential socioeconomic, food, and associated with depression among low-income for Food and Obesity/non-communicable dis- built environments with glycemic control in per- supplemental nutrition assistance program par- eases Research, Monitoring and Action Support): sonswithdiabetes inNewYorkCityfrom 2007– ticipants and income-eligible nonparticipants. J overview and key principles. Obes Rev 2013; 2013. Am J Epidemiol 2018;187:736–745 Nutr 2015;145:622–627 14(Suppl. 1):1–12 194. Kern DM, Auchincloss AH, Stehr MF, et al. 209. Arenas DJ, Thomas A, Wang J, DeLisser HM. 179. Glanz K, Sallis JF, Saelens BE, Frank LD. Neighborhood price of healthier food relative to A systematic review and meta-analysis of de- Nutrition Environment Measures Survey in unhealthy food and its association with type 2 pression, anxiety, and sleep disorders in US stores (NEMS-S): development and evaluation. diabetes and insulin resistance: the Multi-Ethnic adults with food insecurity. J Gen Intern Med Am J Prev Med 2007;32:282–289 Study of Atherosclerosis. Prev Med 2018;106: 2019;34:2874–2882 180. U.S. Department of Agriculture, Economic 122–129 210. Kinsey EW, Dupuis R, Oberle M, Cannuscio Research Service. Food environment atlas. Ac- 195. Coleman-Jensen A, Gregory C, Singh A. CC, Hillier A. Chronic disease self-management cessed 10 July 2019. Available from https://www Household food security in the United States within the monthly benefit cycle of the Supple- .ers.usda.gov/data-products/food-environment- in2013(EconomicResearchReportNo.173).U.S. mental Nutrition Assistance Program. Public atlas/2019 Department of Agriculture, Economic Research Health Nutr 2019;22:2248–2259 181. HerforthA,AhmedS.Thefoodenvironment, Service, 2014. Accessed 25 October 2020. Avail- 211. Seligman HK, Bindman AB, Vittinghoff E, its effects on dietary consumption, and potential able from https://www.ers.usda.gov/webdocs/ Kanaya AM, Kushel MB. Food insecurity is as- for measurement within agriculture-nutrition in- publications/45265/48787_err173.pdf?v50 sociated with diabetes mellitus: results from the terventions. Food Secur 2015;7:505–520 196. Seligman HK, Lyles C, Marshall MB, et al. A National Health Examination and Nutrition Ex- 182. Lytle LA, Sokol RL. Measures of the food pilot food bank intervention featuring diabetes- amination Survey (NHANES) 1999-2002. J Gen environment: a systematic review of the field, appropriate food improved glycemic control Intern Med 2007;22:1018–1023 2007-2015. Health Place 2017;44:18–34 among clients in three states. Health Aff (Mill- 212. Berkowitz SA, Karter AJ, Corbie-Smith G, 183. McKinnon RA, Reedy J, Morrissette MA, wood) 2015;34:1956–1963 etal. Food insecurity, food “deserts,” and glycemic Lytle LA, Yaroch AL. Measures of the food envi- 197. Berkowitz SA, Seligman HK, Choudhry NK. control in patients with diabetes: a longitudinal ronment: a compilation of the literature, 1990- Treat or eat: food insecurity, cost-related med- analysis. Diabetes Care 2018;41:1188–1195 2007. Am J Prev Med 2009;36(Suppl.):S124–S133 ication underuse, and unmet needs. Am J Med 213. Barnard LS, Wexler DJ, DeWaltD, Berkowitz 184. Turner C, Aggarwal A, Walls H, et al. Con- 2014;127:303–310.e3 SA. Material need support interventions for di- cepts and critical perspectives for food envi- 198. Walker RJ, Gebregziabher M, Martin-Harris abetes prevention and control: a systematic re- ronment research: a global framework with B, Egede LE. Quantifying direct effects of social view. Curr Diab Rep 2015;15:574 implications for action in low- and middle-income determinants of health on glycemic control in 214. Seligman HK, Bolger AF, Guzman D, Lopez´ countries. Glob Food Secur 2018;18:93–101 adultswith type2 diabetes. Diabetes TechnolTher A, Bibbins-Domingo K. Exhaustion of food budg- 185. Ahern M, Brown C, Dukas S. A national 2015;17:80–87 ets at month’s end and hospital admissions