Dear editor

We read with interest the article by Lian et al¹ on acupuncture-related therapies for post-stroke sleep disorders (PSSD). While valuable, several methodological concerns merit discussion.

First, the study’s definition of PSSD requires scrutiny, as it did not adequately account for the distinct clinical entities encompassed by PSSD, which include post-stroke insomnia, hypersomnia,² obstructive sleep apnea (OSA),³ and restless legs syndrome (RLS),⁴ each with unique pathophysiology. Lian et al excluded OSA and RLS, asserting that these conditions are unrelated to stroke. However, both are prevalent among post-stroke patients and significantly impact sleep quality and recovery.^3,4

While the omission of continuous positive airway pressure therapy and dopaminergic agents likely stems from this exclusion, the claim that “Current Western medical treatments primarily rely on pharmacotherapy, including benzodiazepines…” remains misleading. Benzodiazepines are generally not recommended for post-stroke insomnia due to adverse effects and risk of motor deficits recurrence.² Instead, CBT-i is considered the first-line treatment for stroke-related insomnia.² The failure to acknowledge these established therapeutic approaches undermines the clinical relevance of the study’s framework.

Second, the classification of acupuncture therapies requires further clarification. Electro-acupuncture was treated as a standalone intervention rather than a needling technique under the broader acupuncture umbrella. Moxibustion, which uses heat from burning Artemisia vulgaris, was grouped under acupuncture, despite being fundamentally different from needle-based methods.⁵ These conceptual inconsistencies may have biased the analysis.

Third, the exclusive use of the PSQI as the primary outcome is problematic. PSQI is a general sleep quality measure and does not capture features specific to PSSD subtypes such as post-stroke OSA or post-stroke RLS. Objective metrics like apnea-hypopnea index or RLS questionnaires like International RLS Study Group Rating Scale would be more appropriate. Additionally, two included RCTs measured only serotonin and norepinephrine levels without PSQI data, and employed transcutaneous electrical acupoint stimulation which does not involve needle penetration. Including studies that do not meet predefined eligibility criteria may bias the meta-synthesis, potentially over- or underestimating acupuncture’s true therapeutic effect.

Fourth, Lian et al’s study pooled clinically heterogeneous PSSD conditions but lacked subtype-stratified analyses, obscuring potential treatment response variations across subtypes. By aggregating mechanistically distinct conditions under a single diagnostic category and applying non-specific outcome measures, the study limited clinically meaningful conclusions regarding acupuncture’s efficacy for specific PSSD presentations.

Fifth, the article failed to assess the safety of acupuncture, despite existing evidence indicating a 3–6% incidence of adverse events in post-stroke insomnia treatment.⁶ This omission hinders understanding of acupuncture’s risk-benefit balance.

Sixth, although inconsistency testing was unnecessary due to the absence of closed loops, Lian et al neither assessed between-trial heterogeneity nor performed subgroup analysis, meta-regression, or sensitivity analysis to explore its sources, which likely produced artificially narrow confidence intervals, masking the true uncertainty around the estimated treatment effects.⁷

Seventh, the outdated 2008 Cochrane RoB tool was used instead of the updated RoB 2.0 tool which was developed to improve assessment reliability.⁸ Furthermore, the article did not grade the strength of evidence for outcomes derived from the network meta-analysis. Following the PRISMA for Network Meta-Analyses statement⁹ would enhance methodological rigor and transparency.

In conclusion, while Lian et al’s study offers preliminary insights into the role of acupuncture therapies in managing PSSD, the methodological limitations outlined above must be addressed to ensure more robust and clinically translatable findings. Future studies should prioritize (1) delineating which PSSD subtypes respond to acupuncture—whether as monotherapy or adjunctive therapy—and (2) characterizing its safety profile in this vulnerable population. These constructive critiques aim to guide subsequent investigations and foster scholarly discourse to enhance methodological quality in this field.