Itching

Chronic pruritus is itch that lasts six weeks or longer, in the definition the 2025 European guideline on chronic pruritus uses. It can come from skin disease or from disease elsewhere in the body, including chronic kidney disease and liver and bile-duct disease. The guideline cites data from the Dialysis Outcomes and Practice Patterns Study putting moderate to severe itch at up to 30% of people on dialysis; this itch is called chronic kidney disease-associated pruritus or uraemic pruritus. Most of the acupuncture research on itch outside the skin clinic is in that group, and this page reports it. Atopic dermatitis and urticaria covers the trials in eczema and hives, including the Munich laboratory studies of allergen-provoked itch.

The Cochrane review

The most recent Cochrane review that covers itch is a 2016 review of acupuncture and related interventions for symptoms of chronic kidney disease by Kim and colleagues at Pusan National University in Korea. It searched to January 2016 and included 24 trials with 1,787 participants, adults with kidney disease at stage 3 to 5, and covered needling, acupressure, ear acupressure, transcutaneous electrical acupoint stimulation (electrodes on the skin over the points), far-infrared heat and moxibustion. Six of the 24 measured uraemic pruritus. The authors judged every trial at high or unclear risk of bias for allocation concealment, which means a reader cannot tell whether the people enrolling patients knew the next assignment.

The pruritus results came one trial at a time, because the trials used different scales and comparisons. Against sham needling, one trial of 40 people found itch 20.2 points lower on a 0 to 45 scale after four weeks. Against an oral antihistamine, one trial found more people improved with acupuncture (risk ratio 1.38), and against oral calcitriol another found no difference (risk ratio 1.00). Acupressure against an unspecified control lowered itch scores at six weeks, by less at 18 weeks. The authors rated the evidence for manual acupressure as an add-on in people on haemodialysis very low quality, and wrote that incomplete reporting of harms did not let them assess safety.

Later reviews of uraemic pruritus

Lu and colleagues, at Taipei Tzu Chi Hospital and MacKay Memorial Hospital in Taiwan, pooled 40 trials published from 2002 to 2022 of needling, ear acupressure, injection into points, heat on points and electrical stimulation. They report higher response rates with acupuncture, ear acupressure and point injection combined with massage than with controls, and list small trials and no long-term follow-up among their limitations. The points used most often were SP10 (Xuehai), in 9 trials, then LI11 (Quchi), SP06 (Sanyinjiao) and ST36 (Zusanli), in 7 each.

The same group's 2024 network analysis, which compares several methods at once through shared comparators, found itch on a 10-point visual analogue scale 2.63 points lower when acupuncture was added to conventional treatment than with conventional treatment alone. Only two of its trials were double-blinded, and most came from a single country. Zhang and colleagues in Chengdu found acupuncture's response rate similar to that of oral and topical drugs (risk ratio 1.20, not statistically significant) and higher than dialysis alone when the two were combined (risk ratio 1.42).

He and colleagues at the West China School of Nursing searched to December 2025 and compared 12 complementary therapies. Seven of their trials used a sham comparator and 21 added a therapy to usual care. With the 9 trials at high risk of bias set aside, they found itch lower with acupuncture than with controls (standardised mean difference −2.12, a result in units of the spread between patients); with those trials put back, the effect shrank and was no longer statistically significant. On CINeMA, a tool for grading confidence in network analyses, they rated the certainty of every therapy but aromatherapy low or very low.

Itch beyond kidney disease

A 2015 review by Yu and colleagues in Wuhan looked for trials of acupuncture against placebo acupuncture or no treatment for any itch, and found 3 with 70 people between them: one in uraemic pruritus from Taiwan and two in atopic eczema from Germany, two of the three from the same laboratory. Pooled, itch intensity was 19 points lower with acupuncture, with high heterogeneity (I² 80%, the share of variation between trials that chance does not explain) and signs of publication bias. A 1984 crossover study in 25 healthy volunteers injected histamine into the skin and found shorter itching and a smaller red flare with acupuncture than with needling at nearby non-acupuncture sites or no needling, with little effect on peak itch intensity.

