International Journal of Environmental Research and Public Health Article Racial and Sex Differences between Urinary Phthalates and Metabolic Syndrome among U.S. Adults: NHANES 2005–2014 Rajrupa Ghosh 1, Mefruz Haque 1, Paul C. Turner 2, Raul Cruz-Cano 1 and Cher M. Dallal 1,* ���������� ������� Citation: Ghosh, R.; Haque, M.; Turner, P.C.; Cruz-Cano, R.; Dallal, C.M. Racial and Sex Differences between Urinary Phthalates and Metabolic Syndrome among U.S. Adults: NHANES 2005–2014. Int. J. Environ. Res. Public Health 2021, 18, 6870. https:// doi.org/10.3390/ijerph18136870 Academic Editor: Paul B. Tchounwou Received: 1 June 2021 Accepted: 20 June 2021 Published: 26 June 2021 Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affil- iations. Copyright: © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/). 1 Department of Epidemiology and Biostatistics, School of Public Health, University of Maryland, College Park, MD 20742, USA; rajrupa@terpmail.umd.edu (R.G.); haque.mefruz@gmail.com (M.H.); raulcruz@umd.edu (R.C.-C.) 2 Maryland Institute for Applied Environmental Health, School of Public Health, University of Maryland, College Park, MD 20742, USA; pturner3@umd.edu * Correspondence: cdallal@umd.edu; Tel.: +1-301-405-7065 Abstract: Phthalates, plasticizers ubiquitous in household and personal care products, have been associated with metabolic disturbances. Despite the noted racial differences in phthalate exposure and the prevalence of metabolic syndrome (MetS), it remains unclear whether associations between phthalate metabolites and MetS vary by race and sex. A cross-sectional analysis was conducted among 10,017 adults from the National Health and Nutritional Examination Survey (2005–2014). Prevalence odds ratios (POR) and 95% confidence intervals (CIs) were estimated for the association between 11 urinary phthalate metabolites and MetS using weighted sex and race stratified multivariable logistic regression. Higher MCOP levels were significantly associated with increased odds of MetS among women but not men, and only remained significant among White women (POR Q4 vs. Q1 = 1.68, 95% CI: 1.24, 2.29; p-trend = 0.001). Similarly, the inverse association observed with MEHP among women, persisted among White women only (POR Q4 vs. Q1 = 0.53, 95% CI: 0.35, 0.80; p-trend = 0.003). However, ΣDEHP metabolites were associated with increased odds of MetS only among men, and this finding was limited to White men (POR Q4 vs. Q1 = 1.54, 95% CI: 1.01, 2.35; p-trend = 0.06). Among Black men, an inverse association was observed with higher MEP levels (POR Q4 vs. Q1 = 0.43, 95% CI: 0.24, 0.77; p-trend = 0.01). The findings suggest differential associations between phthalate metabolites and MetS by sex and race/ethnicity. Keywords: phthalates; metabolic syndrome (MetS); race; sex 1. Introduction Phthalates are found in a wide range of household, consumer, and personal care prod- ucts [1]. Low molecular weight (LMW) phthalates, including diethyl phthalate (DEP) and dibutyl phthalate (DBP), are used as solvents in personal care products such as fragrances, lotions, and cosmetics, as well as in pharmaceutical coatings [1,2], whereas high molecular weight (HMW) phthalates, including Di(2-ethylhexyl) phthalate (DEHP) metabolites, are commonly used as plasticizers in food packaging materials, consumer products, and medical devices [1,2]. Plasticizers are able to leach into products or the environment, facilitating human exposure via ingestion, inhalation, and dermal contact [1,3]. Given the multiple and widespread use of phthalates, the majority of the US population is continuously exposed [2]. Women and racial ethnic minorities (Black and Mexican/Hispanic individuals) are reported to have higher urinary levels of phthalate metabolites than their White counter- parts [4–6]. Increased use of personal care products, such as cosmetics, fragrances, and hair care products, may explain the higher phthalate metabolite levels among women [7,8]. Potential explanations for this include increased frequency of specific hair care [9] and feminine hygiene product [4] use among Black women, and personal care product use [10] among Latina women. Int. J. Environ. Res. Public Health 2021, 18, 6870. https://doi.org/10.3390/ijerph18136870 https://www.mdpi.com/journal/ijerph https://www.mdpi.com/journal/ijerph https://www.mdpi.com https://orcid.org/0000-0001-7715-1198 https://doi.org/10.3390/ijerph18136870 https://doi.org/10.3390/ijerph18136870 https://creativecommons.org/ https://creativecommons.org/licenses/by/4.0/ https://creativecommons.org/licenses/by/4.0/ https://doi.org/10.3390/ijerph18136870 https://www.mdpi.com/journal/ijerph https://www.mdpi.com/article/10.3390/ijerph18136870?type=check_update&version=1 Int. J. Environ. Res. Public Health 2021, 18, 6870 2 of 18 Phthalates are classified as xenoestrogens and have been negatively associated with several health conditions [11], including metabolic disorders such as obesity and insulin resistance [12–15]. The association between phthalate exposure and dyslipidemia is in- consistent, with two studies reporting either positive [16] or null associations [17]. These disorders contribute to metabolic syndrome (MetS), defined as having at least three of the