CoChat

The DOI matches its Crossref record; 63 of the 64 references it cites were found in a registry or an index; 9 of the 24 sources its claims cite were checked for support.

Irregular sleep/wake patterns are associated with poorer academic performance and delayed circadian and sleep/wake timing

Open access

Abstract

The association of irregular sleep schedules with circadian timing and academic performance has not been systematically examined. We studied 61 undergraduates for 30 days using sleep diaries, and quantified sleep regularity using a novel metric, the sleep regularity index (SRI). In the most and least regular quintiles, circadian phase and light exposure were assessed using salivary dim-light melatonin onset (DLMO) and wrist-worn photometry, respectively. DLMO occurred later (00:08 ± 1:54 vs. 21:32 ± 1:48; p < 0.003); the daily sleep propensity rhythm peaked later (06:33 ± 0:19 vs. 04:45 ± 0:11; p < 0.005); and light rhythms had lower amplitude (102 ± 19 lux vs. 179 ± 29 lux; p < 0.005) in Irregular compared to Regular sleepers. A mathematical model of the circadian pacemaker and its response to light was used to demonstrate that Irregular vs. Regular group differences in circadian timing were likely primarily due to their different patterns of light exposure. A positive correlation (r = 0.37; p < 0.004) between academic performance and SRI was observed. These findings show that irregular sleep and light exposure patterns in college students are associated with delayed circadian rhythms and lower academic performance. Moreover, the modeling results reveal that light-based interventions may be therapeutically effective in improving sleep regularity in this population.

Overview

CoChat Summary

TL;DR

In 61 Harvard undergraduates studied for about 30 days, more irregular sleep/wake patterns were associated with later circadian timing and worse academic performance. The paper argues that the timing differences were largely explained by differences in light exposure rather than by detectable differences in circadian physiology.

Key findings

  • Sleep regularity index (SRI) was positively correlated with GPA in the full sample (n = 59): Pearson r = 0.37, p < 0.004; an increase of 10 in SRI was associated with an average GPA increase of 0.10.
  • Irregular sleepers had significantly later melatonin timing than Regular sleepers: DLMO was 00:08 ± 1:54 vs. 21:32 ± 1:48, p < 0.003; the sleep propensity rhythm also peaked later (06:33 ± 0:19 vs. 04:45 ± 0:11, p < 0.005).
  • Irregular sleepers had lower-amplitude light rhythms than Regular sleepers: 102 ± 19 lux vs. 179 ± 29 lux, p < 0.005, with more light during biological night and less daytime light.
  • Sleep duration did not differ significantly between groups (7.16 ± 0.64 h vs. 7.27 ± 0.59 h, p = 0.68), and sleep regularity was uncorrelated with sleep duration (r = 0.05, p = 0.71 in the full sample).

Methods

  • Observational study of full-time Harvard College undergraduates (excluding first-years), age 18+, living in campus housing; 63 enrolled, 61 completed the study. Participants reported self-selected sleep/wake schedules for ~30 days via online diary during Fall 2013.
  • Sleep regularity was quantified with the Sleep Regularity Index (SRI); on day 15, participants were classified using days 1–14, and the highest and lowest quintiles were selected for actiwatch/light and dim-light saliva collection. GPA was self-reported by 59 participants.
  • Main analyses included correlations between SRI and GPA, group comparisons of Regular vs. Irregular sleepers, and a previously validated mathematical model of the human circadian pacemaker to test whether observed DLMO differences could be explained by light exposure patterns. Statistical tests included Wilcoxon rank-sum tests, t-tests, and Pearson correlations.

Limitations

  • The paper states that it cannot establish causality between sleep regularity and academic performance or circadian timing.
  • DLMO was measured on only one day, which the paper notes may limit prediction of individual circadian timing and may contribute to model-data mismatch; this is especially relevant for Irregular sleepers, and this limitation is explicitly stated in the Discussion.
  • The sample was a specific population of Harvard undergraduates living on campus, so the authors note the daytime napping strategy and findings may not generalize to adults who work full time or students not living near campus; this is evident from the text.

