Cumulative environmental exposures and metabolic syndrome: A study of heavy metals and volatile organic compounds

Abstract

Background:

Metabolic Syndrome (MetS), a condition affecting over one-third of the U.S. population, heightens the risk of cardiovascular disease, Type 2 diabetes, and premature mortality. While individual links between heavy metals (HM), volatile organic compounds (VOC), and MetS have been established, the impact when these environmental toxins are combined remains unclear and unexplored. This study investigates how simultaneous exposure to HMs and VOCs influences the risk of MetS.

Methods:

Weighted Quantile Sum regression and Bayesian kernel Machine Regression were performed on data from 6603 participants in the National Health and Nutrition Examination Survey (2011–2020) to determine the impact of HMs and VOCs detected in urine on MetS. Further analyses were performed for individuals placed in subgroups based on age, sex, race/ethnicity, and monthly poverty level index.

Results:

The analyses reveal that combined exposure to HMs and VOCs is associated with an increased risk of MetS; in particular, exposure to cadmium, tin, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine, and N-acetyl-S-(2-carboxyethyl)-L-cysteine significantly elevates the risk of developing MetS. Younger adults (18–50 years), men, Hispanics and non-Hispanic whites, and those with a monthly poverty index > 1.3 (higher socioeconomic status) emerged as the most vulnerable groups.

Conclusion:

These findings emphasize an urgent need to address and tackle the cumulative impact of environmental toxins through a shift in public health efforts to go beyond investigating isolated exposures to address real-world chemical exposures. By understanding these cumulative risks, we can begin to mitigate them and pave the way for more effective interventions, especially for at-risk populations.

1. Introduction

Metabolic Syndrome (MetS) currently affects over one-third of the population of the United States, increasing risks for cardiovascular disease (CVD), stroke, Type 2 diabetes (T2D), and overall mortality (Molokwu et al., 2017). The National Cholesterol Education Program (NCEP) defines MetS as a diagnosis with at least three of the following five hallmarks: abdominal obesity, hyperlipidemia, low high-density lipoprotein cholesterol (HDL-c), hypertension (HTN), and insulin resistance (Alberti et al., 2009). Numerous factors are linked to MetS, most of which stem from lifestyle and environmental exposures. In recent years, studies have shown a strong relationship between certain heavy metals (HM) or volatile organic compounds (VOC) and the occurrence of MetS (Martins et al., 2023; Christensen et al., 2019; Dong et al., 2024). Such substances are encountered daily, often inhaled from the atmosphere or consumed in the diet.

Heavy metals, such as cadmium, lead, and barium, are highly dense, non-biodegradable elements strongly linked to MetS (Xu et al., 2021). Excessive exposure to such metals leads to a deluge of reactive oxidative species (ROS) in the body, increasing oxidative stress, blunting enzyme activation, and diminishing antioxidant defense systems (Fu and Xi, 2020). Lead, commonly used in gasoline and water pipes (Martins et al., 2023), stimulates inflammation in the body, causing atherosclerotic plaques (Prokopowicz et al., 2017). An increase in CVD markers such as C-reactive protein and low-density lipoprotein cholesterol is seen in association with increased cadmium levels (Obeng-Gyasi, 2020). Barium has been observed to impair liver function, leading to alterations in basic metabolism (Wang et al., 2022). Studies have also shown that increased concentrations of HMs are associated with higher blood pressure, lower HDL-c, and higher levels of triglycerides (Bulka et al., 2019).

Volatile organic compounds (VOC), produced naturally or artificially, are commonly present in industrial settings, often as the result of carbon emissions (Dong et al., 2024). Inhalation of such compounds affects cell proliferation, differentiation, and apoptosis (Ebersviller et al., 2012). VOCs can also weaken the immune system and metabolism by interfering with cell signaling (Ebersviller et al., 2012). Recently, the effects of VOCs have been linked to cancer, hormonal disturbances, hypertension, and obesity (Tan et al., 2024; Lei et al., 2023).

Previous studies have shown that HMs and VOCs individually increase the risk of MetS (Xu et al., 2021; Tan et al., 2024; Lei et al., 2023); however, no study has assessed the effect of these exposures when combined on MetS. This study analyzes the relationship between combined exposure to HMs and VOCs and how it contributes to the onset of MetS, using Weighted Quantile Sum (WQS) regression and Bayesian Kernel Machine Regression (BKMR) analysis on data accessed from the National Health and Nutrition Examination Survey (NHANES) database for the years 2011–2020.

2. Materials and methods

The National Health and Nutrition Examination Survey (NHANES) is one of the largest public databases available for national cross-sectional analysis, containing data detailing the characteristics and health status of ~97 % of Americans (National Center for Health Statistics, 2023). NHANES is an excellent data source for the estimation of how cumulative exposure to HMs and VOCs contributes to MetS. It is one of the few data sources consistently collecting data concerning exposure to environmental toxins in the U.S. population and made available for public use to conduct comprehensive research regarding the impact of these environmental toxins on health. Our study used data encompassing patient information from 2011 to 2020. Data, including demographics, diagnoses, and laboratory results, were collected from each participant in the survey. The participants are interviewed, examined and their samples collected at the Mobile Examination Center (MEC) (National Center for Health Statistics, 2023).

