Environmental Toxicant Exposure and Depressive Symptoms

Key Points Question Is there an association between common environmental toxicants and depressive symptoms among US adults? Findings This cross-sectional study of 3427 adults from the National Health and Nutrition Examination and Survey found that 27 environmental toxicant levels measured in blood or urine samples were associated with depressive symptoms, as assessed by the 9-item Patient Health Questionnaire. Systemic inflammation assessed by peripheral white blood cell count mediated these associations. Meaning These findings suggest that many common environmental toxicants are associated with depressive symptoms, providing potential targets for intervention measures and mechanistic research.


Introduction
Depression is one of the key diseases covered by the World Health Organization's Mental Health Gap Action Programme.An estimated 3.8% of the population is affected by depression globally. 1,2Mental illnesses have led to the loss of several billion dollars annually in the US.The international community has invested in a concerted effort to improve mental health, but satisfactory outcomes have yet to be achieved as of 2020. 3 Therefore, the identification of modifiable risk factors for depression is critical. 4ychiatric disorders have complex, multifaceted, and interrelated environmental origins. 5With increasing environmental pollution, many pollutants may impact human disease. 6][10][11][12] However, there are inherent limitations to this approach.For example, some toxicants that show an association with depression when studied in isolation may not prove robust or as clinically relevant when considered in combination with other toxicants. 13posome studies provide a scientific framework to uncover the biological consequences of exposure to a wide range of risk factors, allowing for the avoidance of problems related to selective reporting and confounding by coexposures. 14,15To date, exposome studies have uncovered the associations of multiple behavioral risk factors and social factors with depression, 16,17 while exposome studies exploring the associations of environmental toxicants with depression remain limited. 8In addition, exploration of the associations of environmental toxicants with depressive disorders has been limited by sample size and the lack of comprehensive measurements of environmental toxicants.Therefore, a comprehensive exposome analysis of environmental toxicants is essential for investigating the association between these toxicants and depression.
Previous studies have identified the activation of inflammatory pathways as one of the key mechanisms involved in the pathogenesis of depression and have revealed several inflammatory biomarkers associated with depression, including total white blood cells (WBCs) and C-reactive protein. 18,191][22][23] However, there are significant knowledge gaps in addressing multivariate mediation involving not only high-dimensional mediators but also multiple toxicants with inherent covariance and grouping structures.Therefore, a framework for the analysis of high-dimensional mediators is required to explore a mixed mediation setting involving multiple exposure categories and groups of endogenous biomarker mediators. 24,25sed on a no-hypothesis approach, the purpose of this study was to comprehensively examine the association between environmental toxicant exposure and depressive symptoms in a population of US adults using an exposome approach based on data from the National Health and Nutrition Examination Survey (NHANES) to provide a theoretical basis for policy-driven exposure reduction and depression prevention interventions.In addition, a mediation analysis framework was implemented to examine the mediating function of systemic inflammation.

Study Design and Dataset Generation
NHANES is a nationally representative, cross-sectional study that collects health survey data from the US resident and ambulatory civilian population, with a questionnaire administered in the home followed immediately by a standardized health survey administered in a specially equipped mobile examination center. 26In this study, we included survey results from 3427 participants who had environmental toxicants characterized and who completed the 9-item Patient Health Questionnaire Race and ethnicity were categorized as Mexican American, non-Hispanic Black, non-Hispanic White, other Hispanic, and other race (which comprised participants who identified as non-Hispanic multiracial).These data were collected as a confounding factor owing to the differences among individuals of different races and ethnicities regarding susceptibility to depression and ability to metabolize environmental toxins.

JAMA Network Open | Psychiatry
NHANES was approved by the Research Ethics Review Board of the US Centers for Disease Control and Prevention National Center for Health Statistics, and written informed consent was obtained from all adult participants.We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cross-sectional studies.

Outcome Definition
The PHQ-9, a depression screening instrument, was administered to determine the frequency of depressive symptoms over the past 2 weeks.The instrument incorporates Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition) depression diagnostic criteria. 27Responses were categorized as "not at all," "several days," "more than half the days," or "nearly every day," with scores ranging from 0 to 3. The total score is based on the sum of the points for each item and ranges from 0 to 27.Participants can be classified into 4 categories of depression based on the questionnaire scores: asymptomatic (0-4), mild (5-9), moderate (10-14), moderately severe (15-19), and severe (20-27).We considered a score of 5 or greater to be positive for depression, but additionally grouped the participants based on PHQ-9 scores of at least 10 and at least 15 for sensitivity analyses.

