Primary School Educational Attainment among Children Living Near Toxic Sites in Low- and Middle-Income CountriesAbstract Number:2296 Anne Riederer*, Darby Jack, and Jack Caravanos Anne Riederer* Blacksmith Institute, United States, E-mail Address: [email protected] Search for more papers by this author , Darby Jack Mailman School of Public Health, Columbia University, United States, E-mail Address: [email protected] Search for more papers by this author , and Jack Caravanos Hunter College, City University of New York, United States, E-mail Address: [email protected] Search for more papers by this author AbstractEarly educational attainment drives labor productivity and economic growth in countries worldwide. Exposure to lead and other environmental neurotoxicants can adversely affect children’s cognitive development, resulting in lower IQ and other adverse educational outcomes. We combined data from Blacksmith Institute’s Toxic Sites Inventory Program (TSIP) and the national Demographic and Health (DHS) surveys for selected low- and middle-income countries to test the hypothesis that primary school-aged children living near sites severely contaminated with lead and other neurotoxicants would have lower educational attainment than a control group of children matched on age and region. For each site, TSIP provides information on location, history/description, size, and other data. For each household, DHS provides information on each child’s age and grade level, sex, cluster (i.e., residential community) location, wealth index, and total number of household members; some DHSs also include mother’s education. We classified children as exposed if their residential cluster (center point) was =5 km from a TSIP site. In a preliminary analysis using the Philippines 2003 and 2008 DHS and TSIP data on 12 urban lead-contaminated sites, we used mixed multivariable regression to explore the association between exposure status and completed years of primary school. In a model controlling for sex, wealth index, number of household members, region and DHS year, and accounting for intra-cluster and intra-region correlation, we found that 8 year olds in the exposed communities (n=120) completed significantly fewer primary grades than the control children (n=589) (ß=-0.24; -0.47, -0.01, 95% CI; p=0.045). The association was not statistically significant for 9 year olds. In this presentation, we will share results of analyses using the full dataset (approximately 17 countries), including sub-analyses controlling for mother’s education where available.