Does CPAP treatment change Alzheimer's disease biomarkers in people with sleep apnea, and are findings consistent across biomarker types?

Which Alzheimer's biomarkers does CPAP actually move?

The clearest evidence points to markers of ongoing brain injury and inflammation. In a six-month follow-up of people with sleep apnea who started CPAP, those who stuck with the treatment showed reductions in two blood markers: GFAP, a protein released by activated astrocytes (support cells that react to brain injury), and NfL, a protein released when nerve fibers are damaged [1]. The same adherent group also showed lower levels of IL-6, an inflammatory signal [1]. These are not the classic amyloid or tau proteins that define Alzheimer's disease pathology, but they reflect the downstream damage and inflammation that accompany it.

Amyloid markers tell a less consistent story. A three-month randomized study of people with mild sleep apnea and subjective cognitive decline found that the plasma Aβ42/Aβ40 ratio—a measure comparing two forms of amyloid protein—significantly decreased in the CPAP group compared with controls [3]. That direction is notable because a lower ratio is generally considered a more adverse Alzheimer's profile, yet the same study reported improved subjective cognition [3]. The authors suggested CPAP may still play a preventive role, but the biomarker direction and the cognitive direction did not align neatly, and they called for further research [3].

Why injury and inflammation markers may respond more readily than amyloid

One plausible explanation is that CPAP works by correcting the repeated drops in oxygen and sleep fragmentation caused by sleep apnea, and those stresses may drive ongoing injury and inflammation rather than the initial buildup of amyloid plaques. In a study comparing people with and without sleep apnea, those with apnea had higher plasma NfL and GFAP, and over time their NfL and GFAP rose faster—about 2.37 versus 0.72 pg/mL per year for NfL and 15.19 versus 4.54 pg/mL per year for GFAP [1]. If apnea accelerates these injury processes, then treating it could slow them, which is consistent with the reductions seen in adherent CPAP users [1].

Amyloid deposition, by contrast, may accumulate over years or decades and could be less responsive to a few months of breathing support. The six-month study found that CPAP adherence was associated with favorable short-term trajectories across biomarkers, but the authors described this as suggesting CPAP may be a modifiable contributor rather than proving it reverses amyloid pathology [1]. The three-month amyloid finding [3] and the broader biomarker improvements [1] come from different populations, durations, and biomarker panels, so they do not yet form a single consistent picture.

Do cognitive benefits line up with biomarker changes?

Cognitive outcomes generally lean in a favorable direction for CPAP, though the evidence is not uniform. In the six-month study, adherent CPAP users showed improved cognition, with a 0.59-point gain on the MoCA and a 1.48-point improvement on the ADAS-Cog [1]. A twelve-month pilot study of people with mild cognitive impairment and sleep apnea found that CPAP users had significantly slower global decline on the MMSE and improvements in overall cognition and specific cognitive domains compared with non-users [5]. A retrospective study of people with mild cognitive impairment or Alzheimer's dementia found that CPAP-adherent patients had a smaller increase in the Clinical Dementia Rating scale than non-adherent patients, although the MMSE change did not differ significantly between groups [2].

These cognitive findings are associations, not proof of cause and effect. In the retrospective study, adherence was not randomly assigned, and the authors noted the clinical potential of CPAP to delay deterioration rather than demonstrating it [2]. The pilot study was small—21 participants—and the authors called for larger, longer trials to confirm whether CPAP has disease-modifying potential [5]. The three-month randomized study did find improved subjective cognitive scores alongside the amyloid ratio change, but the authors themselves described the results as suggestive and warranting further research [3].

Who was studied, and does that change the answer?

The populations differ in ways that matter. The six-month study included people with Alzheimer's disease and compared CPAP adherent versus non-adherent participants [1]. The twelve-month pilot included people with mild cognitive impairment confirmed by Alzheimer's biomarkers and sleep apnea, with CPAP-compliant users mostly having moderate to severe apnea and non-users having mild apnea or intolerance [5]. The three-month randomized trial focused on mild sleep apnea with subjective cognitive decline—a group where CPAP use is itself debated [3]. The retrospective study included people with mild cognitive impairment or Alzheimer's dementia and at least one year of follow-up [2].

Duration also varies from three months to twelve months, and the biomarker panels are not the same across studies. Only the six-month study measured GFAP and NfL together with amyloid and tau markers [1]; the three-month study measured only plasma amyloid ratios [3]; the twelve-month pilot focused on cognition and sleep architecture rather than blood biomarkers [5]; and the retrospective study reported cognitive scales without biomarker changes [2]. A separate study of adults with Down syndrome found that sleep disruption was associated with greater amyloid and tau burden, but it did not analyze whether CPAP changed biomarkers—and half of those using OSA treatment still had sleep-disordered breathing, suggesting treatment was often suboptimal [4]. That study therefore cannot confirm or refute the CPAP-biomarker link; it mainly underscores that sleep problems are common and under-detected in a high-risk group [4].

Sources used in this answer

Related research questions