MetS in individuals was identified according to the definition proposed by the NECP (Alberti et al., 2009). Abdominal obesity was defined using waist circumference ≥ 102 cm in men and ≥ 88 cm in women. Hypertriglyceridemia was defined as serum triglyceride levels ≥ 150 mg/dL or if individuals were using medication for it. Dyslipidemia was defined as serum HDL cholesterol levels < 40 mg/dL in men, < 50 mg/dL in women, or if individuals were using medication for it. Hypertension was defined as systolic blood pressure of ≥ 130 mmHg, diastolic blood pressure of ≥ 85 mmHg, or if the individuals were using antihypertensive medications. Impaired glucose levels were identified as a fasting plasma glucose of ≥ 100 mg/dL or if individuals were using glucose lowering medications.

Physical measurements were recorded and samples for labs (serum triglycerides, HDL, fasting glucose, and other standard biochemistry profiles) were collected at the MEC. Trained personnel recorded blood pressure, and three consecutive measurements were taken and recorded for each individual. For the analysis in this study, only the third systolic and diastolic blood pressure measurements were used. Medication history for different conditions was collected using a standard questionnaire and recorded by trained personnel. Medications for hypertriglyceridemia and dyslipidemia were considered if the participants were taking any cholesterol medications. Antihypertensive medication was identified if the participants were taking any blood pressure medications. Glucose lowering medication was evaluated to see whether the participants took any insulin or diabetic pills to lower blood sugar.

Urine samples were collected at the MEC and then processed and analyzed at the Division of Laboratory Sciences, National Center for Environmental Health, Centers for Disease Control and Prevention in Atlanta, GA, to determine exposures to HMs and VOCs and their concentrations. Each VOC and HM analyzed had a specific lower limit of detection (LLOD), which is included in Supplementary Tables 1 and 2. For the compounds with concentrations lower than LLOD, the value was calculated as LLOD divided by the square root of two (NCfH, 2024).

The NHANES data are de-identified, and the Institutional Review Board (IRB) at LSU Health Sciences Center in Shreveport determined this study to be exempt from IRB oversight.

3. Statistical analysis

The complete initial sample dataset included participants aged ≥ 18 years who provided data from 2011 to 2020. Any entries with incomplete demographic data, and those participants who were not measured for HM and VOC exposure, were removed. Participants having HM and VOC exposure concentrations with a variance < 0.1 were further excluded. We then determined the correlation among the HMs and VOCs, and HMs and VOCs with a correlation of > 0.6 were removed. This resulted in a total final sample size (N) of 6603 participants included in our study for analysis.

After accounting for HMs and VOCs with a concentration variance > 0.1 and a correlation < 0.6, the following VOCs were included in our study: 2-methyl hippuric acid, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine, 2-aminothiazoline-4-carboxylic acid, N-acetyl-S-(benzyl)-L-cysteine, N-acetyl-S-(n-propyl)-L-cysteine, N-acetyl-S-(2-carboxyethyl)-L-cysteine, N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine, N-acetyl-S-(2-carbamoyl-2-hydroxyethyl)-L-cysteine, N-acetyl-S-(2-hydroxyethyl)-L-cysteine, N-acetyl-S-(2-hydroxypropyl)-L-cysteine, and phenylglyoxylic acid. HMs included in the study were barium, cadmium, cobalt, cesium, molybdenum, lead, tungsten, manganese, tin, arsenic, and mercury.

Baseline characteristics were analyzed with categorical variables presented as frequencies and continuous variables presented as mean values. The categorical variables were compared using the Chi-squared test and the continuous variables were compared using a t-test. The association between MetS and exposure (individual HMs and VOCs) was determined using univariate logistic regression for urinary concentrations of the compounds, with further analysis of the concentrations being separated into quartiles based on exposure. The results are presented as odds ratio (OR) and corresponding 95 % confidence interval. A p-trend value was also calculated between the quartile-divided concentrations of HMs and VOCs.

The Weighted Quantile Sum (WQS) regression model is a statistical analysis method commonly used to evaluate the impact of combined exposures on an outcome (Zhang et al., 2019). In this study, a WQS index was calculated to determine the joint effect of HMs and VOCs on MetS using quartiles of mixtures. The WQS regression was performed by allocating 60 % of data to validation and 40 % to testing using bootstrap sampling (n = 1000). Bayesian Kernel Machine Regression (BKMR) analysis was also performed to determine the combined effects of HMs and VOCs, which uses a nonparametric Bayesian framework for variable selection and accounts for nonlinear and nonadditive dose-response relationships associated with exposure in the mixture (Bobb et al., 2018). The combined effect of HMs and VOCs in the BKMR model was analyzed for every five-percentile increase in the concentration of the mixture. The response function of individual HMs and VOCs was also analyzed in the BKMR model by fixing each chemical at specific percentiles (25th, 50th, and 75th percentiles); 30,000 iterations were performed for the BKMR model. The BKMR model was also diagnosed for model fitting with convergence analysis.

The univariate logistic regression for individual HMs and VOCs, p-trend analysis, the WQS regression, and the BKMR analysis (with convergence analysis) were also performed for multiple subgroup analyses. The subgroup analysis was stratified based on age, sex, race/ethnicity, and monthly poverty level index to determine which demographics were at more risk of having MetS with respect to HM and VOC exposure.

The statistical analyses were performed in the open-source software R (version 4.4.1) (Team, 2024) using gWQS (Stefano Renzetti, 2023) and bkmr (Bobb, 2022) packages for the respective analyses. The results were significant when the p-value was < 0.05.

4. Results

4.1. Baseline characteristics

In the bivariate analysis, age and race/ethnicity were significantly different (p < 0.01) in people with and without MetS. People with MetS were uniformly distributed among the age groups, while most people without MetS were ≤ 50 years old. Non-Hispanic Whites had the highest prevalence of MetS. Among environmental exposures (i.e., HMs and VOCs), higher concentrations of cadmium, tin, N-acetyl-S-(2-carboxyethyl)-L-cysteine, and N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine were significantly (p < 0.01) higher in individuals with MetS, whereas levels of barium and N-acetyl-S-(2-hydroxyethyl)-L-cysteine were significantly (p <0.01) lower in people with MetS. The results are shown in Table 1.

