Urinary phenols and parabens metabolites associated with cardiovascular disease among adults in the United States

The field of environmental health has begun to examine the effects of higher-order chemical combinations. The current literature lacks studies exploring associations between multiple organic chemical mixtures and cardiometabolic diseases (CVDs). This study aimed to evaluate associations between urinary phenols, parabens metabolites, and total and individual CVDs among a nationally representative sample of adults in the US. This cross-sectional study analyzed 7 urinary chemicals detected among the general population from the 2005–2016 National Health and Nutrition Examination Survey (NHANES, n=10,428). Multivariate logistic regression and weighted quantile sum (WQS) regression were applied to examine relationships between phenols and parabens metabolites, alone and in combination, and total and individual CVDs prevalence. Compared with the lowest quartile, URBPA (OR: 1.52; 95% CI: 1.20–1.91; P=0.001) levels in the highest quartile were independently associated with increased total CVD. The WQS index of phenols and parabens mixtures were independently correlated with total CVD (adjusted odds ratios [OR]: 1.16; 95% confidence interval [CI]:1.06–1.28; P=0.002), angina (adjusted OR: 1.30; 95% CI: 1.07–1.59; P=0.009), and heart attack (adjusted OR: 1.30; 95% CI: 1.12–1.51, P<0.001). Urinary bisphenol A (URBPA, weight=0.636) was the most heavily weighted component in the total CVD model. Restricted cubic spline regression demonstrated positive correlations and nonlinear associations between URBPA and both total CVD (P for nonlinearity=0.032) and individual CVD (heart attack; P for nonlinearity=0.031). Our findings suggested that high combined levels of phenols, and parabens are associated with an increased CVD risk, with URBPA contributing the highest risk.


