Editorial: Digital medicine and artificial intelligence.

VM is now an established diagnostic entity at the interface of headache medicine, neuro-otology, and functional vestibular disorders (; ). The 2022 Bárány Society/International Headache Society update defines VM on the basis of recurrent vestibular symptoms of moderate or severe intensity, a current or previous history of migraine, a temporal relationship between vestibular episodes and migrainous features, and exclusion of alternative vestibular diagnoses (). This framework has substantially improved diagnostic consistency; nonetheless, VM remains underrecognized because patients often present with heterogeneous vestibular complaints, variable attack duration, and incomplete overlap between headache and vestibular symptoms (; ). From a therapeutic perspective, the most difficult clinical question is not whether preventive therapy should be offered, but which agent is most appropriate for a given patient (). In real-world practice, treatment selection is rarely driven by a single symptom or by network meta-analysis rankings alone. Rather, clinicians must account for attack pattern, headache burden, interictal dizziness, emotional distress, visual dependence, functional disability, age, frailty, weight, and cognitive vulnerability (; ). That problem is particularly relevant for venlafaxine and topiramate, two commonly used but biologically and clinically distinct options (). This review therefore focuses on a specific and clinically relevant issue: how venlafaxine and topiramate can be selected more rationally in VM when treatment is aligned with phenotype rather than prescribed as interchangeable migraine preventives. The aim is not to propose a rigid algorithm unsupported by evidence, but to synthesize current data into a pragmatic framework that is cautious, guideline-consistent, and applicable to daily practice.