A palpitation is a heartbeat a person notices: racing, pounding, fluttering, or skipping and adding beats. The NHS lists stress and anxiety, exercise, lack of sleep, caffeine, alcohol and some medicines among the common causes, and calls the skipped or extra beats ectopic beats. A premature ventricular complex (PVC) is an early beat from the heart’s lower chambers, the ventricles, and a premature atrial beat is one from the upper chambers. Cai and colleagues, at Guangdong Provincial Hospital of Chinese Medicine and RMIT University in Melbourne, note that PVCs can pass unnoticed or cause palpitations, dizziness, chest pain or fatigue. Their 2022 review is the only one below limited to premature beats, and it took in adults with PVCs and no ischaemic or structural heart disease, the group the 2022 ESC guideline calls idiopathic PVCs. Trials of acupuncture for PVCs count premature beats on a 24-hour Holter, a portable ECG worn for a day, and Cai and colleagues counted a result as effective when that count fell by half or more. A smaller literature covers supraventricular tachycardia (SVT), bursts of fast rhythm that start above the ventricles. Atrial fibrillation has its own page, Atrial fibrillation.
When should palpitations be checked by a doctor?
The NHS advises seeing a GP when palpitations keep coming back, happen more often, last longer than a few minutes, or come with a known heart condition or a family history of heart problems, and calling 999 when they do not stop or come with chest pain, shortness of breath or fainting.
Is there a Cochrane review?
Cochrane has no review of acupuncture for palpitations, premature beats or SVT. An overview by Ci and colleagues in the Journal of Traditional Chinese Medicine, published in December 2025, searched the Cochrane Library among eight databases to December 2023 and found 10 reviews of acupuncture for heart rhythm disorders, none of them from Cochrane. The overview rated the methodological quality of all 10 very low. Of the 34 outcomes those reviews reported, it graded none high, 4 moderate, 13 low and 17 very low on GRADE, the scale reviews use to say how far a result can be trusted.
Systematic reviews
The reviews share trials, so they are not independent confirmations of one another.
| Review | Searched to | Trials (people) | Comparison | Pooled result | Certainty (GRADE) |
|---|---|---|---|---|---|
| Cai 2022, PVCs | November 2022 | 9 (847) | Body acupuncture against sham; acupuncture or ear acupressure added to usual care against usual care | Higher effective rate against sham (3 trials, 233 people, risk ratio 2.68) and as an add-on (4 trials, 264 people, risk ratio 1.21); no significant difference for ear acupressure | Moderate for both body acupuncture comparisons; low for ear acupressure |
| Li 2017, mixed arrhythmias | 2016 | 13 (797) | Acupuncture against drugs, sham or bed rest | Higher response rate when added to antiarrhythmic drugs in PVCs (286 people); no significant difference from drugs in paroxysmal SVT (203 people) | Not graded |
| Liu 2018, mixed arrhythmias | February 2016 | 9 (638) | Acupuncture alone or added to drugs, mostly against drugs | Higher response rate when added to antiarrhythmic drugs in PVCs | Not graded |
| Ning 2023, mixed arrhythmias | June 2022 | 11 (804) | Acupuncture added to Chinese herbal medicine, against the herbal medicine alone | Higher effective rate; fewer premature beats in 24 hours | Not graded |
| Wen 2014, SVT | Not stated in the abstract | 5 (323) | Acupuncture against no treatment, diltiazem or propafenone | Lower heart rate than no treatment; higher effective rate than diltiazem; no significant difference from propafenone | Described as low; GRADE not stated |
Cai and colleagues set aside trials whose combined treatments left the share owed to acupuncture unclear, such as acupuncture plus herbs plus a drug against the drug alone. Three of their nine trials compared body acupuncture with sham. The sham result varied widely between trials; when Cai and colleagues dropped the one trial with unclear allocation concealment, the two trials rated at low risk of bias in every domain gave a risk ratio of 3.83, meaning people given acupuncture were nearly four times as likely to halve their premature beats. They graded that comparison moderate rather than high because the trials were small. The trials most often needled PC06 (Neiguan), HT07 (Shenmen), GV10 (Lingtai), GV11 (Shendao) and ST36 (Zusanli), for 15 to 30 minutes over about four weeks. Only one trial measured quality of life, and the symptom scores differed from trial to trial, so the review could pool neither.