five following criteria: elevated abdominal obesity, elevated triglycerides, high blood pressure, elevated fasting glucose levels, and reduced high-density lipoprotein (HDL) levels [18]. MetS affects about 35% of US adults [19], with noted differences by sex and race, and is associated with an increased risk of several chronic diseases, including cardio- vascular diseases [20], diabetes [21], and cancer [22]. Peroxisome proliferator–activated receptors (PPARs), functionally activated by both a lower or higher concentration of phthalates, play a critical role in adipogenesis and lipid storage [23–25]. The interaction of phthalates with PPARs affects glucose metabolism and leads to adipogenesis and disturbs metabolic homeostasis, subsequently resulting in metabolic dysfunctions [23–25]. Hence, the alteration of the lipid and carbohydrate cell signaling pathways has been hypothesized as a mechanism of action for phthalates. While prior studies have investigated the relationship between urinary phthalates and individual components of MetS [12,14,16,26], only one study to date has investigated associ- ations between urinary phthalate concentrations and MetS among US adults [16]. This prior study included participants from the National Health and Nutritional Examination Survey (NHANES) cycles 2001–2010, and observed sex differences in associations between specific urinary phthalate metabolites and MetS [16]. However, despite evidence of racial/ethnic disparities in both phthalate exposure and the prevalence of MetS [4,5,19], it remains un- clear whether racial differences exist in the association between phthalate metabolites and MetS. The objective of this analysis was to assess whether phthalate metabolite concentra- tions are differentially associated with MetS among Black, Mexican/Hispanic, and White individuals who participated in NHANES cycles 2005–2014. 2. Materials and Methods Study Design and Population: This cross-sectional analysis utilized data from the National Health and Nutrition Examination Survey (NHANES) across five survey year cycles conducted during 2005–2014. NHANES is a nationally representative survey ad- ministered by the Centers for Disease Control and Prevention National Center for Health Statistics (NCHS) [27,28]. It uses a complex, multistage probability sampling strategy that oversamples certain subgroups of the non-institutionalized civilian resident population of the United States, residing in the 50 states and in Washington D.C. [27,28]. Additional details on NHANES are published elsewhere [27,28]. The initial analytic population consisted of 16,264 individuals with data on phthalate metabolites during the 2005–2014 survey year cycles. Urinary phthalate metabolites were measured by NHANES in a randomly selected subsample in each survey cycle. Participants aged <18 years (n = 4610) were excluded from the analysis, as the criteria for classifying metabolic syndrome varies for children [29]. Additionally, the following exclusions were applied: participants undergoing dialysis (n = 44), those who reported being pregnant or were lactating (n = 301), those with missing values for phthalate metabolites (n = 261), observations with a sample weight = 0 (n = 124), and individuals who selected “other race- including multiracial” (n = 907) due to heterogeneity. The final analytic sample consisted of 10,017 adults ≥ 18 years, for whom urinary phthalate measurements were available for the 2005–2014 survey years. Data Collection and Covariate Information: Data on demographics, 24-h dietary in- take (total caloric intake (kcal)), and smoking status were collected via questionnaires through in-person interviews. The family poverty–income ratio (PIR) compares family income to US census-defined poverty levels, with a value of 1.0 indicating the federal poverty threshold and above 1.0 indicating income above poverty. Smoking status (cur- rent/former/never) was determined via responses to two questions regarding having Int. J. Environ. Res. Public Health 2021, 18, 6870 3 of 18 smoked at least 100 cigarettes in their lifetime (yes/no) and their current smoking status (yes/no). Blood collection and physical measurements were taken in the mobile examination cen- ter by trained health technicians, including height, weight, waist circumference, and blood pressure, following standard NHANES procedures. A random subsample of NHANES participants were asked to fast for ≥8 h prior to blood collection. Fasting status was docu- mented using a questionnaire prior to blood draw. Fasting samples were used to measure the blood glucose and serum triglycerides. High blood pressure (yes/no) was defined by either an average of three consecutive blood pressure measurements or if current blood pressure medication use was indicated on the interview questionnaire. High blood glucose (yes/no) was defined by either plasma glucose levels or if current insulin/diabetes medica- tion use was indicated in the interview questionnaire. Details on standardized NHANES procedures have been previously described [30–34]. Measurement of exposure: Phthalate metabolites were measured in spot urine sam- ples using high performance liquid chromatography-electrospray ionization-tandem mass spectrometry (HPLC-ESI-MS/MS), as previously described [30]. The specific phthalate metabolites measured varied across the survey years, with a total of 16 phthalate metabo- lites measured across the 2005–2014 survey years. Phthalate metabolites were retained in this analysis if (i) the metabolite was measured in all of the survey cycle years, (ii) less than 50% of the values for a specific metabolite were missing across the survey cycle years, and (iii) more than 50% of the samples were at or above the analytical limit of detection (LOD). The following 11 phthalate metabolites were retained: mono-ethyl phthalate (MEP), mono-n-butyl phthalate (MnBP), mono-isobutyl phthalate (MiBP), mono carboxynonyl phthalate (MCNP), mono carboxyoctyl phthalate (MCOP), mono-(3-carboxypropyl) ph- thalate (MCPP), and mono-benzyl phthalate (MBzP), and the di(2-ethylhexyl) phthalate (DEHP) metabolites of mono-(2-ethyl-5-carboxypentyl) phthalate (MECPP), mono-(2-ethyl- 5-hydroxyhexyl) phthalate (MEHHP), mono-2-ethylhexyl phthalate (MEHP), and mono- (2-ethyl-5-oxohexyl) phthalate (MEOHP). Different limits of detection (LOD) were used by NHANES across the survey cycles (see Table S1). To estimate the LOD in this analysis, the highest reported LOD for any given metabolite across all survey years was used. The proportion of samples below the LOD was less than 10% for all of the metabolites, with the exception of MEHP (40%) (Table S1). The concentration of metabolites below the LOD were replaced with the LOD for that metabolite divided by the square root of two [35]. In addition to analyzing the individual metabolites, the following summary measures were generated and analyzed: low molecular weight (ΣLMW) metabolites (MEP + MnBP + MiBP), DEHP (ΣDEHP) metabolites (MEHP + MECPP + MEHHP + MEOHP), and high molecular weight (ΣHMW) metabolites (MCNP + MCOP + MCPP + MBzP + ΣDEHP). Measurement of outcome: Metabolic syndrome (MetS) was defined using the National Cholesterol Education Program’s Adult Treatment Panel III report criteria [18]. Participants were classified as having MetS if they met at least three of the following five criteria: waist circumference ≥102 cm (men) or ≥88 cm (women), fasting serum triglycerides ≥150 mg/dl, high density lipid (HDL) cholesterol <40 mg/dl (men) or <50 mg/dl (women), blood pressure ≥130/85 mm Hg or treatment for hypertension, and fasting blood glucose ≥ 100 mg/dl or treatment for diabetes [18]. Among those with MetS (n = 2403), 68% had three, 24.4% had four, and 7.6% had all five of the MetS components. The majority of those with MetS had a high waist circumference (93.1%), followed by high blood pressure (83.6%) and low HDL levels (70.6%). Individuals were not excluded if data on specific MetS components were missing, given that Mets is diagnosed based on the presence of a minimum of three components. In the overall population, the availability of MetS component data was as follows: waist circumference (98.3%), fasting serum triglycerides (46.0%), HDL (98.0%), high blood pressure (97.4%), and fasting blood glucose (50.2%). As fasting samples were only collected in a subsample of the NHANES population, glucose and triglyceride status was unknown for 49.8% and 54.0% of the population, respectively. However, among those who were classified as not having MetS in our analytic Int. J. Environ. Res. Public Health 2021, 18, 6870 4 of 18 population but had met two of the criteria above (n = 2702), 20.2% were missing glucose and 21.4% triglycerides. Statistical Analysis: Distributions of the demographic, behavioral, and metabolic vari- ables were assessed by estimating the frequencies and weighted percentages (categorical variables) and weighted means and standard errors (continuous variables). Distributions were examined by sex and race, and further by the presence of MetS status. Indepen- dent sample t-tests (continuous variables) and chi-square tests (categorical variables) were utilized for bivariate analyses to evaluate race and sex differences. When used as a con- tinuous variable, individual phthalate metabolites, ΣLMW, ΣHMW, and ΣDEHP were log (natural log) transformed to account for non-normal right skewed data. Weighted geometric means (GM) and 95% confidence intervals (CI) were estimated for the levels of individual phthalate metabolites, ΣLMW, ΣHMW, and ΣDEHP for the following analytic populations: overall, sex stratified, and race and sex stratified. Correlations between the phthalate metabolites were estimated by Spearman correlation coefficients. MetS was analyzed as a dichotomous outcome, while phthalate metabolites or sums of metabolites were categorized based on quartile distributions in the overall population. Multivariable logistic regression models were used to estimate the prevalence odds ratios (POR) and 95% confidence limits (CI) for the association between urinary phthalate metabolites and MetS overall, as well as stratified by (1) sex and (2) sex and race. Creatinine was included as a covariate in all models to adjust for urine concentration and flow rate. Based on the existing literature, the following a priori potential confounders were included in the final models: age (continuous), race (White/Black/Hispanic or Mexican), sex (male/female), education (