Sleep ultradian cycling: Statistical distribution and links with other sleep variables, depression, insomnia and sleepiness-A retrospective study on 2,312 polysomnograms.
Level 4 - case-series / case-control
Retrospective cross-sectional analysis of hospital polysomnography records
PubMed 30819535 · doi:10.1016/j.psychres.2018.12.141
What was done
Researchers analyzed polysomnograms from 2,312 unmedicated patients admitted for medical or psychiatric reasons to Erasme University Hospital between 2003 and 2014. They tested the statistical distribution of sleep ultradian cycle metrics (number of cycles and mean cycle duration) using Q-Q plots and examined associations with sleep architecture variables (REM latency, REM duration, REM/NREM ratio), Major Depressive Disorder diagnosis, Beck Depression Inventory short version scores, and Insomnia Severity Index scores.
What was found
The number of cycles per night and mean cycle duration showed a normal distribution, with Q-Q plots close to linearity. Both variables showed predominant links with REM sleep-related metrics (REM latency, REM sleep duration, and REM/NREM ratio). Neither cycle count nor mean cycle duration was associated with a diagnosis of Major Depressive Disorder or depression intensity on the Beck Depression Inventory. Major Depressive Disorder diagnosis was significantly associated with the Insomnia Severity Index and correlated with depressive symptom intensity. The abstract reported no numerical values, effect sizes, or p-values.
Why it matters
Confirming the normal distribution of sleep cycle metrics in a large clinical cohort validates parametric testing for these variables in sleep research. It also indicates that macro-level ultradian cycle count and duration are not direct markers of major depression.
Limits
The study design is retrospective and limited to a single-center inpatient cohort admitted for mixed medical and psychiatric reasons, which may not generalize to healthy populations. The abstract does not provide exact numerical values, effect sizes, confidence intervals, or p-values, and it does not detail specific medical comorbidities or age distributions.
Cited by
- supports Human sleep cycles repeat approximately every 90 minutes across the night.