Reliability and validity of the Japanese version of the Ullanlinna Narcolepsy Scale and Swiss Narcolepsy Scale for screening Japanese individuals with narcolepsy type 1

Kosuke Tanioka,Toshiyuki Hirasawa,Mariko Yanagihara,Noboru Takeuchi,Hideaki Nakayama,Akira Usui,Yuichi Inoue

DOI: https://doi.org/10.1016/j.sleep.2023.07.015

Abstract:Objective: To validate the Japanese versions of the Ullanlinna Narcolepsy Scale (J-UNS) and Swiss Narcolepsy Scale (J-SNS) for screening narcolepsy in the Japanese population and to discuss strategies for their use in hypersomniac individuals. Methods: We selected 451 outpatients with excessive daytime sleepiness (EDS) already diagnosed according to the International Classification of Sleep Disorders third edition. They responded to both scales twice at 1-month intervals. After eliminating individuals who met the exclusion criteria, validity and reliability analyses were performed on 408 and 381 participants, respectively. Results: Patients with narcolepsy type 1 (NT1) displayed higher J-UNS and lower J-SNS scores than those with NT2 and other sleep disorders. The intraclass correlation coefficients and weighted κ coefficient for scale scores in the total participants and patients with NT1 were ≥0.70 and ≥ 0.40, respectively, indicating high reliability. Furthermore, both the sensitivity and specificity of these scales upon using the original cut-off scores (14 for UNS and 0 for SNS) for detecting NT1 were 0.70 or ≥0.70, suggesting high validity. Additionally, the receiver operating characteristic curve analysis revealed that the best cut-off score did not change for the J-SNS but that for the J-UNS, it increased to 18. In our study, the scale's sensitivity and specificity changed from 96% to 82% and 58%-78%, respectively. Conclusions: Both scales revealed satisfactory screening abilities for NT1 in the Japanese population. However, it may be better to use J-UNS cut-off scores of 18 for a population with EDS.