Acute stiff neck

An acute stiff neck is sudden neck pain with painful, restricted movement. NICE’s Clinical Knowledge Summary calls it acute torticollis, or wry neck, and describes its typical features as a sudden onset of severe pain on one side, restricted and painful neck movements, and tenderness with palpable spasm on that side. It tells clinicians to explain that the condition usually resolves within 24 to 48 hours, occasionally takes up to a week, and often recurs. In East Asian medicine the name is luo zhen (落枕) in Chinese and nakchim in Korean. Kwon and Lee, reviewing the acupressure research, describe it as neck pain and stiffness that comes on after sleeping, and note that the World Health Organization’s standard terminology translates it as “stiff neck”, a term that loses the link to sleep.

Cochrane has no review of acupuncture for this presentation. A 2024 meta-analysis of 10 Chinese trials found acupuncture, alone or added to massage and other conventional care, ahead of conventional care alone, and rated every result very low in certainty; no trial in it had a sham control. A 2025 German guideline advises against acupuncture for acute non-specific neck pain. Neck pain that lasts for months is a different body of research, which Neck pain reports.

No Cochrane review of the acute presentation

The Cochrane review of acupuncture for neck disorders by Trinh and colleagues, updated in May 2016, set out to cover acute, subacute and chronic neck pain. Its 27 trials included three on whiplash, ranging from acute to chronic, and none on acute neck pain without an injury: the other groups were chronic myofascial, arthritic, chronic non-specific, radicular, and subacute or chronic mechanical neck pain. Cochrane withdrew the update in November 2016, and no Cochrane review of acupuncture for acute neck pain or torticollis was found for this page.

Systematic reviews

At Chongqing Medical University, Yu and colleagues searched Chinese and English databases for randomised trials of acupuncture for stiff neck. In two of the 10 trials acupuncture replaced the control treatment; in the other eight it was added to it, with one trial using warm-needle moxibustion and one electroacupuncture. Illness had lasted one to five days on average, and courses ran three to seven days. Against conventional care the acupuncture groups had a higher total effective rate, the share of patients each trial graded as improved (risk ratio 1.12, 95% CI 1.04 to 1.21); pain 0.93 points lower on a visual analogue scale (8 trials); lower Neck Disability Index scores (2 trials); and more lateral flexion and rotation (2 trials). The pain results disagreed with one another far more than chance allows (heterogeneity of 95%), and Egger’s test found evidence of publication bias for both main outcomes. In subgroups the differences lost significance where patients averaged 40 or older or treatment ran longer than three days. The authors list among their limitations that every outcome was subjective, that no trial compared acupuncture with a placebo, and that the trials lacked long follow-up.

The 15 studies Kwon and Lee found were case series and case studies without control groups. They pressed points in the neck and shoulder, most often GB20 (Fengchi), or points elsewhere, most often GB39 (Xuanzhong); eight studies used a single point, either GB39, SI11 (Tianzong), BL57 (Chengshan) or TB03 (Zhongzhu). Across the studies, symptoms resolved completely after one session in a mean of 72% of people with points near the neck and 69% with distant points, and the reviewers called the evidence inconclusive because of the studies’ low methodological quality. Acupressure describes the method.

The Agency for Healthcare Research and Quality’s 2020 review of treatments for acute pain, by Chou and colleagues, found acupressure might be effective for acute musculoskeletal pain, and named neck pain among the conditions on which research is still needed to compare treatments.

Trials

Acupuncture with movement

• Motion style acupuncture, 2026. Across four Korean medicine hospitals, Lee and colleagues randomised 128 adults whose neck pain had lasted under four weeks to motion style acupuncture treatment or conventional acupuncture, two or three times a week for two weeks, with blinded assessors. Both groups had needles at TB15 (Tianliao), SI15 (Jianzhongshu) and LI16 (Jugu); in the motion style group the physician moved the patient’s neck, actively and passively, while the needles stayed in. At week 3, pain on movement was 15.2 mm lower on the visual analogue scale with the guided movement, and the difference lasted to week 9. Adverse events were mild and similar in the two groups. The authors note that without a sham or minimal acupuncture group the trial cannot separate the effect of needling from expectation, and that eight weeks of follow-up cannot show whether the method prevents chronic pain.
• A distal point and massage, 2024. In Beijing, Li and colleagues randomised 162 people with stiff neck attributed to the levator scapulae muscle to acupuncture, tuina (Chinese therapeutic massage) or both, three sessions on alternate days. Acupuncture meant one needle at SI03 (Houxi) on the affected side, angled toward PC08 (Laogong) and kept in for two to three minutes while the patient moved the neck and shoulders. Pain fell in all three groups; the combination scored lowest at days 1 and 3, and by day 7 tuina alone scored lower than acupuncture alone. No serious adverse reactions occurred. This is the one trial in Yu’s review that concealed allocation.

