Letters to the Editor
Vol. 46: Issue 5 - October 2026
Absent cervical VEMP responses in iron deficiency anaemia: saccular hypofunction or insufficient sternocleidomastoid activation?
Article
Dear Editor,
We read with great interest the article by Atalay et al. reporting cervical vestibular evoked myogenic potential (cVEMP) findings in 34 patients with iron deficiency anaemia (IDA) and 34 healthy controls 1. The authors are to be commended for addressing a question that, to our knowledge, had not previously been examined, and for the clarity of their laboratory-based case definition. The central observation, that cVEMP responses were absent in 30.8% of IDA ears compared with 5.8% of control ears, is striking. Precisely because the finding is novel and may enter the literature as evidence of “peripheral vestibular hypofunction” in IDA, we would like to raise 4 methodological points that, in our view, limit this interpretation and that the authors may be able to resolve with their existing data.
First, the unit of analysis. All comparisons were performed on 136 ears drawn from 68 participants, yet the 2 ears of one individual are not independent observations, particularly for a systemic exposure such as anaemia that affects both labyrinths simultaneously. Treating correlated ears as independent doubles the apparent sample size and inflates the significance of the chi-square and Mann-Whitney tests. The appropriate approaches are either a participant-level analysis (e.g. proportion of patients with at least one, or with bilaterally, absent responses) or a both-ear model using generalised estimating equations 2. The Cover figure shows a bilaterally absent recording, which suggests that the 21 absent ears may be concentrated in a smaller number of patients. We would ask the authors to report how many patients had unilateral and how many had bilateral absence; this single number would allow readers to judge the robustness of the primary result.
Second, the amplitude asymmetry ratio (AR) appears to restate the first finding rather than to add to it. Table II gives n = 34 per group for AR, which is only possible if patients with a unilaterally absent response were assigned an AR of 100%. In the IDA group, with 21 absent ears, such values would dominate the distribution, and the reported mean of 59.8% with a standard deviation of 40.2% is consistent with a bimodal sample rather than a genuine shift in inter-aural symmetry. If AR were recomputed only in participants with bilaterally present responses, we suspect the difference from controls (25.5 ± 27.1%) would diminish substantially. The summary presents increased absence and increased asymmetry as 2 independent lines of evidence; we believe they are one.
Third, and most importantly, the study did not monitor or normalise sternocleidomastoid (SCM) electromyographic activity. Participants were instructed to turn the head against gravity, but no target electromyographic (EMG) level, rectification, or background-activity correction is described. cVEMP amplitude scales approximately linearly with tonic SCM contraction, and international consensus guidelines recommend monitoring the level of muscle activation and correcting raw amplitude for it, because unmonitored recordings substantially widen the range of normal values and increase false “absent” responses 3,4. This is not a theoretical concern in IDA: fatigue and reduced exercise tolerance are cardinal symptoms of the condition, and an anaemic patient lying supine may simply generate a weaker or less sustained contraction over 150 sweeps. Two internal observations favour this explanation over saccular injury. Among ears with a present response, P1 and N1 latencies, P1-N1 amplitudes and thresholds (93.9 vs 94.5 dB nHL) were essentially identical between groups; a lesion of the saccule or inferior vestibular nerve severe enough to abolish the reflex in one-third of ears would be expected to elevate thresholds and reduce amplitudes in the remaining ears, whereas a purely effort-dependent artefact would produce exactly the all-or-none pattern observed. Reporting the mean rectified EMG per group, or at least confirming that the device provided a visual EMG target, would go a long way towards excluding this confounder.
Fourth, 2 confounders that could readily have been addressed were not. Vitamin B12 deficiency, which the authors themselves cite as a cause of absent cVEMP responses and reduced amplitudes 5, frequently coexists with iron deficiency and was not listed among the exclusion criteria, nor was diabetes mellitus. Moreover, no dose–response analysis was attempted; a simple correlation between haemoglobin or ferritin and response presence or amplitude would have strengthened the causal argument considerably and is available from the data collected. In this context, the interpretation of “ischaemic damage” to the labyrinth in patients whose anaemia was defined only by a haemoglobin below 12-13.5 g/dL and was not stratified by severity seems premature; iron-dependent mitochondrial and enzymatic pathways in hair cells, which the authors also mention, do not require tissue hypoxia and would be a more parsimonious hypothesis. Finally, a single 500 Hz stimulus was used, and recent data show that responses absent at 500 Hz may be recovered at 750-1000 Hz when the resonance properties of the otolith organs are altered 6; testing at a second frequency would help to distinguish true areflexia from frequency tuning.
None of these points detracts from the originality of the hypothesis. On the contrary, we believe the question deserves a definitive answer, and the authors are well placed to provide it: a prospective design measuring EMG-normalised cVEMP and ocular VEMP at 2 frequencies before and 3 months after iron repletion, analysed at the participant level with adjustment for B12 status and correlated with ferritin, would establish whether the abnormality is real, reversible and attributable to iron. We look forward to the authors’ response and to further work on this topic.
Conflict of interest statement
The author declares no conflict of interest.
Funding
This letter did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
History
Received: September 3, 2026
Accepted: September 21 2026
References
- Atalay F, Yaşar M, Uyanık T. Vestibular evoked myogenic potentials in patients with iron deficiency anaemia. Acta Otorhinolaryngol Ital. 2026;46:156-160. doi:https://doi.org/10.14639/0392-100X-N3038
- Chen C, Zhang N, Curhan G. Both-ear method for the analysis of audiometric data. Ear Hear. 2022;43:1447-1455. doi:https://doi.org/10.1097/AUD.0000000000001216
- Papathanasiou E, Murofushi T, Akin F. International guidelines for the clinical application of cervical vestibular evoked myogenic potentials: an expert consensus report. Clin Neurophysiol. 2014;125:658-666. doi:https://doi.org/10.1016/j.clinph.2013.11.042
- Rosengren S, Colebatch J, Young A. Vestibular evoked myogenic potentials in practice: methods, pitfalls and clinical applications. Clin Neurophysiol Pract. 2019;4:47-68. doi:https://doi.org/10.1016/j.cnp.2019.01.005
- Özdemir D, Mehel D, Küçüköner Ö. Vestibular evoked myogenic potentials in patients with low vitamin B12 levels. Ear Nose Throat J. 2021;100:NP231-NP235. doi:https://doi.org/10.1177/0145561319878952
- Şenol B, Ocak E, Aydoğan Z. Evaluation of the effect of aging on frequency in cervical vestibular evoked myogenic potentials. Eur Arch Otorhinolaryngol. 2026;283:3647-3654. doi:https://doi.org/10.1007/s00405-026-10090-6
Downloads
License
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License.
Copyright
Copyright (c) 2026 Società Italiana di Otorinolaringoiatria e chirurgia cervico facciale
How to Cite
- Abstract viewed - 171 times
- PDF downloaded - 26 times

