Diurnal variation in variables related to cognitive performance: a systematic review.
Level 2 - randomized trial
Systematic review of experimental within-subject and observational chronobiological studies, not randomized controlled treatment trials
PubMed 37589927 · doi:10.1007/s11325-023-02895-0
What was done
A systematic review was conducted to evaluate time-of-day changes in cognitive performance and assess adherence to chronobiological research standards. Database and citation searches identified 2,391 candidate records. Studies were included if they evaluated healthy adult males at a minimum of two time points (including morning and evening) using objective cognitive performance measures in peer-reviewed journals. Eleven studies met all inclusion criteria.
What was found
Cognitive performance varied across the day, with the direction and magnitude of changes depending on the specific cognitive domain. Diurnal differences ranged from 9.0% to 34.2% for reaction time, 7.3% for alertness, and 7.8% to 40.3% for attention. Where time-of-day effects were detected, reaction time and accuracy were highest in the evening. However, several studies and variables found no significant time-of-day differences, reflecting inconsistent findings across the literature.
Why it matters
This review demonstrates that diurnal cognitive variations are domain-specific and highlights that time of day must be controlled in cognitive testing. It also reveals that chronobiological research in this domain currently lacks standardisation.
Limits
The review was restricted exclusively to healthy adult males, precluding generalisation to females or clinical cohorts. Only 11 studies met eligibility criteria, and none met all standard methodological requirements for chronobiological research. Substantial variation in testing protocols, task selection, and confounding control limited direct comparability.
Cited by
- contradicts The human brain has a narrow window in the first half of the day during which cognitive and problem-solving abilities peak.