Spasticity after stroke is raised muscle tone in the affected arm or leg, and the reviews on this page mostly score it on the Modified Ashworth Scale (MAS), an assessor's rating of muscle tone. Cochrane has not published a review of acupuncture, electroacupuncture or dry needling for spasticity after stroke. The reviews that exist split by comparison. A 2010 review of sham-controlled trials by Kong, Ernst and colleagues found three chronic-stage trials that measured the MAS, and none showed a benefit; Stroke recovery reports that finding. Later reviews pooled trials that added acupuncture to rehabilitation and compared it with rehabilitation alone, and they report lower MAS scores at low or moderate certainty. Reviews of dry needling of the spastic muscles report short-term reductions, some of them against a sham.
What Cochrane reviews report
The nearest Cochrane review is the 2016 review of acupuncture for stroke rehabilitation by Ai Yang and colleagues, of 31 trials with 2,257 people in the subacute and chronic stages. Its abstract reports results for daily activities, neurological function, motor function, cognition, depression, swallowing and pain, and none for spasticity or muscle tone.
Reviews of acupuncture and electroacupuncture
| Review | Trials and people | Comparison | What the authors report | Certainty |
|---|---|---|---|---|
| Park 2014, Korea | 8 trials, 399 people | Controls without acupuncture | No effect on the MAS or on the H/M ratio, a nerve-conduction measure of reflex excitability, at the end of treatment; the H/M ratio fell immediately after the first session | Quality of all trials inadequate |
| Lim 2015, Korea, searched to July 2013 | 5 trials | Usual care or placebo acupuncture | Lower spasticity at the wrist, knee and elbow; long-term studies needed | Not stated in the abstract |
| Cai 2017, searched to September 2016 | 22 trials, 1,425 people | Electroacupuncture added to drugs and rehabilitation, against those alone | Upper-limb MAS SMD −0.57; lower-limb MAS SMD −0.88, with high heterogeneity; no add-on effect on upper-limb motor function | Not stated in the abstract |
| Xue 2022, Chengdu, searched to June 2022 | 88 trials, 6,431 people | Added to rehabilitation against rehabilitation; acupuncture against rehabilitation | MAS SMD −0.73 added to rehabilitation; −0.22 in place of it | Low (added); moderate (in place) |
| Xing 2026, Tianjin, network meta-analysis | 28 trials, 2,626 people | 14 acupuncture methods, each added to rehabilitation | Luan's three-needle technique with meridian sinew cluster needling ranked first on the MAS for the upper limb | Not graded in the abstract |
| Xiong 2026, Beijing, network meta-analysis, searched to December 2025 | 34 trials, 3,383 people | Rehabilitation, conventional treatment or sham | MAS and Clinic Spasticity Index SMD −1.03, I² 88%; warm needling and high-dose acupuncture ranked first | Moderate to low |
A standardised mean difference (SMD) expresses the gap between groups in units of the spread of scores; I² measures how much the trials disagree beyond chance. In Xue's review, most comparisons were open: two trials against sham, described rather than pooled, reported acupuncture more effective, and the authors rated the trials at high risk of bias or some concerns and asked for sham-controlled trials with objective measures. The points the 88 trials used most often were LI04 (Hegu), LI15 (Jianyu), LI11 (Quchi), TB05 (Waiguan) and LI10 (Shousanli) in the arm, and ST36 (Zusanli), GB34 (Yanglingquan), SP06 (Sanyinjiao), LV03 (Taichong) and SP10 (Xuehai) in the leg.
All 34 trials in Xiong's review were run in China, 30 of them published in Chinese, and all reported positive results, which the authors flagged as a possible source of publication or language bias. They defined high dose as at least 9 points, at least two sessions a week for 8 weeks or more, and the needling sensation called deqi, and described their network analysis as exploratory. Xing and colleagues wrote that their rankings were not yet strong enough to change guidelines. A 2019 overview of systematic reviews by Khan and colleagues, of 18 reviews of non-drug treatments for spasticity, rated the evidence for acupuncture added to routine care after stroke as moderate. Electroacupuncture and Scalp acupuncture describe two of the methods these trials used.
