Hip pain

Hip osteoarthritis has its own Cochrane review of acupuncture, separate from the knee since 2018, and it is small: six trials with 413 people, only two of them against sham acupuncture, the control procedure that imitates needling. Other hip pain has less still. Pain after hip replacement has been studied in a handful of trials of needling around surgery, and greater trochanteric pain syndrome, pain over the bony point on the outside of the hip, in one trial of dry needling. The guidelines that rate acupuncture for hip osteoarthritis disagree, as they do for the knee, and the two that grade the hip evidence separately grade it no higher than the knee’s.

Sham acupuncture covers the controls these trials used and why none of them is inert.

The Cochrane review

Manheimer and colleagues published the review in 2018, searching to March 2018. It came from splitting their 2010 review of peripheral joint osteoarthritis into a hip review and a knee review. Four of the six trials enrolled only people with hip osteoarthritis and two enrolled people with hip or knee osteoarthritis. Participants were mostly women in their sixties with two to eight years of hip pain, and every result was measured four to nine weeks after randomisation. Certainty is graded under GRADE, the standard scale from high to very low.

The review judged both shams believable, and both at risk of weak effects of their own: one placed non-penetrating needles on the real points and the other put real needles in the wrong places. Four trials reported adverse events and none reported a serious one attributed to acupuncture; two reported minor bruising, bleeding or pain where needles went in. The authors concluded that acupuncture probably has little or no effect on pain or function relative to sham, and that the benefit in the unblinded trial is likely due at least partly to participants’ greater expectations of it.

Earlier and wider reviews

The 2010 Cochrane review that the hip review replaced covered knee, hip and hand osteoarthritis: 16 trials with 3,498 people, of which 3 enrolled only hip osteoarthritis and 1 a mix of hip and knee. It pooled the joints together and did not meta-analyse the hip trials on their own; Knee osteoarthritis reports its results, which the knee trials dominate.

The Acupuncture Trialists’ Collaboration pooled raw patient data in its 2018 update by Vickers and colleagues: 39 trials with 20,827 patients across four chronic pain conditions, osteoarthritis among them. Acupuncture did better than sham and than no acupuncture for each condition, by about 0.2 and 0.5 standard deviations respectively, and effects shrank by about 15% at one year. Its osteoarthritis trials include White 2012, which enrolled people with hip or knee osteoarthritis, but the analysis reports no result for the hip on its own. Chronic pain covers it in full.

A 2023 review by Park and colleagues took hip pain of any cause, searching eight databases to August 2022, and included 12 trials with 806 patients, some of them around hip surgery. It found no serious adverse events and called the evidence weak given low quality and small samples.

The hip trials

The largest trial with hip patients is the German ARC trial by Witt and colleagues, published in 2006. It randomised 712 people with chronic knee or hip osteoarthritis to up to 15 sessions of acupuncture over three months, added to routine care, or to routine care alone; another 2,921 who declined randomisation had acupuncture and were followed alongside. At three months the WOMAC index, the standard osteoarthritis questionnaire, had improved by 16.7 points more in the acupuncture group, and the gain held to six months. Treating doctors chose the points. The trial had no sham arm, and its hip patients are the single trial behind the Cochrane review’s routine-care comparison.

Two trials compared needling with sham. Fink and colleagues, in 2001, treated 67 people at a German university rehabilitation department, the real group at points including GB30 (Huantiao), GB31 (Fengshi), BL54 (Zhibian) and GB34 (Yanglingquan), the sham group with needles placed away from the classical points without manipulation. Both groups improved at two weeks and two months, with no difference between them. White and colleagues, in 2012, randomised 221 people awaiting hip or knee replacement to real acupuncture, a non-penetrating Streitberger needle or mock electrical stimulation, each with an empathic or a neutral consultation, eight sessions over four weeks. Real and placebo acupuncture differed by 2.7 mm on a 100 mm pain scale, which was not significant, while one practitioner’s patients reported 10.9 mm more relief than another’s, which was.

