Prenatal PM2.5 exposure in the second and third trimesters predicts neurocognitive performance at age 9-10 years: A cohort study of Mexico City children

Abstract

Introduction: Prenatal exposure to fine particulate matter air pollution (PM2.5) is an important, under-studied risk factor for neurodevelopmental dysfunction. We describe the relationships between prenatal PM2.5 exposure and vigilance and inhibitory control, executive functions related to multiple health outcomes in Mexico City children.

Methods: We studied 320 children enrolled in Programming Research in Obesity, GRowth, Environment and Social Stressors, a longitudinal birth cohort study in Mexico City. We used a spatio-temporal model to estimate daily prenatal PM2.5 exposure at each participant's residential address. At age 9-10 years, children performed three Go/No-Go tasks, which measure vigilance and inhibitory control ability. We used Latent class analysis (LCA) to classify performance into subgroups that reflected neurocognitive performance and applied multivariate regression and distributed lag regression modeling (DLM) to test overall and time-dependent associations between prenatal PM2.5 exposure and Go/No-Go performance.

Results: LCA detected two Go/No-Go phenotypes: high performers (Class 1) and low performers (Class 2). Predicting odds of Class 1 vs Class 2 membership based on prenatal PM2.5 exposure timing, logistic regression modeling showed that average prenatal PM2.5 exposure in the second and third trimesters correlated with increased odds of membership in low-performance Class 2 (OR = 1.59 (1.16, 2.17), p = 0.004). Additionally, DLM analysis identified a critical window consisting of gestational days 103-268 (second and third trimesters) in which prenatal PM2.5 exposure predicted poorer Go/No-Go performance.

Discussion: Increased prenatal PM2.5 exposure predicted decreased vigilance and inhibitory control at age 9-10 years. These findings highlight the second and third trimesters of gestation as critical windows of PM2.5 exposure for the development of vigilance and inhibitory control in preadolescent children. Because childhood development of vigilance and inhibitory control informs behavior, academic performance, and self-regulation into adulthood, these results may help to describe the relationship of prenatal PM2.5 exposure to long-term health and psychosocial outcomes. The integrative methodology of this study also contributes to a shift towards more holistic analysis.

Keywords: Air pollution; Exposure windows; Go/No-Go; Latent class analysis; Neurodevelopment; Particulate matter.

Figure A1.

Associations of daily prenatal PM2.5 levels to mean standard deviation of reaction time (SDRT) on three Go/No-Go tasks (Happy, Neutral, and Letter), accounting for prenatal maternal blood lead concentrations.

Figure A2.

Associations of daily prenatal PM2.5 levels to odds ratio of membership in Class 2 (vs Class 1), accounting for prenatal maternal blood lead concentrations.

Figure A3.

Associations of daily prenatal PM2.5 levels to mean standard deviation of reaction time (SDRT) on three Go/No-Go tasks (Happy, Neutral, and Letter), without accounting for maternal smoking and household tobacco exposure.

Figure A4.

Associations of daily prenatal PM2.5 levels to odds ratio of membership in Class 2 (vs Class 1), without accounting for maternal smoking and household tobacco exposure.

Figure A5.

Sex-specific associations of daily prenatal PM2.5 levels to mean standard deviation of reaction time (SDRT) on three Go/No-Go tasks: Happy, Neutral, and Letter.

Figure 1.

Trial procedure of the Go/No-Go task.

Figure 2.

Latent class analysis (LCA) of child Go/No-Go performance at age 9-10 years. Orange: Class 1 (high performance; n = 220). Blue: Class 2 (low performance; n = 100).

Figure 3.

Child Go/No-Go task performance by latent class. A) Hit rate (HR); B) False alarm rate (FAR); C) Mean reaction time (MRT); D) Standard deviation of reaction time (SDRT). Orange: Class 1 (high performance; n = 220). Blue: Class 2 (low performance; n = 100).

Figure 4.

Associations of daily prenatal PM2.5 levels to odds ratio of membership in Class 2 (vs Class 1. Class 1 = High Performance (Reference) Group. Class 2 = Low Performance Group. X-axis represents days of gestation and y-axis shows changes in odds of Class 2 membership (relative to Class 1) associated with a 10μg/m³ increase in PM2.5 exposure. Black lines graph the expected change in odds of Class 2 membership at each day of gestation; gray shadings show pointwise 95% CIs. A critical window (CW) is defined as a range of days over which the 95% CI excludes zero.

Figure 5.

Associations of daily prenatal PM2.5 levels to mean standard deviation of reaction time (SDRT) on three Go/No-Go tasks: Happy, Neutral, and Letter.