for study of the association between food environ- 199. Walker RJ, Williams JS, Egede LE. Pathways hypoglycemia. Health Aff (Millwood) 2014;33: ments and county-level health outcomes. J Rural between food insecurity and glycaemic control in 116–123 Health 2011;27:367–379 individuals with type 2 diabetes. Public Health 215. Seligman HK, Davis TC, Schillinger D, Wolf 186. Haynes-Maslow L, Leone LA. Examining the Nutr 2018;21:3237–3244 MS. Food insecurity is associated with hypogly- relationship between the food environment and 200. Seligman HK, Schillinger D. Hunger and cemia and poor diabetes self-management in a adult diabetes prevalence by county economic socioeconomic disparities in chronic disease. low-income sample with diabetes. J Health Care and racial composition: an ecological study. BMC N Engl J Med 2010;363:6–9 Poor Underserved 2010;21:1227–1233 Public Health 2017;17:648 201. Orr CJ, Keyserling TC, Ammerman AS, 216. Seligman HK, Jacobs EA, Lopez´ A, Tschann J, 187. Auchincloss AH, Diez Roux AV, Mujahid MS, Berkowitz SA. Diet quality trends among adults Fernandez A. Food insecurity and glycemic con- Shen M, Bertoni AG, Carnethon MR. Neighbor- with diabetes by socioeconomic status in the trol among low-income patients with type 2 hood resources for physical activity and healthy U.S.: 1999-2014. BMC Endocr Disord 2019;19:54 diabetes. Diabetes Care 2012;35:233–238 care.diabetesjournals.org Hill-Briggs and Associates 21

217. Berkowitz SA, Baggett TP, Wexler DJ, endocrinologists in the United States, 2012. BMC review of randomized controlled trials. Curr Diab Huskey KW, Wee CC. Food insecurity and met- Health Serv Res 2015;15:541 Rep 2014;14:558 abolic control among U.S. adults with diabetes. 233. Lopez-DeFede` A, Stewart JE. Diagnosed 249. Norris SL, Chowdhury FM, Van Le K, et al. Diabetes Care 2013;36:3093–3099 diabetes prevalence and risk factor rankings, Effectiveness of community health workers in 218. Berkowitz SA, Karter AJ, Lyles CR, et al. Low by state, 2014–2016: a ring map visualization. the care of persons with diabetes. Diabet Med socioeconomic status is associated with in- Prev Chronic Dis 2019;16:E44 2006;23:544–556 creased risk for hypoglycemia in diabetes pa- 234. Rutledge SA, Masalovich S, Blacher RJ, 250. Shah M, Kaselitz E, Heisler M. The role of tients: the Diabetes Study of Northern California Saunders MM. Diabetes self-management edu- community health workers in diabetes: update (DISTANCE). J Health Care Poor Underserved cation programs in nonmetropolitan countiesd on current literature. Curr Diab Rep 2013;13: 2014;25:478–490 United States, 2016. MMWR Surveill Summ 163–171 219. Heerman WJ, Wallston KA, Osborn CY, et al. 2017;66:1–6 251. Islam N, Nadkarni SK, Zahn D, Skillman M, Food insecurity is associated with diabetes self- 235. DeVoe JE, Tillotson CJ, Wallace LS. Usual Kwon SC, Trinh-Shevrin C. Integrating community care behaviours and glycaemic control. Diabet sourceof care as a healthinsurancesubstitutefor health workers within Patient Protection and Med 2016;33:844–850 U.S. adults with diabetes? Diabetes Care 2009; Affordable Care Act implementation. J Public 220. Seligman HK, Smith M, Rosenmoss S, 32:983–989 Health Manag Pract 2015;21:42–50 Marshall MB, Waxman E. Comprehensive dia- 236. Kang H, Lobo JM, Kim S, Sohn M-W. Cost- 252. Centers for Disease Control and Prevention. betes self-management support from food related medication non-adherence among U.S. Addressing Chronic Disease Through Community banks: a randomized controlled trial. Am J Public adults with diabetes. Diabetes Res Clin Pract Health Workers: A Policy and Systems-Level Health 2018;108:1227–1234 2018;143:24–33 Approach; Second Edition. Atlanta, GA, National 221. Palar K, Napoles T, Hufstedler LL, et al. 237. Lessem SE, Pendley RP. QuickStats: per- Center for Chronic Disease Prevention and Health Comprehensive and medically appropriate