Long naps and slower processing — and sleep apnoea is not the explanation

Design

Cross-sectional analysis of baseline data from a sleep ancillary study, with actigraphy and home apnoea screening

Population

686 older adults; 223 with sleep apnoea and 463 without

Primary outcome

Association of actigraphy-measured napping minutes with global cognition and its domains

Effect

Processing speed beta -0.0973 (95% CI -0.1664 to -0.0282); no significant apnoea-by-napping interaction

Daytime napping in older people gets read two ways: as harmless rest, or as an early sign of something. This analysis of baseline data from the US POINTER sleep ancillary study measured it properly rather than asking about it — 686 participants with home apnoea screening, wrist actigraphy and sleep diaries, with a nap defined as at least 30 continuous minutes of sleep outside the main sleep period.

More napping minutes were associated with slower processing speed (beta -0.0973, 95% CI -0.1664 to -0.0282). Global cognition and the other domains did not show the same association. Of the 686, 223 had sleep apnoea and 463 did not, and the interaction between apnoea and napping was not significant — so there is no evidence here that undiagnosed apnoea is what links the naps to the cognitive finding. One subgroup signal appeared, in participants who both had sleep apnoea and belonged to an under-represented ethno-racial group, where more napping was associated with worse executive function (beta -0.2493, -0.4943 to -0.0042).

The authors call the work exploratory and they are right to. This is cross-sectional: slower processing speed could cause more napping as easily as the reverse, and a third factor — early neurodegeneration, depression, medication burden — could produce both. What it does justify is asking about daytime sleep in a cognitive assessment and taking a long habitual nap as a reason to look further rather than as reassurance, including at the medication list and at mood.

• Ask about daytime sleep duration as part of any cognitive assessment.
• Review sedating medications and mood before attributing long naps to age.
• Screen for sleep apnoea on its own merits — this study gives no reason to screen because of napping.
• Do not advise older patients to stop napping on the strength of a cross-sectional association.

Why it matters

It removes the most popular explanation for why a napping older patient might be cognitively slower.

Don't overread it

Cross-sectional data cannot say whether napping affects cognition or slower cognition leads to more napping.

The statistics, in plain English

The confidence interval for the main finding runs from -0.1664 to -0.0282: it excludes zero, so the association is unlikely to be chance, but its upper end is very close to zero, so the effect may be small. The non-significant interaction is worth reading correctly — it means the data gave no evidence that apnoea changes the napping-cognition relationship, not that it has been shown to make no difference. The subgroup finding in a doubly-defined group has a confidence interval that nearly touches zero (-0.0042) and comes from one of many comparisons, which is exactly the shape of a result that does not replicate.

Read the rest in the app

You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.

Get Daily Doctor — free

Scan to keep reading on your phone. No account needed to start.