Exposure Definition
In this study, a total of 89 environmental toxicants in 13 categories were measured in the participants who were selected by NHANES investigators based on the subsample A weights.After the questionnaire survey, blood and urine samples were collected from the participants for the evaluation of the exposure level of environmental toxicants.Toxicants were assessed by either the measurement of hemoglobin adducts for reactive organic compounds such as acrylamide, glycidamide, ethylene oxide, and formaldehyde or by assaying inorganic elemental toxicants and metals in the urine.Detailed information and measurement methods for the 89 environmental toxicants and the contaminants included in each toxicant category are available in the eMethods and eTable 1 in Supplement 1.Following consideration of missing data and values below the lower limit of detection (eMethods in Supplement 1), a total of 62 types in 10 categories of environmental toxicants were included in the primary analysis.The concentrations of hemoglobin adduct were expressed as the number per milligram of hemoglobin, with analyses adjusted for hemoglobin concentration.The urine toxicants were adjusted per milligram of urine creatine to adjust for urinary dilution.

Statistical Analysis
Data were analyzed from July 1, 2023, to January 31, 2024.We performed statistical analyses of the exposome using 2 methods, including the exposome-wide association study (ExWAS), which considers the effects of each environmental toxicant independently, and the deletion-substitutionaddition (DSA) algorithm, which considers all environmental toxicants simultaneously. 28More details are described in the eMethods in Supplement 1.We conducted sensitivity ExWAS analyses to enhance the robustness of our findings.First, we performed sensitivity analyses based on different cutoff points (Ն10 and Ն15) of the PHQ-9.We also conducted ExWAS analysis based on the data that did not impute missing values and did not exclude participants with missing values for more than one-third of all environmental toxicants.
We developed a mediation analysis framework consisting of 2 analyses, with the objective of examining whether the associations between environmental toxicants and depressive symptoms were mediated by natural log-transformed total WBC count as a surrogate for systemic inflammation.
The initial analysis performed was a 1-way pairwise mediation, while the second method involved exposure dimension reduction followed by pairwise mediation analysis.Additionally, a reverse mediation analysis was conducted to confirm the absence of reverse mediation.Exposure dimension reduction followed by pairwise mediation analysis is another approach to mediation analysis that aims to reduce the dimensionality of exposures to reduce the number of mediating role models.The approach of the second option is to use exposure-class risk scores (ERS) for mediation analysis.More details about the mediation analysis framework are available in the eMethods in Supplement 1.
We  1).The baseline data before the imputation of missing values are shown in eTable 2 in Supplement 1 and have a distribution similar to the results before the interpolation process.
Among 89 toxicants, a total of 62 with sufficient data were included.Of these, 10 were analyzed dichotomously based on detection limits, and the remaining 50 were log(ln) transformed to approximate a normal distribution (eTable 1 in Supplement 1).As shown in eFigure 2 in Supplement 1, there were different degrees of correlation among the 62 environmental toxicants investigated in this study.The correlation coefficients ranged from −0.227 to 0.988, with most toxicants having a correlation coefficient of less than 0.800.
The ExWAS analysis accounted for potential confounding factors and revealed 27 environmental toxicants in 6 categories with a positive association with depressive symptoms.These categories include acrylamide and glycidamide, ethylene oxide, metals (2 types), nicotine metabolites (3 types), PAH (6 types), and VOC metabolites (14 types).In particular, individuals with detectable levels of MHBMA2 had a risk of depressive symptoms 1.74 (95% CI, 1.38-2.18)times higher than those with undetectable levels.Furthermore, total nicotine equivalent-2 (TNE-2) and total hydroxycotinine, which are nicotine metabolites, were also associated with depressive symptoms.
For each 1-U increase in IQR, the likelihood of depressive symptoms increased by 42% (95% CI, 26%-59%) for TNE-2 and 41% (95% CI, 26%-59%) for hydroxycotinine (Table 2 and   Sensitivity analyses were conducted to confirm the robustness of our observed connections. Using PHQ-9 scores of 10 or more and 15 or more as cutoff values for grouping, our findings revealed that when using a PHQ-9 score of 10 or more as the cutoff, a total of 23 environmental toxicants across 7 categories were associated with depressive symptoms.When the cutoff was increased to a PHQ-9 score of 15 or more, 8 environmental toxicants across 3 categories remained associated with depressive symptoms.Notably, 4 environmental toxicants across 2 categories were consistently associated with depressive symptoms, regardless of the cutoff value, including nicotine metabolites and thallium.It is noteworthy that some environmental toxicants, such as thallium and arsenocholine, were negatively correlated with depressive symptoms at increasing PHQ-9 cutoff values.Furthermore, MHBMA2 continued to be associated with depressive symptoms (eFigure 3 and eTable 4 in Supplement 1).In addition, we also conducted an ExWAS analysis based on the data that did not impute missing values and did not exclude participants with missing values for more than one-third of all environmental toxicants.The results suggest that a total of 29 environmental toxicants across 6 categories were associated with depressive symptoms based on the data that did not impute missing values, and a total of 28 environmental toxicants across 6 categories were associated with depressive symptoms based on the data that did not exclude participants with A mediation analysis framework was used to further explore the mechanisms associated with the prevalence of depressive symptoms due to environmental toxicants.We found that total WBC count had a mediating role in associations of 33 environmental toxicants with depressive symptoms, as shown in Figure 3 and eTable 6 in Supplement 1.There were 17 direct associations between environmental toxicants and depressive symptoms in 33 mediated models, and inflammatory biomarkers mediated 5% to 19% of these associations in 17 mediated models.Reverse mediation analysis revealed that inflammatory biomarkers played a mediating role in 31 inverse associations of environmental toxicants and depressive symptoms (eFigure 5 and eTable 7 in Supplement 1) .
However, for the associations of environmental toxicants and depressive symptoms, 9 intermediate models met the criterion of P < .05/50,while in reverse associations of environmental toxicants and depressive symptoms, none met the criteria.This suggests that the total WBC count, possibly acting as a surrogate for systemic inflammation, may mediate the association between environmental toxicants and depressive symptoms.In addition, we assessed the mediating role of total WBC count in ERS and depressive symptoms associations and found that total WBC count mediated the association of ERS (nicotine metabolites, VOC metabolites, PAHs, and metals) with depressive symptoms, as shown in Figure 3 and eTable 8 in Supplement 1. HPMA, hydroxypropylmercapturic acid; HPMMA, 3-hydroxy-1-methylpropylmercapturic acid; MA, mandelic acid; MHA, methylhippuric acid; MHBMA, monohydroxybutenyl-mercapturic acid; PAH, polycyclic aromatic hydrocarbon; TNE-2, total nicotine equivalent-2; and VOC, volatile organic compound.