The three trials against active treatment were small and unblinded. Haslam’s 2001 trial randomised 32 people awaiting hip replacement to six sessions of acupuncture at points including GB29 (Juliao), GB30 and GB34, or to advice and exercises, and reported greater improvement on a modified WOMAC with acupuncture immediately after treatment and eight weeks later. Stener-Victorin and colleagues in 2004 gave 45 people electroacupuncture, hydrotherapy or neither, all with patient education, using local points including GB29, GB30 and ST31 (Biguan); both active groups improved and the education group did not. A 2010 Chinese trial with 60 people compared acupuncture with diclofenac and is known here only through the Cochrane review.

Pain after hip replacement

Usichenko and colleagues, in 2005, placed permanent press needles at four ear points (lung, shenmen, thalamus and hip) or at four non-acupuncture points on the rim of the ear, kept for three days after hip replacement, in 54 patients who completed the study. Those with the ear points used 37 mg of the opioid piritramide in 36 hours against 54 mg with sham; pain scores and analgesic side effects did not differ. A 2025 review by Tanghe and colleagues of acupuncture before or during hip and knee replacement found eight studies, seven of them randomised, too varied to pool; they reported less fentanyl during surgery and less patient-controlled analgesia afterwards, with no consistent change in patient-reported outcomes. Postoperative pain covers acupuncture after surgery generally.

Greater trochanteric pain syndrome

No review of acupuncture for greater trochanteric pain syndrome was found. A 2025 network meta-analysis by Wang and colleagues of conservative treatments included 19 trials with 1,701 participants and ranked exercise first; one of its trials tested dry needling, which inserts needles into muscle trigger points rather than acupuncture points. In that trial, Brennan and colleagues in 2017 randomised 43 people (50 hips) to dry needling or a cortisone injection over six weeks and found dry needling no worse on pain or function. A 2024 review of dry needling around the hip found seven trials with 273 patients across hip osteoarthritis, greater trochanteric pain syndrome, piriformis syndrome and healthy athletes, could not pool them, and found strong evidence lacking.

What the guidelines say

NICE, which writes clinical guidance for the NHS in England, published its osteoarthritis guideline, NG226, on 19 October 2022. Recommendation 1.3.8 says not to offer acupuncture or dry needling to manage osteoarthritis, with no distinction between joints. Its rationale says the evidence was predominantly for the knee, showed a lack of benefit and some evidence of harm, and that acupuncture was not cost-effective.

The American College of Rheumatology and the Arthritis Foundation, in their 2019 guideline by Kolasinski and colleagues, conditionally recommended acupuncture for knee, hip and hand osteoarthritis. Its text notes that the most positive trials with the largest effects were in the knee and says nothing separately about the hip. The College’s 2026 update, approved by its board on 8 September 2026 and published as a summary ahead of the full paper, again conditionally recommends acupuncture for the knee, hip and hand, and rates the certainty of evidence moderate for knee and hip.

The Royal Australian College of General Practitioners, in its 2018 guideline for knee and hip osteoarthritis, suggests not offering traditional, laser or electroacupuncture, a conditional recommendation against. It rates the evidence low for the knee and very low for the hip, where it found no significant effect of traditional acupuncture against sham. The College states that the guideline is being updated. Clinical guidelines sets these positions beside other bodies’.

What the reviews could not settle

The sham comparison rests on two trials with 120 people. Cochrane rated that evidence moderate, yet its confidence interval still runs from a moderate benefit to none, and both shams may have had effects of their own. The guidelines read the same few hip trials differently: the ACR rates their certainty moderate and recommends acupuncture conditionally, the RACGP rates it very low and advises against, and NICE judged osteoarthritis from mostly knee evidence. Every hip result in the Cochrane review was measured within nine weeks, so no review established how long any effect lasts in the hip. No sham-controlled trial of electroacupuncture for the hip was found, though NICE asked for research on electroacupuncture in osteoarthritis. After hip replacement the trials measured opioid use more consistently than pain, and greater trochanteric pain syndrome has one dry-needling trial and no trial of acupuncture at acupuncture points.

AcuiQ’s hip pain page lists the protocols individual studies prescribed, each with its citation.