Does acupuncture help arthritis pain

For osteoarthritis, the evidence depends sharply on which type of needling you mean – electroacupuncture shows a real but modest edge over sham; manual acupuncture alone does not.

The 2021 randomized trial of 480 people with knee osteoarthritis is the strongest single piece of evidence here [5][6]. Intensive electroacupuncture (EA) produced a 13 per 100 response-rate advantage over sham at 8 weeks (60% vs 47%, p=0.023), and that gap held through 26 weeks. Manual acupuncture (MA) did not reach statistical significance at 8 weeks (59% vs 47%, p=0.05) but was close. Both real treatments beat sham on pain scores and function at most timepoints, and the differences were about 0.5 to 1 point on a 10-point pain scale – noticeable but not huge. The sham group also improved substantially (47% response), which signals a large placebo component.

Earlier trials and reviews tell a different story for basic needle acupuncture. The 1994 trial of OA knee found no meaningful difference between real and sham acupuncture on any pain measure [1]. The 2007 meta-analysis in the Annals of Internal Medicine concluded that any short-term benefit of manual acupuncture for knee OA was not clinically meaningful compared to sham [2]. A 2006 study found that real and sham acupuncture both beat conservative therapy equally, and were indistinguishable from each other [4]. The Cochrane review for hip OA (2018) said acupuncture "likely has little to no effect" compared to sham [3].

For rheumatoid arthritis, the Cochrane review from 2005 found only limited evidence of benefit for acupuncture or electroacupuncture, with the studies too small and methodologically weak to draw firm conclusions [7].

The adverse event rates across these trials are low. In the 2021 trial, about 1 in 10 people had a mild event (bruising, post-needling pain) regardless of group [5][6]. A systematic review of 58 trials found that sham acupuncture itself produced a nocebo effect: adverse events were about 5 per 100 higher in sham than in no-treatment groups [8].

The table below shows the main comparison from the most rigorous modern trial:

My call: Intensive electroacupuncture offers a modest, real but not large, pain reduction for knee osteoarthritis that lasts at least 6 months. Manual acupuncture alone has not consistently beaten sham in well-controlled trials. For rheumatoid arthritis, the evidence is too thin to recommend it. The harms are mostly minor bruising and soreness, but sham acupuncture also produces a nocebo effect. Confidence: moderate for electroacupuncture in knee OA; low for manual acupuncture in knee OA; not clear for any form in rheumatoid arthritis.