Abstract
BACKGROUND: The health benefits of secondary education have been little studied. We undertook country-level longitudinal analyses of the impact of lengthening secondary education on health outcomes amongst 15-24 year olds. METHODS: Exposures: average length of secondary and primary education from 1980 to 2013.Data/Outcomes: Country level adolescent fertility rate (AFR), HIV prevalence and mortality rate from 1989/90 to 2013 across 186 low-, middle- and high-income countries.Analysis: Longitudinal mixed effects models, entering secondary and primary education together, adjusted for time varying GDP and country income status. Longitudinal structural marginal models using inverse probability weighting (IPW) to take account of time varying confounding by primary education and GDP. Counterfactual scenarios of no change in secondary education since 1980/1990 were estimated from model coefficients for each outcome. FINDINGS: Each additional year of secondary education decreased AFR by 8.4% in mixed effects models and 14.6% in IPW models independent of primary education and GDP. Counterfactual analyses showed the proportion of the reduction in adolescent fertility rate over the study period independently attributable to secondary education was 28% in low income countries. Each additional year of secondary education reduced mortality by 16.9% for 15-19 year and 14.8% for 20-24 year old young women and 11.4% for 15-19 year and 8.8% for 20-24 year old young men. Counterfactual scenarios suggested 12% and 23% of the mortality reduction for 15-19 and 20-24 year old young men was attributable to secondary education in low income countries. Each additional year of secondary education was associated with a 24.5% and 43.1% reduction in HIV prevalence amongst young men and women. INTERPRETATION: The health benefits associated with secondary education were greater than those of primary education and were greatest amongst young women and those from low income countries. Secondary education has the potential to be a social vaccine across many outcomes in low and middle income countries.