Introduction
Cardiovascular diseases (CVDs) is the leading cause of death and disability worldwide and is associated with substantial burdens on medical and health resources (GBD 2016Causes of Death Collaborators 2017GBD 2016 DALYs andHALE Collaborators 2017). The American Heart Association has predicted that approximately 43.9% of the US population will have CVD by 2030 (Benjamin et al. 2017). Therefore, exploring the modifiable environmental risk factors that contribute to CVD might provide insights into potential preventive measures. The environment is awash with synthetic chemicals from a variety of sources. Estimates suggest that >30,000 synthetic chemicals are in current use, among which at least 5,500 are produced at quantities >100 tons per year (Hartung 2009). Persistent chemical use is common in food processing, packaging, the production of chemical necessities, and the Ting Yin and Xu Zhu contributed equally to this work. pharmaceutical industry, and the most commonly used chemicals include phenols and parabens. To date, numerous epidemiological and clinical studies have described the adverse effects of ubiquitous exposure to multiple phenols and parabens due to widespread use in food processing and packaging applications (Calafat et al. 2015). As a consequence, these chemicals have contaminated the surrounding environment and can be transformed into corresponding metabolites, which can result in high concentrations in animal and human tissues and bodily fluids (Wilhelm et al. 2003;Ilaria et al. 2020). CVD has been associated with higher concentrations of phenols and parabens among individuals with lower socioeconomic status (Ruiz et al. 2018), lower body mass index (BMI) , and exposure to cigarette smoke, although not all studies have reported consistent results (Quiros-Alcala et al. 2018;Julia B. et al. 2020). Among phenolic chemicals, bisphenol A (BPA) is produced at the highest volume and is used extensively and globally in the production of polycarbonate plastics. Regulatory bans have been enacted to reduce BPA use and promote the increased use of replacement compounds (Ilaria et al.2020). Benzophenone-3 (BP3) is a chemical ultraviolet filter that is commonly used to absorb ultraviolet rays. Nonoccupational exposures to dichlorophenol (DCP) occur through the inhalation of contaminated air or the ingestion of contaminated food or chlorinated water (Rachel D. et al. 2015). 24-DCP and 25-DCP are byproducts produced by wastewater treatment, waste incineration, and wood pulp bleaching and as metabolites of some organochlorine pesticides (Dodson et al. 2012). Triclosan (TCS) is an antimicrobial chemical that is widely used in products such as antibacterial soaps, toothpaste, pens, diaper bags, and medical devices. TCS has been associated with the development of bacterial resistance via target site modification in a dosedependent manner (Nietch et al. 2013). Some animal-based studies have revealed that the byproducts of phenol decomposition can affect mitochondrial function (Juyoung et al. 2016), disrupt calcium signaling (Cherednichenko et al. 2012), damage reproductive and thyroid hormone production (Stephanie K. et al. 2018), alter immunological parameters (Yueh et al. 2014), and potentially reduce cardiovascular function (Cherednichenko et al. 2012;Ahn et al. 2008). Parabens have a lipophilic character and are widely used as additives in germicides and as anticorrosive agents, particularly butyl-phydroxybenzoate (BUP), ethyl 4-hydroxybenzoate (EPB), methyl 4-hydroxybenzoate (MPB), and propyl 4hydroxybenzoate (PPB) (Ferguson et al. 2016). The lipophilicity of these exogenous chemicals allows for toxic hydrophilic species to permeate lipophilic membranes, inducing CVD (Harold and Harold I. 2013). Additionally, the accumulation of lipophilic phenols and parabens due to prolonged exposure and the total serum load of lipophilic species might be determining factors that trigger CVD (Harold and Harold I. 2013). Single-chemical pollutant analyses examining individual phenols and parabens have indicated that these singular chemicals are associated with increased risks of obesity (Traci N. et al. 2020;Lesliam et al. 2018;van der Thomas P. et al. 2020), type 2 diabetes (Andrea et al. 2019;Robert M. and Rebecca A. 2019), hypertension (Andreia et al. 2019), and cardio-related risk factors (Symielle A. et al. 2020).
Although studies have shown various correlations between exposure to individual metabolites as separate chemicals (Chris et al. 2020), each has been insufficient to explain the real-world pathological outcomes. The field of environmental health has increasingly moved toward studying the effects of higher-order chemical combinations (Goodson et al. 2015). Recent evidence has documented an increase in the occurrence of acute exacerbations in cardiovascular events and an increase in CVD-associated chronic morbidity following exposure to multiple synthetic chemicals (Bhatnagar 2006;Calafat et al. 2010). Weighted quantile sum (WQS) has been recommended as a multiple regression method that can be used to analyze the synergistic effects of multiple chemical exposures on CVD outcomes to identify the effects of mixed exposures and define the most influential chemicals within the mixture (Carrico et al. 2015).
The aim of the present study was to explore the effects of multiple exposures to phenols and parabens on CVD outcomes by analyzing the demographic characteristics and corresponding urinary chemical marker data collected in the National Health and Nutrition Examination Survey (NHANES) from 2005 to 2016.

Study population and design
The NHANES is administered by the National Center for Health Statistics (NCHS) at the US Centers for Disease Control and Prevention (CDC). NHANES uses a multistage, stratified, probability sampling design to obtain unbiased national health information (detailed sample design: https:// wwwn.cdc.gov/nchs/nhanes/tutorials/module2.aspx). Data from NHANES surveys for 2005-2016 include exposures to multiple synthetic chemicals among the noninstitutionalized US population (NCHS) (Fangjian and W. 2016). None of the authors of the current study were involved in the collection of the data or the production of the database in use.
A total of 60,936 subjects were enrolled in the initial analysis. Participants missing data for phenols and parabens (n = 45,209) and those participants younger than 20 years or missing data regarding CVD outcomes (n = 5,299) were excluded from the present analysis. In total, 10,428 subjects with complete data regarding exposure to multiple phenols and parabens, primary covariates, and CVD outcomes were included for further analysis (Fig. S1).
The concentrations and distributions of the 7 remaining urinary chemical metabolites were analyzed. The urinary metabolites BPA; BP3; 24-DCP; and 25-DCP were extracted using online solid-phase extraction (SPE) and measured by high-performance liquid chromatography and tandem mass spectrometry (MS/MS). The urinary metabolites TCS, MPB, and PPB were extracted using online SPE and measured by high-performance liquid chromatography-isotope dilution MS/MS (34). All concentrations below the LOD were replaced by the LOD value divided by the square root of 2 (LOD/√2). The detailed protocol is described elsewhere (NHANES 2011-2012: Environmental Phenols & Parabens Data Documentation, Codebook, and Frequencies (cdc.gov)). All urinary measurements were adjusted for creatinine to account for differences in kidney function.

