Exposure to obesogenic endocrine disrupting chemicals and obesity among youth of Latino or Hispanic origin in the United States and Latin America: A lifecourse perspective

Summary Following a 2019 workshop led by the Center for Global Health Studies at the Fogarty International Center on the topic of childhood obesity prevention and research synergies transpiring from cross‐border collaborations, we convened a group of experts in the United States and Latin America to conduct a narrative review of the epidemiological literature on the role of obesogenic endocrine disrupting chemicals (EDCs) in the etiology of childhood obesity among Latino youth in the United States and Latin America. In addition to summarizing and synthesizing results from research on this topic published within the last decade, we place the findings within a lifecourse biobehavioral framework to aid in identification of unique exposure‐outcome relationships driven by both biological and behavioral research, identify inconsistencies and deficiencies in current literature, and discuss the role of policy regulations, all with the goal of identifying viable avenues for prevention of early life obesity in Latino/Hispanic populations.


| INTRODUCTION
The World Health Organization estimates that obesity worldwide has tripled since the 1970s, 1 with 62% of adults categorized as overweight or obese in the Americas. The high prevalence of obesity has spared no age group, including young children. 2,3 Obesity during early life is concerning due to the short-term metabolic and psychosocial consequences of excess adiposity, as well as the lifelong implications for cardiometabolic disease risk. [4][5][6] In the United States, children and adolescents of Latino/Hispanic descent have approximately twofold higher prevalence of obesity than their White counterparts. 7 While the root causes of this discrepancy are multifaceted, socioeconomic and ethnic inequities are key drivers. In Latin America, many countries are undergoing the epidemiological transition, related to a process of modernization and urbanization that coincides with increased life expectancy. 8 The industrial advancements occur in concomitance with reduced physical activity and higher intake of processed foods, foods high in calories, unhealthy fats and refined carbohydrates, and exposure to harmful chemicals used in food processing and wrapping 9 -a parallel process known as the nutrition transition. 10 Together, these shifting patterns result in rising rates of obesity and obesity-related chronic disease. 11 Beyond diet and other lifestyle factors, Latino/Hispanic populations have higher prevalence of some genetic polymorphisms that place them at greater risk of obesity-related disease-a topic reviewed in detail elsewhere. 12 Thus, in all populations but especially those in Latin America, it is important to identify risk factors and preventive strategies as early as possible. 13 Nearly two decades ago, Baillie-Hamilton postulated the hypothesis that synthetic chemicals disrupt weight-control mechanisms and may be a cause of obesity. 14 These chemicals, known as obesogenic endocrine disrupting chemicals (EDCs) or "obesogens," 15,16 have been of increasing interest in recent years.
Some EDCs are short-lived, with half-lives of days to months (e.g., bisphenol A [BPA] and phthalates), while others are long-lived and can persist in the environment or in vivo for years (e.g., perfluoroalkyl and polyfluoroalkyl subtances). 17 Although EDCs were initially developed for beneficial uses such as improvements in agricultural output (e.g., pesticides), enhanced safety of common household items (e.g., flame retardants in upholstery), and personal care (e.g., phthalates in soaps and shampoos), excessive exposure to these chemicals is a negative aspect of industrialization and globalization. Because of their endocrine-mimicking properties, EDCs can disrupt hormone biosynthesis, lipid metabolism and adipogenesis to promote weight gain and obesity 15,18,19 through numerous pathways, including but not limited to the following: activating peroxisome proliferator activated receptors, 20,21 disrupting the biosynthesis and function of sex steroid hormones, [22][23][24] altering thyroid function, 25,26 and impacting mechanisms that control appetite and satiety, 20,27,28 as summarized in Table S1.
Moreover, as shown in Figure 1, the impact of EDC exposure may be larger during vulnerable developmental life stages characterized by rapid growth and development, including the in utero period (exposure to the fetus via the mother's environment), infancy (exposure to the infant via the mother during breastfeeding, as well as the infant's own exposure via food/formula containers and household products), early childhood (exposure to the child via diet and other aspects of the environment), and puberty (exposure to the child via diet and environmental exposure). 29 The objectives of this narrative review are threefold. Our first goal is to review current evidence regarding the effects of early life EDCs exposure on obesity and body composition among youth in Latin America and US youth of Latino/Hispanic descent. Our second goal is to place study findings within a lifecourse framework given the relevance of developmental stages vulnerable to environmental exposures. Our final goal is to discuss avenues for preventive efforts, with a specific focus on policy regulation given the potential for wideranging impact.