Discussion
In this cross-sectional profiling of environmental toxicants to uncover risk factors for depressive symptoms, we found that 6 of 10 categories of toxicants were associated with the risk of depressive symptoms.Furthermore, men and individuals younger than 65 years appear more vulnerable to Environmental toxicants -log 10 (P value) environmental toxicants than women and older individuals.Systemic inflammation may mediate the association between environmental toxicants (or ERS) and depressive symptoms.This suggests that more attention should be paid to exposure to environmental toxicants in the population to reduce the risk of depression.
Our findings indicate that of the 10 categories of toxicants we surveyed, 6 were associated with the risk of depressive symptoms: nicotine metabolites, VOC metabolites, PAH, acrylamide and glycidamide, ethylene oxide, and metals.Our finding that nicotine exposure was associated with depressive symptoms is supported by previous studies. 11,29Although many smokers report that smoking can alleviate the negative effects of depression, a longitudinal birth cohort study found an association when it was assessed prospectively. 30A meta-analysis suggested that nicotine has a negative impact on childhood brain development during adolescence and may have long-term negative effects on the brain and behavior. 31,32In addition, our study is the first, to our knowledge, to systematically examine the association between VOC metabolites and depressive symptoms, which may reflect the metabolism of VOCs in the body and specific toxic byproducts. 12Our findings suggest that most of the VOC metabolites, especially MHBMA2 and CEMA, are associated with depressive symptoms, which is also consistent with previous studies. 12,33previous meta-analysis 34 revealed that PAH compounds were associated with an increase in depressive symptoms among adults, with 1-napthol showing the strongest overall association, consistent with our findings.Experimental studies have also indicated that PAH exposure can lead to brain development defects through oxidative stress and can result in specific anxiety-related behavioral disorders. 35In addition, previous studies have shown associations between depressive symptoms and 4 other categories of toxicants, including acrylamide, 21,36 ethylene oxide, 37 and heavy metals.[38][39][40][41] Overall, ample evidence supports our exposome research on toxicants.Implementing targeted measures based on these findings could play a pivotal role in preventing and treating depression.
Sensitivity analyses showed that our results remained relatively robust.However, we found a positive association of arsenocholine and thallium with depressive symptoms, and there are several possible reasons for this.First, the use of antidepressants can lead to the development of obesity, which can cause some lipophilic environmental toxins to accumulate in adipose tissue, reducing their concentration in the blood or urine. 42Second, magnesium therapy can lead to rapid recovery from major depression. 43However, magnesium supplementation may affect the metabolism of some metals in the body, which may explain the finding in this study that exposure to thallium and cesium is protective against major depression.In addition, arsenocholine is found in seafood and is less harmful, and levels in the body can be altered by dietary changes.More evidence is needed with further studies. 44rthermore, we observed differential susceptibility to toxicants across various demographic characteristics, with a greater association among men and young individuals than among women and older individuals.Previous research has shown that men may be more vulnerable to the effects of certain environmental toxicants than women due to the impact of testosterone on neurotoxicants.
6][47] In addition, young individuals may have more routes of exposure to environmental toxicants, such as occupational exposures, cosmetics, and home renovations, than older individuals. 48Furthermore, the brain's development can continue until early adulthood, and during this developmental process, the brain may be more susceptible to environmental pollution. 49erefore, it is crucial to pay more attention to the exposure of young people to environmental toxicants.
Environmental exposure-induced oxidative stress leading to inflammation is a critical factor contributing to a range of human diseases. 20Depression is associated with chronic systemic inflammation, cell-mediated immune activation, and chronic inflammation.Therefore, inflammation could act as an intermediary factor in the link between environmental chemical exposure and the onset of depression.In our study, we used a mediation analysis framework with distinct penalization and estimation algorithms for each type of analysis.However, the consistent findings throughout these methods highlight the importance of the associations, increasing our confidence in the actual mediating pathways.Using both methods, we found that inflammation plays a mediating role in the associations between depressive symptoms and VOCs, nicotine, metal, and PAH.Previous evidence also supports our findings. 22,23For example, imbalances in specific metal levels can weaken the structure, regulation, and catalytic functions of enzymes, proteins, receptors, and transport proteins, which, induced by oxidative stress, can lead to inflammation and decreased metal proteins, contributing to the development of various neurological disorders. 22Additionally, inflammation has been identified as a crucial factor linking nicotine and depression.Omega-3 fatty acids prevent nicotine withdrawal-induced exacerbation of anxiety and depression in rats by modulating oxidative stress and the inflammatory response. 23Further research is necessary to explore the role of inflammation in the association between other toxicants and depression.