Assessment of CVD outcomes
Participants responded to questions related to medical conditions and CVD outcomes prior to undergoing a physical examination, which was documented by a computer-assisted personal interviewing (CAPI) system. Responses to followup questions related to CVD-specific medications and healthcare were also ascertained. Self-reported diagnosed and borderline diabetes/high blood pressure assessments were combined as "diabetes/hypertension" for the purposes of this analysis.
Participants aged ≥20 years were asked, "Has a doctor or other health professional ever told you that you have [congestive heart failure, coronary heart disease, angina, heart attack, or stroke]?" Cardiovascular disease (CVD) was defined as any reported diagnosis of congestive heart failure (CHF), coronary heart disease (CHD), angina, heart attack, or stroke. The data for each CVD outcome were collected for further analyses of associations with the 7 assessed urinary chemical metabolites.

Covariates
Potential confounders of the CVD outcome measure, including demographic characteristics (age, sex, educational levels, race/ethnicity, and poverty income ratio [PIR]), lifestyle characteristics (BMI, smoking, and alcohol use), comorbidities (diabetes mellitus and hypertension), and urine creatinine (Ucrea) levels, were collected. BMI was calculated as the ratio of weight in kilograms to height in meters squared (kg/m 2 ). Ucrea data were collected by interview or laboratory detection performed by NHANES.
Covariate details were as follows: (1) age was modeled continuously; (2) sex was dichotomized as male or female; (3) education level was coded as below/above high school and high school (or K-12 [6-19 years old], ≤ high school/ GED, some college, ≥ bachelor's degree); (4) the family PIR was assessed as the ratio of family income to the poverty threshold; poverty was characterized as a family PIR below 1 (Lesliam et al. 2018); (5) a smoker was identified as someone who smoked at least 100 cigarettes in his lifetime, and alcohol use was categorized as never drinker, former drinker, and current drinker; (6) diabetes mellitus (DM) and hypertension were assessed and determined by the questionnaire administered by NHANES personnel (Lesliam et al. 2018); Ucrea (mg/dL) concentrations were measured from the urine samples using an automated colorimetric method based on a modified Jaffe reaction on a Beckman Synchron AS/ASTRA clinical analyzer (Beckman Instruments, Inc., Brea, CA) and were used to correct for urine dilution in our statistical models.