food centage of adults aged $45 years who reduced Promotion, Division for Heart Disease and Stroke support is associated with improved HIV and or delayed medication to save money in the Prevention, April 2015. Accessed 25 October 2020. – diabetes health. J Urban Health 2017;94:87 99 past 12 months among those who were pre- Available from https://www.cdc.gov/dhdsp/docs/ 222. Richardson AS, Ghosh-Dastidar M, Beckman scribed medication, by diagnosed diabetes chw_brief.pdf R, et al. Can the introduction of a full-service status and agedNational Health Interview 253. National Association of Chronic Disease ’ supermarket in a food desert improve residents Survey, 2015. MMWR Morb Mortal Wkly Directors. Community programs linked to clinical – economic status and health? Ann Epidemiol 2017; Rep 2017;66:679 services community health workers: reim- – 27:771 776 238. Patel MR, Piette JD, Resnicow K, Kowalski- bursement/advocacy. Accessed 10 June 2020. 223. Zhang YT, Mujahid MS, Laraia BA, et al. Dobson T, Heisler M. Social determinants of Available from https://www.chronicdisease.org/ Association between neighborhood supermar- mpage/domain4_chw_ra health, cost-related non-adherence, and cost- ket presence and glycated hemoglobin levels 254. Egbujie BA, Delobelle PA, Levitt N, Puoane reducing behaviors among adults with diabetes: among patients with type 2 diabetes mellitus. T, Sanders D, van Wyk B. Role of community findings from the National Health Interview Am J Epidemiol 2017;185:1297–1303 health workers in type 2 diabetes mellitus self- Survey. Med Care 2016;54:796–803 224. Gary-Webb TL, Egnot NS, Nugroho A, management: a scoping review. PLoS One 2018; 239. Herkert D, Vijayakumar P, Luo J, et al. Cost- Dubowitz T, Troxel WM. Changes in perceptions 13:e0198424 related insulin underuse among patients with of neighborhood environment and cardiometa- 255. Palmas W, March D, Darakjy S, et al. Com- diabetes. JAMA Intern Med 2019;179:112–114 bolic outcomes in two predominantly African munity health worker interventions to improve 240. Piette JD, Wagner TH, Potter MB, Schillinger American neighborhoods. BMC Public Health glycemic control in people with diabetes: a sys- D. Health insurance status, cost-related medica- 2020;20:52 tematic review and meta-analysis. J Gen Intern tion underuse, and outcomes among diabetes 225. Dubowitz T, Ghosh-Dastidar M, Cohen DA, Med 2015;30:1004–1012 patients in three systems of care. Med Care 2004; et al. Diet and perceptions change with super- 256. Gary TL, Batts-Turner M, Yeh HC, et al. The – market introduction in a food desert, but not 42:102 109 effects of a nurse case manager and a community because of supermarket use. Health Aff (Mill- 241. Rosenthal E. When high prices mean need- health worker team on diabetic control, emer- – wood) 2015;34:1858–1868 less death. JAMA Intern Med 2019;179:114 115 gency department visits, and hospitalizations 226. Baird MD, Schwartz HL, Hunter GP, et al. 242. Doucette ED, Salas J, Wang J, Scherrer JF. among urban African Americans with type 2 Does large-scale neighborhood reinvestment Insurance coverage and diabetes quality indica- diabetes mellitus: a randomized controlled trial. work? Effects of public–private real estate in- tors among patients with diabetes in the US Arch Intern Med 2009;169:1788–1794 vestment on local sales prices, rental prices, and general population. Prim Care Diabetes 2017; 257. Kangovi S, Mitra N, Grande D, Huo H, Smith – crime rates. Hous Policy Debate 2020;30:164– 11:515 521 RA, Long JA. Community health worker support 190 243. Ali MK, Shah MK. Age and age-old dispar- for disadvantaged patients with multiple chronic 227. Kazemian P, Shebl FM, McCann N, ities in diabetes care persist. JAMA Intern Med diseases: a randomized clinical trial. Am J Public – Walensky RP, Wexler DJ. Evaluation of the cas- 2019;179:1386 1387 Health 2017;107:1660–1667 cade of