| METHODS
This narrative review considers recent epidemiological studies examining the impact of exposure to EDCs on obesity and related measures of fat mass and distribution from birth through adolescence.
We focused on original research studies published from January 1, 2010, to January 31, 2020, and restricted articles to human studies of EDC exposure in relation to obesity-related outcomes. We considered both cross-sectional and prospective study designs.
Given our focus on Latino populations, we only included studies of youth in Latin American countries or those in the United States where Latino or Hispanic participants comprised ≥15% of the sample.
We conducted the literature search for this narrative review in duplicate. In the first step, two independent researchers identified available evidence using a search strategy comprised of separate keywords (i.e., substance and outcome). The search was conducted using Pubmed/Medline and SCOPUS, given our interest in human health and the biomedical sciences, with the following keywords: "environmental obesogen," "endocrine disrupting compound," or "poly- and "infant growth, weight gain, overweight, obesity, birthweight, adiposity, childhood obesity." We conducted the search in titles, abstract, and full article text in Pubmed and supplemented with additional studies identified through a review led by Liu and Peterson. 30 After compiling results of the search, we removed duplicates, followed by studies where outcomes were assessed in persons >18 years of age, those that focused on metals as the EDCs of interest given that metals operate through distinct mechanisms from those summarized above (specifically, heavy metals tend to accumulate in organs and disrupt organ function, as opposed to interfering with hormone function), 31  In utero exposure is associated with: • # birthweight • # adiposity in girls but "adiposity in boys DDT is a pesticide once widely used to control insects in agriculture and mitigate infectious diseases for which insects are vectors. Due to its detrimental effects to wildlife, DDT was banned in many countries since the early 1970s. DDE is a metabolite of DDT that is found widely in the environment. Key routes of exposure are ingestion and inhalation.
In utero exposure is associated with: • # birthweight • " adiposity and overweight/obesity in boys during late childhood & adolescence • Sex-specific differences in BMI trajectory from birth through 9 years

Short-lived
Phthalates Group of short-lived chemicals that give plastics flexibility but are also included in solvents, textiles, adhesives, detergents, clothing, and personal care products. Key routes of exposure include ingestion and dermal contact.
In utero exposure is associated with: • " birthweight among infants of women with hyperglycemia during pregnancy • " adiposity and overweight/obesity during late childhood and adolescence • Some evidence of sex-specific associations: " adiposity in girls but # adiposity in boys during mid-childhood Concurrent exposure is associated with: • " adiposity in girls during mid-childhood

Phenols
Short-lived chemicals (the most common being bisphenol A) used to make plastics, and included as a disinfectant in household cleaning and consumer products (e.g., mouthwashes, gargles, and throat sprays). Key routes of exposure include ingestion and dermal contact from household products.
In utero exposure is associated with: • # adiposity in girls during early and middle childhood Concurrent exposure is associated with: • " adiposity in girls but # adiposity in boys during adolescence Organophosphate pesticides (OPs) Short-lived chemicals widely-used as an insecticide in agriculture, home gardens, and veterinary practice. These insecticides work by damaging enzymes that control nerve signals in insects, and can cause acute neurological effects in humans. Human exposure occurs through ingestion, inhalation, and dermal contact, with higher exposure in agricultural communities.
In utero exposure is associated with: • " birthweight among infants of mothers with the non-susceptible genotype for an enzyme involved in OP detoxification (PON1 gene) The most commonly assessed obesity-related outcomes were birthweight (an established marker of the intrauterine environment, an indicator of neonatal adiposity, 56 and a bellwether for future obesity 57 ), body mass index (BMI), and waist circumference.
Nine studies obtained direct measures of adiposity via bioelectrical impedance analysis (n = 7) or air displacement plethysmography (n = 2). Two studies followed participants through puberty (i.e., CHAMACOS and ELEMENT). Prenatal exposure to EDCs was primarily measured in maternal serum and urinary samples during pregnancy (the majority of which were collected during early and/or mid-pregnancy), and only one study used serum from cord blood (Table 1). In the sections below, we summarize findings regarding exposure to EDCs during each sensitive period of interest within the lifecourse in relation to obesity-related outcomes in Latino/Hispanic youth. EDCs and their sources, as well as metrics of obesity-related outcomes reviewed in this paper, are summarized in Figure 1.

