Regarding: “comparative effects of pharmacological interventions in the prophylactic treatment of tension-type headache: systematic review and network meta-analysis”

Regarding: "comparative effects of pharmacological interventions in the prophylactic treatment of tension-type headache: systematic review and network meta-analysis"We read with interest the systematic review and network meta-analysis (NMA) by Tao et al. examining pharmacological interventions for tension-type headache (TTH) prophylaxis 1.The dose-stratified approach advances beyond prior reviews that pooled heterogeneous dosing regimens.However, several methodological concerns merit discussion.First, the network combines trials using oral placebos (for amitriptyline, tizanidine, mirtazapine) with those using injectable placebos (for BTX-A and lidocaine), yet treats all placebos as a single common comparator node.Evidence suggests that invasive placebos exert systematically stronger effects than oral placebos 2, which the authors themselves acknowledge.This placebo heterogeneity threatens the transitivity assumption underlying NMA, potentially inflating SUCRA rankings for injectable treatments.Given that BTX-A 100 U and lidocaine 25 ml feature prominently in the results, a sensitivity analysis restricting the network to oral-placebo-controlled trials, or node-splitting oral from injectable placebos, would substantially strengthen the interpretability of comparative rankings.Second, while the dose-stratified strategy is conceptually sound, splitting drugs into numerous dose-specific nodes (e.g.seven separate BTX-A doses) creates an extremely sparse network where many nodes are supported by only a single small trial.This fragmentation inflates heterogeneity estimates ( 2 reaching 19.13 at 24 weeks), widens credible intervals until most comparisons become uninformative, and renders SUCRA rankings unreliable, as demonstrated in simulation studies 3.A hierarchical dose-response model borrowing strength across doses of the same drug would more principally balance granularity with statistical precision.Alternatively, grouping doses within pharmacologically plausible ranges could yield more robust estimates for clinical decision-making.Third, the included trials span over three decades and 18 countries, during which diagnostic criteria evolved from ICHD-I to ICHD-II and clinical practice changed substantially.The analysis does not explore whether diagnostic version, publication era, or geographic region contributed to heterogeneity through meta-regression, despite the high 2 values.Furthermore, concomitant analgesic use, medication overuse status, and psychiatric comorbidity burden-strong prognostic factors in chronic TTH 4-are neither captured nor discussed.Amitriptyline's apparent superiority may partly reflect its dual action on comorbid depression and sleep disturbance rather than a purely analgesic mechanism.Finally, reporting aggregate adverse event rates without severity stratification substantially limits clinical utility.Conflating mild anticholinergic effects with treatment-limiting events under a single odds ratio (e.g.OR 9.53 for amitriptyline 100 mg) hinders the benefit-risk appraisal and shared decision-making that the authors advocate 5.Extracting discontinuation rates due to adverse events would better inform tolerability assessments.In summary, this NMA addresses an important evidence gap, but its conclusions would be strengthened by formally addressing placebo heterogeneity between oral and injectable comparators, employing hierarchical dose-response modeling to mitigate network sparsity, conducting meta-regression to explore key heterogeneity sources, and refining the safety analysis by adverse event severity.