Model-written summary. Check it against the paper before relying on it.

Integrity signals

Registry

Crossref · 10.1038/s41598-017-03171-4

Cites

63 of 63 references verified0 citation errors · 0 not checkable

Statistics

Nothing to recompute.

Figures

Figure 1 Two dimensions of sleep: duration and regularity. (A) Average daily sleep duration vs. sleep regularity index (SRI) for all participants (n = 61) assessed across the whole study interval. Participants identified for the Irregular (n = 12) and Regular (n = 12) groups at the study mid-point are red circles and blue squares, respectively. Other individuals are white triangles. As explained in the Methods, the SRI values differed slightly by end of study, so those identified as most extreme at the study midpoint did not necessarily remain most extreme at end of study; however, the differences between Regular and Irregular groups remained highly significant (see Results). Error bars indicate mean and standard deviation for Regular and Irregular groups in both SRI and sleep duration. Sleep patterns for four participants collected using daily diaries are shown using double-plotted raster diagrams, where black bars indicate episodes of sleep and gray bars indicate missing data. Four examples are displayed: (B) an Irregular long sleeper, (C) an Irregular short sleeper, (D) a Regular long sleeper, and (E) a Regular short sleeper.
Figure 1 Two dimensions of sleep: duration and regularity. (A) Average daily sleep duration vs. sleep regularity index (SRI) for all participants (n = 61) assessed across the whole study interval. Participants identified for the Irregular (n = 12) and Regular (n = 12) groups at the study mid-point are red circles and blue squares, respectively. Other individuals are white triangles. As explained in the Methods, the SRI values differed slightly by end of study, so those identified as most extreme at the study midpoint did not necessarily remain most extreme at end of study; however, the differences between Regular and Irregular groups remained highly significant (see Results). Error bars indicate mean and standard deviation for Regular and Irregular groups in both SRI and sleep duration. Sleep patterns for four participants collected using daily diaries are shown using double-plotted raster diagrams, where black bars indicate episodes of sleep and gray bars indicate missing data. Four examples are displayed: (B) an Irregular long sleeper, (C) an Irregular short sleeper, (D) a Regular long sleeper, and (E) a Regular short sleeper.CC BY · PubMed Central
Figure 2 Sleep/wake and light/dark cycles differ between Regular and Irregular groups. Gray lines show individual data (n = 12 for each sleep/wake panel and n = 11 for each light/dark panel). Colored lines show group mean and standard deviation in one-hour bins, with data for each individual averaged across the whole study interval (i.e., multiple days). Dark gray bars indicate clock night (22:00 to 10:00). Left panels: Sleep/wake rhythm (percentage of time asleep) for (A) Regular and (B) Irregular sleepers. Right panels: Normalized light levels for (C) Regular and (D) Irregular sleepers.
Figure 2 Sleep/wake and light/dark cycles differ between Regular and Irregular groups. Gray lines show individual data (n = 12 for each sleep/wake panel and n = 11 for each light/dark panel). Colored lines show group mean and standard deviation in one-hour bins, with data for each individual averaged across the whole study interval (i.e., multiple days). Dark gray bars indicate clock night (22:00 to 10:00). Left panels: Sleep/wake rhythm (percentage of time asleep) for (A) Regular and (B) Irregular sleepers. Right panels: Normalized light levels for (C) Regular and (D) Irregular sleepers.CC BY · PubMed Central
