A longitudinal cohort study presented as an abstract for the American Thoracic Society's 2026 International Conference found that the overall severity of obstructive sleep apnea did not predict changes in a blood test used as a stand-in for brain amyloid, the protein at the center of Alzheimer's disease. Researchers at the Icahn School of Medicine at Mount Sinai and NYU Grossman School of Medicine tracked 64 cognitively normal adults, ages 55 to 90, drawing blood plasma twice, an average of 3.16 years apart, and measuring the ratio of two amyloid fragments, Aβ42 and Aβ40, whose decline is read as a sign of amyloid building up in the brain. None of six separate measures of sleep apnea burden, including the standard apnea-hypopnea index, time spent with low oxygen, and how often sleep was interrupted by arousals, showed a statistically significant relationship with how that blood ratio changed over time, either on its own or after accounting for age, sex, and the ApoE4 gene variant linked to Alzheimer's risk.
The question mattered because obstructive sleep apnea is common in older adults, and earlier work from the same NYU research group, using brain imaging rather than blood tests, had reported that more severe apnea tracked with faster amyloid accumulation over time. The new analysis drew its 64 participants from NYU's Alzheimer's Disease Research Center; just over half had no meaningful apnea by the study's cutoff, about a quarter had mild apnea, and one in five had moderate to severe apnea, none of it treated during the study period. Plasma samples were analyzed by mass spectrometry at Araclon Biotech in Spain, and the researchers used both simple correlations and adjusted statistical models to look for a link with each sleep measure.
One narrower result complicates a flat reading of no effect. Among the roughly half of participants who did not have apnea, a higher rate of sleep arousals was tied to a steeper annual decline in the blood ratio, a pattern consistent with more amyloid buildup, in the unadjusted analysis (Spearman's rho of negative 0.398, p equals 0.045). That association disappeared once the researchers adjusted for age, sex, and ApoE4 status, and no similar pattern turned up among participants who did have apnea. The study's authors said the absence of an overall effect may partly reflect how few participants had moderate to severe apnea, and suggested the arousal finding in apnea-free adults was worth testing in a larger sample.