Asia Health Policy Program working paper # 38
February 10, 2014
Despite a large number of observational studies consistently reporting the association between shorter sleep duration and higher body weight, causality has yet to be established at a population level. This study aims to estimate the population-level causal effect of sleep duration on adolescent body weight, using an instrumental variable (IV) approach that exploits a unique natural experiment in the context of South Korea’s highly competitive secondary education. In March 2011, amid growing concerns over the negative consequences of late-night tuition at private tutoring institutes (hagwon), authorities in 3 of the 16 administrative regions in South Korea decreed adjusting the closing hours of hagwon to 10 p.m. This policy change caused a substantial and plausibly exogenous variation in the sleep duration of the “marginal student,” whose sleep duration is most likely to be affected by the policy. The IV estimation results on a sample of general high school 10th- and 11th-graders in the 2009−2012 Korea Youth Risk Behavior Web-based Survey show that a 1-hour increase in sleep duration led to a 0.56 kg/m2 reduction in body mass index, or a 4.3 percentage-point decrease in overweight/obesity. Short sleep duration among adolescents may be an important contributor to increased body weight at the population level.