Figure 3 Melatonin secretion is delayed in Irregular sleepers. Top panel: (A) Timing of salivary dim light melatonin onset (DLMO) for Regular (n = 12) and Irregular (n = 12) groups. Individuals in each group are shown as dots, with y-axis position jittered for visibility. Groups means (triangles) and standard deviations (error bars) are shown. Middle two panels: Time course for salivary melatonin concentration are shown for Regular participants in (B) and Irregular participants in (C), along with average sleep midpoint times for the nighttime sleep block for each individual (black dots), with y-axis position jittered for visibility. Gray lines show individuals in 1-h bins. Colored lines with error bars show group mean and standard deviation in 1-h bins. Bottom panel: (D) Actual timing of DLMO vs. model prediction for timing of DLMO, assessed using saliva. Data points correspond to individuals in the Regular (blue square) and Irregular (red circle) groups. Error bars show mean and standard deviation for each group. Differences in group averages are displayed. The dashed lines show linear regressions for each group.
Figure 3 Melatonin secretion is delayed in Irregular sleepers. Top panel: (A) Timing of salivary dim light melatonin onset (DLMO) for Regular (n = 12) and Irregular (n = 12) groups. Individuals in each group are shown as dots, with y-axis position jittered for visibility. Groups means (triangles) and standard deviations (error bars) are shown. Middle two panels: Time course for salivary melatonin concentration are shown for Regular participants in (B) and Irregular participants in (C), along with average sleep midpoint times for the nighttime sleep block for each individual (black dots), with y-axis position jittered for visibility. Gray lines show individuals in 1-h bins. Colored lines with error bars show group mean and standard deviation in 1-h bins. Bottom panel: (D) Actual timing of DLMO vs. model prediction for timing of DLMO, assessed using saliva. Data points correspond to individuals in the Regular (blue square) and Irregular (red circle) groups. Error bars show mean and standard deviation for each group. Differences in group averages are displayed. The dashed lines show linear regressions for each group.CC BY · PubMed Central
Figure 4 Example of model inputs and outputs for one participant. Variables are shown for days 280–300 of a 300-day simulation for one participant from the Regular group. (A) Binned light levels in lux. (B) The two circadian pacemaker variables, x (blue) and x c (red). (C) The predicted salivary melatonin concentration. (D) The clock-time of Dim Light Salivary Melatonin Onset (DLSMO) on each day.
Figure 4 Example of model inputs and outputs for one participant. Variables are shown for days 280–300 of a 300-day simulation for one participant from the Regular group. (A) Binned light levels in lux. (B) The two circadian pacemaker variables, x (blue) and x c (red). (C) The predicted salivary melatonin concentration. (D) The clock-time of Dim Light Salivary Melatonin Onset (DLSMO) on each day.CC BY · PubMed Central
Figure 5 Correlations between sleep regularity index (SRI), grade point average (GPA), and timing of melatonin secretion. Panels (A, B and C) show the relationships between the variables: SRI, GPA, and salivary DLMO. Dashed lines show the linear fits, with r-values and p-values shown for each linear (Pearson) correlation. Each data point represents an individual, with colors indicating whether the individual was a member of the Regular (blue), Irregular (red), or neither group (black). Note that DLMO was only assessed in the Irregular and Regular participants.
Figure 5 Correlations between sleep regularity index (SRI), grade point average (GPA), and timing of melatonin secretion. Panels (A, B and C) show the relationships between the variables: SRI, GPA, and salivary DLMO. Dashed lines show the linear fits, with r-values and p-values shown for each linear (Pearson) correlation. Each data point represents an individual, with colors indicating whether the individual was a member of the Regular (blue), Irregular (red), or neither group (black). Note that DLMO was only assessed in the Irregular and Regular participants.CC BY · PubMed Central

