Gastroparesis is delayed emptying of the stomach with no blockage to explain it. The 2021 European consensus of United European Gastroenterology (UEG) and the European Society for Neurogastroenterology and Motility (ESNM), written by Schol and colleagues, defines it as symptoms that come with delayed gastric emptying in the absence of mechanical obstruction, and names nausea and vomiting as its cardinal symptoms. Gastroenterologists separate it by cause: diabetic gastroparesis, postsurgical gastroparesis after an operation on the stomach or nearby organs, and idiopathic gastroparesis, where no cause is found. The acupuncture literature follows the same split unevenly: almost all of the long-term trials are in diabetes, the postsurgical trials study people still in hospital after an operation, and idiopathic gastroparesis has one small trial.
The Cochrane review
The 2018 Cochrane review by Kim and colleagues searched to March 2018, without language limits, and included 32 randomised trials with 2,601 participants. 31 were in diabetic gastroparesis and one in postsurgical gastroparesis. Acupuncture was manual in 24 trials and electrical in 8, and most trials measured outcomes only in the short term, usually over about four weeks. Most trials counted the share of people whose symptoms had “improved”, a category each trial defined in its own way; one used a validated symptom scale.
- Against sham. In the one sham-controlled trial, 60 inpatients in China, symptom scores on the Gastroparesis Cardinal Symptom Index (GCSI, 0 to 45) did not differ between real and sham needling at 12 weeks (mean difference 0.17 points, 95% CI −0.62 to 0.96). The authors rated this low certainty.
- Against prokinetic drugs. In 12 trials with 963 people, more people improved with acupuncture than with domperidone, mosapride or cisapride, drugs that speed stomach emptying (risk ratio 1.25, 95% CI 1.17 to 1.33). Very low certainty.
- Added to other treatment. In 17 trials with 1,404 people, acupuncture added to drugs or routine care improved more people than the same treatment alone (risk ratio 1.22, 95% CI 1.16 to 1.28). Very low certainty.
- Harms. Seven trials reported adverse events, among them minor bleeding and bruising, dizziness, dry mouth, loose stools and fatigue; the other 25 did not say whether any occurred.
The authors found a funnel plot that pointed to publication bias, said the benefits should be read with great caution, and drew no conclusion for gastroparesis other than diabetic. No trial measured quality of life. PubMed lists no later version of the review as of October 2026.
Later reviews
The reviews since 2013 split by cause. The postsurgical reviews mostly pool trials in which acupuncture was added to the hospital’s standard care, typically a nasogastric tube to drain the stomach, nothing by mouth and prokinetic drugs, so they cannot separate needling from the attention and expectation that come with it. Most report a “total effective rate”, the share of patients the trial’s clinicians judged improved or cured.
| Review | Trials and people | Comparison | Finding | Certainty |
|---|---|---|---|---|
| Yang and colleagues, 2013, diabetic | 14 trials, 948 people | Prokinetic drugs (13 trials) and sham electroacupuncture (1) | More people responded with acupuncture (risk ratio 1.20, 95% CI 1.12 to 1.29); no difference in the emptying of solid food | Low to very low |
| Cheong and colleagues, 2014, postsurgical | 16 trials reviewed, 7 pooled (370 people) | Usual care or drugs | Higher total effective rate with acupuncture alone (risk ratio 1.27) and with acupuncture plus drugs (1.37) | Low-quality trials, in the authors’ words |
| Fan and colleagues, 2023, after thoracic or abdominal surgery | 15 trials, 759 people | Acupuncture added to prokinetic drugs, against the drugs alone | Higher total effective rate (12 trials, 578 people, risk ratio 1.18, 95% CI 1.09 to 1.27); the gain held after abdominal surgery (9 trials) but not after thoracic surgery (1 trial) or mixed thoracic and abdominal surgery (2 trials) | Moderate for response and recovery rates; very low for symptom scores |
| Xv and colleagues, 2025, postsurgical | 12 trials, 709 people | Acupuncture added to conventional treatment, against conventional treatment alone; no sham | Response 16 percentage points higher (95% CI 11 to 21); shorter time with a gastric tube; less fluid drained | Not graded; no trial blinded its patients |
Fan and colleagues reported a single acupuncture-related adverse event, mild bleeding under the skin and pain that settled on its own. In the Xv review, four trials reported no adverse events in either group and the rest gave no safety data, which its authors list as the main reason to doubt how far the findings transfer.
A 2026 systematic review by Tan and colleagues in Australia gathered 68 studies of acupuncture, electroacupuncture, moxibustion and herbal medicine, 4,566 patients in all, and compared them in a narrative synthesis without pooling. It describes the alternative therapies as effective and names heterogeneous designs, incomplete reporting and publication bias as its limits.