Strengths and Limitations
To the best of our knowledge, the present study is the largest and most recent exposome study of environmental toxicant exposure that simultaneously considers the associations between a comprehensive set (10 classes and 62 types) of toxicants and depressive symptoms.We used a no-hypothesis approach, which is characterized by the systematic testing of many variables in relation to a single outcome, to identify all toxicants associated with the development of depressive symptoms.
Our study has several limitations that warrant consideration.First, the emergence of a plethora of newly synthesized chemicals has resulted in the existence of thousands of chemical compounds. 15e environmental toxicants encompassed in this study represent only a fraction of this diversity, highlighting the need for further identification of environmental toxicants that may have stronger associations with depressive symptoms.Second, several environmental toxicants were excluded due to detection limits, underscoring the necessity of enhancing the sensitivity of detection instruments.
Third, our cross-sectional study did not directly establish a causal relationship between environmental toxins and depressive symptoms; furthermore, assessing depressive symptoms solely through scales does not constitute a diagnosis of depression.Fourth, there may be some unadjusted confounders, such as genetic risks of depression and social factors.Fifth, despite demonstrating the unidirectionality of mediation using bidirectional mediation, which is an effective method for analyzing mediation in cross-sectional studies, 50 prospective research is also required for additional validation.Finally, instead of using inflammatory biomarkers directly, the total WBC count was used as a surrogate for systemic inflammation in this study.