Statistical analyses
A total of 7 urinary phenols and parabens metabolites (BPA, BP3, 24-DCP, 25-DCP, TCS, MPB, and PPB) were analyzed for their contributions to the risk of categorized CVD outcomes (CHD, CHF, heart attack, stroke, and angina).
The urinary chemical metabolite concentrations were logtransformed to normalize their distributions for further analysis. The Kolmogorov-Smirnov test was used to assess the normality of continuous variables. Continuous variables are expressed as the mean (standard deviation [SD]) and were compared using unpaired t tests. Categorical or dichotomous variables are expressed as absolute values (percentages) and were compared using χ 2 statistics. Pearson's correlation coefficients were calculated between all urinary phenols and parabens exposures. Multivariable logistic regression models were used to calculate odds ratios (ORs) and 95% confidence intervals (CIs) to assess the prevalence of total and individual CVDs associated with urinary phenols and parabens metabolites. The metabolite levels of urinary phenols and parabens were divided into quartiles, and the lowest quartile was used as the reference category.
Regression models were adjusted for potential confounders, including age, sex, race, education levels, and poverty, which a previous study (Traci N. et al. 2020) reported as being associated with urinary metabolite concentrations. Ucrea was used to adjust urine concentrations (Ferguson et al. 2016).
The initial adjustment included the following covariates: age, sex, and education. Advanced models were additionally adjusted for smoking, drinking, BMI, diabetes and hypertension history, and Ucrea concentration. Fully adjusted models were adjusted as described for advanced models with the addition of the other 6 urinary chemicals. Associations between urinary chemical metabolite concentrations and CVDs were calculated. The ORs and 95% CIs were used in the multivariable logistic regression.
WQS regression was used to assess associations between combinations of chemical exposures and CVDs. Each exposure variable was assigned a weight within the index, which indicates the importance of the variable within the mixture (Bhatnagar 2006;Han et al. 2016). WQS models were run using the gWQS v.2.0.0 in R v.3.6.1. Within the gWQS function, the reported results were estimated by specifying deciles for exposure weighting, using 40% of the data used as the test set and the remaining 60% used as the validation set, 1,000 bootstrap repetitions, the random seed set to 2016, and a binomial distribution used for the general linear model. In each case, the weights were estimated by pooling only the effects in the positive or negative direction as constrained. Individual chemical weights ≥0.1 were considered the most important. The value of 0.1 was chosen for ease of comparison across models, including those containing different numbers of chemicals. Restricted cubic splines with knots were used at the 5th, 10th, 50th, and 90th percentiles for multiple phenols and parabens distributions, with the highest weight proportions from the above results, to further explore the dose-response curves between multiple phenols, parabens, and total CVD. Significance was set at P <0.05 (two-sided).

Study population characteristics according to total CVD status
The final analysis included 10,428 adults (men, n=5,084; women, n=5,344, mean age 49 years). The demographic characteristics of participants included in the final analysis are presented in Table 1. Based on the occurrence of CVD, the population was divided into a CVD group (N = 1,119, 10.7%) and a non-CVD group (N = 9,309, 89.3%). The CVD group was older, with a higher proportion of men, a lower average education level, a less Hispanic composition, a higher poverty level, a higher average BMI, a higher incidence of smokers, and a higher incidence of both DM and hypertension, which were significantly different from those in the non-CVD group (all P < 0.05).

Phenols and paraben metabolites and correlation analysis
The distributions and limits of the 7 detected chemical metabolites are shown in Table S1. The detection rate for URTCS was 74.5%, and the detection rates for all other chemical metabolites were above 89%. Correlation analyses of the urinary chemical metabolites showed that some of the metabolites were moderately correlated with the other 6 metabolites (Pearson's r ≥ 0.3). Strong positive correlations were found between URMPB and URPPB (correlation coefficient, r = 0.82, P < 0.001) and between UR24-DCP and UR25-DCP (r = 0.83, P < 0.001). A moderate positive correlation was identified between Ucrea and URBPA (r =0.56, P < 0.001; Fig. 1).

Associations between urinary phenols, paraben metabolites, and total CVD
Our findings showed that in model 1, URBP3, URBPA, and URTCS were associated with an increased prevalence of total CVD. The 7 urinary metabolites were further included as covariates to prevent false positives induced by multiple corrections in models 2 and 3. Compared with the lowest quartile, the URBPA (OR: 1.52; 95% CI: 1.20-1.91, P for trend = 0.001) levels in the highest quartile were independently associated with an increased prevalence of total CVD. No significant correlations were identified between the other 6 urinary metabolites and total CVD (Table 2).

