Psychometric Evaluation of the Arabic version of Insomnia Severity Index (ISI) during COVID-19 pandemic

Background Insomnia, a prevalent global sleep disorder, necessitates reliable assessment tools like the Insomnia Severity Index (ISI). However, its structural validity varies across cultures, and no validated Arabic version exists for Bahrain’s unique demographic composition and in the context of the COVID-19 pandemic. This study addresses this gap by evaluating the psychometric properties of the Arabic ISI in this context. Methods A cross-sectional study enrolled 646 Bahraini adults during the COVID-19 pandemic (February 2020–August 2021). Participants were randomly split into subsamples for exploratory factor analyses (EFA, n = 321) and confirmatory factor analyses (CFA, n = 325). Statistical analyses included internal consistency (Cronbach’s α, McDonald’s ω), EFA, CFA, and criterion validity via correlation with the Generalized Anxiety Disorder Scale (GAD-7). Results The Arabic ISI demonstrated excellent internal consistency (α = 0.889, ω = 0.888). EFA revealed a unidimensional structure explaining 60.44% of variance, with strong factor loadings (0.68–0.87). The model's fitness indexes were suitable for the structural model (RMSEA = 0.084, CFI = 0.974). Criterion validity was supported by a moderate-to-strong correlation with GAD-7 (r = 0.613, p < 0.001). Conclusion The Arabic ISI with a unidimensional structure is a reliable and valid tool for assessing the severity of insomnia in Bahrain during COVID-19.