Spring Forward and Fall Back in Health: The Effect of Daylight Saving Time on Acute Myocardial Infarction

Since its inception, the enduring and contentious debate over the practice of daylight saving time (DST) has centered on its potential effect on human health. However, surprisingly little reliable quantitative evidence is available to date. In this paper, we address this issue by examining a relationship between DST and acute myocardial infarction (AMI) in Indiana. Existing studies explored a relationship between DST and AMI by looking at changes in hospital admissions over the days shifting in or out of DST. However, the estimated DST effects based on such a research design are likely to erroneously pick up a seasonality effect, since the timing of DST is by construction commensurate with seasonal transition, due to its original objective to save energy using extended daylight during summer. In an effort to circumvent such an identification issue and establish a causal relationship, we 1) exploit a unique experience in Indiana that allows us to employ a regression-discontinuity difference-in-differences (RDDD) strategy, which directly tests the seasonality effect (and correctly controls for it if any) by using counties that did not practice DST during the same transitional periods as counterfactual, 2) use one of the most comprehensive dataset ever used in this literature that covers the universe of AMI patient-level information from 2002 through 2012, and 3) offer the first evidence on the effect on disease severity. We find evidence that DST has a significant effect on AMI among working-age population. In particular, we find that DST led to sharp increases in admissions among inpatients. The estimated effects are larger in magnitude among males than females and after the spring transition relative to the autumn transition.