Inspired by Sun et al 2024.[1]
OA – osteoarthritis
OAK – OA of the knee
MA – manual acupuncture
EA – electroacupuncture
WA – warm acupuncture
MM – mild moxibustion
SA – sham acupuncture
VAS – visual analogue scale
WOMAC – Western Ontario and McMaster Universities OA index
BMI – body mass index
ARC – Acupuncture in Routine Care studies (part of the Modellvorhaben Akupunktur)
MCID – minimum clinically important difference
ITT – intention to treat– key to acronyms
This paper was published in the journal Pain and Therapy (IF 4.1) – an open access journal from Springer. The journal has been published since 2012
The research comes from Hangzhou, which I have mentioned on here several times before – see, for example: ST36 EA and anti-tumour effects from 9th March 2021, or simply type Hangzhou into the search box on the top right.
This is a moderately large multicentre trial (n=357) but includes 6 parallel arms. 4 arms received some form of needling, 2 arms received moxibustion and one arm received celecoxib. That is too many arms I hear you retort! Yes, one of the acupuncture arms also received moxibustion, the so-called warm acupuncture (WM) group.
The acupuncture points used were familiar ones around the knee – SP10, SP9, GB34, ST34, and the 2 points known as Xiyan (eyes of the knee). The latter can also be described as ST35 (on the joint line lateral to the patella tendon) and Neixiyan (on the joint line medial to the patella tendon). Those of you who have heard me talk on the topic know that I am not a fan of these points because they typically penetrate into the knee joint, which does not have many nerves on the inside, but does have a good environment for culturing bacteria and is not far from the skin surface. Anyway, I’ll get over it for the sake of this paper.
Treatments were performed for 30 minutes, 3 times a week for 4 weeks. The needles used were 0.30mm diameter and 50mm in length, but the needles were only inserted 25mm to 30mm. Having said that the distance to the joint space at the Xiyan points is typically less than 10mm. The MA group received quite vigorous manual stimulation at all 6 points. The EA group received MA plus additional EA across the Xiyan points at 2/100Hz at a tolerable intensity. The WA group received MA plus additional moxibustion on the needles at the Xiyan points. The MM group only received indirect moxibustion at the Xiyan points with no direct skin contact. The SA group received superficial insertion (1-2mm) of fine short needles (0.18mm by 25mm) at non-acupuncture points very close to the real points and no manual stimulation. The celecoxib group received 0.2g of the drug per day orally for 4 weeks.
The primary outcome measures were VAS pain (average pain intensity over the previous 3 days) and the physical function score of the WOMAC. The primary end point was at 4 weeks, immediately following the treatment phase.
The groups appeared to be well matched at baseline in terms of demographics and other relevant statistics (height, weight, BMI, comorbidities).
Baseline VAS was between 5 and 6 (0 to 10 scale) and baseline WOMAC physical function score was between 50 and 60. The latter is a score from 0 to 68, but sometimes it is adjusted up to a percentage. If unadjusted it seems a bit high in comparison to the VAS. If adjusted to a percentage it seems a bit low for the population. I guess I have to assume it is unadjusted.
For comparison, the baseline WOMAC physical function score from the ARC OA hip and knee study was 46.5. 73.9% were OAK and the total number of patients included was 3 553.[2]
The change in VAS from baseline to 4 weeks was 2 to 3 and the change in WOMAC physical function scores was 12 to 30. The MCID for VAS pain is general taken as 1 to 2 and that for WOMAC physical function score as 6 to 8, so you can see why I am searching to see whether or not the WOMAC figures have been adjusted. Also, the sample size calculation in the paper quotes this being based on previous research where the change in WOMAC physical function scores was 8.8, 8.9, 12, 10, 5.8, and 13 for MA, EA, WA, MM, and celecoxib respectively. In the current study the figures for the change scores are 20, 32, 27, 27, 13, and 17.
For comparison again, the change from baseline in WOMAC physical function score in the acupuncture group of the ARC study at 3 months (OAK only) was 16.3. This was a change of about 35%.[2] The percentage change in the current study was 39% at 4 weeks in the MA group, but 56% in the EA group.
I guess whether they are adjusted or not does not matter for the between group comparisons, which have EA coming out on top. In the ITT, EA was better than MA and celecoxib, but not better than WA. WA was not better than MM, which seems a surprise, since the former (WA) is a combination of MM plus MA.
References
1 Sun J, Liang Y, Luo K-T, et al. Efficacy of Different Acupuncture Techniques for Pain and Dysfunction in Patients with Knee Osteoarthritis: A Randomized Controlled Trial. Pain Ther. Published Online First: 17 February 2025. doi: 10.1007/s40122-025-00713-x
2 Witt CM, Jena S, Brinkhaus B, et al. Acupuncture in patients with osteoarthritis of the knee or hip: A randomized, controlled trial with an additional nonrandomized arm. Arthritis Rheum. 2006;54:3485–93. doi: 10.1002/art.22154
The BMAS Blog 
You must be logged in to post a comment.