Neurocognitive Analysis of Low-level Arsenic Exposure and Executive Function Mediated by Brain Anomalies Among Children, Adolescents, and Young Adults in India

Abstract

Importance: Arsenic, a contaminant of groundwater and irrigated crops, is a global public health hazard. Exposure to low levels of arsenic through food extends well beyond the areas with high arsenic content in water.

Objective: To identify cognitive impairments following commonly prevalent low-level arsenic exposure and characterize their underlying brain mechanisms.

Design, setting, and participants: This multicenter population-based cohort study analyzed cross-sectional data of the Indian Consortium on Vulnerability to Externalizing Disorders and Addictions (cVEDA) cohort, recruited between November 4, 2016, and May 4, 2019. Participants aged 6 to 23 years were characterized using deep phenotyping measures of behavior, neuropsychology, psychopathology, brain neuroimaging, and exposure to developmental adversities and environmental neurotoxins. All analyses were performed between June 1, 2020, and December 31, 2021.

Exposure: Arsenic levels were measured in urine as an index of exposure.

Main outcomes and measures: Executive function measured using the cVEDA neuropsychological battery, gray matter volume (GMV) from T1-weighted magnetic resonance imaging, and functional network connectivity measures from resting state functional magnetic resonance imaging.

Results: A total of 1014 participants aged 6 to 23 years (589 male [58.1%]; mean [SD] age, 14.86 [4.79] years) were included from 5 geographic locations. Sparse-partial least squares analysis was used to describe a negative association of arsenic exposure with executive function (r = -0.12 [P = 5.4 × 10-4]), brain structure (r = -0.20 [P = 1.8 × 10-8]), and functional connectivity (within network, r = -0.12 [P = 7.5 × 10-4]; between network, r = -0.23 [P = 1.8 × 10-10]). Alterations in executive function were partially mediated by GMV (b = -0.004 [95% CI, -0.007 to -0.002]) and within-network functional connectivity (b = -0.004 [95% CI, -0.008 to -0.002]). Socioeconomic status and body mass index moderated the association between arsenic and GMV, such that the association was strongest in participants with lower socioeconomic status and body mass index.

Conclusions and relevance: The findings of this cross-sectional study suggest that low-level arsenic exposure was associated with alterations in executive functioning and underlying brain correlates. These results indicate potential detrimental consequences of arsenic exposure that are below the currently recommended guidelines and may extend beyond endemic risk areas. Precision medicine approaches to study global mental health vulnerabilities highlight widespread but potentially modifiable risk factors and a mechanistic understanding of the impact of low-level arsenic exposure on brain development.

Figure 1.. Association of Arsenic Exposure With Executive Function and Altered Gray Matter Volume

A, B, Scatterplots show associations of arsenic exposure with cognition (A) and structural (B) factor scores derived from sparse-partial least squares analysis (to convert µg/L to µmol/L, multiply by 0.0133). Arsenic levels represent residues after covariate regression. The shaded areas in the plots indicate the CIs around the fitted values of the model. C, Brain regions show associations of arsenic exposure with gray matter volume. Colors are for visual representation of the brain areas only. IFG indicates inferior frontal gyrus; INS, insula; ITG, inferior temporal gyrus; L, left; R, right; RAC, rostral anterior cingulate.

Figure 2.. Association of Arsenic Exposure With Altered Brain Connectivity

A, B, Scatterplots depict the association between arsenic exposure and within-network functional connectivity (WNFC) (A) and between-network functional connectivity (BNFC) (B) factor scores. The shaded areas in the plots indicate the CIs around the fitted values of the model. Scores were derived from sparse-partial least squares analysis. To convert µg/L to µmol/L, multiply by 0.0133. Arsenic levels represent residues after covariate regression.

Figure 3.. Association of Socioeconomic Status (SES) and Body Mass Index (BMI) With Arsenic Exposure

Functions are graphed for 2 levels: high (ie, 1 SD above the mean) and low (ie, 1 SD below the mean). To convert µg/L to µmol/L, multiply by 0.0133. Arsenic levels represent residues after covariate regression. The shaded area in the plot indicates the confidence interval around the fitted values of the model (ie, the level of uncertainty around the estimate). The graph is for descriptive purposes only. All inferential analyses maintained continuous values.