Shorter total sleep time on the night before cognitive testing was associated with higher odds of mild cognitive impairment (MCI) reversion after 1 year, suggesting acute sleep disruption may transiently impair cognition and contribute to MCI misclassification.
Key Findings
Results
Shorter prior-night total sleep time (TST) was associated with higher odds of MCI reversion at 1-year follow-up.
Logistic regression assessed whether prior-night TST and sleep efficiency (SE) predicted reversion, adjusting for age and sex.
Sleep was objectively measured using actigraphy at baseline.
Only prior-night TST showed a significant association with reversion; sleep efficiency showed no significant association.
Random-night sleep measures and 14-day average sleep measures showed no significant associations with MCI reversion.
Results
Approximately 27.6% of MCI participants reverted to normal cognition at 1-year follow-up.
16 out of 58 community-dwelling older adults with MCI reverted to normal cognition.
Mean participant age was 75 years.
MCI was defined as cognitive performance ≥ 1.5 standard deviations below norms in at least one domain, with preserved daily functioning.
Participants completed cognitive testing at both baseline and 1-year follow-up.
Results
Single-domain MCI cases were more likely to revert than multi-domain MCI cases.
Greater susceptibility to acute factors such as poor sleep was suggested as an explanation for higher reversion rates in single-domain MCI.
This finding implies that single-domain MCI diagnoses may be less stable and more influenced by transient state factors like sleep disruption.
The pattern is consistent with the hypothesis that acute sleep disruption may transiently impair cognition and contribute to MCI misclassification.
Results
Sleep efficiency (SE) on the night before cognitive testing showed no significant association with MCI reversion.
Both TST and SE were examined as predictors of MCI reversion using logistic regression.
SE was measured objectively via actigraphy.
Unlike TST, SE on the prior night did not predict whether a participant would revert to normal cognition at follow-up.
Chronic and random-night SE measures also showed no associations with reversion.
Results
Chronic and random-night sleep measures were not associated with MCI reversion, suggesting the effect is specific to acute prior-night sleep disruption.
14-day average sleep measures (reflecting chronic sleep patterns) showed no significant associations with MCI reversion.
Random-night sleep measures also showed no significant associations.
Only the specific night immediately before cognitive testing was associated with reversion outcomes.
This pattern suggests that the mechanism operates through transient, state-dependent cognitive impairment rather than long-term sleep health.
Discussion
The authors propose that brief screening of prior-night sleep before cognitive testing could improve diagnostic reliability in MCI.
Sleep screening before testing may help identify low-reliability assessments.
Actigraphy-derived metrics are suggested as a practical tool for this purpose.
The authors frame acute sleep disruption as a potential contributor to MCI misclassification rather than true cognitive recovery.
Improving diagnostic stability in MCI research is cited as a practical implication of these findings.
What This Means
This research suggests that how well an older adult sleeps the night before a memory and thinking test can influence whether they appear to have mild cognitive impairment (MCI) or normal cognition. In a study of 58 older adults with MCI, those who slept fewer hours the night before their baseline cognitive test were more likely to appear cognitively normal when retested a year later — a pattern called 'reversion.' Nearly 28% of participants reverted to normal cognition at the one-year follow-up, which is a substantial proportion.
Importantly, the association was specific to the single night immediately before testing. Average sleep over two weeks, or sleep on random nights, was not linked to reversion. This specificity suggests the effect is not about overall sleep health, but rather that a poor night of sleep on the eve of testing can temporarily worsen cognitive performance, leading to an MCI diagnosis that may not reflect a person's true, stable cognitive status. People with impairment in only one cognitive domain (single-domain MCI) appeared especially susceptible to this effect compared to those with impairment across multiple domains.
This research suggests that the timing of sleep relative to cognitive testing matters for the accuracy of MCI diagnosis. A person might be classified as having MCI partly because they happened to sleep poorly the night before. The authors propose that routinely checking how well someone slept before administering cognitive tests — using simple tools like wrist-worn activity monitors — could help clinicians and researchers identify cases where a diagnosis may be unreliable, ultimately improving the accuracy of MCI detection and research.
Chang J, Latona G, Rizzo M, Phatak V, Murman D. (2026). Overnight sleep affects the stability of neuropsychological classification in mild cognitive impairment.. Alzheimer's & dementia : the journal of the Alzheimer's Association. https://doi.org/10.1002/alz.71128