Table 1.

Baseline characteristics of study participants with corresponding frequencies for demographic and mean concentrations for heavy metals and volatile organic compounds.

			With Metabolic Syndrome	Without Metabolic Syndrome	p-value
DEMOGRAPHICS
Age	18–50 years		636 (27.7 %)	2990 (69.5 %)	< 0.01
	51–60 years		508 (22.1 %)	571 (13.3 %)
	61–70 years		638 (27.8 %)	404 (9.4 %)
	70 + years		517 (22.5 %)	339 (7.9 %)
Sex	Female		1132 (49.2 %)	2176 (50.6 %)	0.32
	Male		1167 (50.8 %)	2128 (49.4 %)
Race/ Ethnicity	Non-Hispanic White		909 (39.5 %)	1544 (35.9 %)	< 0.01
	Non-Hispanic Black		590 (25.7 %)	1031 (24.0 %)
	Non-Hispanic Asian		190 (8.3 %)	545 (12.7 %)
	Hispanic		542 (23.6 %)	994 (23.1 %)
	Other Race/Ethnicity		68 (3.0 %)	190 (4.4 %)
Income	Monthly poverty level index ≤ 1.30		757 (32.9 %)	1462 (34.0 %)	0.26
	Monthly poverty level index 1.30–1.85		371 (16.1 %)	613 (14.2 %)
	Monthly poverty level index > 1.85		1103 (48.0 %)	2087 (48.5 %)
	Don’t know		54 (2.3 %)	106 (2.5 %)
	Refused		14 (0.6 %)	36 (0.8 %)
COMPOUNDS IN URINE
Heavy Metals (μg/L)		Barium	1.63 (3.11)	1.80 (3.08)	0.03
		Cadmium	0.41 (0.47)	0.29 (0.39)	< 0.01
		Cobalt	0.56 (1.38)	0.52 (0.78)	0.28
		Cesium	5.03 (3.38)	4.93 (3.41)	0.22
		Molybdenum	49.18 (45.91)	49.91 (47.34)	0.54
		Lead	0.53 (0.67)	0.50 (0.93)	0.22
		Tungsten	0.11 (0.21)	0.12 (0.54)	0.08
		Manganese	0.16 (0.61)	0.16 (0.45)	0.68
		Tin	1.55 (4.44)	1.11 (3.27)	< 0.01
		Arsenic	17.51 (51.09)	16.31(42.91)	0.34
		Mercury	0.49 (2.17)	0.45 (1.14)	0.5
Volatile Organic Compounds (ng/mL)		2-methyl hippuric acid	69.05 (393.09)	82.62 (932.77)	0.41
		N-acetyl-S-(N-methyl carbamoyl)-L-cysteine	282.64 (1018.06)	239.81(320.42)	0.05
		2-aminothiazoline–4-carboxylic acid	173.29 (211.24)	180.56 (212.91)	0.18
		N-acetyl-S-(benzyl)-L-cysteine	13.29 (28.75)	13.80 (43.36)	0.57
		N-acetyl-S-(n-propyl)-L-cysteine	11.59 (29.94)	13.00 (38.87)	0.10
		N-acetyl-S-(2-carboxyethyl)-L-cysteine	180.32 (233.49)	156.88 (208.83)	< 0.01
		N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine	411.13 (295.41)	384.00 (289.49)	< 0.01
		N-acetyl-S-(2-carbamoyl–2-hydroxyethyl)-L-cysteine	13.12 (15.69)	13.93 (16.26)	0.05
		N-acetyl-S-(2-hydroxyethyl)-L-cysteine	1.50 (2.60)	1.94 (3.58)	< 0.01
		N-acetyl-S-(2-hydroxypropyl)-L-cysteine	75.66 (298.89)	67.90 (165.43)	0.17
		Phenylglyoxylic acid	303.70 (893.21)	284.56 (321.22)	0.20

^*Values in bold are significant (p < 0.05)

4.2. Single HM and VOC exposures

The correlation between the onset of MetS and individual HM and VOC exposure was analyzed using univariate logistic regression, as shown in Table 2. Based on these results, cadmium (OR: 3.60), cesium (OR: 1.18), tin (OR: 2.23), lead (OR: 1.47), N-acetyl-S-(N-methyl carbamoyl)-L-cysteine (OR: 1.68), N-acetyl-S-(2-carboxyethyl)-L-cysteine (OR: 1.53), N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine (OR: 1.41), and phenylglyoxylic acid (OR: 1.24) had a positive association with MetS, as their odds were significantly increased (p < 0.05) in the highest quartile (Q4) when compared to the lowest quartile (Q1). In contrast, barium (OR: 0.73), cobalt (OR: 0.83), 2-aminothiazoline-4-carboxylic acid (OR: 0.85), N-acetyl-S-(n-propyl)-L-cysteine (OR: 0.82), and N-acetyl-S-(2-hydroxyethyl)-L-cysteine (OR: 0.58) had a negative association with MetS, since their odds significantly (p < 0.05) decreased in Q4 compared to Q1.

Table 2.

Association between single heavy metal and volatile organic compound exposure and metabolic syndrome for the total population.