WQS regression analysis of urinary phenols and paraben metabolite mixtures and total and individual CVD
The adverse relationships between urinary phenol and paraben mixtures and the prevalence rates of total and individual CVDs were analyzed using WQS regression analysis. The WQS index of phenols and parabens mixtures were independently correlated with total CVD (adjusted OR: 1.16; 95% CI: 1.06-1.28, P = 0.002), angina (adjusted OR: 1.30; 95% CI: 1.07-1.59, P = 0.009), and heart attack (adjusted OR: 1.30; 95% CI: 1.12-1.51, P < 0.001) but was not correlated with the prevalence rates of CHF, CHD, or stroke (Table 3).
URBPA (weight = 0.636) was the most heavily weighted component in the total CVD model. UR2,5-DCP (weight = 0.320) was the only other component with a weight >0.1. We also calculated the weights of each environmental chemical for the individual CVD models (Fig. 2). The URBPA weights were 0.347, 0.262, 0.376, 0.615, and 0.588 for the CHF, CHD, angina, heart attack, and stroke models, respectively. UR2,5-DCP was also heavily weighted, with values of 0.116, 0.328, 0.453, 0.204, and 0.588 for the CHF, CHD, angina, heart attack, and stroke models, respectively (Table S2).

Associations between URBPA and the prevalence rates of individual and total CVD
The relationships between URBPA and individual CVDs were further analyzed using multiple logistic regression, as shown in Table 3. After adjusting for multivariate models, URBPA levels in the highest quartile remained significantly associated with increased risks of heart attack (OR: 1.54; 95% CI: 1.18-2.02, P for trend = 0.003) and stroke (OR: 1.52; 95% CI: 1.12-2.07, P for trend = 0.029). In a restricted cubic spline with a multivariate logistic regression model, the URBPA level demonstrated nonlinear associations and positive correlations with total CVD (P for nonlinearity = 0.032; Fig. 3) and heart attack (P for nonlinearity = 0.031; Fig. 4). In addition, URBPA levels had linear associations and positive correlations with other individual CVDs (angina, P for nonlinearity = 0.326; stroke, P for nonlinearity = 0.870; and CHF, P for nonlinearity = 0.689; Fig. S2).

