The Long-term Effect of Acupuncture for Migraine Prophylaxis — A Randomized Clinical Trial

Zhao L, et al. · JAMA Internal Medicine, 2017 · doi:10.1001/jamainternmed.2016.9378

RCT · n=249GRADE: ModerateExpert levelNNT ≈ 4–5✓ every figure fact-checked against the source

Clinical bottom line

A 24-week, 3-arm RCT (true acupuncture vs sham vs waiting list) in episodic migraine without aura. True acupuncture reduced mean monthly attack frequency by 3.2 vs 1.4 (waiting list) at week 16, with the effect maintained through week 24. Adequately powered for the primary outcome; assessor-blinded.

Risk of bias (Cochrane RoB 2)

Domain	Judgement	Note
Randomization	Low	Central computer-generated sequence, concealed allocation
Blinding (participants)	Some concerns	Sham credibility assessed; practitioners unblinded by design
Outcome measurement	Low	Blinded assessors; diary-based standardized outcome
Selective reporting	Low	Pre-registered protocol; outcomes match registration

What this changes at the chair

For episodic migraine prophylaxis, a structured preventive course (weekly, reassessed ~session 8) is supported by moderate-certainty evidence. Effect size vs usual care is clinically meaningful; counsel patients honestly on the contribution of non-specific effects and individual variability.

Limitations the abstract won't tell you

Single-country recruitment limits generalizability; the sham arm used shallow needling at non-points, which may carry mild physiological activity — a conservative bias direction. Migraine-days (not attacks) would be the modern regulatory endpoint.

AI appraisal for licensed professionals — verify against the original before clinical use. · qiboo.ai