Overcoming Inequities in Obesity: What Don’t We Know That We Need to Know?

The Global Picture

Recognizing the obesity epidemic as a global problem is fundamental to understanding both the scope and complexity of the challenges and the type of transformative thinking that can set the tone for the next wave of our efforts (Kumanyika, 2019a; Swinburn et al., 2019). The United States is definitely not alone in the inability to curb the obesity epidemic. All countries are affected, although the United States has, and apparently will have for the coming decades, the highest prevalence of obesity compared to other high-income countries (Organization for Economic Development and Cooperation, 2017). The World Health Organization estimates that obesity worldwide has tripled since the 1970s, that 1.9 billion adults aged 18 and older, 41 million children younger than 5 years, and 340 million youth aged 5 to 19 years are overweight or have obesity (World Health Organization, 2018). The Lancet Commission on Obesity concluded that no country has successfully addressed the problem and the rise in obesity globally has been unabated. The Commission, of which I was a member, makes a strong case for developing novel and bold solutions using a syndemic approach that considers common drivers of malnutrition (both obesity and undernutrition) and climate change as major, interrelated threats to population health (Swinburn et al., 2019). Continuing to view obesity and related solutions in isolation of these other problems may become untenable.

U.S. Overall Trends and Ethnic Disparities

Efforts to combat the U.S. obesity epidemic have been ongoing in earnest for nearly two decades, informed by a substantial body of knowledge and guided by numerous programmatic and policy recommendations (Institute of Medicine, 2012, 2014). I say “in earnest” because the view of obesity as primarily a clinical or “cosmetic” problem may have delayed recognition of the epidemic when its signs first appeared (Kumanyika, 2001). Figure 1 provides a reminder of the trends over the past two decades as reported by the National Health and Nutrition Examination Survey (NHANES)—increases in obesity prevalence among adults and children with no sign of a downturn at the national level (National Center for Health Statistics, 2017). A downturn in obesity prevalence between 2003-2004 and 2013-2014 among children aged 2 to 5 years appeared to be an exception (Ogden et al., 2016), but in 2015 to 2016 prevalence in this age group in NHANES had returned to the 2003 to 2004 level (Hales, Fryar, Carroll, Freedman, & Ogden, 2018).

OPEN IN VIEWER

Figure 1.

Trends in obesity prevalence in U.S. adults and youth, 1988-1994 through 2013-2014.

Figure 2 shows the overall prevalence and disparities between non-Hispanic Black and Hispanic versus non-Hispanic White children and adults as of 2015-2016 (Hales, Carroll, Fryar, & Ogden, 2017). Disparities are reflected in prevalence of 50% or more in Black and Hispanic women compared to 38% in non-Hispanic White women, and the higher prevalence in Hispanic versus non-Hispanic White men. In youth, disparities are evident in boys, particularly for Hispanic versus other boys, as well as girls. Twenty percent to nearly 30% of Black and Hispanic children have obesity compared to fewer than 15% in non-Hispanic White children. Data for American Indians/Alaska Natives and Native Hawaiians and other Pacific Islanders are not shown, but disparities in obesity prevalence affect these populations as well (Bullock, Sheff, Moore, & Manson, 2017; National Center for Health Statistics, 2016; Nigg, Shor, Tanaka, & Hayes, 2011).

OPEN IN VIEWER

Figure 2.

Obesity prevalence in U.S. adults (age-adjusted) and youth, total and by race/ethnicity and gender, 2015-2016.

Taken together, these data reflect an overall lack of progress in stabilizing or lowering obesity prevalence. There have been promising findings of stabilizing or decreasing trends in younger children in some states and in children participating in the Supplemental Nutrition Program for Women, Infants, the Children (WIC) (Pan et al., 2019; Robert Wood Johnson Foundation, 2018). There are also encouraging reports of weight loss efficacy or effectiveness in clinical trials or community programs (Liao et al., 2016; Look AHEAD Research Group, 2014). However, such findings are not consistent and do not add up to an impact overall. Some well-designed and conducted intervention and observational child obesity studies suggest that although substantial efforts are being made, they may be inadequate in relation to the scope and nature of the problem. Dietz describes the disappointing, largely null results of two major, National Institutes of Health (NIH)–funded multicenter intervention trials with low-income Hispanic and African American children (The Girls Health Enrichment Multisite Studies [GEMS] and the Childhood Obesity Prevention and Treatment Research Consortium [COPTR] studies; Dietz, 2019). He also comments on the mixed results of the large, NIH-funded Healthy Communities Study (HCS), an observational study of progress in addressing childhood obesity nationwide (Arteaga et al., 2015), and he calls for novel approaches.