Group	Compound	Continuous	Q1	Q2	Q3	Q4	p for trend
Heavy Metals (μg/L)	Barium	0.98 (0.96,1)	Reference	0.91 (0.79,1.05)	0.85 (0.74,0.98)	0.73 (0.63,0.85)	< 0.001
	Cadmium	2.04 (1.79,2.33)	Reference	2.03 (1.72,2.38)	3.24 (2.77,3.80)	3.60 (3.08,4.21)	< 0.001
	Cobalt	1.03 (0.98,1.08)	Reference	1.17 (1.02,1.35)	1.17 (1.01,1.34)	0.83 (0.72,0.96)	< 0.001
	Cesium	1.01 (0.99,1.02)	Reference	1.33 (1.15,1.54)	1.18 (1.02,1.36)	1.18 (1.02,1.36)	< 0.001
	Molybdenum	1 (1,1)	Reference	1.18 (1.02,1.36)	1.33 (1.15,1.53)	1.03 (0.89,1.20)	< 0.001
	Lead	1.03 (0.93,1.1)	Reference	1.36 (1.17,1.57)	1.54 (1.33,1.78)	1.47 (1.27,1.70)	< 0.001
	Tungsten	0.8 (0.62,1.05)	Reference	1.21 (1.05,1.39)	1.19 (1.03,1.37)	0.95 (0.83,1.10)	< 0.001
	Manganese	1.02 (0.93,1.13)	Reference	1.07 (0.93,1.24)	0.91 (0.79,1.05)	1.02 (0.89,1.18)	0.003
	Tin	1.03 (1.02,1.05)	Reference	1.49 (1.28,1.73)	1.84 (1.59,2.14)	2.23 (1.93,2.59)	< 0.001
	Arsenic	1 (1,1)	Reference	1.07 (0.93,1.24)	1.07 (0.93,1.23)	1.08 (0.93,1.25)	< 0.001
	Mercury	1.01 (0.98,1.04)	Reference	1.07 (0.92,1.23)	1.01 (0.87,1.16)	1.07 (0.92,1.23)	< 0.001
Volatile Organic Compounds (ng/mL)	2-methyl hippuric acid	1 (1,1)	Reference	1.08 (0.94,1.25)	0.96 (0.83,1.11)	0.9 (0.78,1.03)	0.05
	N-acetyl-S-(N-methyl carbamoyl)-L-cysteine	1 (1,1)	Reference	1.55 (1.34,1.8)	1.7 (1.46,1.97)	1.68 (1.45,1.95)	< 0.001
	2-aminothiazoline–4-carboxylic acid	1 (1,1)	Reference	0.94 (0.82,1.09)	1 (0.87,1.15)	0.85 (0.74,0.99)	0.07
	N-acetyl-S-(benzyl)-L-cysteine	1 (1,1)	Reference	1.15 (1,1.33)	1.13 (0.98,1.3)	1.12 (0.97,1.29)	0.16
	N-acetyl-S-(n-propyl)-L-cysteine	1 (1,1)	Reference	0.96 (0.83,1.11)	0.89 (0.77,1.02)	0.82 (0.71,0.95)	0.003
	N-acetyl-S-(2-carboxyethyl)-L-cysteine	1 (1,1)	Reference	1.42 (1.23,1.65)	1.43 (1.23,1.66)	1.53 (1.32,1.77)	< 0.001
	N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine	1 (1,1)	Reference	1.39 (1.20,1.61)	1.34 (1.15,1.55)	1.41 (1.22,1.63)	< 0.001
	N-acetyl-S-(2-carbamoyl–2-hydroxyethyl)-L-cysteine	1 (0.99,1)	Reference	0.91 (0.79,1.05)	1 (0.87,1.16)	0.9 (0.78,1.04)	0.355
	N-acetyl-S-(2-hydroxyethyl)-L-cysteine	0.95 (0.93,0.97)	Reference	0.88 (0.76,1.01)	0.80 (0.70,0.93)	0.58 (0.50,0.67)	< 0.001
	N-acetyl-S-(2-hydroxypropyl)-L-cysteine	1 (1,1)	Reference	1.01 (0.87,1.16)	1.03(0.89,1.18)	0.93 (0.81,1.08)	0.42
	Phenylglyoxylic acid	1 (1,1)	Reference	1.34 (1.15,1.54)	1.38 (1.20,1.60)	1.24 (1.07,1.43)	0.004

^*Values in bold are significant (p < 0.05)

Analyses stratified by sex, age, race/ethnicity, and income were also performed to further investigate demographic differences. Unless otherwise stated, all results listed hereafter are considered significant (p<0.05). In terms of sex (Supplementary Table 3), both men and women exhibited a positive association for cadmium, lead, tin, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine, N-acetyl-S-(2-carboxyethyl)-L-cysteine, and N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine. Only men had a positive correlation for cesium, N-acetyl-S-(benzyl)-L-cysteine, and phenylglyoxylic acid.

When subjects were categorized based on age (Supplementary Table 4), cadmium had a direct association with MetS across all age divisions. In those 18–70 years of age, there was also a positive relationship between MetS and individual exposures to lead and tin, along with tungsten, in those 51–70 years old. Among VOCs, increased concentrations of N-acetyl-S-(N-methyl carbamoyl)-L-cysteine were positively associated with MetS in all age groups except in those aged 61–70 years; 2-methyl hippuric acid, N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine, N-acetyl-S-(2-carbamoyl-2-hydroxyethyl)-L-cysteine, and phenylglyoxylic acid also had positive associations for those aged 18–50 years.