| EDC exposure during the prenatal and perinatal periods
The majority of the studies focused on EDC exposure during the prenatal and perinatal periods, with exposure assessment typically occurring during early or mid-pregnancy. While exposures around the time of conception and during pregnancy occur within the prenatal period, and those that occur immediately before or after birth take place within the perinatal period, we discuss findings for these two periods together given wide variability in the week of gestation at which EDC exposure was assessed, and the fact that assessments of exposure via maternal tissues in late pregnancy and at delivery may also reflect EDC exposure earlier in pregnancy. We summarize key findings below.

| Prenatal and perinatal exposure à adiposity at birth
We identified six studies that assessed prenatal and perinatal EDC exposure in relation to outcomes at birth. Birthweight was the most widely used indicator of adiposity, which makes sense given the wealth of literature linking both low [58][59][60][61] and high birthweight 57 to obesity and obesity-related outcomes later in life. 30

DDT and DDE
Heggeseth et al. 39 characterized four distinct BMI growth trajectory classes for 249 participants from the CHAMACOS cohort and then examined sex-specific associations of DDT and DDE with odds of membership in each trajectory class. The investigators found that higher maternal serum concentrations of DDT and DDE during pregnancy were associated with a BMI gain pattern characterized by a stable increase from 2 to 5 years, followed by a rapid increase in BMI gain from 5 to 9 years of age in boys. 39 In girls, higher prenatal DDT exposure was associated with stable growth from 2 to 9 years of age. 39 Moreover, while DDT and DDE exposure was not related to obesity status at 7 years of age in this cohort, 37 investigators noted a sex-specific effect of these chemicals on adiposity at later ages among boys only. Specifically, starting at age 9 years, the odds of being overweight or obese doubled per 10-fold increase in prenatal concentrations of DDT and DDE in boys. 38 By 12 years of age, prenatal exposure to DDT and DDE was associated not only with overweight/ obesity but also with continuous BMI z-score and waist circumference in boys, even after accounting for pubertal status. 41 On the other hand, in a highly exposed population in Chiapas, Mexico, Cupul-Uicab et al. found no association of DDE exposure with BMI growth patterns from birth through 18 months in 789 children. 32 These null findings could be due to the fact that health effects of toxicants may not become apparent until middle to late childhood.