Reviews of dry needling
- 2020, narrative. Valencia-Chulián and colleagues found 16 studies, 7 of them randomised, which generally reported less spasticity and pain and more range of movement; control groups, muscles and measures varied too much to pool, and the authors asked for sham dry needling as the control.
- 2020. Núñez-Cortés and colleagues included 6 randomised trials with 221 people and found less spasticity in most muscles tested, at low certainty, with every trial measuring only short-term effects.
- 2021. Fernández-de-las-Peñas and colleagues pooled 7 trials: spasticity SMD −1.01 at short-term follow-up, mainly in the leg (MAS mean difference −1.05), with no effect at 4 weeks and none on motor function. They rated the evidence for the leg as moderate, downgraded for imprecision.
- 2026. Yuan and colleagues, searching to April 2026, pooled 11 trials with 276 people: MAS mean difference −0.65 overall, −0.45 against sham and −0.76 when dry needling was added to rehabilitation, largest at the ankle plantar flexors (−1.05). They rated certainty moderate to low, for small samples and blinding.
Dry needling describes the method.
Sham-controlled trials
- Leg, 2004. Fink and colleagues, at Hannover Medical School in Germany, randomised 25 people about five years after stroke, with a stiff inturned foot, to acupuncture or a purpose-built placebo needle. After 4 weeks the ankle MAS averaged 3.3 in both groups. The H-reflex, a measure of spinal nerve excitability, rose with real needling.
- Arm, 2005. Wayne and colleagues, at the New England School of Acupuncture in Massachusetts, randomised 33 people with chronic arm weakness to traditional Chinese acupuncture, including electroacupuncture, or sham, for up to 20 sessions. The intention-to-treat analysis found no difference in any outcome, spasticity included; among participants who kept to the protocol, wrist spasticity improved more with real acupuncture, in a small, unadjusted post hoc analysis.
- Dry needling, arm, 2024. Panahi and colleagues, at Shiraz University of Medical Sciences in Iran, gave 24 people 12 sessions of neurorehabilitation with 4 sessions of dry needling or sham needling of the wrist and finger flexors. Both groups improved; dry needling did better on muscle thickness, reflex torque, motor function and dexterity, but not on spasticity.
Kong 2010 counted three chronic-stage sham-controlled trials of spasticity, and all three found no benefit on the MAS. Sham acupuncture explains what such controls can and cannot show.
What guidelines say
NICE's guideline NG236 on stroke rehabilitation in adults, published in October 2023 and current, recommends stretching, splints and managing triggers for spasticity, botulinum toxin A for focal spasticity of the arm, a trial of electrical stimulation (NMES, FES or TENS) for focal spasticity, and oral baclofen for generalised spasticity. Its spasticity recommendations do not mention acupuncture or dry needling.
The Brazilian practice guidelines for stroke rehabilitation, part II, published in 2022, state that acupuncture and electroacupuncture combined with conventional care may be beneficial for spasticity, and recommend that acupuncture, electroacupuncture and functional bandaging should be considered (class IIa, level A); the same guideline rates acupuncture not useful for recovering gait (class III, level A). A 2025 review of 19 stroke rehabilitation guidelines by Zhu and colleagues, which Stroke recovery also reports, counted the Brazilian Academy of Neurology's 2022 guideline as the one, of the 19, that strongly recommended acupuncture or electroacupuncture for spasticity. Clinical guidelines sets out other bodies' positions.
What the reviews could not settle
The comparison with sham rests on a handful of trials of 25 to 33 people for acupuncture, which found no difference on the MAS, and on small dry needling trials whose pooled effect against sham Yuan rated moderate to low in certainty. The larger pooled results compare acupuncture added to rehabilitation with rehabilitation alone, where participants knew whether they were needled, so they cannot separate needling from extra treatment time and attention. Xue's review noted that the MAS is a subjective scale and that blinding of the assessors was often unclear. Xiong's review found too little long-term follow-up to judge lasting effects, and the dry needling reviews measured effects over days to weeks. Difficulty swallowing covers another problem after stroke.
AcuiQ's pages on post-stroke spasticity, upper-limb spasticity, lower-limb spasticity, spastic paralysis after stroke and hemiplegia list the protocols individual studies prescribed, each with its citation.