Abstract
AIMS: To assess the association of sleep factors (sleep duration, trouble sleeping, sleep disorder) and combined sleep behaviours with the risk of clinically relevant depression (CRD). METHODS: A total of 17 859 participants (8806 males and 9053 females) aged 20-79 years from the National Health and Nutrition Examination Survey (NHANES) 2007-2014 waves were included. Sleep duration, trouble sleeping and sleep disorder were asked in the home by trained interviewers using the Computer-Assisted Personal Interviewing (CAPI) system. The combined sleep behaviours were referred to as 'sleep patterns (healthy, intermediate and poor)', with a 'healthy sleep pattern' defined as sleeping 7-9 h per night with no self-reported trouble sleeping or sleep disorders. And intermediate and poor sleep patterns indicated 1 and 2-3 sleep problems, respectively. Weighted logistic regression was performed to evaluate the association of sleep factors and sleep patterns with the risk of depressive symptoms. RESULTS: The total prevalence of CRD was 9.5% among the 17 859 participants analysed, with females having almost twice as frequency than males. Compared to normal sleep duration (7-9 h), both short and long sleep duration were linked with a higher risk of CRD (short sleep: OR: 1.66, 95% CI: 1.39-1.98; long sleep: OR: 2.75, 95% CI: 1.93-3.92). The self-reported sleep complaints, whether trouble sleeping or sleep disorder, were significantly related with CRD (trouble sleeping: OR: 3.04, 95% CI: 2.59-3.56; sleep disorder: OR: 1.83, 95% CI: 1.44-2.34). Furthermore, the correlations appeared to be higher for individuals with poor sleep pattern (OR: 5.98, 95% CI: 4.91-7.29). CONCLUSIONS: In this national representative survey, it was shown that there was a dose-response relationship between sleep patterns and CRD.