Consecutive day effects between sleep quality and affective symptoms among youth in the Brazilian High-Risk Cohort study
Varidel, M. R., Borgnolo, L., An, V., Carpenter, J. S., Hickie, I. B., Pan, P. M., da Silva, F., Crouse, J. J., et al.
psychiatry and clinical psychology
Quick verdict
Claims cumulative consecutive-night sleep effects on youth affect via causal EMA analysis; directionally plausible but key evidence is statistically weak.
Summary
This preprint analyses smartphone ecological momentary assessment (EMA) data from 88 young people (mean age 17.8; 63.6% female) from the Brazilian High-Risk Cohort, collected April 2020 to May 2021 during COVID-19. Participants rated sleep quality each morning (1-5) and six affective symptoms (anxiety, irritability, sadness, overall mood, inactivity, tiredness; 1-7, higher=poorer) four times daily for up to 28 days, averaged to daily means. Using Bayesian mixed-effects models with person-mean centring, random intercepts and slopes, the authors estimate bidirectional next-day effects, incremental effects of two-night-lagged sleep quality (two-step sequential regression after Loh & Ren 2023), and simulate four two-night sleep-quality scenarios. Better sleep quality predicted lower next-day symptoms on five of six measures (largest: tiredness ATE -0.33), while affect predicted sleep only for anxiety (-0.11) and irritability (-0.12). Lagged sleep quality showed smaller two-sigma effects (e.g., tiredness -0.17). Simulated stable-good versus stable-poor sleep favoured good sleep (overall mood delta -2.10), but all mixed-scenario contrasts were null. Key limitations: small, sparse sample (454 observations in the lagged analysis; median 3 complete days/person for day-level analysis), self-rated sleep biased by mood, COVID-era data, untestable confounding assumptions, and consecutive-day findings that do not survive the authors' own Bonferroni threshold. A novel, carefully executed but statistically fragile first look at consecutive-day sleep-affect dynamics.
Review complete
Assessment at a glance
5
risk flags identified
● high 1
● medium 3
● low 1
5 strengths · 5 weaknesses · 5 author questions
Levels reflect verification complexity — not a quality score. Evolve never rates your paper.
Key claims under review
- Within-person next-day effects between sleep quality and affective symptoms are bidirectional, with sleep quality exerting roughly twice the influence of affect on sleep (e.g., sleep-to-anxiety ATE -0.23 vs. anxiety-to-sleep -0.11).
- Sleep quality two nights prior has an incremental effect on current-day affective symptoms beyond the prior night's sleep (e.g., tiredness -0.17, sadness -0.14), implying cumulative consecutive-day effects.
- Simulated two-night interventions show stable good sleep improves all six affective symptoms versus stable poor sleep (e.g., overall mood delta -2.10, SE 1.03), with mixed improving/worsening scenarios intermediate.
- These dynamics are interpreted as a self-reinforcing sleep-affect feedback loop that circadian-targeted, sustained sleep interventions could break in youth.
Strengths
Novel question with clear gap
First EMA study (per authors) to quantify incremental effects of two-nights-lagged subjective sleep quality on affect, complementing experimental sleep-restriction literature (Dinges 1997; Booth 2021) that used duration, not quality.
Careful within-person modelling
Person-mean centring, random intercepts and random slopes, Bayesian estimation (brms/rstanarm) with explicit convergence criteria (R-hat<1.01, >=100 effective samples/chain) and uncertainty propagated across estimation steps via simulation.
Transparent evidence thresholds
Consistent, interpretable direction of effects
Selection-bias checking
Weaknesses
Headline consecutive-day claim exceeds the evidence
Sparse data strain the models
Causal language outruns design
Self-report circularity
Asymmetric adjustment across directions
Full Review Report
$1.99 · one-timeUnlock the complete preliminary analysis Evolve produced for this paper.
- All 5 strengths & 5 weaknesses, analyzed in detail
- 5 risk flags with severity, section & reviewer questions
- Methods & statistical evidence assessment
- 5 questions a peer reviewer would ask
- Full six-dimension evaluation report & recommendations
One-time purchase · this paper only
One-time purchase, this paper only. AI-generated preliminary analysis — not peer review.
Frequently asked questions
What did this study find about sleep and mood in young people?
In 88 Brazilian youth (ages ~14-22) completing up to 28 days of smartphone daily reports, better sleep quality was followed by lower next-day anxiety, sadness, fatigue, inactivity and overall negative mood; the effect of sleep on mood was larger than the reverse. Sleep quality from two nights earlier also showed smaller additional effects, which the authors interpret as cumulative ('consecutive-day') effects, though these weaker findings did not pass strict statistical correction.
Does this study prove that improving sleep improves mental health?
No. It is an observational smartphone study using causal-inference-style regression and simulated interventions. The simulated 'stable good sleep vs stable poor sleep' comparison favoured good sleep, but the more nuanced consecutive-night scenarios were statistically inconclusive, sleep was self-rated (and mood can bias those ratings), and unmeasured confounding cannot be ruled out. Randomised trials (e.g., sleep-improvement interventions) remain the stronger evidence, and they do generally support sleep's benefit for mental health.
Who was studied and when?
A subsample of the Brazilian High-Risk Cohort (BHRC): 88 young people with a mean age of 17.8 years (63.6% female), recruited in Sao Paulo and Porto Alegre, who completed morning sleep ratings and four affect surveys per day for up to 28 days between April 2020 and May 2021 - a period overlapping the COVID-19 pandemic, which the authors note limits generalisability.
What are the main limitations of the paper?
A small analytic sample (88 of 271 eligible) with sparse adherence (median 3 complete days per person for the core analysis; 454 observations in the lagged analysis), a single subjective sleep-quality item that mood can bias, no objective sleep or circadian measures, day-level averaging that discards within-day dynamics, COVID-era data collection, and key consecutive-day results that reach only conventional significance rather than the study's stricter multiple-comparison threshold.
About this review
This is an AI-generated preliminary review by Evolve.Pub, produced from the openly available text of the paper. It is not peer review and does not represent the authors' views. The canonical record of the paper is at the publisher link above.
- Review based on extracted preprint text; Figure 1 and supplementary materials (S2 mathematical detail, S3 selection flow, S4-S6 fixed-effects tables, S7 interaction analyses) were not available, so modelling detail and some judgments could not be verified.
- Table values were transcribed through text extraction; narrative and tables were cross-checked, but minor transcription errors are possible.
- This is a non-peer-reviewed medRxiv preprint (posted September 16, 2026); the DOI (10.64898/...) is unusual and could not be verified offline.
Go deeper on this paper
Wrote a paper? Stress-test it before reviewers do.
Upload your manuscript for the same structured scrutiny — reasoning, statistics, evidence, figures, and reproducibility.