The HCS results are especially relevant to the recent natural history of childhood obesity initiatives as it relates to disparities (Kumanyika, 2018). The HCS team identified nearly 10,000 past or current community policies and programs in a sample of 130 U.S. communities. Communities assessed included some that were selected specifically because they were known to be actively engaged in child obesity prevention efforts. The policies and programs identified in the 130 communities were categorized by type and setting and analyzed for associations with children’s behavior and weight status. The study team measured children’s current weight status and collected prior weight and height data from medical records. The HCS identified many strategies that were associated with favorable changes in children’s dietary and physical activity behaviors and weight change trajectories, but apparently not among ethnic minority and low-income children. Analyses of interaction effects by region, race/ethnicity and family income showed that the clear, positive associations of policies and programs with behavior and weight outcomes applied only for White children in higher income families in the Northeast.

Thus, the sense that something is missing from current approaches is undeniable and must be a continued source of inquiry. What clues are we missing? What it is that is not adding up?

What We Know and What We Don’t Know

To say that we are not yet on the pathway to success is in no way intended to diminish the value of what we have learned and done so far. To the contrary, the fact that what we know is insufficient can be seen as an opportunity for a critical analysis of where to go next. Our current approaches may be on target in many respects. These approaches fit with well-established theoretical frameworks that call for this type of complex problem to be addressed at individual, interpersonal, community, organizational and policy and environmental levels (Stokols, 1992, 1996). They are designed to follow through on well-conceived and comprehensive recommendations for obesity prevention approaches that address these levels (Institute of Medicine, 2012; Kumanyika, Brownson, & Cheadle, 2012). What is missing may be adequacy of (a) translation—how well these recommendations are being translated into programs and policies that fit the heterogeneity and diversity of population groups and contexts, and (b) quality of implementation—the extent to which these programs are implemented in ways that achieve the desired reach, dose, and sustainability. My impression is that limitations of both types are relevant. In addition, the lack of a systems perspective may be an overarching concern with the way strategies are being applied. Although it is clear from what we already know that we need comprehensive policy, systems, and environmental (PSE) approaches that address interrelated, mutually reinforcing drivers that operate as a system (Institute of Medicine, 2012), many efforts involve separate or parallel actions rather than combinations.

Translation may be a particular issue for individually focused studies from an equity perspective. Our current approaches may be limited by insufficient knowledge of or attention to the potential influences on obesity prevention and management that affect diverse ethnic groups. This situation could result from an overreliance on the principle that efficacy studies are a gold standard and a tendency to underestimate the difficulty of achieving the same outcomes in natural settings. Even the “best-case” efficacy studies may be affected by differences in the contexts for behavior change among the highly selected participants recruited for these studies. We know that in NIH-sponsored diabetes and hypertension prevention trials, trajectories of weight loss and maintenance differ by race-ethnicity, with smaller initial weight losses among Black participants (Wingo, Carson, & Ard, 2014). This suggests that, on average, Black participants experience the interventions differently from their White counterparts. However, studies that compare weight loss only treatments with treatments that combine weight loss and sodium reduction or drug therapy (metformin) reported smaller weight losses in the combined interventions, with minimal ethnic differences (Kumanyika, Espeland, et al., 2002; West, Prewitt, Bursac, & Felix, 2008). More focused explorations of when and why ethnic differences in intervention uptake or response occur might be fruitful.

The concept that targeting interventions to high risk populations and tailoring them to the relevant circumstances of individuals and groups in these populations is well-established but is complex and not yet well understood (Kong, Tussing-Humphreys, Odoms-Young, Stolley, & Fitzgibbon, 2014; Liu et al., 2012). Culturally tailored interventions have not necessarily achieved better weight loss results than standard interventions. This may be because, in these interventions, contextual factors overwhelm potential intervention effects. Co-creating intervention designs with communities outside of the constraints of approaches “proven” in efficacy studies could be worth exploring—not only by revealing the nature and nuances of the relevant contexts but also by fostering community ownership and indicating what intervention assumptions are not met and associated implications.

