Association between endolymphatic hydrops severity and audio-vestibular outcomes in Ménière’s disease: A systematic review and meta-analysis of correlation coefficients

Ménière’s disease (MD) is characterized by episodic vertigo, fluctuating sensorineural hearing loss (SNHL), tinnitus, and aural fullness, with endolymphatic hydrops (EH) considered its key pathological substrate. Despite advances in EH assessment (magnetic resonance imaging [MRI]-based grading and physiologic tests), the strength and consistency of associations between EH severity and audio-vestibular measures remain uncertain. This study aimed to review systematically and meta-analyze correlation coefficients quantifying the relationship between EH severity and audiometric and vestibular measures in adults with MD. From inception through September 2025, we searched PubMed, Scopus, Web of Science, and Cochrane for publications associating EH with audio-vestibular outcomes in patients with MD. We independently screened and abstracted relevant data, then assessed bias using National Institutes of Health (NIH) tools. Correlations were Fisher Z-transformed and combined using random-effects models. Heterogeneity (I², τ²) and subgroups (cochlear/vestibular and frequency) were assessed. Sixteen studies were included. EH severity showed a moderate correlation with poorer hearing with pure-tone audiometry (PTA): cochlear r = 0.41 (95% CI 0.34–0.48; I²=0%) and vestibular r = 0.42 (0.33–0.51; I²=30.9%). Caloric canal paresis (CP%) correlated weakly (cochlear r = 0.14 [0.01–0.26]; vestibular r = 0.22 [0.10–0.34]; both I²=0%). Clinical stage correlated weak–moderately (cochlear r = 0.33 [0.22–0.43; I²=27.8%]; vestibular r = 0.36 [0.22–0.49; I²=38.4%]). Frequency-specific PTA associations were strongest at low/mid bands (cochlear r = 0.49/0.45; vestibular r = 0.51/0.42) and weaker at high frequencies (cochlear r = 0.30; vestibular r = 0.34). Disease duration showed weak correlations (cochlear r = 0.20 [I²=0%]; vestibular r = 0.33 [I²=30.2%]). The EH severity grades demonstrate moderate correlation with poorer hearing and weak correlation with caloric paresis, and clinical stage shows weak-moderate correlation. Relations are strongest at low and mid frequencies and are more consistent for cochlear than vestibular EH. These observations support EH as an imaging correlate of auditory involvement in MD, particularly for low- and mid-frequency hearing thresholds. However, correlations do not establish causality, and standardized imaging with longitudinal designs is needed to clarify temporal relationships and prognostic utility.