Word syllable duration for AOS (Bunker et al., 2024)

<b>Purpose:</b> Neurogenic speech and language disorders—such as acquired apraxia of speech (AOS) and aphasia with phonemic paraphasia (APP)—are often misdiagnosed due to similarities in clinical presentation. Word syllable duration (WSD)—a measure of average syllable length in multisyllabic words—serves as a proxy for speech rate, which is an important and arguably more objective clinical characteristic of AOS and APP. This study reports stability of WSD over time for speakers with AOS (and aphasia).<b>Method:</b> Twenty-nine participants with AOS and aphasia (11 women and 18 men, <i>M</i>_age = 53.5 years, <i>SD</i> = 13.3) repeated 30 multisyllabic words (of three-, four-, and five-syllable lengths) on three occasions across 4 weeks. WSDs were calculated for each word and then averaged across each list (i.e., word length), as well as across combined lists (i.e., all 30 words) to yield four WSDs for each participant at each time point. Stability over time was calculated using Friedman’s test for the group and using Spearman’s rho for the individual level. Effects of time and word length were examined using robust mixed-effects linear regression.<b>Results:</b> Friedman’s tests and correlations indicated no significant difference in WSDs across sampling occasions for each word length separately or combined. WSD correlated positively with AOS severity and negatively with intelligibility but was not correlated with aphasia severity. Regression analyses confirmed WSD to be stable over time, while WSD calculated from only five tokens (i.e., WSD-5) was less stable over time.<b>Conclusions:</b> Results indicate that WSD can be a stable measure over time, at the individual and group level, providing support for its use in diagnosis and/or as an outcome measure, both clinically and for research. In general, WSD outperformed WSD-5, suggesting that it may be better to calculate WSD from more than five tokens. Stability of WSD in other populations and suitability for differential diagnosis need to be determined. Currently, differentiating disorders by speaking rate, alone, is not recommended.<b>Supplemental Material S1.</b> Robust linear mixed-effects regression models without controlling for aphasia severity (WAB-R AQ) or AOS severity (ASRS 3.0).Bunker, L. D., Bailey, D. J., Poss, E., Mauszycki, S., &amp; Wambaugh, J. L. (2024). Stability over time of word syllable duration for speakers with acquired apraxia of speech. <i>Journal of Speech, Language, and Hearing Research, 67</i>(10S), 4038–4052. https://doi.org/10.1044/2024_JSLHR-23-00007<b>Publisher Note:</b> This article is part of the Special Issue: Select Papers From the 8th International Conference on Speech Motor Control.

研究目的：神经源性言语语言障碍——如获得性言语失用症（acquired apraxia of speech, AOS）与伴音素性错语的失语症（aphasia with phonemic paraphasia, APP）——常因临床表现相似而出现误诊。词音节时长（word syllable duration, WSD）即多音节词的平均音节长度，可作为言语速率的代理指标，而言语速率是AOS与APP重要且可以说更具客观性的临床特征。本研究旨在报告合并失语症的AOS患者的WSD随时间的稳定性。
研究方法：29名合并失语症的AOS患者（女性11名，男性18名，年龄均值为53.5岁，标准差为13.3）在4周内的3个测试时点，重复朗读30个涵盖3、4、5音节长度的多音节词。研究人员先计算每个词的WSD，再按各词长分组（即对应词长的词汇列表）以及合并所有30个词的总列表分别取均值，最终每个被试在每个测试时点可获得4个WSD值。组水平的时间稳定性采用弗里德曼检验进行分析，个体水平则采用斯皮尔曼rho相关分析。采用稳健混合效应线性回归检验时间与词长的主效应。
研究结果：弗里德曼检验与相关分析结果显示，无论按单独词长分组还是合并所有词长分组，各采样时点的WSD均无显著差异。WSD与AOS严重程度呈正相关，与言语可懂度呈负相关，但与失语症严重程度无显著相关。回归分析证实WSD随时间保持稳定，而仅基于5个朗读样本计算的WSD（即WSD-5）的时间稳定性较差。
研究结论：研究结果表明，无论是个体水平还是组水平，WSD均为随时间保持稳定的测量指标，为其在临床与研究中用于诊断或作为结局指标提供了支持。总体而言，WSD的表现优于WSD-5，提示WSD的计算应基于超过5个朗读样本。WSD在其他人群中的稳定性以及用于鉴别诊断的适用性尚待进一步明确。目前不推荐仅通过言语速率来鉴别各类言语障碍。
补充材料S1：未控制失语症严重程度（西方失语症评定量表修订版言语商WAB-R AQ）或AOS严重程度（言语失用评定量表3.0 ASRS 3.0）的稳健线性混合效应回归模型。
Bunker, L. D., Bailey, D. J., Poss, E., Mauszycki, S., & Wambaugh, J. L. (2024). 获得性言语失用症患者的词音节时长随时间稳定性研究. 《言语、语言与听力研究杂志》, 67(10S), 4038–4052. https://doi.org/10.1044/2024_JSLHR-23-00007
出版者说明：本文属于特刊：第8届国际言语运动控制会议精选论文。