Table 1 lists an array of factors that can influence effectiveness of weight loss programs differentially for higher vs. lower risk populations (Kumanyika, 2019a). These influences relate to experiences and opportunities that affect food intake, physical activity, weight control motivations or metabolic pathways related to weight loss or gain and are conditioned by race/ethnicity, language or literacy level or various stressors. We need to know how these factors converge to predispose to and perpetuate above-average obesity risks in ethnic minority populations. The table entries prompt thinking about potentially protective influences, such as cultural assets that buffer effects of obesogenic environmental contexts or factors that enable people to be healthy within otherwise adverse circumstances. The list may seem tedious to consider, especially when we think we already have solid interventions on the shelf. On the other hand, it reminds us that lives of ethnic minority populations are shaped and lived differently from those of the majority White population, and in ways that directly or indirectly affect eating and physical activity patterns and weight control. Using an equity lens to understand the context for and dynamics of obesity prevention and control encourages reflection about when approaches that differ radically from standard approaches might be needed. Without such a lens, these multiple and interrelated factors that predispose to or perpetuate obesity remain latent. Hence, rather than attributing the lack of success of an intervention to factors beyond a person’s control, providers may blame people for not adhering to their recommendations, or people may blame themselves.

OPEN IN VIEWER

Table 1.

Potential Influences on Obesity in U.S. Racial/Ethnic Minority Populations.

Variable	Health implications (obesity)
Racial/ethnic category	• Cultural food preferences • Neighborhood access (segregation; discrimination) • Targeted marketing of unhealthy foods • Mobility (freedom of movement) • Historical and ongoing trauma
Socioeconomic status; social position	• Neighborhood access (poverty) • Food purchasing power • Food insecurity • Activity patterns • Housing • Access to health care • Transportation
Language/literacy	• Access to nutrition information • Access to quality education • Better social integration
Stress	• Eating and physical activity • Embodiment • Constant need to cope • Sleep
Migration and migration stress	• Adverse circumstances prior to migration • Abrupt exposure to obesogenic environment • Loss of connections with home environment • Downward social mobility
Cultural assets and protection	• Preservation of traditional healthy behaviors • Buffering from aggressive promotion of unhealthy foods and beverages • Coping mechanisms, including faith
Structural empowerment and resilience	• Ability to benefit from new opportunities • Social capital and social support

Source. Adapted and reproduced with permission from the National Academy of Sciences, courtesy of The National Academies Press, Washington, DC. Common threads in obesity risk among racial/ethnic and migrant minority populations, In. National Academies of Sciences, Engineering, and Medicine, 2019.

Implementation quality may be especially important for PSE change interventions. The HCS data did not include measures of implementation, but the authors speculated that limited adoption and implementation capacity might have been part of the explanation for the lack of impact in high-risk populations (Kumanyika, 2018). Relevant to using a systems perspective, the intensity variable used for HCS analyses indicated that the more obesity-related policies and programs being applied in a given setting, the better the outcomes. However, I have been skeptical about the ability of PSE change interventions that only address eating and physical activity to get us where we need to be at either the population or individual level. Rather, I am advocating for a more integrative and people-oriented perspective (Kumanyika, 2017). Asking what it takes for an intervention to be successful from a people perspective (using an equity lens as mentioned above) can stimulate awareness of hidden assumptions that can then be considered and addressed. This approach would be expected to identify ways to combine obesity-related PSE approaches—synergistically—with other interventions that address important contextual factors such as social and economic resources, or various aspects of community capacity, to facilitate the uptake and receptivity to a set of programs or policies and their effectiveness and sustainability. This people-oriented approach also brings in core health education and health promotion strategies. Community capacity includes knowledge and behavioral skills of residents as well as collective agency and ability to effect changes.

Conclusion

The problem of inequities in obesity cannot be viewed separately from the overall national and global epidemics. However, within our efforts to solve the obesity epidemic overall, we must intensify our efforts to understand what it will take to address disparities and achieve equity. I wrote, in 1994,

to the extent that the need for interventions on obesity in minority populations offers a chance to develop more constructive treatment paradigms, there is an opportunity here to benefit the field of obesity research as a whole. The most pressing mandate is for solutions to address energy imbalance at the population level. It is only in the context of such solutions that effective clinical approaches can be developed for minority populations with a high obesity prevalence. (Kumanyika, 1994)

This conclusion still resonates for me today. Tackling this problem in the highest risk groups may provide the insight, mandate, and motivation to develop innovative and effective solutions that can be applied more broadly. This expands the paradigm from starting with the easiest or simplest and then moving to the more complex to include approaches that embrace the complexity from the beginning and prioritizing those most at risk. The Lancet Commission on Obesity has set the stage for this type of future-oriented thinking. I look forward to the time when reversing the obesity epidemic can be added to the list of our greatest public health achievements.