The Association of Ambient PM2.5 Exposure with Anemia Prevalence Amongst the Adult Indian Population

Anemia, characterized by a decrease in red blood cells and often accompanied by diminished hemoglobin levels, is highly prevalent in India. Exposure to ambient PM_2.5 has been identified as a potential risk factor to anemia. However, such evidence is confined to the developed world and lacking in developing countries. Here using health data from the fourth wave of National Family Health Survey (NFHS) and satellite-based PM_2.5 exposure, we examine the association of ambient PM_2.5 exposure and anemia prevalence in both male and female adults of age group 15-49 yrs across 640 districts in India.

We use hemoglobin concentration as the primary data input from the NFHS-4 survey to identify the anemia prevalence. Anemia case is detected for hemoglobin content below 12 g/dL for female and below 13 g/dL for male, measured during the survey and adjusted for the altitude effect. In India, 23% male and 54% female is found to have anemia. For the PM_2.5 exposure, we use a satellite-based dataset since the ground-based measurements are inadequate in India and do not cover 45% districts. This national database (at 1-km resolution) was generated by converting MODIS-MAIAC aerosol optical depth to surface PM_2.5 using a dynamic scaling factor from MERRA-2 reanalysis data. The data is calibrated against the existing ground-based network.

We match the exposure of the survey clusters using the geocode information and estimate chronic exposure (for 10 years previous to the survey time) as the exposure metric. A linear mixed model is used to examine the association between the anemia prevalence and PM_2.5 exposure, adjusted for the confounding factors such as nutritional status (BMI), smoking status, biomass fuel use (to account for household air pollution exposure), wealth index, residence type (urban/rural), age and sex. The analysis (in an ecological study design) is carried out to estimate the increase in anemia prevalence due to 10 μg m^-3 increase in ambient PM_2.5 exposure.