diabetes care in the United States, 2005- 244. Brown AF, Gregg EW, Stevens MR, et al. 258. Kangovi S, Mitra N, Norton L, et al. Effect of 2016. JAMA Intern Med 2019;179:1376–1385 Race, ethnicity, socioeconomic position, and community health worker support on clinical 228. Danaei G, Friedman AB, Oza S, Murray CJL, quality of care for adults with diabetes enrolled outcomes of low-income patients across primary Ezzati M. Diabetes prevalence and diagnosis in managed care: the Translating Research Into care facilities: a randomized clinical trial. JAMA in US states: analysis of health surveys. Popul Action for Diabetes (TRIAD) study. Diabetes Care Intern Med 2018;178:1635–1643 Health Metr 2009;7:16 2005;28:2864–2870 259. Peek ME, Cargill A, Huang ES. Diabetes 229. American Diabetes Association. Economic 245. Sequist TD, Adams A, Zhang F, Ross-Degnan health disparities: a systematic review of health costs of diabetes in the U.S. in 2017. Diabetes D, Ayanian JZ. Effect of quality improvement on care interventions. Med Care Res Rev 2007; Care 2018;41:917–928 racial disparities in diabetes care. Arch Intern 64(Suppl.):101S–156S 230. Liese AD, Ma X, Reid L, et al. Health care Med 2006;166:675–681 260. Ricci-Cabello I, Ruiz-Perez´ I, Nevot-Cordero access and glycemic control in youth and young 246. Heisler M, Smith DM, Hayward RA, Krein SL, A, Rodr´ıguez-Barranco M, Sordo L, Gonçalves DC. adults with type 1 and type 2 diabetes in South Kerr EA. Racial disparities in diabetes care pro- Health care interventions to improve the quality Carolina. Pediatr Diabetes 2019;20:321–329 cesses, outcomes, and treatment intensity. Med of diabetes care in African Americans: a system- 231. Ngo-Metzger Q, Sorkin DH, Billimek J, Care 2003;41:1221–1232 atic review and meta-analysis. Diabetes Care Greenfield S, Kaplan SH. The effects of financial 247. Hunt CW, Grant JS, Appel SJ. An integrative 2013;36:760–768 pressures on adherence and glucose control review of community health advisors in type 2 261. Porterfield D, Jacobs S, Farrell K, et al. among racial/ethnically diverse patients with diabetes. J Community Health 2011;36:883–893 Evaluation of the Medicaid Coverage for the diabetes. J Gen Intern Med 2012;27:432–437 248. Little TV, Wang ML, Castro EM, Jimenez´ J, National Diabetes Prevention Program Demon- 232. Lu H, Holt JB, Cheng YJ, Zhang X, Onufrak S, Rosal MC. Community health worker interven- stration Project: Final Report. RTI International, Croft JB. Population-based geographic access to tions for Latinos with type 2 diabetes: a systematic November 2018. Accessed 25 October 2020. 22 Social Determinants of Health and Diabetes Diabetes Care

Available from https://cdn.ymaws.com/www 278. World Health Organization. Social deter- 295. ZhangX,NorrisSL,GreggEW,BecklesG.Social .chronicdisease.org/resource/resmgr/diabetes_ minants of health: evidence on social determi- support and mortality among older persons with dpp_materials/medicaid_demonstration_proje nants of health. Accessed 25 October 2020. diabetes. Diabetes Educ 2007;33:273–281 .pdf Available from https://www.who.int/social_ 296. Williams DR, Lawrence JA, Davis BA. Racism 262. Hill-Briggs F, Lazo M, Renosky R, Ewing C. determinants/themes/en/ and health: evidence and needed research. Annu Usability of a diabetes and cardiovascular disease 279. Berger-Schmitt R. Considering social cohe- Rev Public Health 2019;40:105–125 education module in an African American, di- sion in quality of life assessments: concept and 297. Reskin B. The race discrimination system. abetic sample with physical, visual, and cognitive measurement. Soc Indic Res 2002;58:403–428 Annu Rev Sociol 2012;38:17–35 impairment. Rehabil Psychol 2008;53:1–8 280. Berger-Schmitt R, Noll HH. Conceptual 298. Whitaker KM, Everson-Rose SA, Pankow JS, 263. Lilly CL, Bryant LL, Leary JM, et al.; MSHA. framework and structure of a European system et