When analyzing race/ethnicity outcomes (Supplementary Table 5), cadmium and tin exposures were positively associated with MetS in all races/ethnicities analyzed. Cesium, lead, mercury, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine, N-acetyl-S-(2-carboxyethyl)-L-cysteine, N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine, and phenylglyoxylic acid were also individually correlated with MetS in Non-Hispanic Whites. Non-Hispanic Blacks had no additional positive correlations. In non-Hispanic Asians, mercury was positively correlated with MetS. Lastly, in Hispanics, lead, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine, N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine, and N-acetyl-S-(2-carbamoyl-2-hydroxyethyl)-L-cysteine exposures were positively related to MetS.

In terms of income (Supplementary Table 6), cadmium, tin, and 2-aminothiazoline-4-carboxylic acid were positively associated with MetS in all income tiers, while N-acetyl-S-(2-hydroxyethyl)-L-cysteine was negatively associated with MetS. Those with a monthly poverty index below 1.30 had additional positive correlations between MetS and tungsten. Those with monthly poverty level index ranging from 1.30 to 1.85 had positive relationships between MetS and molybdenum, lead, N-acetyl-S-(2-carboxyethyl)-L-cysteine, and N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine concentrations. Lastly, those with a monthly poverty level index above 1.85 had positive associations between MetS and concentrations of cesium, molybdenum, lead, N-acetyl-S-(2-carboxyethyl)-L-cysteine, N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine, and phenylglyoxylic acid.

4.3. HMs and VOCs Co-exposure and MetS

The WQS regression model revealed a statistically significant effect of concurrent exposures to VOCs and HMs on the risk of MetS (OR: 1.60, CI: 1.49–1.71, p <0.001) (Fig. 1). Subgroup analyses further highlighted variations in risk across different demographic factors. For age, the study showed a statistically significant (p < 0.001) increase in the odds of developing MetS among individuals aged 18–50 years (OR: 1.61, CI: 1.39–1.86). In contrast, no significant associations were observed for other age groups (p >0.05) (Fig. 1). When stratified by income, the WQS regression revealed a significant positive association with MetS at all income levels (p < 0.001). Participants with a monthly poverty level index ≤ 1.3 had odds of 1.55 (CI: 1.36–1.77), which increased slightly for those with an index of 1.3–1.85 (OR: 1.79, CI: 1.44–2.22), and remained elevated for individuals with an index > 1.85 (OR: 1.67, CI: 1.50–1.87) (Fig. 1).

Fig. 1.

Odds ratios of metabolic syndrome associated with combined heavy metal and volatile organic compound exposure by WQS regression.

Race/ethnicity-based subgroup analysis identified Hispanics as having the highest odds of developing MetS due to combined exposure (OR: 1.88, CI: 1.62–2.20, p < 0.001), closely followed by Non-Hispanic Whites (OR: 1.86, CI: 1.64–2.11, p < 0.001) (Fig. 1). In contrast, Non-Hispanic Blacks (OR: 1.45, CI: 1.25–1.70, p < 0.001) and Non-Hispanic Asians (OR: 1.63, CI: 1.21–2.33) exhibited comparatively lower risks. Finally, sex-based analyses indicated that men (OR: 1.65, CI: 1.47–1.82, p < 0.001) were at a higher risk of developing MetS than women (OR: 1.50, CI: 1.36–1.66, p < 0.001) (Fig. 1). Across most WQS models, cadmium, tin, and N-acetyl-S-(N-methyl carbamoyl)-L-cysteine emerged as the most influential compounds contributing to MetS risk. The weights for the WQS regression models are available in Supplementary Figures 1 to 7.

Using the BKMR model for the total population and subgroups, Figs. 2 and 3 illustrate the cumulative association between mixtures of HMs and VOCs and an elevated risk of MetS. The analysis demonstrated a pronounced increase in risk when exposure levels of these contaminants exceeded the 50th percentile, with the risk escalating progressively as the percentile increased. The BKMR analysis was also performed for participants placed in subgroups based on race/ethnicity, sex, age, and monthly poverty level index. In the subgroup analysis, participants belonging to age groups 18–60 years (≥55th percentile), men (≥60th percentile), Hispanics and Non-Hispanic Asians (≥60th percentile), and with a monthly poverty level index ≥ 1.3 (≥55th percentile) showed an increased risk of MetS as the percentile concentration of the exposure mixture increased. Additional figures showing an association between mixtures of HMs and VOCs and MetS are available in Supplementary Figures 8 and 9. Convergence analysis performed for all the BKMR models (including models in the subgroup analyses) showed that the models converged sufficiently with 30,000 iterations. Plots for the convergence analyses are available in Supplementary Figure 10 to 13.

Fig. 2.

Effect of the mixture of heavy metals and volatile organic compounds on metabolic syndrome analyzed using BKMR models in the total population and subgroups stratified by age and sex.

Fig. 3.

Effect of heavy metals and volatile organic compounds mixture on metabolic syndrome analyzed using BKMR models in subgroups stratified by race/ethnicity and income.

Further examination of individual compounds within the BKMR framework highlighted cadmium and N-acetyl-S-(2-carboxyethyl)-L-cysteine as having a positive association with MetS risk when other chemicals were fixed at their 25th, 50th, and 75th percentile concentrations. Similar results were seen when the BKMR analysis was performed for the age groups 18–50 years and Hispanics. Non-Hispanic Asians had an increased risk from cadmium only. For the age group 51–60 years, positive correlations with MetS were associated with tungsten and tin when other chemicals were fixed at their 25th, 50th, and 75th percentiles. For individuals with a monthly poverty level index of 1.3–1.85, cadmium and tin were found to be positively associated with an increased risk of MetS, and for men with a monthly poverty level index > 1.85, N-acetyl-S-(2-carboxyethyl)-L-cysteine was an additional risk. Estimated risk plots for individual compounds within the BKMR models can be seen in Figs. 4–7.

