Validade discriminativa e acurácia diagnóstica de parâmetros posturográficos da Horus em indivíduos com e sem vestibulopatia

Abstract

Introduction: Efficient postural balance results from the joint participation of somatosensory, visual, vestibular, neuromotor, and cognitive systems. Changes in the vestibular system can cause dizziness, vertigo and postural instability. Posturography can be applied to assess postural oscillations in individuals with and without vestibular dysfunction (VD). However, there is a lack of studies investigating the discriminatory validity and diagnostic accuracy of Horus® posturographic parameters in differentiating between individuals with and without VD under different sensory conditions. Objective: To evaluate the discriminative validity and diagnostic accuracy of Horus® posturographic parameters across seven sensory conditions in individuals with and without VD. Methodology: A methodological study of discriminative validity was conducted through cross-sectional clinical research and based on COSMIN and STARD consensus. The sample constituted of individuals aged 40 to 79 years, of both sexes, divided into two groups: Group 1 (with VD) and Group 2 (without VD). Clinical characterization questionnaires, cognition (Leganés Cognitive Test); physical activity level (International Physical Activity Questionnaire); and Horus® posturography were applied using the parameters ellipse confidence/stability limit (CE/SL ratio), mean anteroposterior and mediolateral velocity (AP/MV and ML/MV), and Sensory Analysis (SA). To assess discriminative validity, non-parametric multivariate analyses (MANOVA) and Mann-Whitney tests were performed, controlled with Holm's adjustment, and rank-biserial correlation was used as an effect size estimate. Additionally, the sensitivity and specificity of the variables CE/SL, ML/MV, AP/MV from C4, Vestibular Function, and Composite Equilibrium Index (CEI) were determined using ROC curves, Youden Index, as well as calculating the area under the curve (AUC) and the 95% CI. For all statistical analyses, was used a significance level of 5% (p<0.05). Results: The final sample comprised 153 individuals, 78 from the group with VD (G1=78) and 75 without VD (G2=75), with average ages of 59.36 and 56.48 years, respectively. Regarding the discriminative validity of the posturographic parameters from Horus®, significant differences were found between individuals with and without VD for all dependent variables of CE/SL and ML/MV (p<0.001), with moderate to large differences (ES: 0.28) and in the majority of dependent variables of AP/MV and SA (p<0.005), with slight to moderate differences (ES: > 0.11 and < 0.43). However, in condition 5, "optokinetic to the right on an unstable surface," "left visual dependence," and "tunnel visual dependence" did not show significant differences. The results of the ROC curve analyses demonstrated that the CE/LS variable showed an AUC of 0.726 (p< 0.0001), with a sensitivity of 52.6% and a specificity of 84%; the ML/MV showed an AUC of 0.711 (p<0.0001), with a sensitivity of 53.85% and a specificity of 80.00%, in contrast, the AP/MV presented a lower AUC of 0.5942 (p =0.04), with a sensitivity of 67.9% and a specificity of 46.67%. The Vestibular Function obtained an AUC of 0.676 (p=0.0003), with a sensitivity of 70% and a specificity of 60.8%, while the CEI stood out with an AUC of 0.773 (p< 0.0001), sensitivity of 79.5%, and specificity of 64%. Conclusions: The parameters CE/SL, ML/MV, and AP/MV across the seven conditions and SA are moderately capable of discriminating between individuals with and without VD. Six hypotheses were accepted, indicating they discriminated against individuals with and without VD. Additionally, the results of the ROC curve analyses for the variables CE/SL, ML/MV, AP/MV from C4, vestibular function, and CEI demonstrated that the instrument has a moderate to good discriminative ability to identify individuals with and without VD, with high specificity for certain variables but moderate sensitivity, and the CEI appears to have a good combination of sensitivity and specificity.