Insomnia is the most prevalent sleep complaint, but remains largely an unidentified public health issue. The Insomnia Severity Index (ISI) is a brief self-report questionnaire to assess insomnia, long-established both in clinical and research settings. The present study aimed to analyse the reliability, validity, and accuracy of the ISI European Portuguese version. After the translation protocol, 1,274 participants (65.54% female), with a mean (SD, range) age of 37.52 (16.82, 18-95) years, completed the ISI. This sample included 250 patients with insomnia from a Sleep Medicine Centre, presenting a diagnosis of insomnia disorder (Diagnostic and Statistical Manual of Mental Disorders, Fifth Edition; International Classification of Sleep Disorders, Third Edition), and 1,024 individuals from the community. A group of 30 patients with obstructive sleep apnea (OSA) was also recruited. Cronbach's α was 0.88 (internal consistency), and corrected item-total correlations ranged from 0.56 to 0.83. An exploratory factor analysis (oblique rotation) revealed a two-factor solution for both clinical and community samples. The ISI total score significantly differentiated insomnia disorder, no insomnia, and OSA subgroups with a large effect size (η 2 = 0.42). The correlation between ISI and Pittsburgh Sleep Quality Index supported concurrent validity (0.82), and discriminant validity was confirmed by a moderate correlation between ISI and Beck Depression Inventory Second Edition (0.32). The area under the curve was 0.88, and the optimal cut-off to detect clinical insomnia was 14 (82.1% sensitivity, 79.5% specificity). In conclusion, the Portuguese version of the ISI is a reliable and valid measure of insomnia in clinical and non-clinical populations. Our present study also contributes to relevant data for the international literature regarding the cut-off score of the scale for the detection of clinical insomnia.
Keywords: Portuguese validation; clinical accuracy; insomnia; insomnia severity index.
© 2020 European Sleep Research Society.