Fig. 4.

Effect of single heavy metals and volatile organic compounds mixture on metabolic syndrome analyzed using BKMR models in the total population and in subgroups stratified by sex, when other heavy metals and volatile organic compounds were fixed at their corresponding 25th (red), 50th (green), and 75th (blue) percentiles.

Fig. 7.

Effect of single heavy metals and volatile organic compounds mixture on metabolic syndrome analyzed using BKMR models in subgroups stratified by income, when other heavy metals and volatile organic compounds are fixed at their corresponding 25th (red), 50th (green), and 75th (blue) percentiles.

N-acetyl-S-(2-hydroxyethyl)-L-cysteine, phenylglyoxylic acid, cobalt, and cesium were negatively associated with MetS when concentrations of other compounds were fixed at their 25th, 50th, and 75th percentiles in the model. For the age group 18–50 years, N-acetyl-S-(2-hydroxyethyl)-L-cysteine and cobalt, and the age group 51–60 years, lead was the only exposure found to be negatively associated with MetS. When individuals were analyzed based on monthly poverty level indices, a monthly poverty level index between 1.3 and1.85 showed N-acetyl-S-(2-hydroxyyethyl)-L-cysteine to have a negative association, with cesium also showing a negative effect for people with a monthly poverty level index of > 1.85 when other concentrations of other compounds are fixed at their 25th, 50th, and 75th percentiles. N-acetyl-S-(2-hydroxyyethyl)-L-cysteine and barium showed a negative association for men. None of the compounds had a negative effect in Non-Hispanic Asians, while in Hispanics, N-acetyl-S-(2-hydroxyyethyl)-L-cysteine, cesium, and cobalt showed a negative impact. Additional risk plots for individual compounds within the BKMR models can be seen in Supplementary Figures 14 to 16.

5. Discussion

Our study sheds light on the importance of considering combined exposures to VOCs and HMs on the risk of developing MetS. The analysis showed that increased individual serum levels of cadmium, lead, tin, 2-methyl hippuric acid, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine, and N-acetyl-S-(3,4-dihydroxy butyl)-L-cysteine were associated with higher odds for MetS. N-acetyl-S-(2-hydroxyethyl)-L-cysteine was the only compound negatively correlated with MetS. These correlations are seen most prominently in those 18–60 years old, men, Hispanics or non-Hispanic Whites, and those with a monthly poverty level index of > 1.3. The significant findings indicated by the WQS regression highlight the importance of the combined impact of VOCs and HMs on the increased likelihood of developing MetS, especially among younger age groups (18–50 years), men, Hispanics and Non-Hispanic Whites, and those with a monthly poverty level index > 1.3. Cadmium, tin, and N-acetyl-S-(N-methyl carbamoyl)-L-cysteine had the greatest effect on MetS among the mixtures in the WQS models. BKMR analysis further supports the findings from the WQS regression models, revealing a clear positive relationship between combined exposure to HMs and VOCs with the incidence of MetS. The risk of MetS increased specifically when the concentration exceeded the 50th-60th percentile, with risk continuing to increase as the concentration of the exposure mixture increased. Elevated levels of compounds including cadmium, tin, and N-acetyl-S-(2-carboxyethyl)-L-cysteine were positively associated with MetS risk, while cesium, cobalt, and N-acetyl-S-(2-hydroxyethyl)-L-cysteine showed inverse associations in the BKMR models. It also showed that those 18–60 years old, male, Hispanic, and with a higher income (monthly poverty level index >1.3) were more at risk. Interestingly, non-Hispanic Whites did not show an increased association with MetS, as seen in the other analyses. The significant associations identified through WQS regression and BKMR analysis highlight the intricate interplay between these chemical exposures and individual demographic factors. Notably, cadmium, tin, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine, and N-acetyl-S-(2-carboxyethyl)-L-cysteine consistently emerged as major contributors to MetS risk across multiple subgroups, providing evidence of their dominant influence within the exposure mixture of HMs and VOCs analyzed in our study.

HM and VOC exposure occurs through exposure to a multitude of sources, most of which are encountered through routine daily activities. Common water sources for human and agricultural use are contaminated with HMs from mining activities, industrial effluent discharge, domestic sewage, and run-off from agricultural farms (Ali et al., 2019). Soil is also contaminated with HMs by the continued and increasing use of fertilizers and settling of emissions from internal combustion engines, especially in areas with heavy vehicular traffic (Ali et al., 2019). These HMs tend to persist in the environment and are transferred up the food chain, affecting plants and animals along the way and eventually reaching humans (Ali et al., 2019). This leads to bioaccumulation and biomagnification of these HMs in the fruits, vegetables, and animals consumed by humans (Ali et al., 2019; Angon et al., 2024), thus exposing individuals to increasing concentrations of HMs. At the same time, individuals are also increasingly exposed to VOCs both outdoors and indoors. Exposure to VOCs occurs more frequently indoors, in environments such as households, schools, and workplaces (Batterman et al., 2014). Common sources of VOC exposure include outdoor vehicular emissions and indoor sources such as moth repellants, chlorinated solvents, cleaning products, odorants, and cooking related activities (Batterman et al., 2014). We observed that the sources of HM and VOC exposure are related to activities and products that individuals commonly use, which makes them susceptible to toxin exposure in their environment. Unfortunately, exposure to either HMs or VOCs is not exclusive; since individuals are exposed to the above-mentioned sources simultaneously, concurrent exposure to HMs and VOCs occurs. This emphasizes the need to assess the risk of concurrent/combined exposure to environmental toxins on health outcomes. We examined the impact of combined exposure to HMs and VOCs on the risk of MetS. MetS incorporates multiple disease conditions, which makes it a suitable surrogate for assessing the overall health status of individuals. Therefore, our study helps to answer the prevailing question of whether combined exposure to environmental toxins, particularly HMs and VOCs, has an impact on the health of an individual, using MetS as a proxy for the health status of the population.

