Sleep health · 2026
Differential impact of school start times: Exploring heterogeneous treatment effects of delaying school start times on sleep and BMI in the START study.
СтатьяНазвание и аннотация: Перевод в очереди
Обработка статьи · этапы, ошибки и управление
Загрузка состояния…
Ошибки и ограничения
История · было → стало
Состояние обновляется каждые 5 секунд, пока блок открыт. Процент показывается только там, где известен объём работы. Пауза останавливает продолжение этапа; текущий запрос к источнику может завершиться.
Billmyer Emma, Petersen Ashley, Erickson Darin J, Wahlstrom Kyla, Laska Melissa N, Widome Rachel
Аннотация
PURPOSE: We aimed to demonstrate a method for identifying subgroups that may have experienced heterogeneous effects of a delayed high school start time policy. This approach could suggest how the implementation of a policy might be prioritized.
METHODS: In the START study, students (n = 2230) from 5 schools in the Twin Cities, Minnesota metropolitan area were followed over 3 waves of annual data collection. All schools started early (7:30 or 7:45 am) at baseline. Two "policy change schools" had set start times later by roughly 1 hour by follow-up 1; 3 "comparison schools" maintained their early start time throughout the study. We employed generalized linear mixed model trees, a data-driven method that detects potential treatment-subgroup interactions while accounting for the clustered nature of the data. The stability of variable selection was assessed using a subsampling approach.
RESULTS: Among Asian and Black students whose parents did not complete college (a subgroup identified by the generalized linear mixed model trees), those who attended policy change schools gained over 1 hour and 15 minutes of sleep per night, relative to comparison school students. The other subgroups had lesser treatment effects. With body mass index, minimal differences were observed between subgroups identified by generalized linear mixed model trees.
CONCLUSION: We showed how the generalized linear mixed model trees method can enable identification of groups that may experience differing impacts from a population-based policy. However, this analysis was meant primarily as a demonstration and the results should be seen as preliminary and hypothesis generating, due to the small number of participants in some subgroups.
METHODS: In the START study, students (n = 2230) from 5 schools in the Twin Cities, Minnesota metropolitan area were followed over 3 waves of annual data collection. All schools started early (7:30 or 7:45 am) at baseline. Two "policy change schools" had set start times later by roughly 1 hour by follow-up 1; 3 "comparison schools" maintained their early start time throughout the study. We employed generalized linear mixed model trees, a data-driven method that detects potential treatment-subgroup interactions while accounting for the clustered nature of the data. The stability of variable selection was assessed using a subsampling approach.
RESULTS: Among Asian and Black students whose parents did not complete college (a subgroup identified by the generalized linear mixed model trees), those who attended policy change schools gained over 1 hour and 15 minutes of sleep per night, relative to comparison school students. The other subgroups had lesser treatment effects. With body mass index, minimal differences were observed between subgroups identified by generalized linear mixed model trees.
CONCLUSION: We showed how the generalized linear mixed model trees method can enable identification of groups that may experience differing impacts from a population-based policy. However, this analysis was meant primarily as a demonstration and the results should be seen as preliminary and hypothesis generating, due to the small number of participants in some subgroups.