Comprehensive assessments of long-term sleep habits in epidemiological study: Validity and reliability of sleep factors questionnaire (SFQ) among Chinese women

Abstract Objective Epidemiological studies suggested that poor sleep is a potentially novel risk factor for several health outcomes currently; however, there are no validated questionnaires that can systematically measure sleep parameters within these studies. We evaluated the reliability and validity of 17-item sleep factors questionnaire (SFQ), which was developed to comprehensively assess long-term sleep habits for the Jiujiang Breast Cancer Study (JBCS), Jiujiang, China. Methods The participants included 100 women aged 18–74 years, who were randomly selected from the JBCS project, and completed a SFQ at baseline and again 1 year later, and 4 quarterly 30 consecutive days (a total of 120 days) sleep diaries over this same year. Reliability was tested by comparing the 2 SFQs; validity by comparing the average measures between the SFQ and the 4 sleep diaries. Results Validity analysis showed moderate correlation (γ = 0.41) for sleep duration with the adjusted concordance correlation coefficient (CCC adj ) of 0.54; the weighted κ statistics indicated an excellent agreement for night/shift work and sleep medication use; fair-to-moderate for sleep quality, light at night (LAN), nighttime sleeping with light on, sleep noise and nap time; slight-to-fair for sleep quality and nighttime wakings frequency. Reliability analysis showed excellent correlation for night/shift work and sleep medication use; fair-to-moderate for LAN, nighttime wakings frequency, insomnia frequency, sleep noise and nap time; but slight-to-fair for insomnia frequency and nighttime sleeping with light on; the CCC adj for sleep duration was 0.61. Conclusions The SFQ showed reasonable reliability and validity for sleep assessments in most domains.