Li and colleagues and Liu and colleagues covered every kind of arrhythmia and pooled response rates that their trials defined differently, which Cai and colleagues gave as a reason for their narrower question. The reviews draw on a small pool of trials: Li 2017 and Wen and colleagues report an identical heart-rate difference against no treatment, 18.8 beats a minute with the same confidence interval, Li under sinus tachycardia and Wen under SVT. The Ci overview picked out one PVC result as moderate certainty, Li 2017’s add-on comparison with a risk ratio of 1.15. Ning and colleagues compared acupuncture plus herbal medicine with herbal medicine alone, which says nothing about acupuncture against conventional drugs.
Trials
A sham-controlled trial by Huang and colleagues in Guangzhou, published in Chinese in 2025, randomised 64 people with frequent PVCs classed as functional, meaning without an underlying heart disease, to wrist-ankle acupuncture, a style that needles set zones near the wrists and ankles, at the zones called upper 1 and upper 2 on both wrists, or to sham at the same places. Each group had 60-minute sessions on weekdays for 4 weeks. Both groups had fewer PVCs on Holter afterwards, and the acupuncture group had fewer than the sham group; 29 of 32 met the trial’s effective-rate threshold against 15 of 32 with sham. The abstract does not describe the sham technique or how long anyone was followed after treatment.
The add-on trials in the Cai review gave acupuncture with mexiletine, amiodarone or guideline-recommended care and compared it with the same drugs alone, with no sham. Cai and colleagues rated the blinding of participants and staff unclear in every one of them; they judged the Holter count an objective outcome that missing blinding would affect little, and symptom scores more open to that bias.
Safety
Five trials in the Cai review reported 11 adverse events with acupuncture and 8 in control groups, mostly pain, bruising or a rash at the needle site, all mild and resolving without treatment. A safety review by Liang and colleagues, published in June 2026 and searched to November 2025, pooled 5 randomised trials of acupuncture for arrhythmias with 356 participants; it found no significant difference in adverse events between acupuncture and control groups and no serious event attributed to acupuncture, and rated the certainty of that finding low.
What guidelines say
The 2022 ESC guideline on ventricular arrhythmias and the prevention of sudden cardiac death, by Zeppenfeld and colleagues for the European Society of Cardiology, was published on 26 August 2022 and updates the 2015 edition. For idiopathic PVCs that cause symptoms or weaken the heart, it recommends beta-blockers, calcium channel blockers or flecainide, and catheter ablation as first-line treatment for PVCs from the right ventricular outflow tract or the left fascicles. It does not mention acupuncture. The 2017 AHA/ACC/HRS guideline on ventricular arrhythmias, by Al-Khatib and colleagues, published in Circulation in 2018, does not mention acupuncture, and nor does the 2019 ESC guideline on SVT by Brugada and colleagues. The ESC still lists the 2022 and 2019 documents as its guidelines on these subjects.
The Chinese document on this question is a Delphi consensus published in 2021 by Li and colleagues at Beijing University of Chinese Medicine, a survey of experts rather than a cardiology society guideline. Its 30 panellists, two-thirds of them with at least 20 years of acupuncture practice, voted through three rounds on 35 statements and agreed on 32. The consensus recommends acupuncture as an adjunct to regulate heart rate and rhythm, and names PC06, BL15 (Xinshu), CV17 (Danzhong) and HT07. The panel agreed in the first round that acupuncture improves symptoms and quality of life in every arrhythmia except atrial fibrillation, but reached agreement on rate and rhythm only in the third. The four reviews it rested on were all rated very low on GRADE. Clinical guidelines compares other bodies’ positions.
What the reviews could not settle
The sham-controlled evidence for PVCs rests on the three sham-controlled trials in the Cai review and the Huang trial. Cai and colleagues measured the effect at the end of treatment, which in most trials lasted about four weeks, so how long a reduction lasts is unknown. Cai and colleagues found no core outcome set for PVC trials and no validated symptom measure, so the trials cannot show on a shared scale whether fewer beats on a Holter means fewer palpitations. Sham acupuncture describes how the kinds of sham differ. None of the reviews above reports a result for premature atrial beats on their own. For SVT, Wen 2014 pooled 5 trials against no treatment or drugs, and none of those trials reported adverse events. Anxiety covers the trials for one of the causes the NHS lists.
AcuiQ’s palpitations, tachycardia and arrhythmia pages list the protocols individual studies prescribed, each with its citation.