Cross-sectional study of 2,795 cognitively unimpaired, community-dwelling adults aged 40–70 (Healthy Brain Project) examining whether self-reported OSA was linked to poorer cognition (Cogstate Brief Battery) and higher dementia risk (CAIDE score), and whether APOE ε4 moderated those links.
OSA was associated with poorer memory (Cohen's d = −0.22) and markedly higher CAIDE dementia risk scores (d = 0.84), but not with attention. The memory gap was attenuated after adjusting for vascular risk and was driven by untreated OSA. APOE ε4 did not moderate OSA–cognition associations; however, OSA alone—not APOE ε4 alone—drove the elevated CAIDE scores.
- OSA defined by self-report only; no objective severity measure (e.g., AHI) available, and ~7% diagnosed OSA likely underestimates true prevalence. - Cross-sectional design prevents causal or longitudinal inference. - Small number of APOE ε4 homozygotes (n = 74) may have underpowered interaction tests.
Routine OSA screening in middle-aged adults (40–70) may flag those at elevated dementia risk—particularly through modifiable vascular burden—before cognitive symptoms appear. Treating OSA may attenuate memory deficits, though vascular risk persists regardless of treatment status.
Explore related topics
