Four reviews have counted the adverse events children had after acupuncture, meaning any unwanted effect. The most recent of them found that nearly all the studies it read set out to test whether treatment worked and gave adverse effects little attention, so the reviews could list events but not measure how likely each is. The first systematic review, by Adams and colleagues in the University of Alberta’s department of paediatrics, was published in Pediatrics in 2011. It searched 18 databases in any language up to September 2010 for original research on needle acupuncture in anyone from birth to 17 years, screened 9,537 references and kept 37 reports. This page reports what each counted and where the counting stops.
What the reviews counted
| Review | What it read | What it found |
|---|---|---|
| Jindal et al. 2008, US National Institutes of Health | 31 articles, including 23 randomised trials and 8 reviews | 1.55 adverse events per 100 treatments |
| Adams et al. 2011, Pediatrics | 37 reports: trials, cohort studies and case reports | 279 events: 253 mild, 1 moderate, 25 serious |
| Yang et al. 2015, Pediatric Research | 24 systematic reviews covering 142 randomised trials and 12,787 children | 6 of the 24 reviews reported adverse events; none reported a death |
| Bailey et al. 2026, scoping review | 14 studies published from July 2015 to June 2025 in which 1,412 children and young people up to 21 received needle acupuncture | No serious events; 8 studies reported minor events only and 6 reported none |
The serious events
Of the 279 events Adams and colleagues found, 146 came from randomised trials, 95 from cohort studies and 38 from case reports and case series. The 25 serious events were 12 cases of thumb deformity, 5 infections, and one each of cardiac rupture, pneumothorax (a collapsed lung), nerve impairment, subarachnoid haemorrhage (bleeding around the brain), intestinal obstruction, coughing up blood, reversible coma and an overnight stay in hospital. The authors write that many of the serious events might have been caused by substandard practice, and that their results support adult studies finding acupuncture safe when appropriately trained practitioners give it.
The 253 mild events were pain, bruising, bleeding and symptoms getting worse. Where the studies gave a denominator, 168 of 1,422 children, 11.8%, had a mild event. The 2026 scoping review by Aiko Bailey and colleagues at the University of Washington found the same kinds of event a decade on: minor bruising, local pain or numbness, tingling where a needle went in, and minor bleeding. A trial of 98 infants with colic recorded a single drop of blood.
What differs from adults
Anatomy. WHO’s 2020 benchmarks for the practice of acupuncture list needling an infant’s fontanelle, the soft gap between the skull bones, before it has closed as a contraindication. The 2008 NIH review adds advice from the sources it read to use few needles, or delay treatment, for children who have overeaten, are overtired or are very weak.
Fear of needles. A study from Children’s Hospital Los Angeles starts from the observation that children and their carers often turn acupuncture down because they expect needling to hurt. It surveyed 230 patients aged 8 to 21 after 1,380 sessions in an outpatient pain clinic. On a 0 to 10 faces scale the mean needling pain was 1.3, and 57.7% reported no needling pain at their first session. An earlier telephone survey of 47 families referred by the pain service at Boston Children’s Hospital found that 67% of the children rated treatment as pleasant; 85% had also received heat or moxa and 26% cupping. Both surveyed children who had been treated, so neither measures how many turned treatment down.
Methods without a needle. Laser acupuncture shines low-level laser light on a point instead of inserting a needle; Laser acupuncture describes it. A 2020 Mayo Clinic review of 21 randomised trials, which describes laser acupuncture as attractive for people with a fear of needles, older people and children, found adverse events in 6 trials: tingling, pain flare-ups and passing fatigue, all mild and gone within 24 hours. A 2026 review of laser acupuncture for autistic children, again from the University of Alberta, found 2 trials with 76 children and no adverse effects reported by parents in either; its authors argue that needles can be a source of healthcare trauma and that non-needle options should come first. Acupressure uses no needle, and press needles, the Bailey review notes, barely penetrate the skin and are less likely to hurt.
Moxibustion. In a 2014 review of moxibustion case reports, most of the people burned were newborns, children or people with reduced sensation, and 37 of the cases were newborns burned by a folk method used for jaundice. Burns from moxibustion and cupping reports those cases.
Children having cancer treatment. Two hospital series looked at children whose treatment had lowered their platelets, the blood cells that make blood clot. A Columbia University Medical Center review of 32 patients and 237 sessions found no bleeding, including at sessions given during severe thrombocytopenia (a low platelet count). The paediatric cancer unit at Hospital Sant Joan de Déu in Barcelona reviewed 196 patients and 1,107 sessions using semi-permanent needles, alone or with ordinary filiform needles, and recorded no infections and one superficial bruise on the ear of a patient with a very low platelet count. Bleeding and anticoagulants covers bleeding risk in adults.
Paediatric tuina
Tuina is Chinese therapeutic massage, and paediatric tuina applies it to infants and young children. A 2024 umbrella review from Hong Kong Polytechnic University read 22 systematic reviews of paediatric massage in children under 5. Only one of the 22 reported adverse events: a review of massage for newborn jaundice listed anaemia, diarrhoea, fever and skin rash, without showing that massage caused them. The authors ask future trials to report adverse events in full.
What the reviews could not count
None of the 14 studies in the 2026 review was designed to assess safety, and its authors write that nearly all paediatric studies focused on whether treatment worked. The adult studies on After a treatment followed large groups of patients for that purpose; none of the four reviews reports a study of that kind in children. The Yang overview could draw safety data from only 6 of its 24 reviews. Bailey and colleagues searched one database in English only and left out press needles, and note that research has looked at physical harm and not at psychological harm. Their review also ran to age 21, and its authors suggest future studies look at children before puberty separately. They ask future studies to report the practitioner’s qualifications, the needle size and the technique alongside any harm, since Adams and colleagues tied many serious events to substandard practice; Choosing a practitioner sets out how practitioners are trained and registered.
AcuiQ’s infant colic, bedwetting, cerebral palsy and tic disorder pages list the protocols individual studies prescribed for children, each with its citation. They record what was prescribed, not what happened to the children afterwards.