al. Experiences of discrimination and incident Evaluation of the effectiveness of a problem- of social indicator. In EuroReporting Working type 2 diabetes mellitus: the Multi-Ethnic Study solving intervention addressing barriers to Paper #9. Mannheim,Centre for Survey Research of Atherosclerosis (MESA). Am J Epidemiol 2017; cardiovascular disease prevention behaviors in and Methodology, 2000 186:445–455 3 underserved populations: Colorado, North 281. Chuang Y-C, Chuang K-Y, Yang T-H. Social 299. Bacon KL, Stuver SO, Cozier YC, Palmer JR, Carolina, West Virginia, 2009. Prev Chronic Dis cohesion matters in health. Int J Equity Health Rosenberg L, Ruiz-Narvaez´ EA. Perceived racism 2014;11:E32 2013;12:87 and incident diabetes in the Black Women’sHealth 264. Boulware LE, Ephraim PL, Hill-Briggs F, et al. 282. van Dam HA, van der Horst FG, Knoops L, Study. Diabetologia 2017;60:2221–2225 Hypertension self-management in socially dis- Ryckman RM, Crebolder HF, van den Borne BH. 300. Sarkar U, Piette JD, Gonzales R, et al. Prefer- advantaged African Americans: the Achieving Social support in diabetes: a systematic review ences for self-management support: findings from Blood Pressure Control Together (ACT) random- of controlled intervention studies. Patient Educ a survey of diabetes patients in safety-net health ized comparative effectiveness trial. J Gen Intern Couns 2005;59:1–12 – Med 2020;35:142–152 systems. Patient Educ Couns 2008;70:102 110 283. Taylor SE. Social support: a review. In The 301. National Academies of Sciences, Engineer- 265. Brown DS, Delavar A. The Affordable Care Handbook of Health Psychology. Friedman MS, Act and insurance coverage for persons with ing, and Medicine. A Framework for Educating Ed. New York, Oxford University Press, 2011, pp. Health Professionals to Address the Social Deter- diabetes in the United States. J Hosp Manag 189–214 Health Policy 2018;2:2 minants of Health. Washington, DC, The National 284. Strom JL, Egede LE. The impact of social Academies Press, 2016 266. Casagrande SS, McEwen LN, Herman WH. support on outcomes in adult patients with Changes in health insurance coverage under the 302. Alley DE, Asomugha CN, Conway PH, Sanghavi type 2 diabetes: a systematic review. Curr DM. Accountable health communitiesdaddressing Affordable Care Act: a national sample of U.S. Diab Rep 2012;12:769–781 social needs through Medicare and Medicaid. N Engl adults with diabetes, 2009 and 2016. Diabetes 285. Ford ME, Tilley BC, McDonald PE. Social Care 2018;41:956–962 J Med 2016;374:8–11 support among African-American adults with 267. Lee J, Callaghan T, Ory M, Zhao H, Bolin JN. 303. Gottlieb LM, Tirozzi KJ, Manchanda R, Burns diabetes, part 2: a review. J Natl Med Assoc The impact of Medicaid expansion on diabetes AR, Sandel MT. Moving electronic medical records 1998;90:425–432 management. Diabetes Care 2020;43:1094–1101 upstream: incorporating social determinants of 286. Thoits PA. Social support an psychological 268. KaufmanHW, Chen Z,FonsecaVA, McPhaul health. Am J Prev Med 2015;48:215–218 well-being: theoretical possibilities. In Social MJ. Surge in newly identified diabetes among 304. Adler NE, Stead WW. Patients in contextd Support: Theory, Research, and Application. medicaid patients in 2014 within Medicaid ex- EHR capture of social and behavioral determi- Sarason IG, Sarason BR, Eds. Hingram, MA, pansion states under the Affordable Care Act. nants of health. N Engl J Med 2015;372:698–701 Kluwer, 1985, pp. 53–72 Diabetes Care 2015;38:833–837 305. Giuse NB, Koonce TY, Kusnoor SV, et al. 287. FlorCR,BaldoniNR,AquinoJA,etal.Whatisˆ 269. Zhang JX, Bhaumik D, Huang ES, Meltzer Institute of Medicine measures of social and the association between social capital and di- DO. Change in insurance status and cost-related abetes mellitus? A systematic review. Diabetes behavioral determinants of health: a feasibility medication