Post-traumatic stress disorder

Post-traumatic stress disorder (PTSD) is a mental health condition that can follow a traumatic event such as combat, assault, abuse or a disaster. The acupuncture trials for it compare acupuncture with five different things: a waitlist, on which people receive no treatment until the study ends; usual care; paroxetine, an antidepressant; group cognitive behavioural therapy (CBT); and sham acupuncture, a control procedure made to resemble it. Most are small, and most of the US trials enrolled veterans or serving soldiers. One trial, published in JAMA Psychiatry in 2024, compared acupuncture with a sham in 93 combat veterans. Two guidelines, from the US Department of Veterans Affairs and Department of Defense (VA/DoD) and from the UK’s National Institute for Health and Care Excellence (NICE), have weighed the trials and recommended neither for nor against acupuncture.

No Cochrane review of acupuncture for PTSD

A Europe PMC search of the Cochrane Database of Systematic Reviews on 11 October 2026 found Cochrane reviews of psychological therapies, internet-based CBT and repetitive transcranial magnetic stimulation for PTSD, and none of acupuncture. The evidence on this page comes from systematic reviews in other journals, the trials they pooled, and the evidence reviews behind the two guidelines.

What the trials measure

Most trials report a PTSD severity score. The Clinician-Administered PTSD Scale (CAPS, in its current version CAPS-5) is a structured interview scored by a trained assessor; the PTSD Checklist (PCL) is a questionnaire the person fills in. Many trials also score depression, anxiety and sleep. Effects are often given as Cohen’s d, the difference between two means divided by the spread of the scores; the 2021 design paper for the 2024 trial below treats a d of 0.8 or more as a large effect.

Reviews

Kim and colleagues searched 23 databases and screened 136 articles. Grant and colleagues used GRADE, the system that rates confidence in a pooled result from high to very low. Their standardised mean difference (SMD) expresses the gap between groups in units of the spread of scores, and they pooled every comparator in one estimate, so a trial against a waitlist and a trial against a sham count alike. At the end of treatment the review found no significant difference on depression, anxiety or sleep quality. Across the 7 trials the review found little risk of serious adverse events; some participants had pain, superficial bleeding or a haematoma, a bruise from bleeding under the skin, where a needle went in.

Tang and colleagues, at Guangzhou University of Chinese Medicine, pooled trials in which the acupuncture group often received a drug as well, and in two of them psychotherapy too. Heterogeneity means the trials disagreed with each other more than chance explains. Five of the eight carried a high risk of bias for blinding, and the authors note that most were run in China. Kwak and colleagues found the ear points Shenmen and Kidney used most often in the military studies.

Trials

• Against waitlist and group CBT, 2007. A pilot trial by Hollifield and colleagues at the University of Louisville randomised 84 people, according to NICE’s evidence review, to 24 one-hour acupuncture sessions twice weekly over 12 weeks, to group CBT, or to a waitlist. Two-thirds were women, and for 62% the first trauma came before the age of 12. On a self-reported PTSD scale, acupuncture had a large effect against the waitlist (d 1.29), similar to group CBT (d 1.42), and both held at 3 months. NICE rated the comparison with the waitlist very low quality and found no significant difference between acupuncture and the CBT group on any outcome, also at very low quality.
• Against paroxetine, 2012. A trial by Wang and colleagues randomised 138 survivors of the Wenchuan earthquake in China to electroacupuncture, in which a small current passes through the needles, at GV20 (Baihui), Sishencong (EX-HN1), GV24 (Shenting) and GB20 (Fengchi), or to paroxetine. The authors reported larger falls in CAPS, depression and anxiety scores with electroacupuncture. NICE reanalysed 127 of the participants and found no significant difference on the CAPS at the end of treatment, 3 months or 6 months, at low quality.
• Added to usual care, 2014. A trial by Engel and colleagues at Walter Reed National Military Medical Center and the Uniformed Services University randomised 55 US service members to usual PTSD care plus eight one-hour acupuncture sessions twice weekly, or to usual care alone. At 12 weeks the PCL had fallen by 19.8 points against 9.7, and the CAPS by 35.0 against 10.9. Depression, pain and functioning also improved more with acupuncture.
• Against sham, for sleep, 2018. A trial by Huang and colleagues at a VA medical center randomised 60 veterans with mild traumatic brain injury and persistent poor sleep, two-thirds of whom had PTSD, to up to 10 sessions of real or sham acupuncture. The Pittsburgh Sleep Quality Index, a sleep questionnaire, improved by 4.4 points against 2.4 with sham (P = .04), and the authors reported the effect in people with and without PTSD.
• Against sham, 2024. A trial by Hollifield and colleagues at the Tibor Rubin VA Medical Center in Long Beach, California, randomised 93 combat veterans aged 18 to 55, 85 of them men, between 2018 and 2022. Both groups had up to 24 one-hour sessions over 15 weeks, and 71 completed them. The design paper sets out the prescription: on alternate visits, needles on the front at LV03 (Taichong), PC06 (Neiguan), HT07 (Shenmen), ST36 (Zusanli) and SP06 (Sanyinjiao) on both sides and at XH03 (Yintang), or on the back at GB20 (Fengchi) and BL14 (Jueyinshu), BL15 (Xinshu), BL18 (Ganshu), BL20 (Pishu), BL21 (Weishu) and BL23 (Shenshu), plus three points chosen for each person’s pattern in traditional Chinese medicine. The sham placed needles 1 to 2 cm from reference points chosen as unrelated to PTSD, less than a quarter of an inch deep, with a stimulator that did not work, as the design paper describes. CAPS-5 scores fell in both groups (d 1.17 with acupuncture, 0.67 with sham), and 7.1 points more with acupuncture (d 0.63, P = .005) in the intention-to-treat analysis, which counts everyone as randomised. Fear-potentiated startle, the eyeblink startle reflex recorded while a learned threat cue is shown without its consequence, a laboratory measure of fear extinction, improved with acupuncture and not with sham. The authors report low withdrawal rates.
• Same trial, secondary outcomes, 2025. A secondary analysis of the 2024 trial by Hsiao and colleagues found that anxiety, depression and sleep scores improved in both groups, with a large effect for acupuncture (d 1.3 for anxiety) and a moderate one for sham (d 0.9), and no statistical difference between them.

