Adaptation and Validation of the Chinese Version of the Sleep Quality Questionnaire

Runtang Meng,Tsukasa Kato,Stefanos Mastrotheodoros,Lu Dong,Daniel Yee Tak Fong,Fang Wang,Menglu Cao,Xinliang Liu,Chenmin Yao,Jinhong Cao,Yong Yu,Yi Luo,Chuanhua Yu,David Gozal
DOI: https://doi.org/10.1007/s11136-022-03241-9
2023-01-01
Quality of Life Research
Abstract:Sleep quality is essential to health. The current study aimed to adapt and validate the Sleep Quality Questionnaire (SQQ) into Chinese language. The Chinese version of the SQQ (SQQ-C) was created following the guidelines for cross-cultural adaptation. Compliant with the COSMIN methodology, baseline data (N = 13,325) examined three validity domains and internal consistency, including content validity using the content validity index (CVI) and the cognitive debriefing and focus group (relevance, comprehensiveness and comprehensibility), construct validity using structural validity and cross‑sectional measurement invariance, and criterion validity using concurrent/convergent validity. Follow-up data (N = 3410) gathered within a mean of 168 (167–207) h interval were used to additionally assess longitudinal measurement invariance and test–retest reliability using intraclass correlation coefficient (ICC). Scale-level CVI/Average was equal to 0.922; Item-level CVIs ranged from 0.889 to 1.000 (excellent), except for item 2 (0.556-fair). A panel of local experts and local participants during cognitive debriefing and focus group stated that it had sufficient relevance and comprehensibility but a slight deficiency in comprehensiveness. Confirmatory factor analysis indicated a stable two-factor structure encompassing Daytime Sleepiness Subscale and Sleep Difficulty Subscale from baseline to follow-up data. The SQQ-C-9 (without item 2) outperformed the SQQ-C-10 (full form). The SQQ-C-9 provided evidence of measurement invariance (strict) across subgroups (cohorts, gender, and age) and across time. The SQQ-C was negatively correlated with the Chinese Nonrestorative Sleep Scale and the Chinese Sleep Condition Indicator. Cronbach's alpha (α), McDonald's Omega (ω), and ICC, respectively, ranged from 0.712 to 0.838, 0.723 to 0.840, and 0.738 to 0.764 for total scale and each subscale. The SQQ-C exhibits adequate psychometric properties and a stable two-factor structure, and should enable valuable assessments of sleep quality in clinical and research settings.
What problem does this paper attempt to address?