Shuta Tomisato ^1* Takanori Mori ¹ Kazumi Asano ¹ Daichi Iimura ² Yasuto Yada ³ Saburo Takahashi ⁴ Koichiro Wasano ^1,5* Takeyuki Kono ¹ Hiroyuki Ozawa ¹

• ¹ Department of Otorhinolaryngology, Head and Neck Surgery, School of Medicine, Keio University, Tokyo, Japan
• ² Institute of Human Sciences, University of Tsukuba, Ibaraki, Japan
• ³ Department of Language Sciences, Tokyo Metropolitan University, Tokyo, Tôkyô, Japan
• ⁴ Sumiyoshi Elementary School, Fuchu, Tokyo, Japan
• ⁵ Department of Otolaryngology, Head and Neck Surgery, Tokai University School of Medicine, Isehara, Kanagawa, Japan

Background: Cluttering is a speech disorder distinct from stuttering. Despite this distinction, there is no established method to clearly differentiate the two disorders. This study aimed to use objective criteria to differentiate cluttering from stuttering in Japanese speakers. Methods: Participants were 32 consecutive native-Japanese speakers who visited the Keio University Hospital between July 2020 and January 2023 with a chief complaint of speech disfluency. One physician and two speech-language-hearing therapists concurred on a stuttering or cluttering diagnosis of the 32 patients based on recordings of the Kitsuon kensa-ho test. The frequencies of stuttering-like disfluencies (SDF) and normal disfluencies (NDF) were calculated from the Kitsuon kensa-ho, and the ratio of disfluencies (RDF) was calculated as the ratio of SDF to NDF. Differences between the cluttering and stuttering groups in the RDF and the mean articulatory rate (MAR) for oral reading and a monologue task were tested using the Mann-Whitney U test. ROC curves were used to determine the sensitivity and specificity that well-distinguished subjects with cluttering from those with stuttering; the experts' diagnosis was the gold standard. Results: Of the 32 participants, 12 (38%) were diagnosed with cluttering and 20 (62%) with stuttering. The cluttering and stuttering groups were comparable in demographic characteristics. The RDF on monologue task had the highest sensitivity in diagnosing cluttering, and the MAR on monologue task had the highest specificity. Adopting provisional criteria of a monologue RDF of 1.2 or greater and a monologue MAR of 7.5 or greater produced a sensitivity of 0.92 and a specificity of 0.95.We conclude that combining monologue RDF and monologue MAR well-distinguished cluttering from stuttering. This method provides new objective diagnostic criteria, which can aid clinicians, therapists, and basic researchers.