Sham acupuncture sets out what a minimal-needling sham can and cannot rule out.

What guidelines say

The VA/DoD Clinical Practice Guideline for the Management of PTSD and Acute Stress Disorder, version 4.0, published in June 2023 and current, reviewed evidence to 4 May 2022. Its recommendation 27 finds insufficient evidence to recommend for or against acupuncture, alongside other mind-body practices including yoga, tai chi or qigong, hypnosis and Transcendental Meditation. The Work Group cited the 2018 Grant review, whose three trials with a clinician-rated outcome suggested benefit against usual care, sham and paroxetine. It named low study quality, the mix of styles such as whole-body and ear acupuncture, and comparison conditions that did not control for the nonspecific parts of acupuncture as concerns, and rated its confidence in the evidence very low. It found little evidence of harm beyond discomfort with needles and mild local reactions such as bleeding. The 2017 edition had found insufficient evidence to recommend acupuncture as a primary treatment for PTSD. The 2023 guideline’s strong recommendations are for three trauma-focused psychotherapies (cognitive processing therapy, eye movement desensitisation and reprocessing (EMDR), and prolonged exposure) and for paroxetine, sertraline or venlafaxine. The 2024 sham-controlled trial was published after its evidence cut-off.

NICE’s guideline NG116, published in December 2018, does not mention acupuncture in its recommendations. For adults it recommends trauma-focused CBT and, after non-combat trauma more than 3 months earlier, EMDR. The evidence review behind it included 2 acupuncture trials with 222 people, the 2007 waitlist trial and the 2012 paroxetine trial. The committee judged that small single studies could not make it confident the benefits were true effects, and made no recommendation. NICE’s April 2025 surveillance review decided not to update the guideline at that time. Clinical guidelines sets out other bodies’ positions on acupuncture.

What the reviews could not settle

The reviews pooled trials against waitlists, drugs and usual care, often with acupuncture added to another treatment, and the one review that graded its evidence rated it low to very low. The 2024 sham-controlled trial ran at a single VA site, in combat veterans aged 18 to 55, and its design excluded people who had not responded to two evidence-based PTSD treatments, people in active psychotherapy for PTSD, and anyone who had had acupuncture in the past year. It measured symptoms to the end of treatment, so how long the difference lasts is unknown, and its sham produced a moderate improvement of its own and did not differ from acupuncture on anxiety, depression or sleep. NICE’s evidence review found one trial against a trauma-focused therapy, group CBT in the 2007 trial of 84 people, and rated it very low quality. The 2012 paroxetine trial reads as favouring electroacupuncture in its authors’ analysis and as no different in NICE’s. Anxiety and insomnia cover the trials for two problems that often accompany PTSD.

AcuiQ’s PTSD, post-traumatic stress disorder and trauma pages list the protocols individual studies prescribed, each with its citation.