Trials by cause
Diabetic gastroparesis
- Electroacupuncture against sham, 2008. A pilot trial in Taiwan by Wang and colleagues gave 19 people with type 2 diabetes four sessions over 2 weeks of electroacupuncture at
ST36(Zusanli) andLI04(Hegu), or sham electroacupuncture. In the 9 who had real treatment, the time for half a solid meal to leave the stomach fell from 144 to 99 minutes and GCSI scores fell from baseline; neither changed with sham. - Manual acupuncture against sham, crossover, 2015. In a crossover trial by Li and colleagues, 25 people had a week of real and a week of sham acupuncture, a month apart, and 21 finished. Real acupuncture reduced the food retained in the stomach at 2 and 4 hours, and the GCSI score, more than sham did. Blood glucose and HbA1c did not differ.
- Point combinations, 2020. A trial at three centres led by Wu Xuefen at Hunan University of Traditional Chinese Medicine randomised 99 people, excluding postsurgical gastroparesis, to
CV12(Zhongwan) with ST36,PC06(Neiguan) with ST36, or a non-acupoint with ST36, five days a week for three weeks. GCSI scores fell in all three groups and most with CV12 and ST36; barium left the stomach faster in all three, with no difference between them. - Needleless electrical stimulation against sham, 2025. A single-centre crossover trial in Texas by Zhu and colleagues taught 26 people to stimulate ST36 and PC06 through skin electrodes twice a day for 4 weeks, alternating with 4 weeks of stimulation at sham points; 18 finished. Nausea, retching, vomiting, fullness and bloating scores fell from baseline with real stimulation and did not change with sham. One author is employed by the company that made the device.
Postsurgical gastroparesis
A trial in Shanghai by Sun and colleagues randomised 63 people with gastroparesis after abdominal surgery for primary liver cancer to daily acupuncture at ST36 and other points, or 20 mg of metoclopramide injected into muscle three times a day. The trial reported 90.6% cured with acupuncture and 32.3% with metoclopramide, after fewer treatments. The trial had no sham arm; of the 15 trials in the Fan review, one described blinding its patients.
Idiopathic gastroparesis
A trial in Texas by Song and colleagues gave 18 people with idiopathic gastroparesis one hour of needleless electrical stimulation at acupoints, timed to their breathing, and on another visit the same at non-acupoints, after a liquid meal that brought on nausea. Nausea eased equally with both. Real stimulation restored more normal electrical rhythm in the stomach, about 24% against 6% with sham. This was a single session, not a course of treatment. Electroacupuncture explains how the needled and needleless forms differ.
What guidelines say
The American College of Gastroenterology’s 2022 guideline by Camilleri and colleagues states in its recommendation 16 that acupuncture alone or combined with prokinetic drugs may be beneficial for symptom control in diabetic gastroparesis, and that it cannot be recommended for other causes (conditional recommendation, very low quality of evidence). Its evidence table cites the Cochrane review, the Yang review and several of the trials above, and the text beneath it says the available evidence does not support electroacupuncture, acupuncture or herbal medicines in clinical practice. The ACG still lists the 2022 document as its gastroparesis guideline in October 2026.
The European consensus put the statement “Acupuncture is effective for gastroparesis” to a Delphi vote of 40 experts from 19 countries, in which a statement needed at least 80% agreement to be endorsed. It was not endorsed: 3% agreed, and the evidence was graded B. The consensus text cites the Cochrane review’s higher improvement rates against usual or drug treatment and calls that conclusion uncertain, and the panel lists acupuncture among the treatments on which it reached no consensus either way. Functional dyspepsia, whose symptoms overlap with those of gastroparesis, reports the sham-controlled trials in that condition.
What the reviews could not settle
The Cochrane review rests its sham comparison on one trial of 60 people, and its drug comparisons on short trials in China that counted unvalidated “improved” categories, with signs of publication bias. The postsurgical reviews mostly pool trials in which acupuncture was added to care and the patients knew it, so the share of the effect due to the needling itself is unknown; Sham acupuncture sets out what a sham control can and cannot rule out. Idiopathic gastroparesis has one single-session trial, and 28 of the 32 Cochrane trials measured only short-term outcomes. Harms went unreported in most trials. A trial registered in 2020 by Chen and colleagues plans to compare 4 weeks of needleless electrical stimulation at ST36 and PC06 with sham in 90 people with delayed gastric emptying after distal gastrectomy, with a year of follow-up; its results do not appear in PubMed as of October 2026. Postoperative ileus covers the related trials on the bowel after surgery.
AcuiQ’s gastroparesis, diabetic gastroparesis and postoperative gastroparesis pages list the protocols individual studies prescribed, each with its citation.