Sleep Does Not Explain Racial Gaps in Grades and Absences Among Georgia Ninth Graders

A new study of ninth-grade students in semi-rural Georgia has found striking racial and ethnic differences in academic performance and school attendance, but, perhaps surprisingly, neither sleep duration nor chronotype explained those gaps. The research, published in the Journal of Clinical Sleep Medicine, offers a careful statistical test of a hypothesis that has gained traction in recent years: that disparities in adolescent sleep might be an important, modifiable driver of educational inequities. The answer, at least in this sample, appears to be no.

The study, led by Henok Mulatu Teferi of Emory University’s Rollins School of Public Health together with co-first author Gina Marie Mathew of Stony Brook University and colleagues, analyzed data from 406 ninth graders enrolled at two racially and ethnically diverse high schools. Ninth grade is a pivotal year in the American education system; it marks the transition into courses that count toward graduation and a period when absenteeism patterns can harden into chronic disengagement. Prior research has documented racial and ethnic disparities in both grade point averages and school absences, and a large body of work links sleep, in terms of its duration, timing, and quality, to how well adolescents perform in school. What remained unclear was whether sleep differences between groups actually account for a meaningful share of the achievement and attendance gaps.

To answer that question, the researchers employed counterfactual-based mediation analysis, a statistical framework that asks how much of the total effect of an exposure on an outcome flows through an intermediate variable, the mediator. In this case, race and ethnicity, categorized as Asian, Black, Hispanic, or White, served as the exposure. Two sleep measures acted as candidate mediators: average sleep duration and mid-sleep on free days corrected for sleep debt, abbreviated MSFsc, a widely used proxy for chronotype developed by chronobiologist Till Roenneberg and colleagues. Chronotype captures whether a person’s internal circadian clock leans toward morningness or eveningness, a trait that shifts markedly later during adolescence as puberty advances.

The outcomes were analyzed with separate statistical models. Grade point average, treated as a continuous measure, was modeled using linear regression, while school absences, which are count data, were modeled using Poisson regression, an approach suited to non-negative integer outcomes. All models adjusted for sociodemographic covariates. This dual-model design allowed the team to estimate, separately for GPA and absences, whether the sleep measures carried any of the association between race or ethnicity and academic outcomes.

The total effects were substantial and statistically significant. Compared with White students, Black students had lower GPAs, with a regression coefficient of negative 2.70 and a 95 percent confidence interval spanning negative 5.06 to negative 0.42. Asian students, by contrast, had higher GPAs than their White peers, with a coefficient of 5.46 and a confidence interval of 2.44 to 8.10. Attendance patterns diverged as well. Asian students were less likely to be absent than White students, with a relative risk of 0.51, and Hispanic students also showed lower absence rates, with a relative risk of 0.59. In other words, Asian students in this sample were roughly half as likely to miss school as White students, and Hispanic students about 40 percent less likely.

Yet when the researchers tested whether sleep duration or MSFsc mediated these differences, the mediation effects were not statistically significant for either GPA or absences. The sleep measures did not carry the association between race or ethnicity and academic outcomes. This null mediation result is the study’s central finding, and it carries real weight precisely because the underlying disparities were clearly present in the data. The authors conclude that factors beyond these two sleep dimensions likely contribute to the observed academic inequities.

The result complicates a popular narrative. Sleep is undeniably important for adolescents: the National Sleep Foundation recommends eight to ten hours per night for teenagers, and meta-analytic reviews have consistently linked shorter sleep duration, poorer sleep quality, and greater sleepiness to weaker school performance. Later school start times have been shown in actigraphic studies to lengthen adolescent sleep, and delayed circadian phase in puberty is well established. Given documented racial and ethnic sleep disparities among United States children and adolescents, and evidence that discrimination is associated with disrupted sleep, the idea that sleep might partially explain achievement gaps was plausible and, importantly, testable. Interventions targeting sleep, such as adjusted start times or sleep hygiene education, would be comparatively tractable compared with the deeper structural forces shaping educational outcomes.

But this study suggests that such interventions, at least as measured here, would not close the GPA or attendance gaps on their own. Several technical considerations temper the interpretation. The design was cross-sectional, capturing a single snapshot rather than tracking students over time, so causal claims about mediation are inherently limited. The sleep measures were not objective; actigraphy or polysomnography might reveal differences that self-reported or estimated duration and chronotype proxies miss. The authors themselves call for longitudinal studies using objective sleep assessments. Additionally, the sample comprised 406 students from two schools in semi-rural Georgia, a setting that may not generalize to urban districts or other regions, and the mediation framework tested only two candidate mediators out of a vast space of possible pathways.

The broader implications point researchers toward other explanatory candidates. Chronic absenteeism has been linked to school climate, disciplinary practices, socioeconomic conditions, health access, and family circumstances, and prior work on racial disparities in discipline suggests that school-level factors, not just individual behavior, shape attendance and achievement. Sleep quality, sleep disorders, environmental disruptions such as noise and light exposure, and stress from discrimination may all matter in ways that simple duration and chronotype measures do not capture. The study, funded by the Eunice Kennedy Shriver National Institute of Child Health and Human Development and the National Heart, Lung, and Blood Institute, does not diminish the case for healthy adolescent sleep; it simply shows that the arithmetic of educational inequity cannot be reduced to hours slept or a preference for evenings.

For educators and policymakers, the message is twofold. Sleep remains a legitimate target for improving adolescent health and functioning across all groups, and later start times still offer documented benefits for sleep duration. But closing racial and ethnic gaps in grades and attendance will require attention to the social and structural determinants that this study’s null mediation result implicitly highlights. The Georgia ninth graders in this dataset arrived at school with differences in performance and presence that their sleep clocks, however measured, could not account for, and that finding should redirect both scientific inquiry and intervention design toward the fuller set of forces shaping who thrives in American high schools.

Subject of Research: Racial and ethnic disparities in academic performance and absenteeism among adolescents and the role of sleep duration and chronotype

Article Title: Racial/ethnic disparities in grade point averages and absences among 9th graders in Georgia: the role of sleep differences

Article References: Teferi, H. M., Mathew, G. M., Hale, L., Johnson, D. A., & Gazmararian, J. A. (2026). Racial/ethnic disparities in grade point averages and absences among 9th graders in Georgia: the role of sleep differences. Journal of Clinical Sleep Medicine, 22(1), Article 182. https://doi.org/10.1007/s44470-026-00187-4

Image Credits: AI Generated

DOI: 10.1007/s44470-026-00187-4

Keywords: sleep duration, chronotype, grade point average, school absences, racial disparities, adolescents, mediation analysis, Georgia, sleep health, academic achievement, ninth grade, circadian rhythms