Acupuncture for Migraine Without Aura and Connection-Based Efficacy Prediction: A Randomized Clinical Trial
Zhang et al. · JAMA Network Open · 2026
Evidence Level
STRONGOBJECTIVE
To assess the efficacy of verum vs. sham acupuncture in migraine without aura and to identify brain connectivity patterns that predict treatment response
WHO
120 adults (18-65 years) with migraine without aura, 79% women, mean age 36.8 years
DURATION
4-week baseline + 4 weeks of treatment (12 sessions)
POINTS
8 points: GV-20, GV-16, GB-20 bilateral, EX-HN5 bilateral, LI-4 bilateral with deqi stimulation
🔬 Study Design
Verum acupuncture
n=60
Needling of traditional points with deqi stimulation
Sham acupuncture
n=60
Superficial needling at non-acupoint sites without deqi
📊 Results in numbers
Reduction in migraine days per month
Reduction in pain intensity (VAS)
Improvement in disability (HIT-6)
Neuroimaging prediction accuracy
📊 Outcome Comparison
Reduction in monthly migraine days
This study showed that real acupuncture is more effective than sham acupuncture in reducing pain and improving quality of life in people with migraine. In addition, MRI scans can help identify which patients respond best to treatment, paving the way for more personalized medicine.
Article summary
Plain-language narrative summary
This innovative randomized clinical trial combined the study of acupuncture efficacy for migraine without aura with advanced neuroimaging techniques to develop predictive models of treatment response. Conducted in Beijing between 2021 and 2023, the study enrolled 120 adult participants with a confirmed diagnosis of migraine without aura according to international criteria. Participants were randomized to receive verum or sham acupuncture for 4 weeks, with twelve 30-minute sessions. The verum acupuncture protocol used 8 traditional points (Baihui, Fengfu, bilateral Fengchi, bilateral Taiyang, and bilateral Hegu) with stimulation to elicit deqi sensation, while the control group received superficial needling at non-meridian points without deqi.
All participants underwent functional MRI at baseline for analysis of brain connectivity. The results demonstrated significant superiority of verum acupuncture across multiple measures. There was an additional 1-day-per-month reduction in migraine episodes, a 1-point decrease on the visual analog scale for pain, and a 2.9-point improvement on the Headache Impact Test (HIT-6). Quality of life also showed significant improvements across all domains assessed by the Migraine-Specific Quality of Life Questionnaire (MSQ).
The principal innovation of the study lay in the application of connectome-based predictive modeling (CPM), a machine learning technique that analyzes patterns of functional connectivity throughout the brain. This analysis identified specific neural signatures that predict treatment response: hypoconnectivity between the default mode network and subcortical-cerebellar regions predicted pain reduction, while hyperconnectivity between subcortical-cerebellar regions and motor networks predicted improvement in functional disability. These findings suggest distinct neurobiological mechanisms underlying the different therapeutic benefits of acupuncture. The study has several important methodological strengths.
The randomized controlled design with adequate blinding (confirmed by the Bang index) ensures the validity of the clinical results. The pioneering application of CPM to acupuncture research represents a significant advance in understanding the neural mechanisms of treatment. The sample size was adequately calculated, and the acupuncture protocol followed established guidelines, with experienced acupuncturists. Limitations include the exclusive focus on migraine without aura, limiting generalizability to other types of headache.
The absence of post-treatment neuroimaging precluded the analysis of dynamic changes in brain connectivity. The 4-week treatment period may be insufficient to capture all the benefits of acupuncture, which often accumulate over time. The clinical implications are substantial. The results confirm the efficacy of acupuncture for migraine, aligning with prior evidence and international guidelines that recommend acupuncture as preventive treatment.
More importantly, the identification of neuroimaging biomarkers paves the way for personalized medicine in acupuncture. In the future, MRI scans could help clinicians identify patients with the highest likelihood of responding to treatment, optimizing healthcare resources and avoiding unnecessary treatments. This study lays the groundwork for a more precise and individualized approach to treating migraine with acupuncture.