Itch is a common side effect of morphine given into the spinal fluid. A 2018 double-blind trial in Saitama, Japan, gave 30 women having a planned caesarean press needles or sham needles at LI04 (Hegu), PC06 (Neiguan), LI11 and TB06 (Zhigou) on both arms the day before surgery. Itch occurred in 67% of each group.

The kidney trials

The trial behind the Cochrane review's sham comparison is a 2005 study at China Medical University Hospital in Taichung, Taiwan. It randomised 40 people on haemodialysis whose itch had not responded to standard measures to needling at LI11 on one side or at a non-acupoint 2 cm away, three times a week for a month. Itch scores fell from 38.3 to 17.3 with LI11 and stayed at 16.5 three months later, against 38.3, 37.5 and 37.1 in the control group; blood tests did not differ.

A 2026 double-blind trial from the same university gave 60 people on haemodialysis low-level laser at LI11, SP10 and XL03 (Baichongwo, which the authors write EX-LE3) or a sham laser, three times a week for four weeks. The 5-D itch scale, a questionnaire on itch, was 6.9 points lower with the real laser, and no adverse events occurred. Laser acupuncture describes the method.

For acupressure, a 2015 trial at Putuo Hospital in Shanghai taped Vaccaria seeds to six ear points three times a week for six weeks in 32 people, against tape without seeds on the same points in 30, and found visual analogue itch scores of 3.8 against 5.6 at follow-up. A 2016 trial in Turkish dialysis units randomised 75 people to acupressure or electrical stimulation at LI11, or a control group, for 12 sessions, and found less itch in both treated groups than in controls and no difference between them. A 2026 trial in Kerman, Iran, allocated 90 people by dialysis-day clusters to four weeks of ear acupressure or routine care; itch fell from 6.1 to 3.2 on a 10-point scale with acupressure and did not change with routine care. Neither of these two trials used a sham.

What guidelines say

The 2024 guideline of Kidney Disease: Improving Global Outcomes (KDIGO) on the evaluation and management of chronic kidney disease, published in Kidney International in April 2024, lists acupuncture, citing the Lu 2022 review, in a table of management strategies for uraemic pruritus, beside gabapentinoids, ultraviolet B therapy and topical agents. The table sits under a practice point on symptom management rather than a graded recommendation.

The European S2k guideline on chronic pruritus by Weisshaar and colleagues, published in Acta Dermato-Venereologica in August 2025, updates the 2019 edition. Its table of options for kidney-associated itch includes acupuncture at LI11 among treatments with antipruritic effects in controlled studies, citing the 2005 trial, and it recommends difelikefalin, a drug acting on kappa-opioid receptors. Its section on physical treatments describes acupuncture as having much less evidence for itch than for pain and concludes that the authors cannot make a recommendation on physical treatment. Its table for hepatic and cholestatic pruritus, the itch of liver and bile-duct disease, does not list acupuncture.

What the reviews could not settle

The one sham-controlled needling trial that the Cochrane review and the European guideline cite enrolled 40 people at a single hospital, and the reviews since have pooled mostly trials that add acupuncture to dialysis or drugs without a sham, scored on response rates whose definitions vary. The Cochrane review rated its evidence very low quality, He 2026 rated acupuncture low or very low, and He 2026’s acupuncture effect did not survive putting the high-risk trials back. No review compared acupuncture with difelikefalin, the drug the European guideline recommends. The PubMed searches for this page found no randomised trial in people with cholestatic itch, and the one trial in opioid-induced itch found no difference.

AcuiQ lists the protocols individual studies prescribed for itching, uraemic pruritus, pruritus on dialysis and chronic itch, each with its citation. Ear acupuncture covers the ear point systems the acupressure trials used.