Fainting during acupuncture is usually described as a vasovagal reaction, and practitioners call it needle syncope or needle fainting. Every study on this page that recorded where the patient was found them sitting or standing, and every protocol it cites has the patient lie down. A 2017 paper from the University of Minnesota’s children’s hospital proposed the term acupuncture-associated vasovagal response, arguing that “syncope” and “shock” name different physiological states and that phrases pairing “needle” with sickness can themselves make patients more anxious.
MedlinePlus, the US National Library of Medicine’s health encyclopedia, defines fainting as a brief loss of consciousness from a drop in blood flow to the brain, and lists fear, severe pain, standing in one place for too long and low blood sugar among its triggers. How often any reaction of any kind follows treatment is on After a treatment; this page covers fainting alone.
How often studies recorded it
The figures below come from studies that counted every treatment given over a set period, so each has a denominator. They do not all count the same thing: some recorded only full faints, some included feeling faint, and one reported only a broader category of autonomic reactions.
| Study | Setting | What was counted | Result |
|---|---|---|---|
| Chen et al. 1990 | Veterans General Hospital, Taipei, August 1988 to April 1989 | Needle fainting | 55 episodes in 52 patients over 28,285 treatments, about 2 in 1,000 |
| White et al. 2001 | 78 UK doctors and physiotherapists | Fainting among significant adverse events | 6 in 31,822 consultations |
| MacPherson et al. 2001 | 574 members of the British Acupuncture Council, four weeks | Fainting among significant adverse events | 4 in 34,407 treatments |
| Witt et al. 2009 | 229,230 German patients treated by physicians | Vegetative symptoms, the German term for reactions of the autonomic nervous system; fainting is not reported separately | 0.7% of patients over about 10 treatments each |
| Kim et al. 2016 | A Korean medicine hospital, 80,523 patients, retrospective record review | Presyncope, feeling about to faint | 4 episodes |
| Christensen et al. 2017 | Community-style sessions for families and staff at a US children’s hospital, 18 months | Vasovagal symptoms | 5 in 281 treatments, 1.8% |
A 2001 review by Edzard Ernst and Adrian White of nine prospective surveys put feelings of faintness and syncope at 0% to 0.3%. Christensen and colleagues cite published rates from 0.02% to 7% of treatments.
Who faints
The Taipei study is the only one in the table to compare groups. Patients fainted at 27 of 2,855 first visits, 0.94%, against 28 of 25,430 follow-up treatments, 0.11%. Those who fainted at a first visit were younger, with a mean age of 39 against 52, and fewer of them had another medical condition. All 55 episodes happened with the patient upright. The usual signs were pallor, cold sweat, nausea and a slow pulse; nobody lost consciousness completely, and all recovered soon after lying down.
The five patients in the Minnesota series were all sitting, three had never had acupuncture before, and one had been fasting. The authors list the risk factors described in the literature as nervousness or fear, poor rest, fatigue, dehydration, hunger, a first treatment, strong stimulation, electrical stimulation and a large number of needles. Onset was during or after insertion in four cases and during needle retention in one, and they cite reports of onset up to 12 hours after treatment.
The World Health Organization’s 1999 guidelines on basic training and safety in acupuncture give a similar list: nervousness, hunger, fatigue, extreme weakness, an unsuitable position, or too forceful manipulation. The guidelines also single out points that may lower blood pressure, naming LV03 (Taichong).
What practitioners are told to do
The 1999 WHO guidelines tell practitioners to explain the procedure and its sensations before starting, to treat first-time patients lying down with gentle manipulation, and to watch the complexion and check the pulse during treatment. The warning signs they list are feeling unwell, giddiness, a sense that the room is moving, weakness, chest oppression, palpitations and nausea. If those appear, the guidelines say to remove the needles at once, lay the patient flat with the head down and the legs raised, and offer a warm sweet drink; they attribute the symptoms to a brief shortfall in the blood supply to the brain. For severe cases they give first aid first, then name points to press or needle once the patient is stable, among them GV26 (Shuigou, also called Renzhong), PC09 (Zhongchong), GV25 (Suliao), PC06 (Neiguan) and ST36 (Zusanli), or moxibustion at GV20 (Baihui), CV06 (Qihai) and CV04 (Guanyuan). Symptoms that persist call for emergency medical help.
WHO’s 2020 benchmarks for the practice of acupuncture reduce the incident protocol for fainting to four steps: stop needling, withdraw every needle, have the patient lie down and rest, and assess and refer as needed. The Minnesota hospital protocol adds raising the feet, fresh air, loosening tight clothing, care to prevent a fall, and calling emergency services if the patient has not regained consciousness after 5 minutes. Most people recover within 5 minutes, its authors write.
The MedlinePlus first-aid advice for anyone who faints is to raise the feet about 30 cm above the heart and keep the person lying down for 10 to 15 minutes. It advises calling emergency services if the person was injured in a fall, does not become alert within a couple of minutes, is pregnant, is over 50, has diabetes, or has chest pain or an irregular heartbeat.
A faint can cause a second injury. A 2008 review of acupuncture safety in Japan by Hitoshi Yamashita and Hiroshi Tsukayama describes most common reactions as mild and short-lived, and adds a caution about injuries that follow drowsiness and needle fainting.
What the sources cannot tell you
None of the studies on this page tested a position, a warning or a pre-treatment snack against an alternative, so the advice to treat first-time patients lying down rests on observation and guideline consensus rather than on a trial. The two UK surveys relied on practitioners to report what they saw over four weeks. The Taipei study is the only one in the table that gives a first-visit rate, and it comes from one hospital clinic in the late 1980s. The rates in the table cannot be pooled, because each study drew the line between feeling faint and fainting in a different place. AcuiQ records what studies prescribed rather than what happened to the people in them, so none of its protocols carries a fainting rate; What AcuiQ does not do sets out that limit, and Choosing a practitioner covers how the people giving treatment are trained.