A Long-Term Study of Changing High School Start Time

Most Benefits are Lost by Second Year

Posted Feb 15, 2016

An increasing number of school districts across the U.S. are changing school start times from earlier to later. It is important to know how the changes directly affects] students, but in most cases, the changes have been too recent for very much evaluation of the effects of the changes to have been done. Any new study that comes out will be given a lot of scrutiny. One study was published recently in the journal Sleep, (Thacher & Onyper, 2016). The authors studied a group of students at a single high school in upstate New York where school start times were delayed from 7:45 AM to 8:30 AM beginning in the 2012-13 school year. Data from the final two years of the earlier start (2010-11 and 2011-12) were compared with the first two years of the later start (2012-13 and 2013-14).  Outcome measures included attendance, tardiness, discipline, and academic performance. They also measured sleep, sleepiness, circadian preference, and general health.

In the first year, students on average were sleeping 20 minutes longer and while they went to bed at about the same time as before, there was a 20 minute delay in morning rise times. They also showed fewer sleep problems and reported falling asleep more quickly. But at the end of year two, students’ bedtimes were delayed while rise times remained stable. So sleep duration returned to what they were before school started later. Sleep problems were somewhat higher at the end of year two compared with the end of year one. Students did not report significant changes in daytime sleepiness. These results are based on averages, and substantial differences from person to person were noted, with some students choosing to extend their sleep more than others.  When attendance and tardiness were looked at, there was little effect on attendance, but unexcused tardiness (that are often due to oversleeping) dropped substantially.  Discipline violations also decreased markedly over the two-year period. Finally, neither student grades nor standardized test scores showed any change.

It should be emphasized that we are still in the early stages of evaluation of effects of changing start times, and I expect many more studies to appear in the near future.  Making firm conclusions at this point would be comparable to predicting this year’s presidential candidates based on the Iowa caucuses and the New Hampshire primary. We still have a long way to go. The school district in this study was in a rural area with only 14% of students living in families below the poverty line. Further, the district is comprised of 94% Anglo-American students. As some of the studies from our lab have shown, poor sleep seems to adversely affect children more if they are in lower income families. Students in the New York school also were different from many districts in the U.S. in that they walk or get to school in parents’ cars rather than ride the bus. This situation is vastly different from many districts across the country where students often have long bus rides that must begin sometimes up to an hour before the start of school.

Clearly, these results are disheartening to those (including myself), who have hoped that delaying start times would have stronger results across many different measures.  The results may give the right to those in many districts who have argued against delaying start times to say “I told you so!”.  In their discussion of the results, the authors say that there was little “buy-in” in the community and perhaps more education should have been done to help parents and students understand the benefits of getting more sleep.  I agree, and as I have stated more than once in this blog, changing school start times in itself will not necessarily lead adolescents to sleep more or sleep better. Students in the sample were sleeping just over 7 hours on school nights, clearly not enough when recommendations are that they should be sleeping closer to 9 hours. And it is worthwhile to remember that individual differences were apparent, such that some students did increase their sleep times. 

Thacher P.V., & Onyper S.V. (2016). Longitudinal outcomes of start time delay on sleep, behavior, and achievement in high school. Sleep 39, 271–281.