Vitamin D tracked with measured sleep, and B12 ran in opposite directions in men and women

Design

cross-sectional analysis within a longitudinal ageing cohort, using wrist actigraphy

Population

460 participants aged 40 and over (mean 70.9 years, 52.8% female, 22.2% Black)

Primary outcome

actigraphic total sleep time, efficiency, onset latency, wake after sleep onset and wake bout length

Effect

higher vitamin D associated with +23.6 minutes total sleep (95% CI 9.6-37.6) and shorter wake bouts (-0.29 min, -0.50 to -0.09); deficiency with -23.0 minutes (-42.1 to -4.0)

Most of what links vitamin D to sleep in older adults comes from self-reported sleep quality, which correlates with mood, pain and expectation as much as with sleep. This analysis of 460 participants aged 40 and over from a longitudinal ageing study used actigraphy instead, relating serum vitamin D and B12 to total sleep time, sleep efficiency, onset latency, wake after sleep onset and average wake bout length.

Higher vitamin D was associated with 23.6 more minutes of total sleep (95% CI 9.6-37.6) and shorter average wake bouts (-0.29 minutes, -0.50 to -0.09). Vitamin D deficiency was associated with 23 minutes less sleep (-42.1 to -4.0). The associations were stronger among Black participants, a group in whom both lower measured vitamin D and shorter measured sleep are well described. The B12 findings are where caution is needed: higher B12 was associated with shorter wake after sleep onset in women and longer in men, and B12 insufficiency with longer total sleep in men and shorter in women - opposite directions by sex, in subgroups of a 460-person cohort, which is the pattern chance produces.

Twenty-four minutes of sleep is a real amount, and this is measured rather than reported, which makes the vitamin D association more credible than most of what precedes it. But it is cross-sectional, and the causal direction is genuinely uncertain: people who sleep poorly go outside less, and vitamin D is a marker of sunlight exposure and general health as much as a nutrient. No trial has shown that supplementing improves sleep, and this study does not support starting one for that indication.

• Do not prescribe vitamin D for sleep - this is cross-sectional and no trial supports it
• Where you are already replacing deficiency for bone health, no change is implied here either way
• Treat the opposite-direction B12 findings in men and women as likely chance, not as a sex difference to act on
• Ask about outdoor time in an older person sleeping poorly; it plausibly drives both the vitamin level and the sleep
• Note this used actigraphy, so the associations concern measured sleep rather than reported quality

Why it matters

It replaces self-reported sleep with measured sleep in a link that has always rested on questionnaires.

Don't overread it

Cross-sectional associations with no trial evidence - and the sex-divergent B12 findings are subgroup results that should not be acted on.

The statistics, in plain English

An association of 23.6 minutes with a confidence interval of 9.6 to 37.6 is reasonably precise and clinically noticeable, but cross-sectional design means the arrow could point either way - poor sleepers may go outdoors less. The B12 results are a textbook subgroup problem: splitting 460 people by sex and finding effects in opposite directions, with no prior reason to expect it, is the pattern that most often fails to replicate. Stronger associations among Black participants, in a subgroup of 22.2% of the cohort, need the same caution.

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.