Discussion
Data were collected from 10,428 adults as part of a prospective, nationally representative US cohort between 2005 and  [2005][2006][2007][2008][2009][2010][2011][2012][2013][2014][2015][2016] 2016, and logistic regression and WQS regression were used to investigate the associations between urinary phenol and paraben compounds and the prevalence rates of total and individual CVDs. The multipollutant model results were relatively consistent but noninterchangeable. The present study found that the 7 chemical metabolites index was independently correlated with total CVD, angina, and heart attack. URBPA was the most heavily weighted chemical in the total CVD model, and the concentration of URBPA was positively and nonlinear associated with CVD risk. In addition, URBPA concentration also displayed linear associations and positive correlations with individual CVDs, including CHF, angina, and stroke. We investigated available phenol and paraben chemicals, including BPA; BP3; 24-DCP; 25-DCP; TCS; MPB; and  PPB, which are commonly used for daily necessities and industrial and agricultural purposes. In our study, WQS regression analyses were performed at the total population level, and systematic analyses were applied to the 7 metabolic compounds, which allowed correlations to be assessed between the phenol and paraben mixtures and the prevalence rates for total and individual CVDs, which could not be observed using conventional approaches. Combinations of chemical exposures have effects on thyroid signaling, metabolism, oxidative stress, and obesity and affect the regulation of cardiovascular function, as demonstrated in both animal and human studies (Traci N. et al. 2020;Ferguson et al. 2016). However, studies have rarely analyzed the effects of both phenols and parabens on CVDs among the general population using higher-order combinations of chemicals. Our study identified an association between the exposure index and CVD outcomes through a lower dimensional plot, linking these biomarker measurements to exposure and internal concentrations, which provided a better understanding of phenols and parabens metabolisms in humans.
In this study, a 7 chemical mixture index was associated with CVD rather than individual chemicals. Chemicals at low concentrations that individually have no observable effects can act together to produce a joint effect, causing a wide range of outcomes, and these mixtures of various organic chemicals can be found in the water, air, and soil (Benjamin et al. 2017;Wilhelm et al. 2003). Our study used WQS regression to evaluate correlations between CVD events and a chemical mixture Fig. 2 Weights from WQS regression for urinary phenols, parabens index, and the prevalence of total and individual CVDs. Protective models are shown and adjusted for age, sex, race, urinary creatinine, education levels, smoking, poverty, body mass index, diabetes mellitus, and hypertension. WQS, weighted quantile sum; CVD, cardiovascular disease Fig. 3 The restricted cubic spline with a multivariate logistic regression model shown association between URBPA and the prevalence rates of total CVD. URBPA, urinary bisphenol A; CVD, cardiovascular disease Fig. 4 The restricted cubic spline with a multivariate logistic regression model shown association between URBPA and the prevalence rates of heart attack. URBPA, urinary bisphenol A. index comprised of 7 correlated, high-dimensional metabolites. When analyzing interactions between chemical exposures and mixed predictors, the WQS model had high sensitivity and specificity, providing deeper insights into correlations between the phenols and parabens index calculations and simultaneously quantifying the relative importance of each urinary metabolite. Strong positive correlations were identified between MPB and PPB and between 24-DCP and 25-DCP. Other metabolites exhibited moderate correlations. Each chemical exerts specific effects on health through interactions with various cellular components, and similar chemical properties determine the internal metabolic response (Nietch et al. 2013;Symielle A. et al. 2020;Bhatnagar 2006). The most heavily weight chemical has the most influential impact on the accumulated chemical mixture index. The WQS index identifies adverse chemicals based on nonnegligible weights (Chris et al. 2020). Among the 7 metabolites in the mixture, BPA was the heaviest weighted chemical associated with total CVD in our study.
In addition to BPA, TCS was identified as a significant factor in the multivariable logistic regression after adjusting for common cardiovascular risk factors. The other 5 chemicals were negatively associated with total and individual CVDs. In 2017, BPA was included in the European Chemical Agency (ECHA) Candidate List of substances of very high concern and was identified as being highly associated with CVD (GBD 2016 Causes of Death Collaborators 2017; Liu et al. 2017). BPA is an estrogenic chemical that promotes vasodilation by stimulating prostacyclin and nitric oxide synthesis and decreasing the production of vasoconstrictor agents (Han et al. 2016). Other potential physiological effects with implications for cardiometabolic health include improvements in endothelial function, platelet function, and glucose and lipid metabolism (Habauzit and Morand 2012;Cicero et al. 2017). The underlying mechanisms of action are thought to be related to the ability of phenols to modulate insulin resistance, activate K + channels, block the activation of Ca 2+ channels (Kerimi and Williamson 2016;Krga et al. 2016), and activate CD4 + T cells and CD36 to aggravate inflammatory factor expression, which can lead to cardiac fibrosis and atherosclerosis (Katelyn Ann et al. 2019;Stephanie et al. 2019). Through these mechanisms, BPA participates in cardiac hypoperfusion and vascular-related dysfunction (Reventun et al. 2020). TCS and its 2 derivatives (24-DCP and 25-DCP) are priority pollutants that coexist in the aquatic system and have been reported to induce cardiac toxicity in vivo (Danting et al. 2020). Our investigations in the general population provide additional epidemiological evidence to indicate the adverse effects associated with these 7 chemicals and contribute to the foundation for advanced studies of environmentally hazardous substances.
Our study has certain limitations. First, the use of a single urine sample may not efficiently reflect the average level of weekly or monthly exposures to chemical compounds; however, a single urinary analysis provided reliable exposure measures for BPA from a mixture of chemicals, likely due to regular exposure in food and daily consumables. Second, WQS regression required the separation of positive and negative models, which restricted its ability to evaluate the overall impacts of all metabolites (Carrico et al. 2015). However, WQS regression has the ability to address high-dimensional mixtures and inevitably collinear data to explore environmental toxicants that represent public health issues. Finally, the possible causative mechanisms underlying the relationships between mixtures of urinary phenols and parabens metabolites and CVD must be further validated in prospective cohort studies.

Conclusions
Our findings suggested that mixtures of urinary phenols and parabens metabolites were significantly and positively associated with CVDs. The effects of multiple exposures to urinary chemicals on CVD differed in magnitude across the chemicals analyzed, with the greatest influence contributed by URBPA. Additional methods are required when mixture analysis is used because these results show the significance of cooccurring exposures.