Study Objectives
Prospectively examine the association between sleep restriction, perceived sleep quality (PSQ) and upper respiratory tract infection (URTI).

Methods
In 1,318 military recruits (68% males) self-reported sleep was assessed at the beginning and end of a 12-week training course. Sleep restriction was defined as an individualized reduction in sleep duration of ≥2 hours/night compared with civilian life. URTIs were retrieved from medical records.

Results
On commencing training, approximately half of recruits were sleep restricted (52%; 2.1 ± 1.6 h); despite the sleep debt, 58% of recruits with sleep restriction reported good PSQ. Regression adjusted for covariates showed that recruits commencing training with sleep restriction were more likely to suffer URTI during the course (OR = 2.93, 95% CI 1.29–6.69, P = 0.011). Moderation analysis showed this finding was driven by poor PSQ (B = -1.12, SE 0.50, P = 0.023), as no significant association between sleep restriction and URTI was observed in recruits reporting good PSQ, despite a similar magnitude of sleep restriction during training. Associations remained in the population completing training, accounting for loss to follow-up. Recruits reporting poor PSQ when healthy at the start and end of training were more susceptible to URTI (OR = 3.16, 95% CI 1.31–7.61, P = 0.010, vs good PSQ).

Conclusion
Good perceived sleep quality was associated with protection against the raised risk of respiratory infection during sleep restriction. Studies should determine whether improvements in sleep quality arising from behavioral sleep interventions translate to reduced respiratory infection during sleep restriction.