Whiplash

Whiplash is a neck injury caused by a sudden acceleration and deceleration of the head, most often in a rear-end or side-impact car collision, and the symptoms that can follow it, from neck pain and headache to dizziness and poor sleep, are grouped as whiplash-associated disorders (WAD). NICE’s Clinical Knowledge Summary, last revised in November 2023, reports that about half of people recover fully within six months and that up to half still have symptoms a year after the injury. Acupuncture trials have recruited people within days of a crash and months or years after it.

No Cochrane review deals with acupuncture for whiplash alone. The most recent review, a 2024 meta-analysis of eight trials, found less neck pain with acupuncture than in control groups that mixed sham procedures with usual care, at moderate certainty, and no difference in disability. Canadian guidelines from 2016 make no recommendation for acupuncture in whiplash and advise against electroacupuncture for neck pain. Half of the trials in that review compared acupuncture with a sham procedure, and the result of each depends on what that sham was.

Is there a Cochrane review of acupuncture for whiplash?

No. A PubMed search of the Cochrane Database of Systematic Reviews found two reviews that include acupuncture trials in whiplash, and neither is confined to it. Verhagen and colleagues’ review of conservative treatments for whiplash, last updated in 2007 with searches to November 2006, included 23 trials with 2,344 participants across every kind of non-surgical treatment. Its one acupuncture trial compared laser acupuncture with a placebo laser in 50 people with acute whiplash; the reviewers rated it low in quality and noted that it reported no differences between groups without supporting data. Across all treatments they concluded that the evidence neither supports nor refutes passive or active treatments for WAD grades 1 or 2, which cover neck and musculoskeletal complaints. The Cochrane review of acupuncture for neck disorders included three whiplash trials with 205 participants and was withdrawn in November 2016; Neck pain reports it.

Systematic reviews

Lee and colleagues, at Pusan National University, updated Moon and colleagues’ review by adding two trials published after 2014 and pooling the results, which Moon had not done. The standardised mean difference in the table is the measure reviews use to combine trials that scored pain on different scales. Lee’s “acupuncture” covers three methods: needling at acupuncture points, two trials of needling muscle trigger points by physiotherapists, and one of motion style acupuncture, in which the patient moves the neck while needles are in place. By subgroup, pain was lower with conventional acupuncture and with motion style acupuncture, and not with electroacupuncture. Disability scores reached a significant difference only when Cameron’s sham-controlled trial, which favoured the control group on disability, was left out. The reviewers did not search grey literature, could not obtain the full text of one trial, and found the funnel plot slightly asymmetric, a sign of possible publication bias that fewer than 10 trials cannot test well. They name GB20 (Fengchi), GB21 (Jianjing), SI11 (Tianzong), SI14 (Jianwaishu), SI15 (Jianzhongshu) and TB15 (Tianliao), which the paper writes TE15, as the points used most in the trials.

The review by Wong and colleagues was written for the Ontario Protocol for Traffic Injury Management (OPTIMa) Collaboration. It concluded that electroacupuncture is not effective and should not be used to manage neck pain.

Trials

Chronic and subacute whiplash

• Electroacupuncture against a penetrating sham, 2011. In a university clinic in Sydney, Cameron and colleagues randomised 124 people aged 18 to 65 with WAD grade I or II, 85% of them chronic, to 12 sessions over six weeks of real or simulated electroacupuncture, both with needles through the skin and both given by one acupuncturist. Moon’s review notes that the sham needles went into non-acupoints without electrical stimulation. Starting from a mean pain score of 5.6 on a 10 cm scale, pain fell 0.9 cm more with real electroacupuncture at three months and 1.3 cm more at six months. The authors judged the difference probably not clinically significant, and found no difference in disability or quality of life once baseline scores were taken into account.
• One session against relaxation, 2013. In Belgium, Tobbackx and colleagues gave 39 people with chronic WAD one session of acupuncture and one of relaxation therapy, in random order, with blinded assessors. Pressure pain sensitivity fell more after acupuncture, both in the neck and in the calf, away from the painful area. Neck disability, symptom severity and conditioned pain modulation, a test of the body’s own pain inhibition, did not differ between the two sessions. A 2015 analysis of the same 39 patients by De Kooning and colleagues found that the change in pain sensitivity was unrelated to heart rate, skin conductance or other autonomic measures, which the authors read as evidence against stress as its cause.
• Dry needling with exercise, 2015. At a university centre and four physiotherapy practices in Queensland, Sterling and colleagues randomised 80 people whose whiplash symptoms had lasted more than three months to six weeks of dry needling of the posterior neck muscles plus exercise, or sham dry needling plus exercise. Neck Disability Index scores, out of 100, did not differ at 6 or 12 weeks, and were 4.4 points lower with dry needling at six months and 3.8 lower at 12 months. The authors called the effects small and not clinically worthwhile, and found no difference in self-rated recovery.

