Neurocognitive deficits in children with polysomnographically confirmed obstructive sleep apnea: A systematic review and meta-analysis

22/08/2026

AI Summary

PSG-confirmed OSA in school-aged children was associated with a moderate neurocognitive deficit (Hedges' g 0.67, 95% CI 0.44 to 0.90, p < 0.001)
Impairments were observed across executive function, working memory, intelligence and phonological processing
Results were consistent with low-to-moderate heterogeneity and robust to sensitivity analyses; findings support neurocognitive screening and early intervention

AI Summary

BACKGROUND: Children with obstructive sleep apnea (OSA) frequently exhibit neurocognitive difficulties, but the magnitude and consistency of cognitive deficits have not been systematically quantified using polysomnographic criteria.

OBJECTIVES: To estimate the pooled magnitude of neurocognitive deficits in predominantly school-aged children (mean age approximately 5-12 years) with PSG-confirmed OSA versus healthy non-snoring controls.

CONCLUSIONS: PSG-confirmed OSA in school-aged children is associated with a moderate, relatively consistent neurocognitive deficit across diverse populations, test batteries, and severity strata.

Basic summary

Int J Pediatr Otorhinolaryngol. 2026 Aug 21;209:112989. doi: 10.1016/j.ijporl.2026.112989. Online ahead of print.

ABSTRACT

METHODS: Systematic review and random-effects meta-analysis (DerSimonian-Laird) following PRISMA 2020 guidelines. PubMed, Embase, Cochrane CENTRAL, and PsycINFO were searched from database inception to July 2026. Inclusion criteria required: (1) PSG-confirmed OSA (AHI ≥1 event/hour); (2) a directly enrolled non-snoring control group; (3) standardised neuropsychological assessment; (4) predominantly school-aged participants (mean age approximately 5-12 years); (5) ADHD excluded. The primary effect measure was Hedges’ g.

RESULTS: Six studies (n = 436; 249 OSA, 187 controls) met eligibility criteria. The pooled Hedges’ g was 0.67 (95% CI: 0.44-0.90, p < 0.001), a moderate cognitive deficit. Between-study heterogeneity was low-to-moderate (I² = 25%, Q = 6.69, df = 5, p = 0.24). Sensitivity analyses confirmed robustness: leave-one-out estimates ranged from g = 0.58 to 0.77 regardless of which study was excluded. Secondary outcomes showed impairment across executive function, working memory, intelligence, and phonological processing. With few studies, heterogeneity estimates should be interpreted cautiously.

CONCLUSIONS: PSG-confirmed OSA in school-aged children is associated with a moderate, relatively consistent neurocognitive deficit across diverse populations, test batteries, and severity strata. These findings support neurocognitive screening in children with OSA and a rationale for early intervention.

PMID:42628447 | DOI:10.1016/j.ijporl.2026.112989

Document this CPD