Does the rate of cognitive impairment in sleep apnea differ between sleep-lab and home sleep testing studies after accounting for severity?

Severity, not the testing setting, is what separates the findings

The clearest pattern across these studies is a threshold effect. In a cross-sectional study of 387 people with obstructive sleep apnea, only severe apnea (an apnea-hypopnea index of 30 or more events per hour) remained linked to mild cognitive impairment after confounders were controlled; milder degrees of apnea lost statistical significance once other factors were accounted for [1]. A large pooled analysis of nearly 6,000 adults followed for five years found a similar but smaller signal: having at least mild apnea was associated with slightly poorer global cognition than having none, and the estimate was essentially the same for moderate-to-severe apnea [3].

That consistency across very different study designs is the useful takeaway. One was a clinic-based sample using either lab or home testing [1]; the other pooled five community cohorts using home-based sleep monitoring [3]. Both point to apnea severity as the variable that tracks with cognition, which is exactly the variable you would need to hold constant before any lab-versus-home comparison could mean anything.

Why home and lab testing could plausibly give different pictures

Lab and home tests measure breathing events differently, and that difference can shift who gets labeled severe. A home test typically records fewer signals and may underestimate or overestimate events depending on the night and the sensors used. The studies here do not quantify that gap, but they do show how much the answer depends on how apnea and cognition are each defined. In the World Trade Center responder cohort, which used both home tests and in-lab monitoring, 65.5% of participants had apnea by a combined index, yet people without apnea actually had a higher prevalence of cognitive impairment (37% versus 21.5%) [2]. That counterintuitive result is a warning: the measured relationship can flip depending on the population and the cutoffs.

The cognitive test itself may be doing more work than the sleep test

The same responder study found that scores on the Montreal Cognitive Assessment, the tool used to define impairment in most of these studies, were strongly influenced by education, race, ethnicity, and whether the test was taken in English or Spanish [2]. The authors concluded that their unexpected finding of less impairment in people with apnea may reflect how the test scores different groups rather than a real protective effect. That matters for the original question: if the cognitive yardstick behaves differently across populations, then comparing impairment rates between a lab-based clinic sample and a home-based community sample confounds the testing setting with who was tested.

A separate study of older women with mild cognitive impairment hints at another layer. Among women with apnea diagnosed by home testing, greater apnea severity was associated with higher tau burden and elevated CRP only in carriers of the APOE4 gene variant, not in non-carriers [5]. In other words, the same severity score predicted different biological outcomes depending on genetic background, which is a further reason a single impairment rate cannot be cleanly attributed to the testing method.

What would actually settle the lab-versus-home question

None of these studies was built to answer it. The 387-person study drew on both lab and home diagnoses but never compared the two groups' cognitive outcomes [1]; the responder cohort used both methods but reported them together [2]; the pooled cohort analysis used home monitoring only [3]. An older study comparing people with apnea who had low blood oxygen to those who did not found more severe cognitive impairment in the hypoxemic group, which supports severity and oxygen loss as the relevant axis rather than the recording venue [4].

The practical implication for a reader is that the choice between a lab study and a home study should not, on current evidence, be expected to change the headline conclusion about apnea and cognition. What would change it is how severe the apnea is, how cognition is measured, and who is in the sample. A study that randomized or matched participants across testing methods at equal severity would be needed to answer the question directly, and no such study appears here.

Sources used in this answer

Related research questions