Acute whiplash

• Motion style acupuncture in hospital, 2020. At a Korean medicine hospital in Bucheon, Kim and colleagues randomised 97 inpatients admitted after an acute whiplash injury to integrative Korean medicine, which according to Lee’s review already included acupuncture, with or without three days of motion style acupuncture. On day five, mean pain on a 0 to 10 scale was 3.55 with the added treatment and 4.59 without, a difference in change of −1.07 points (95% CI −1.76 to −0.37), and pain eased by two points or more sooner in the motion style group. There was no sham group.
• Trigger point needling with physiotherapy, 2010. In the UK, Tough and colleagues randomised 41 people referred to physiotherapy after a recent whiplash injury to physiotherapy with either trigger point needling or a sham needle, to test whether a full trial was feasible. It met every feasibility criterion, with no serious adverse events. Moon’s review reports that it found no difference between groups in pain or disability at six weeks.
• Laser acupuncture, 2006. In Vienna, Aigner and colleagues treated 45 people with acute whiplash with a cervical collar, paracetamol and the muscle relaxant chlormezanone, and added either laser acupuncture (a 5 mW helium-neon laser on 22 points for 15 seconds each) or a placebo laser, three times a week until symptoms resolved. Neck mobility, duration of pain and collar use, and at one year drug use and chronic problems such as headache and vertigo, did not differ between groups.

Harms

Five of the eight trials in Lee’s review reported adverse events. All were mild except one moderate case of itching, in Kim’s trial, that needed antihistamines; the others attributed to acupuncture were hives, dizziness, more neck pain, bruising, fatigue, sweating and low blood pressure, which resolved within a few days.

What the guidelines say

The OPTIMa Collaboration’s 2016 guideline for neck pain and associated disorders, grades I to III, advises that for grades I and II, whether pain has lasted up to three months or longer, clinicians should not offer electroacupuncture, among other treatments, “in view of evidence of no effectiveness”. It makes no recommendation on acupuncture for grade III, the grade with neurological signs.

The Canadian Chiropractic Guideline Initiative’s 2016 guideline for neck pain-associated disorders and WAD, by Bussières and colleagues, replaced two earlier chiropractic guidelines and recommends multimodal care, exercise and advice. It makes no recommendation on acupuncture, which appears only in a remark listing what multimodal care may include. Lee’s review reports that neither this guideline nor the Australian State Insurance Regulatory Authority’s 2014 guideline for acute WAD recommends acupuncture, and attributes that partly to how little research existed when they were written. No update to either Canadian guideline was found for this page.

In the UK, NICE has no guideline on whiplash. Its Clinical Knowledge Summary advises clinicians to give self-care advice and reassurance, encourage early return to usual activities, discourage soft collars, offer pain relief and consider referral to physiotherapy or a psychologist. It does not mention acupuncture. Clinical guidelines sets these positions beside other bodies’.

What the reviews could not settle

Lee’s pooled pain result combines trials against sham with trials against usual care, so it cannot say how much of the difference is specific to needling. The two sham-controlled trials that followed patients for six months or more, Cameron’s electroacupuncture trial and Sterling’s dry needling trial, both found differences their own authors judged too small to matter. Of the acute trials, Kim’s added acupuncture to a programme that already included it and had no sham group, and only Aigner’s laser trial followed patients for a year to see who developed chronic problems. No trial of needle acupuncture soon after a crash has reported that outcome. The reviews also mix acupoint needling, electroacupuncture, trigger point dry needling and motion style acupuncture, and Lee’s review included one trial it could not read in full.

AcuiQ’s whiplash injury page lists the protocols individual studies prescribed, each with its citation.