Air pollution is a significant global health concern, responsible for millions of deaths annually. Fine particulate matter (PM2.5), a major component of air pollution, is a complex mixture of various chemical species. While numerous studies have explored the impact of PM2.5 mass on child health, research on the individual effects of PM2.5 components, particularly in low- and middle-income countries (LMICs) like India, remains limited. Children under five are especially vulnerable to the adverse health effects of air pollution, facing increased risks of acute respiratory infections (ARI), low birth weight (LBW), and anemia. India, with its rapidly growing population and rising air pollution levels, presents a critical case study. This research aims to address this knowledge gap by quantifying the association between specific PM2.5 components, their contributing sources (sectors), and three key under-five child health outcomes: anemia, ARI, and LBW. The study uses a unique dataset that combines individual-level health information from the National Family Health Survey (NFHS-4) with high-resolution, sector-specific PM2.5 concentration data from a combination of satellite observations and air quality modeling (WRF-CMAQ). This approach allows for a more nuanced understanding of the health impacts of air pollution, beyond simply considering total PM2.5 mass. The findings will be critical for informing targeted pollution control strategies and maximizing health benefits in India and other LMICs.

Existing literature extensively documents the adverse effects of ambient PM2.5 exposure on child health globally. Studies have shown strong links between PM2.5 and increased risks of LBW, ARI, and anemia in children. However, these studies often focus on PM2.5 mass concentration as a single indicator of air pollution exposure, overlooking the potential variations in toxicity among different PM2.5 components. Research highlighting the specific impacts of PM2.5 constituents on child health outcomes is limited, especially in LMICs. Previous studies in high-income countries have begun to explore the effects of individual PM2.5 components, but these results might not be directly transferable to LMIC contexts due to differences in emission sources, population demographics, and socioeconomic factors. The current study aims to address this critical knowledge gap by examining the distinct effects of PM2.5 components in India, one of the most populated and polluted countries in the world.

This study employed a two-stage modeling approach to analyze the relationship between PM2.5 components and child health outcomes in India. The first stage involved using data from the fourth round of the National Family Health Survey (NFHS-4) which provided information on child health outcomes (LBW, anemia, and ARI), socioeconomic factors, and maternal health indicators. This dataset included 259,627 observations, with 177,072 observations remaining after removing missing data and excluding those with missing age. The prevalence of LBW, anemia, and ARI were calculated from the analytical sample. The second stage integrated this health data with highly resolved ambient PM2.5 exposure data. PM2.5 concentration data was obtained at a 1km x 1km spatial resolution, using a combination of satellite-derived PM2.5 measurements from MODIS-MAIAC and outputs from a weather research forecasting (WRF) and community multi-scale air quality (CMAQ) model. This model provided information on the concentration of individual PM2.5 components and their sector-specific sources, allowing researchers to examine the individual contributions from road dust, international transport, industry, agriculture, domestic sources, and others. The PM2.5 data from the model was then combined with the satellite PM2.5 data to create exposure estimates that were subsequently assigned to each child's location. A two-stage model was applied to account for collinearity between PM2.5 components. The first stage regressed PM2.5 mass concentration on each component and extracted residuals to be used as an alternative metric in the second stage. The second stage used a logistic mixed-effects regression model to examine the association between PM2.5 components, residuals, and the three child health outcomes, adjusting for relevant covariates including child sex, maternal education, parity, maternal age and BMI (for LBW), maternal hemoglobin levels and per capita iron intake (for anemia), and household socioeconomic status, residence type, passive smoking, and cooking fuel type (for ARI and anemia). Stratified analyses were performed to assess the differential impact of PM2.5 components on children born with LBW compared to those with normal birth weight. Nonlinearity in exposure-response relationships was explored using penalized cubic smoothing splines. The study also assessed the cumulative effects of PM2.5 components by summing their component-specific regression coefficients weighted by mass fraction and evaluated the potential health benefits of meeting national ambient air quality standards (NAAQS) and World Health Organization (WHO) air quality guidelines (AQG) using attributable fraction calculations.