Studies have shown that exposure to a mixture of heavy metals is associated with an increased risk of MetS. Increased exposure to cadmium has been linked to hepatic oxidative stress and pancreatic β-cell dysfunction, which increases the risk of obesity, cardiovascular disease, diabetes, and MetS (Saedi et al., 2023). In every analysis mentioned in this paper, cadmium had a markedly positive association with the risk of MetS. Cadmium is already considered highly carcinogenic and is associated with a high sulfhydryl group affinity, decreasing cellular antioxidant potential (Peana et al., 2022). Cadmium has also been shown to influence the individual components of MetS. Studies have shown cadmium decreases HDL-c and increases the risk of hypertension by increasing systolic and diastolic blood pressures (Noor et al., 2018; Chen et al., 2013). It can also increase the risk of diabetes mellitus by causing increased insulin resistance (Pedro et al., 2019). Additionally, tin exposure positively affected MetS, which can be explained by its toxic impacts on the immune, reproductive, and nervous systems and genetic components (De Azevedo et al., 2013). Not many previous studies reported tin to be associated with MetS. One study showed an increased risk of MetS from exposure to tin with zinc, but they, too, suggested the need for further epidemiological studies to determine the contribution of tin alone to MetS (Feng et al., 2021).

Lead showed a peculiar relationship with MetS in our study. It was positively associated with MetS in univariate logistic regression. However, in WQS regression and BKMR analysis, the association ranged from weakly positive to negative, with most analyses showing no association. This aligns with reports in the literature, which also indicate contrasting results, with one meta-analysis showing increased risk (Xu et al., 2021) and another showing no risk overall (although men had an increased risk individually) (Hasani et al., 2024). At the same time, lead has been shown to increase the risk of all components of MetS (Park et al., 2019). Also, mercury did not show any strong association with MetS in our study, but has been associated with increased risk in previous studies and with components of MetS (Park et al., 2009, 2017).

In previous studies, VOC mixtures have been shown to increase the risk of MetS. VOCs have been proposed to cause insulin resistance and disrupt glucose homeostasis, as well as increase abdominal obesity (Lei et al., 2023; Wang et al., 2023). In our study, N-acetyl-S-(N-methyl carbamoyl)-L-cysteine and N-acetyl-S-(2-carboxyethyl)-L-cysteine were most positively associated with MetS. A similar pattern has been seen in other studies, too, where N-acetyl-S-(N-methyl carbamoyl)-L-cysteine and N-acetyl-S-(2-carboxyethyl)-L-cysteine were associated with increased odds of MetS along with its components (Tan et al., 2024). Study subjects had higher odds of increased waist circumference, triglycerides, and fasting glucose with lower HDL-c (Tan et al., 2024). N-acetyl-S-(2-carboxyethyl)-L-cysteine is an indicator of acrylamide exposure and its effect on MetS risk is attributable to its links to oxidative stress-induced disruption of glucose metabolism and insulin signaling pathways (Markovic et al., 2018). N-acetyl-S-(2-carboxyethyl)-L-cysteine has shown an association with MetS and its components, related to increased triglycerides and insulin resistance and lower HDL-c (Feng et al., 2022; Feroe et al., 2016). We believe that N-acetyl-S-(2-carboxyethyl)-L-cysteine does have a positive association with MetS, although N-acetyl-S-(N-methyl carbamoyl)-L-cysteine also affects the risk of MetS, as in our study we found it to be associated in both univariate logistic regression and WQS regression analyses. On the other hand, the inverse association observed with N-acetyl-S-(2-hydroxyethyl)-L-cysteine, a biomarker for ethylene oxide exposure, warrants further exploration. This compound’s negative associations may stem from as yet unidentified pathways that mitigate oxidative or metabolic stress, which future studies should aim to elucidate.

The study also divided individuals into subgroups based on demographic and socioeconomic factors including age, sex, race/ethnicity, and monthly poverty level index. The literature shows that these factors are used as covariates for adjustment in their respective models for performing statistical analyses, with few studies commenting on differences in MetS risk based on their demography. A meta-analysis showed that HMs did not increase the MetS risk when the individuals were stratified based on sex, whereas another meta-analysis showed an increased risk in men aged 30–50 years (Xu et al., 2021; Hasani et al., 2024). A study showed that the risk for MetS following cadmium exposure was higher among men, as observed in our study as well (Hasani et al., 2024). However, they also observed an increased risk for MetS in older age groups following cadmium exposure (Hasani et al., 2024), which is in contrast to our study, where we found an increased risk in younger adults (18–50 years). For VOCs, N-acetyl-S-(2-carboxyethyl)-L-cysteine showed an increased risk for MetS in individuals aged < 60 years and in men in logistic regression analysis, with no information about the stratified risk with VOC mixture (Dong et al., 2024; Tan et al., 2024). Our study highlighted the risk based on these demographics, with individuals aged 18–50 years, men, Hispanics and non-Hispanic Whites, and those with a monthly poverty level index > 1.3 having higher odds of experiencing MetS. Age and sex findings align with those of previous studies. Younger adults (18–50 years) may experience heightened metabolic effects due to hormonal activity and occupational exposures. The rise in metabolic syndrome (MetS) among adults aged 20–39 years is cause for concern, reflecting both lifestyle and exposure disparities. The prevalence of MetS in younger adults increased from 16.2 % to 21.3 % between 2011 and 2016, with the greatest increase among Hispanics (up to 40.4 % prevalence), which is likely to be influenced by socioeconomic factors such as poverty and limited healthcare access (Hirode and Wong, 2020).