Strengths
- 1First study to apply connectome-based predictive modeling to acupuncture for migraine
- 2Methodologically rigorous design with adequate randomization and blinding
- 3Standardized acupuncture protocol based on prior evidence
- 4Comprehensive analysis including multiple clinical outcome measures
Limitations
- 1Focused only on migraine without aura, limiting generalizability
- 2Absence of post-treatment neuroimaging for longitudinal analysis
- 3Relatively short treatment period (4 weeks)
- 4Sample may be insufficient for some subgroups
Expert Commentary
Dr. Marcus Yu Bin Pai
MD, PhD · Pain Medicine · Physical Medicine and Rehabilitation · Medical Acupuncture
▸ Clinical Relevance
Acupuncture was already part of the preventive arsenal for migraine in international guidelines, but what this work adds in practical value is the concrete possibility of stratifying patients before initiating treatment. In running a pain clinic, I routinely see patients who have already failed two or three oral prophylactics and arrive at acupuncture as a salvage option. The reduction of one additional migraine day per month and the 2.9-point improvement on the HIT-6 may seem modest in absolute terms, but for the patient with chronic migraine and high functional burden, this delta has real relevance. The data showing superiority over sham, with adequate randomization and a standardized protocol of 12 sessions over 4 weeks, offers robust grounding to support the indication in adult patients with migraine without aura who are seeking a non-pharmacological alternative or an adjunct to ongoing drug treatment.
▸ Notable Findings
The most intriguing finding is not the clinical efficacy itself but the predictive modeling based on brain connectome. The study identified that hypoconnectivity between the default mode network and subcortical-cerebellar regions predicts pain reduction, while hyperconnectivity between subcortical-cerebellar regions and motor networks predicts improvement in functional disability—distinct neural mechanisms for distinct outcomes. This dissociation is neurophysiologically relevant: it suggests that analgesia and functional recovery do not share the same neural substrate, which has implications for how we conceptualize migraine beyond a purely algic phenomenon. The correlation r = 0.29 obtained by the predictive model is preliminary, but it is the first demonstration that baseline functional connectivity patterns can guide therapeutic selection in acupuncture—a movement toward precision medicine in functional neurology.
▸ From My Experience
In my practice, the patient profile that responds best to acupuncture for migraine is the one with high attack frequency, well-lateralized pain pattern, and some component of cervical tension—which may reflect precisely the subcortical connectivity that the study maps. I usually observe perceptible clinical response between the third and fifth sessions, generally expressed by the patient as a reduction in attack intensity before a reduction in frequency. I typically run cycles of 10 to 12 sessions and, when response is satisfactory, maintain monthly sessions for 3 to 6 months. I systematically combine this with sleep hygiene guidance, trigger management and, when indicated, maintenance of the oral prophylactic at a lower dose. The prospect of using neuroimaging to predict response is not yet operationalizable in everyday clinical care, but the logic of identifying the responder profile before initiating treatment—whether by clinical assessment or biomarker—already guides my case selection.
Full original article
Read the full scientific study
JAMA Network Open · 2026
DOI: 10.1001/jamanetworkopen.2025.55454
Access original articleScientific Review
Marcus Yu Bin Pai, MD, PhD
CRM-SP: 158074 | RQE: 65523 · 65524 · 655241
PhD in Health Sciences, University of São Paulo. Board-certified in Pain Medicine, Physical Medicine and Rehabilitation, and Medical Acupuncture. Scientific review and curation of every entry in this library.
Learn more about the author →Medical disclaimer: This content is for educational purposes only and does not replace consultation, diagnosis, or treatment by a qualified professional. Some information may be assisted by artificial intelligence and is subject to inaccuracies. Always consult a physician.
Content reviewed by the medical team at CEIMEC — Integrated Centre for Chinese Medicine Studies, a reference in Medical Acupuncture for over 30 years.
Related articles
Based on this article’s categories