A digital cognitive battery for sleep-disordered breathing: Test-retest reliability and practice effects
Related Experiment Video
Updated: Nov 11, 2026
Association Between Sleep Quality and Cognitive Symptoms in Patients with Major Depressive Disorder
Published on: April 26, 2024
Elena Richert1, Minna Pitkänen2, María K Jónsdóttir3
1Department of Psychology, Reykjavik University, Reykjavik, Iceland; School of Technology, Reykjavik University Sleep Institute, Reykjavik, Iceland; Department of Technical Physics, University of Eastern Finland, Kuopio, Finland.
Objective:
Obstructive sleep apnea (OSA) affects cognitive performance and may contribute to long-term cognitive decline. Longitudinal assessments are needed for measuring changes in cognitive functioning in relation to OSA progression and treatment. This requires standardized, reliable cognitive assessments suitable for repeated use, but the psychometric properties of cognitive tests may vary across populations. We evaluated the suitability of a digital cognitive battery, covering various cognitive functions, for repeated administration in individuals with varying OSA severity, based on its psychometric properties.
Methods:
The battery included 14 cognitive tasks. We assessed test-retest reliability, practice effects, and ceiling and floor effects between two administrations (mean interval 12 weeks) in 46 participants (mean age = 49 ± 16 years, 21 women) with a wide range of OSA severity (mean apnea-hypopnea-index (AHI) = 17.6 ± 17.7).
Results:
Psychometric properties varied considerably between cognitive tasks. Test-retest reliability estimates were good or satisfactory (ICC≥.60) for the Psychomotor Vigilance Test (PVT), Symbol Digit Modalities Task (SDMT), Paced Auditory Serial Addition Task (PASAT), Auditory Verbal Learning Task (AVLT), and subscores of the Tower of London Task (ToL). Practice effects were high for the PASAT in Level 1 (p<.001,r=.63) and the Inspection Time Task (ITT; p<.001,r=.52). Ceiling effects were present in the ITT and ToL, whereas intrusions in the AVLT and lapses in the PVT occurred infrequently and clustered at the lower bound.
Conclusion:
Only a subset of cognitive tasks showed adequate psychometrics for repeated administration, highlighting the need for careful task selection and modification when developing cognitive assessments for repeated application in OSA populations. We include recommendations on cognitive test selection and test design for repeated assessments in OSA populations.