We hypothesize that the demographic patterns found in our study may be associated with increased chances of exposure to these environmental toxins. A previous study on pregnant women showed that Hispanic and Black women with lower incomes had higher exposure to HMs (Geron et al., 2022). VOC exposure was higher in women, older individuals, Mexican Americans, non-Hispanic Whites, and people with lower incomes (Konkle et al., 2020). This shows that the demographics of people exposed to HMs and VOCs differ from those of people who eventually develop MetS. Increased exposure to VOCs and HMs in Hispanics and non-Hispanic Whites, as seen in previous studies, might account for our findings of an increased risk of MetS in these racial/ethnic subgroups. However, there was an increased risk of MetS in men and those in higher-income populations compared to the general exposure profile. This could suggest that the male sex increases the MetS risk, while the risk associated with higher incomes may be associated with some other unknown confounding factors that increase the risk for developing MetS.

The findings of our study help establish the link between exposure to multiple toxins and the risk of adverse health outcomes. Cumulative exposure to HMs and VOCs was associated with an increased risk of MetS. Our study builds on data in the existing literature, which has focused mainly on the risk of MetS or its components following exposure to either HMs or VOCs individually. However, our study provides sufficient evidence that, compared to individual exposure, combined exposure further increases the risk of developing MetS. This is especially important because individuals are exposed to multiple toxins including HMs and VOCs simultaneously rather than exclusively. In addition, our study uses newer statistical methods such as WQS regression models and BKMR models to determine the effect of combined exposure to HMs and VOCs on MetS. This is comparable to the methods used in previous studies, which employed similar statistical approaches to determine the effect of cumulative toxin exposure on health (Jiang and Zhao, 2024). Our study also provided additional information about sections of the population at increased risk of MetS from exposure to HMs and VOCs, which had not been explored in previous studies.

Our study also provides hints at possible mechanisms linking combined exposure to HMs and VOCs to an increased risk of developing MetS. Individually, HM and VOC exposure have shown a positive relationship with MetS and its components. However, concurrent exposure to HMs and VOCs can accelerate this process, further increasing the risk of earlier development of MetS. This is supported by the subgroup analysis performed in our study, which showed that younger adults aged 18–50 years with combined toxin exposure were most at risk of having MetS. Another interesting observation in our study is that the demographic and socioeconomic profiles of the people exposed to HMs and VOCs differ from those of the people who eventually develop MetS. This highlights the importance of exposure to HMs and VOCs acting in concert with traditional demographic and socioeconomic risk factors for MetS to accelerate the cumulative risk. The exact mechanisms are difficult to decipher from our study, but they provide potential areas for future research. This is especially important given that MetS is associated with the future development of diseases associated with increased morbidity and mortality, including cardiovascular, cerebrovascular, and hepatobiliary diseases.

Since it has now been established that combined exposures to these compounds need to be decreased, a future study should analyze exposure prevention efforts and their ability to decrease an individual’s risk for MetS. These efforts are crucial to reducing the risk of MetS at the population level. This is important since current public health policies and environmental regulations are still not robust enough to reduce exposure to HMs and VOCs in the general public and decrease the associated health risk. As evident in our study, individuals in the U.S. still have significant exposure to HMs and VOCs, which increases their risk of developing of adverse health outcomes, particularly MetS. This calls for stronger public health policies from the government and increased research to address gaps and inform policy makers for better environmental regulations to improve health of the society.

6. Conclusion

Our study highlights the impact of combined environmental exposures to HMs and VOCs on the risk of MetS. The elevation of multiple exposures in this joint impact analysis indicates the need to focus on the analysis of cumulative environmental exposure, rather than the individualized toxin exposure heavily analyzed in the literature, as it is not uncommon for people to be exposed to a multitude of toxins, rather than being exposed to just one specific toxin. Our study revealed demographic vulnerabilities, including younger adults (aged 18–50 years), individuals identifying as Hispanic or non-Hispanic White, and those with a monthly poverty index > 1.3, all of whom were found to have the highest risk. The WQS and BKMR models demonstrated a positive association, with risk increasing significantly as their concentrations increased, particularly for cadmium, tin, and N-acetyl-S-(N-methyl carbamoyl)-L-cysteine. Our findings emphasize the critical need to address cumulative environmental exposures rather than isolated individual toxins using stronger public health policies and stricter environmental regulations. Public health interventions must focus on reducing these exposures, especially in high-risk demographic groups, to mitigate the burden of MetS. Future research should prioritize an understanding of the mechanisms driving these demographic disparities. We can develop comprehensive approaches to reducing MetS risk and improving population health outcomes by addressing these gaps.

Supplementary Material

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.ecoenv.2025.118238.

Fig. 5.

Effect of single heavy metals and volatile organic compounds mixture on metabolic syndrome analyzed using BKMR models in subgroups stratified by age, when other heavy metals and volatile organic compounds are fixed at their corresponding 25th (red), 50th (green), and 75th (blue) percentiles.

Fig. 6.

Effect of single heavy metals and volatile organic compounds mixture on metabolic syndrome analyzed using BKMR models in subgroups stratified by race/ethnicity, when other heavy metals and volatile organic compounds are fixed at their corresponding 25th (red), 50th (green), and 75th (blue) percentiles.

Data availability

Data will be made available on request.