The study's key findings highlight the significant and independent contributions of various PM2.5 components to child health outcomes in India. The analysis revealed that an interquartile range (IQR) increase in total PM2.5 exposure was associated with a 15% increase in the odds of LBW (OR: 1.15, 95% CI: 1.12–1.18), a 57% increase in the odds of anemia (OR: 1.57, 95% CI: 1.54–1.59), and a 32% increase in the odds of ARI (OR: 1.32, 95% CI: 1.24–1.40). The analysis of PM2.5 components revealed that NO3−, NH4+, elemental carbon (EC), and organic carbon (OC) were strongly associated with all three health outcomes. For instance, an IQR increase in NO3− exposure was associated with the largest effect on LBW (OR: 1.17, 95% CI: 1.14–1.20), anemia (OR: 1.36, 95% CI: 1.32–1.41), and ARI (OR: 1.52, 95% CI: 1.42–1.61). Sectoral analysis showed that PM2.5 from road dust, international transboundary transport, and the domestic sector significantly influenced LBW, anemia, and ARI. Children born with LBW experienced a considerably higher impact from PM2.5 components exposure compared to those born with normal weight. For example, the effect of NO3− on anemia was more pronounced in LBW children (OR: 1.47, 95% CI: 1.41–1.52) compared to children with normal birth weight (OR: 1.34, 95% CI: 1.30–1.38). Exposure-response relationships showed monotonic increases for some components and inverted U-shaped patterns for others. Importantly, the cumulative effect of PM2.5 components was substantially greater than the effect of PM2.5 mass alone. For every 10 µg/m³ increase in PM2.5 mass, the cumulative ORs for LBW, anemia, and ARI were 1.23 (95% CI: 1.21–1.26), 1.49 (95% CI: 1.45–1.52), and 1.35 (95% CI: 1.29–1.41), respectively, significantly higher than the ORs estimated based on PM2.5 mass alone (1.05, 1.10, and 1.11, respectively). Finally, the study estimated substantial reductions in the prevalence of LBW, anemia, and ARI if India were to meet the NAAQS and WHO-AQG, with greater reductions observed when considering the cumulative effects of PM2.5 components.

The findings underscore the critical need to move beyond PM2.5 mass as a sole metric for assessing the health impacts of air pollution. The substantially higher cumulative effects of PM2.5 components compared to PM2.5 mass indicate that strategies focused solely on reducing total PM2.5 may underestimate the potential health benefits achievable through targeted reduction of specific components. The significant associations between NO3−, NH4+, EC, and OC, and the three child health outcomes highlight the importance of focusing mitigation efforts on reducing emissions from sources contributing to these components. The stronger effect observed in LBW children highlights the vulnerability of this population and emphasizes the importance of prenatal and early-life exposure reduction. The substantial reductions in health outcomes projected with the achievement of NAAQS and WHO-AQG underscore the potential public health benefits from implementing stricter air quality standards. However, the study acknowledges several limitations, such as the assumption of stable residence during early life and potential uncertainties in exposure assessment. Future research should focus on validating the findings through well-designed cohort studies and investigating specific biological mechanisms linking PM2.5 components to these health effects.

This study provides compelling evidence of the significant and differential impacts of PM2.5 components on child health in India. The findings demonstrate that focusing solely on PM2.5 mass underestimates the true health burden and that targeted interventions are needed to address the more toxic species. The substantial health gains projected from stricter air quality standards highlight the urgency of implementing effective pollution control measures. Future research should explore the underlying biological mechanisms and validate these findings through long-term cohort studies.

Several limitations should be considered in interpreting the results. Firstly, the study relied on cross-sectional data, making it difficult to establish definitive causal relationships between PM2.5 exposure and health outcomes. Secondly, the accuracy of exposure assessment depended on the reliability of the combined satellite and model data, and potential biases associated with these data sources could affect the results. Thirdly, the study assumed that children's residence did not change during early life, which may not be entirely accurate. Fourthly, the exposure-response relationships for some components showed nonlinear patterns that may be due to unadjusted confounding or other complex factors which could not be sufficiently explored in this cross-sectional study. Further investigations are required to explore these patterns thoroughly. Finally, the model may have limitations in its ability to accurately capture all sources of PM2.5 emissions and their relative contributions.