Conclusions
In this cross-sectional study, a total of 27 environmental toxicants in 6 categories were found to be associated with the prevalence of depressive symptoms.

Figure 1 ).
According to the DSA algorithm, TNE-2 and N-acetyl-S-(2-carboxyethyl)-L-cysteine (CEMA; a VOC metabolite) were identified as key factors associated with the prevalence of depressive symptoms.The analysis was stratified by age and sex.The findings suggest that more types of environmental toxicants were associated with depressive symptoms among men (20 toxicants) and individuals younger than 65 years (23 toxicants) compared with women and individuals 65 years or older, particularly for nicotine and VOC metabolites (Figure2and eTable 3 in Supplement 1).

Figure 2 .
Figure 2. Adjusted Association Between Environmental Toxicants (62 Exposures) and Depressive Symptoms in 3427 Adults (Exposome-Wide Association Study Analysis) Stratified by Age and Sex

Figure 3 .
Figure 3. Mediation of Association Between Environmental Toxicants and Depressive Symptoms by Inflammation BiomarkersPairwise mediation analysis A

Table 1 .
Participant Characteristics Abbreviations: AA, Associate in Arts; BMI, body mass index (calculated as weight in kilograms divided by height in meters squared); eGFR, estimated glomerular filtration rate; GED, General Educational Development.aIncludes participants who identified as non-Hispanic multiracial.

JAMA Network Open | Psychiatry
JAMA Network Open.2024;7(7):e2420259.doi:10.1001/jamanetworkopen.2024.20259(Reprinted) July 3, 2024 11/16 Downloaded from jamanetwork.comby guest on 07/07/2024 The susceptibility to environmental toxicants varied across populations, especially among men and individuals younger than 65 years, compared with women and older individuals.Systemic inflammation, as assessed by the total WBC, may mediate multiple associations of environmental toxicants and depressive symptoms.This research highlights the significance of preventing and regulating important environmental toxicants to gain fresh insights into preventing and potentially treating depression.Data Collection, Toxicant Detection, and Analysis eTable 1. Characteristics of 89 Environmental Contaminants or Metabolites of Environmental Contaminants eTable 2. Characterization of Participants in This Study Before Data Interpolation eTable 3. Adjusted Association Between Environmental Toxicants (62 Exposures) and Depressive Symptoms in 3427 Adults (ExWAS Analysis) Stratified by Age and Sex eTable 4. Adjusted Association Between Environmental Toxicants (62 Exposures) and Depressive Symptoms in 3427 Adults (ExWAS Analysis), Compared With Threshold-Based Results eTable 5. Sensitivity Analyses of the Adjusted Associations Between the Environmental Toxicants (62 Exposures) and Depressive Symptoms (ExWAS Analysis) eTable 6. Mediation of Association Between Environmental Toxicants and Depressive Symptoms by Inflammation Biomarkers eTable 7. Mediation of the Inverse Association Between Environmental Toxicants and Depressive Symptoms by Inflammation Biomarkers eTable 8. Mediation of the Association Between Exposure-Class Risk Score (ERS) and Depressive Symptoms by Inflammation Biomarkers eFigure 1. Flowchart of Sample Selection eFigure 2. Correlation of the Environmental Toxicants (62 Exposures) eFigure 3. Adjusted Associations Between the Environmental Exposome (62 Exposures) and Depressive Symptoms in 3427 Adults (ExWAS Analysis), Compared With Threshold-Based Results eFigure 4. Sensitivity Analyses of Adjusted Associations Between the Environmental Exposome (62 Exposures) and Depressive Symptoms (ExWAS Analysis), Compared With Threshold-Based Results eFigure 5. Mediation of the Inverse Association Between Environmental Toxicants and Depressive Symptoms by Total WBC by Pairwise Mediation Analysis eReferences