Prenatal PM2.5 and Infant Antibiotics: Birth‑Weight Mediation via Causal Analysis

Background

This study assessed whether birthweight mediates the association between prenatal PM2.5 exposure and respiratory infections in infancy, using community‑dispensed amoxicillin as a proxy outcome.

Methods

We used a population‑based cohort study of all singleton births in Scotland (2009–2018), linking administrative health and demographic registries. Annual PM2.5 exposure was assigned to maternal postcode and weighted by pregnancy months. The outcome was ≥1 amoxicillin prescription during the first year. We conducted a preliminary analysis of causal mediation via g‑computation to obtain the controlled direct effect (CDE) and interventional direct (DE) and indirect (IE) effects. Counterfactual contrasts compared PM2.5 fixed at 5 µg/m³ with the observed distribution.

Results

The final cohort included 498,123 births (72% term; 49% female). Mean maternal age was 29 years (SD 6). Mean and median PM2.5 exposure was 7 µg/m³ (SD 1; IQR 6–8), with nearly all births exposed above 5 µg/m³. Comparing a counterfactual scenario of prenatal PM2.5 set to 5 µg/m³ with the observed (higher) exposure distribution, the CDE showed higher odds of at least one amoxicillin prescription in the first year of life (OR = 1.12; 95% CI 1.10–1.14). Interventional DE estimate showed a comparable effect with an OR = 1.12 (95% CI 1.11–1.14), indicating increased odds at the observed PM2.5 levels. The interventional IE through birthweight was close to null (OR 0.99; 95% CI 0.99–1.01).

Conclusions

Preliminary findings suggest prenatal PM2.5 exposure increases the risk of early-life respiratory infections, while birth weight plays a limited mediating role. Further analyses will explore additional counterfactual scenarios, alternative effect measures and will consider sensitivity analyses.

Key messages

• Prenatal PM2.5 exposure increases the risk of early‑life respiratory infections, while birthweight shows limitated mediating influence.

• Direct PM2.5 effects appear to drive infant antibiotic use, indicating a potential role for policies aimed at reducing prenatal exposure.