non-adherence among older U.S. – Metab Syndr 2018;12:601–605 study. Am J Prev Med 2017;52:199 206 adults with diabetes from 2010 to 2014. J Health 306. Dixon B, Pena~ M-M, Taveras EM. Lifecourse Med Econ 2018;4:7 288. Farajzadegan Z, Jafari N, Nazer S, Keyvanara M, Zamani A. Social capitalda neglected issue in approach to racial/ethnic disparities in childhood 270. TunstallH,MitchellR,GibbsJ,PlattS,Dorling obesity. Adv Nutr 2012;3:73–82 D. Is economic adversity always a killer? Disad- diabetes control: a cross-sectional survey in Iran. Health Soc Care Community 2013;21:98–103 307. Gostin LO, Hodge JG Jr, Levin DE. Legal vantaged areas with relatively low mortality rates. interventions to address us reductions in life – 289. Long JA, Field S, Armstrong K, Chang VW, J Epidemiol Community Health 2007;61:337 343 expectancy. JAMA 2020;324:1037–1038 271. Kawachi I, Berkman L. Social cohesion, social Metlay JP. Social capital and glucose control. J – 308. Thomas SB, Quinn SC, Butler J, Fryer CS, capital, and health. In Social Epidemiology. Berkman Community Health 2010;35:519 526 ~ Garza MA. Toward a fourth generation of dispar- LF, Kawachi I, Eds. New York, Oxford University 290. Mendoza-Nu´nez VM, Flores-Bello C, Correa- ~ ities research to achieve health equity. Annu Rev Press, 2000, p. 174 Munoz E, Retana-U Galde R, Ruiz-Ramos M. Re- Public Health 2011;32:399–416 272. Whitehead M, Diderichsen F. Social capital lationship between social support networks and 309. Angulo AJ (Ed.). Miseducation: A History of and health: tip-toeing through the minefield of diabetes control and its impact on the quality of Ignorance-Making in America and Abroad. Balti- evidence. Lancet 2001;358:165–166 life in older community-dwelling Mexicans. Nutr more, MD, Johns Hopkins University Press, 2016 273. Hawe P, Shiell A. Social capital and health Hosp 2016;33:1312–1316 310. Douglass F. Chapter X, Learning to Read. In promotion: a review. Soc Sci Med 2000;51:871–885 291. Ciechanowski P, Russo J, Katon WJ, et al. 274. Szreter S, Woolcock M. Health by associ- Relationship styles and mortality in patients with Life and Times of Frederick Douglass. New York, ation? Social capital, social theory, and the diabetes. Diabetes Care 2010;33:539–544 Citadel Press, 1983 politicaleconomy ofpublichealth.IntJ Epidemiol 292. Trief P, Sandberg JG, Ploutz-Snyder R, et al. 311. Bell D. Silent Covenants: Brown v. Board of fi 2004;33:650–667 Promoting couples collaboration in type 2 di- Education and the Unful lled Hopes for Racial 275. Gittell R, Vidal A. Community Organizing: abetes: the diabetes support project pilot data. Reform. New York, Oxford University Press, 2004 Building Social Capital as a Development Strategy. Fam Syst Health 2011;29:253–261 312. Rothstein E. The Color of Law: A Forgotten Thousand Oaks, CA, Sage Publications, Inc., 1998 293. Roblin DW. The potential of cellular tech- History of How Our Government Segregated Amer- 276. Portes A. Social capital: its origins and nology to mediate social networks for support of ica.NewYork,W.W.NortonandCo.Inc.,2017 applications in modern sociology. Annu Rev chronic disease self-management. J Health Com- 313. Federal Housing Authority. Underwriting Sociol 1998;24:1–24 mun 2011;16(Suppl. 1):59–76 Manual. Underwriting and Valuation Procedure 277. van Staveren IP, Pervaiz Z, Chaudhary AR. 294. Tang TS, Brown MB, Funnell MM, Anderson Under Title II of the National Housing Act. Wash- Diversity, inclusiveness and social cohesion. ISS RM. Social support, quality of life, and self- ington, DC, Federal Housing Administration, 1936 Working Paper Series/General Series. 2013. Ac- carebehaviorsamong AfricanAmericanswith 314. Eisenhauer E. In poor health: supermarket cessed 25 October 2020. Available from http:// type 2 diabetes. Diabetes Educ 2008;34:266– redlining and urban nutrition. GeoJournal 2001; hdl.handle.net/1765/50480 276 53:125–133