Household air pollution and early childhood development in sub-Saharan Africa

Background Solid fuel use, a major contributor to household air pollution in sub-Saharan Africa, is linked to adverse early life health outcomes. However, associations between solid fuel use and cognitive and social-emotional development are understudied. Objective The overarching goal of this dissertation was to assess associations between type of solid fuel used in households and cognitive and socio-emotional development among children under five years. Methods In aim 1, we quantified 24-hour average PM2.5 personal exposure among caregivers of children under five years (as a proxy for children’s exposure). In aim 2, we evaluated the test-retest reliability of the cognitive and socio-emotional domains of the widely used 10-item UNICEF early childhood development index, and their concurrent validity against corresponding domains of The Ages and Stages Questionnaires. Aims 1 and 2 were conducted in Nakaseke district, Central Uganda. In Aim 3, we employed multi-level logistic regression methods to investigate whether household solid fuel type (wood vs. charcoal) is associated with cognitive and socioemotional development using existing Demographic and Health Surveys data from seven sub-Saharan African countries. Results In aim 1, median 24-hour average personal PM2.5 exposure among caregivers (57.1 μg/m3) was more than twice the World Health Organization guideline for 24-hour average PM2.5 exposure, and the two strongest predictors of PM2.5 concentrations were type of fuel and season. Firewood users had 46% PM2.5 higher concentrations than charcoal users (geometric mean ratio: 1.46, 95% CI: 1.13, 1.89), and dry season was associated with 40% higher PM2.5 concentrations than wet season (geometric mean ratio: 1.40, 95% CI: 1.14, 1.72). In aim 2, 60.9% and 62.3% of children were on track on the socio-emotional domain at the respective visits, and test-retest reliability for the domain was fair (kappa=0.44); delay in that domain was significantly associated with worse scores on corresponding Ages and Stages domains and with severe punishment and lack of school attendance. Over 98% of children were on track on the UNICEF index’s cognitive domain at each visit, limiting ability to meaningfully assess that domain. In aim 3, children who lived in households that cooked with wood had higher odds of poor performance on the UNICEF early childhood development index’s cognitive and socio-emotional items compared to children living in households that use charcoal. Conclusions Prospective studies that incorporate objective pollutant measurements are needed to improve understanding and interventions around this topic. Such studies should include contextually appropriate developmental assessment tools, and clearly delineate biologically relevant exposure windows