Acupuncture for Chronic Pain: Update of an Individual Patient Data Meta-Analysis.

The short version

This meta-analysis pooled the raw, patient-by-patient data of 20,827 people from 39 rigorously screened randomized trials — the largest and most careful analysis of acupuncture for chronic pain ever conducted. Acupuncture is the Traditional Chinese Medicine practice of inserting fine needles at specific body points; classical theory describes it as regulating qi (the body's vital functional energy) flowing through meridians (the mapped channels of that flow), while researchers here tested it purely on outcomes. Only trials with airtight randomization were included, covering four conditions: back and neck pain, osteoarthritis, chronic headache, and shoulder pain. The team obtained original data from trial authors and re-analyzed everything centrally — a gold-standard approach that removes many shortcuts ordinary meta-analyses must take. Acupuncture beat both comparison types for every condition (p < .001 — highly unlikely due to chance). Against no-acupuncture care the effect was about 0.5 standard deviations — in plain terms, starting from pain of 60/100, patients ended near 30 with acupuncture versus 43 without, and half of acupuncture patients achieved 50% pain reduction versus 30% of controls. Against sham (fake) acupuncture the margin was smaller (≈0.2 SD), meaning needle placement is only part of the benefit — but sham could not explain it all. Benefits lasted, fading only about 15% by one year. Neck and shoulder pain responded best (effect sizes 0.83 and 0.57 versus sham), and more sessions produced more relief (p = 0.001). Notably, the pooled effect matches or beats common anti-inflammatory drugs, whose effect sizes run below 0.23 for the same conditions. The authors' bottom line, unchanged since: referring a chronic pain patient for a course of acupuncture is a reasonable clinical option — with real, persistent, not-just-placebo benefit and an excellent safety profile.

Key findings

• This is the largest and most rigorous analysis of acupuncture for chronic pain: raw data from 20,827 patients across 39 randomized trials, re-analyzed centrally.
• Acupuncture outperformed both sham (fake) acupuncture and no-acupuncture care for back/neck pain, osteoarthritis, chronic headache, and shoulder pain (all p < .001).
• In practical terms, from a starting pain of 60/100, patients averaged 30 with acupuncture versus 43 without it, and half achieved at least 50% pain relief.
• Benefits persisted with only about a 15% fade at one year, and more sessions brought more relief (a dose-response, p = 0.001).
• The effect versus sham (~0.2 SD) shows needle placement is only part of the benefit — but the pooled effect equals or exceeds typical NSAID effect sizes for the same conditions.

Common questions