Cerebral amyloid PET in obstructive sleep apnoea: a systematic review and exploratory meta-analysis

This study aimed to evaluate cerebral amyloid-PET findings in adults with obstructive sleep apnoea (OSA), account for overlapping cohorts, and quantify global cortical amyloid burden where outcomes were commensurable. PubMed, Scopus, and Web of Science were searched from inception to 17 July 2026 for adult human studies relating OSA or sleep-disordered breathing to cerebral amyloid PET. One estimate per independent data source was retained. The second author independently verified eligibility, cohort independence, extracted quantitative data, and risk-of-bias judgements for all reports entering any quantitative model. Global outcomes were pooled as Hedges' g using restricted maximum-likelihood random-effects models with modified Hartung-Knapp confidence intervals. Fifteen reports representing 12 independent data sources were included. Tracers, reference regions, partial-volume correction, OSA ascertainment, and cognitive status varied substantially. Three global-outcome cohorts of cognitively unimpaired participants (n = 291) yielded g = 0.42 (95% confidence interval, -0.14 to 0.99; I2 = 0%). In a secondary analysis, adding one cohort with established Alzheimer's disease yielded g = 0.54 (95% confidence interval, 0.22-0.86; n = 471; I2 = 0%). A sensitivity model adding one selected regional 11C-PiB estimate yielded g = 0.47 (95% confidence interval, 0.17-0.78). In a secondary analysis combining cognitively unimpaired adults and patients with established Alzheimer's disease, OSA was associated with higher global cerebral amyloid-PET burden. However, the primary evidence in cognitively unimpaired adults remains imprecise, and longitudinal studies do not consistently demonstrate an association between conventional OSA severity and amyloid accumulation.