Prather · Sleep 2015 · prospective cohort study with experimental viral challenge · n=164

Behaviorally Assessed Sleep and Susceptibility to the Common Cold.

Cited 407 times in the scientific literature.

Level 3 - non-randomized controlled study

Prospective cohort study with experimental viral challenge (sleep exposure was observational, not randomized).

PubMed 26118561 · doi:10.5665/sleep.4968 · record verified 2026-08-30

What was done

A total of 164 healthy men and women aged 18 to 55 wore wrist actigraphs and kept sleep diaries for 7 consecutive days to assess sleep duration and continuity. Participants were then quarantined, challenged with nasal drops containing rhinovirus, and monitored over 5 days for the development of a clinical cold (defined as infection combined with objective signs of illness). Logistic regression models evaluated whether actigraphy-assessed sleep predicted cold incidence, adjusting for prechallenge antibody levels, demographics, season, body mass index, psychological factors, and health practices.

What was found

Actigraphy-measured shorter sleep duration significantly predicted cold development relative to sleeping > 7 h per night: sleeping < 5 h resulted in an OR of 4.50 (95% CI, 1.08–18.69) and sleeping 5 to 6 h resulted in an OR of 4.24 (95% CI, 1.08–16.71). Sleeping 6.01 to 7 h was not associated with significantly higher risk (OR = 1.66, 95% CI, 0.40–6.95). Sleep fragmentation and other diary or actigraphy variables were not strong predictors of cold susceptibility.

Why it matters

This study provides prospective, objective behavioral evidence linking short sleep duration directly to biological susceptibility to viral respiratory infection, demonstrating that the relationship persists after controlling for baseline antibody levels and other confounding factors.

Limits

The sample size of 164 healthy volunteers resulted in wide confidence intervals for odds ratios. Sleep was assessed over a single 7-day period prior to quarantine, which may not capture long-term baseline patterns. Because sleep duration was observed rather than experimentally manipulated, residual confounding remains possible.

Cited by