Phthalates
Of the different classes of toxicants reviewed herein, prenatal exposure to phthalates received the most attention, although the evidence across study populations is inconsistent. These discrepancies in findings with respect to phthalates and phenols across individual studies may arise from differences in nativity of the study populations, as well as challenges of assessing exposure to short-lived chemicals.
In CHAMACOS, a study of 345 participants found that maternal urinary concentrations of the low-molecular weight phthalates diethyl phthalate (DEP) and dibutyl phthalate (DBP), and the high-molecular weight di(2-ethylhexyl) phthalate (DEHP) during pregnancy, were positively associated with BMI z-score, waist circumference z-score, and percent body fat during childhood, with particularly strong associa-  42 as well as with continuous BMI z-score through 14 years of age. 43 We note that in these analyses, the investigators did not adjust for pubertal status as it could be on the causal pathway between EDC exposure and weight status or growth.
In contrast to findings in CHAMACOS, a study of 326 mother-  The current literature focuses largely on EDC exposure during the prenatal and perinatal periods, which is likely related to numerous animal models indicating the importance of this timeframe as a sensitive period for environmental exposures. 18,66,67 We also noted that DDT/DDE, organophosphate pesticides, and phthalates have received quite a bit of attention, which makes sense given the increased use of these chemicals in industrializing and agricultural countries such as those in Latin America. 68 During the childhood and adolescent life stages, the majority of studies focused on exposure during the prenatal and perinatal periods.
The chemical of greatest interest were nonpersistent chemicals like BPA and phthalates, which may become more relevant in older children who regularly consume packaged snacks and beverages, which are major sources of BPA and phthalates. Older children also use personal care products that are fraught with such chemicals (e.g., deodorants, cosmetics, and hair products).
With respect to health outcomes, most studies focused on weight-and length/height-based metrics as proxies for excess adiposity, which are widely used given the low cost and ease of weight and height measurement, especially in younger children. Several cohorts also measured waist circumferences and skinfold thicknesses, which are highly correlated with fat mass in children. 69 One cohort directly measured adiposity via air displacement When considering the associations of prenatal and perinatal EDC exposure on obesity beyond infancy, sex-specific associations began to emerge. We noticed a consistent pattern that prenatal POPs exposure corresponded with higher adiposity in boys but not girls. Exposure to short-lived chemicals, like phenols and phthalates, also was associated with adiposity in a sex-specific fashion, although findings are inconsistent. The lack of consistency in findings could be attributed to the various ways in which in utero exposure was handled in the statistical analyses (e.g., average across pregnancy vs. trimester-specific and assessment of individual chemicals vs. summary scores or latent variables), and the extent to which assessment of short-lived chemicals in these studies is an accurate representation of true exposure.
Finally, when considering the relatively small body of literature on EDC exposure and obesity-related outcomes during childhood and/or adolescence, we noted that despite differences in study design, chemicals of interest, and whether or not sex-specific associations were considered, exposure to short-lived chemicals used as plasticizers and in personal care products (i.e., BPA and phthalates) was associated with higher adiposity in girls and lower adiposity in boys.
Such discrepancies between males and females are likely driven, in part, by differences in the endocrine regulation system and fat deposition in pre-and peripubertal youth. 71 While continued research on EDC exposure during the prenatal and perinatal periods will undoubtedly be beneficial, the prenatal period is but one sensitive period during early life ( Figure 1). Therefore, studies that consider independent and joint effects of EDC exposure across multiple sensitive periods of development will aid in identifying the most relevant life stages within which interventions may have the largest impact on the reduction of obesity-related disease. This is particularly important among youth of Latino/Hispanic origin given evidence of race/ethnic differences in timing and tempo of puberty. 76 Moreover, we found that evidence surrounding EDC exposure during childhood and adolescence in relation to obesity-related health was scant, and not all studies considered sex-specific (in addition to race/ ethnicity-specific) associations, which are likely to emerge and diverge from childhood onward. A better understanding of specific culprits of excess adiposity and metabolic risk will unveil the most effective and timely actions for preventive efforts.
Many studies assessing exposure to EDCs during the prenatal and perinatal periods used a single maternal biospecimen to represent the entire pregnancy, thereby precluding the ability to disentangle  78 Despite implementation of such policies, US Latino/Hispanic populations still incur disproportionate exposures to air pollutants, pesticides, and toxic industrial chemicals. 79,80 Reasons for this disparity are complex. 81 Additional surveillance and research is warranted to better understand sociocultural determinants of environmental health of Latino and Hispanic youth in the United States, and to develop the appropriate intervention efforts and initiatives.
Though it is challenging to make inference on the effectiveness of policy to regulate EDC exposure in Latin America due to a lack of surveillance in these settings (Table S2) labeling of processed food and drink products, and improvement of school food and increased physical activity among schoolchildren. 87 We suggest similar recommendations and policies targeting the use of EDCs in food and food products, especially among pregnant women, infants, children, and adolescents. However, such policies will only have measurable impacts on obesity reduction if they are backed up by relevant actions, including population education, enforcement, and program implementation.

| Research priorities
Key research priorities include detailed biomonitoring of EDC exposure levels in Latin American countries and communities of predominantly Latino/Hispanic populations. Following implementation of policies to regulate use of EDCs in agriculture, personal care items, and household products (an effort for which assessment of feasibility is beyond the scope of this review), there will be need to assess the efficacy of such policies to reduce exposure to chemicals and any subsequent impacts on obesity-related health.

| CONCLUSIONS
The