Vojtech Jenny, Toles Laura E, Buckley Daniel P, Stepp Cara E
Department of Speech, Language, & Hearing Sciences, Boston University, MA.
Department of Otolaryngology-Head and Neck Surgery, The University of Texas Southwestern Medical Center, Dallas.
J Speech Lang Hear Res. 2025 Sep 10;68(9):4220-4235. doi: 10.1044/2025_JSLHR-25-00138. Epub 2025 Aug 19.
Hyperfunctional voice disorders are highly prevalent yet difficult to characterize objectively. Relative fundamental frequency (RFF) has the potential to characterize these disorders but faces limited clinical use due to intersubject variability in mean RFF values. This study examined whether RFF variability offers insights beyond traditional mean measures.
Speech samples were collected from 132 adults: individuals with phonotraumatic vocal hyperfunction (PVH; = 44), nonphonotraumatic vocal hyperfunction (NPVH; = 44), and typical voices (controls; = 44). Two measures of RFF variability-standard deviation and spatiotemporal index (STI)-were calculated along with mean RFF values. While standard deviation captures variability in magnitude, STI incorporates variability in time and magnitude. Permutational analyses of variance were conducted to assess relationships between group (PVH/NPVH/control) and the mean, standard deviation, and STI measures. Significant measures were entered along with demographic parameters into hierarchical multinomial logistic regression models using a training set ( = 102). Final model equations were then applied to an independent test set ( = 30) to predict group membership.
Mean and STI measures showed significant group differences, whereas standard deviation did not. Both mean and STI measures improved model performance after adjusting for demographics. Receiver operating characteristic analysis on the test set yielded acceptable classification (area under curve = 0.78) for group membership.
Variability in RFF, especially when considering both time and magnitude, captures subtle features of vocal hyperfunction that may be overlooked by traditional mean measures. These findings underscore the clinical value of advanced RFF variability metrics in characterizing vocal hyperfunction.
功能亢进性嗓音障碍极为常见，但难以进行客观表征。相对基频（RFF）有潜力对这些障碍进行表征，但由于个体间平均RFF值存在变异性，其临床应用受到限制。本研究探讨RFF变异性是否能提供超越传统均值测量的见解。
收集了132名成年人的语音样本：有发声创伤性嗓音功能亢进（PVH；n = 44）、非发声创伤性嗓音功能亢进（NPVH；n = 44）以及嗓音正常者（对照组；n = 44）。计算了RFF变异性的两种测量指标——标准差和时空指数（STI），同时计算了平均RFF值。标准差反映大小的变异性，而STI则综合了时间和大小的变异性。进行了方差置换分析，以评估组（PVH/NPVH/对照组）与均值、标准差和STI测量指标之间的关系。将显著的测量指标与人口统计学参数一起纳入使用训练集（n = 102）的分层多项逻辑回归模型。然后将最终的模型方程应用于独立测试集（n = 30）以预测组成员身份。
均值和STI测量指标显示出显著的组间差异，而标准差则没有。在调整人口统计学因素后，均值和STI测量指标均改善了模型性能。对测试集进行的受试者工作特征分析得出了可接受的组成员身份分类（曲线下面积 = 0.78）。
RFF的变异性，尤其是在同时考虑时间和大小时，捕捉到了嗓音功能亢进的细微特征，而传统的均值测量可能会忽略这些特征。这些发现强调了先进的RFF变异性指标在表征嗓音功能亢进方面的临床价值。