Verification

Claims the paper makes

7 Valid · 1 Not valid · 9 Not checked · 75 Own results and interpretations

44 more claims

Create a free account to see all 92 claims and every source verdict.

See all claims — free account
  • Probably not a substantial claim. 32
  • It may state two separate claims as one. 21

Create a free account to see these claims.

Citations

Cites

Show all 64 references

Research works printed but not linked: 1

  1. Eskin, A. The sparrow clock: Behavior of the free running rhythm and entrainment analysis. Thesis (The University of Texas, 1969).

Cited by

See all 647

Metadata

Authors
Phillips, A. J. K., Clerx, W. M., O’Brien, C. S., Sano, A., Barger, L. K., Picard, R. W., Lockley, S. W., Klerman, E. B., & Czeisler, C. A.
Affiliations
  • Brigham and Women's Hospital (US)
  • MIT Media Lab (US)
Corresponding author
Phillips, Andrew J. K.
Venue
Scientific Reports
Published
2017-06-12
Volume
7
Issue
1
Pages
3216-3216
ISSN
2045-2322 (ISSN-L)
OpenAlex
W2622433070
Cited by
647
References
64
Citations by year
2026: 106 · 2025: 106 · 2024: 94 · 2023: 85 · 2022: 81 · 2021: 62 · 2020: 56 · 2019: 39 · 2018: 14 · 2017: 4
FWCI
18.98
Type
journal article
Language
en
Open access
Licence
Locations
  • Scientific Reports (publishedVersion, cc-by)
  • Europe PMC (PubMed Central) (publishedVersion, other-oa)
  • PubMed Central (publishedVersion, cc-by)
  • PubMed (publishedVersion)
  • Digital Access to Scholarship at Harvard (DASH) (Harvard University) (submittedVersion, cc-by)
  • DOAJ (DOAJ: Directory of Open Access Journals) (submittedVersion, cc-by-sa)
  • DSpace@MIT (Massachusetts Institute of Technology) (submittedVersion, cc-by-nc)
  • Surrey Research Insight Open Access (The University of Surrey) (submittedVersion)
Funders
  • National Space Biomedical Research Institute (HFP-02802, HFP02801)
  • National Institutes of Health (NIH K24, NIH-R01, NIH K99, HFP02802, AG09975)
  • National Institute on Aging (P01AG009975)
  • National Heart, Lung, and Blood Institute (K24HL105664, R01HL094654, K99HL119618, R01HL128538, R01HL114088, R00HL119618)
  • National Institute of General Medical Sciences (R01GM105018)
  • National Center for Research Resources (UL1RR025758)
Topics
  • Life Sciences › Neuroscience › Endocrine and Autonomic Systems › Circadian rhythm and melatonin
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Sleep and related disorders
  • Social Sciences › Psychology › Experimental and Cognitive Psychology › Sleep and Work-Related Fatigue
Keywords
irregular sleep schedules, sleep regularity, circadian timing, academic performance, dim light melatonin onset, light exposure, sleep timing, university students
MeSH terms
Academic Performance, Adolescent, Circadian Rhythm, Female, Humans, Light, Male, Melatonin, Models, Theoretical, Sleep, Time Factors, Wakefulness, Sleep Disorders, Circadian Rhythm, Young Adult

About these checks

Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.

The claim verdicts this report counts are experimental: their accuracy is not yet measured.

A language model drew the claims from the paper on 2026-10-10, under version 3 of the claim definition (model deepseek-v4.1-flash, prompt extract-v3).

A second model labelled each claim on 2026-10-10 (model jev-1.13-20260917, definition revision gold-guideline-v0.2).

Cited sources are checked with a language model’s help (gpt-6-sol).

Statistics and arithmetic are recomputed, not judged.

Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.

What counts as a claim ›How we measure confidence ›How we check this ›

Are you an author, or do you see a problem with this page? Request a review

How to cite

Cite CoChat for the checks on this page, and the paper for its own findings.

This page
CoChat. (2026). Citation check: Irregular sleep/wake patterns are associated with poorer academic performance and delayed circadian and sleep/wake timing. https://app.cochat.ai/papers/10.1038/s41598-017-03171-4
The paper
Phillips, A. J. K., Clerx, W. M., O’Brien, C. S., Sano, A., Barger, L. K., Picard, R. W., Lockley, S. W., Klerman, E. B., & Czeisler, C. A. (2017). Irregular sleep/wake patterns are associated with poorer academic performance and delayed circadian and sleep/wake timing. Scientific Reports. https://doi.org/10.1038/s41598-017-03171-4

Data sources

Bibliographic data: Semantic Scholar, OpenAlex, Crossref.

Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).

Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.