Uncontrolled studies

• Acute torticollis, 2003. In Jerusalem, Samuels gave 18 people one 20-minute session at SI03 and XU08 (Luozhen, also called Wailaogong), one of the extra points, which the paper codes M-UE-24, and measured head rotation with a compass and protractor. Rotation improved by a mean of 52.9%, more in people seen within 24 hours than after 72. There was no control group.
• Dry needling for wry neck, 2017. In Mangalore, Shanmugam and Mathias gave 21 people with wry neck attributed to a locked facet joint one session of dry needling beside the spine and reported lower pain and wider range of motion immediately, at 24 hours and at one week, again without a control group.

Acute pain trials that included the neck

At Duke University, Eucker and colleagues randomised 236 emergency department patients with acute musculoskeletal pain in the neck, back or limbs to ear acupuncture, acupuncture at other points, or usual care. After one hour pain had fallen 1.6 points more on an 11-point scale with ear acupuncture and 1.2 points more with the other acupuncture than with usual care alone. In the same group’s second trial, of 599 patients offered a month of clinic sessions, pain at one month was similar in the two arms, and 43% of those allocated acupuncture could not attend the clinic. Neither abstract reports neck pain separately.

The sham-controlled trial of distant points for neck pain by Irnich and colleagues in Munich enrolled chronic, not acute, patients. It treated 36 people with chronic neck pain and limited mobility once each with acupuncture at distant points, dry needling of local trigger points and sham laser, and found motion-related pain about a third lower after the distant points than after dry needling or sham.

Trials without published results

Three trials of acute neck pain set out their methods in BMJ Open, and a search of Europe PMC by their registration numbers found no results paper for any of them when this page was checked.

• A 2014 pilot protocol from Heilongjiang University of Chinese Medicine planned to randomise 36 people with stiff neck to neck massage with or without acupuncture at SI03 for three days. Yu’s review includes a 2015 Chinese-language trial of Houxi needling with neck massage in 60 people by Tian, Sun, Zhang and Xun, three of whom wrote the protocol.
• A 2024 protocol from Yunnan University of Chinese Medicine plans to randomise 120 people with acute stiff neck to acupuncture with active exercise, sham acupuncture with active exercise, or exercise alone, in a single 10-minute session. It is the only sham-controlled trial of acute stiff neck found for this page.
• A 2014 protocol from Pusan National University planned a 40-patient emergency department pilot of acupuncture added to standard care for acute non-specific neck pain, ankle sprain or headache.

Harms

Only one of the 10 trials in Yu’s review reported adverse reactions, which its authors say could reflect gaps in monitoring or reporting, and none of Kwon and Lee’s acupressure studies reported any. The 2026 motion style trial recorded mild adverse events at similar rates in both groups. Needling the neck covers the risks of needles placed in and around the neck.

What the guidelines say

The German Society for General Practice and Family Medicine’s guideline on non-specific neck pain, upgraded from the S1 to the S3 level of the German guideline system, was summarised in Deutsches Ärzteblatt in October 2025, with its literature search updated to November 2024. It defines acute neck pain as lasting up to three weeks, and issues a negative recommendation for acupuncture in acute non-specific neck pain, graded B on evidence it rates at level V, while acupuncture combined with activating treatments may be offered for chronic neck pain, on level I evidence. The guideline’s central recommendation is activity and self-management.

The Ontario Protocol for Traffic Injury Management (OPTIMa) Collaboration’s 2016 guideline, for neck pain lasting less than six months, advises that for the two milder grades lasting three months or less clinicians should not offer electroacupuncture, among other treatments, “in view of evidence of no effectiveness”.

NICE’s Clinical Knowledge Summary on acute torticollis, last revised in October 2023, recommends painkillers, heat or cold packs, neck stretches and an offer of physiotherapy referral, and advises against a cervical collar. It does not mention acupuncture, and states that it could find no guideline, systematic review or randomised placebo-controlled trial comparing treatments for acute torticollis, so its advice is extrapolated from other sources. Clinical guidelines compares how guideline bodies treat acupuncture across conditions.

What the reviews could not settle

The one meta-analysis of acupuncture for acute stiff neck pooled 10 small Chinese trials against massage and other conventional care, found its main results heterogeneous and at risk of publication bias, and rated all of them very low in certainty. No published trial of acute stiff neck has compared acupuncture with sham, so the trials cannot say how much of the difference comes from the needling rather than from attention, expectation or the extra treatment; the one sham-controlled trial planned has not reported. The trials also cannot be read against the usual course of the condition: NICE expects most cases to resolve within a few days, and the trials in Yu’s review treated for three to seven days with no untreated group. The distal-point idea behind SI03 and similar points has its sham-controlled support in chronic neck pain, from a single session in 36 people, and the acute trials with control groups compared needling with massage or added movement to needling. None of the reviews or trials above reported recurrence, which CKS says is common.

AcuiQ’s stiff neck, acute neck pain, limited range of motion of the neck and neck pain pages list the protocols individual studies prescribed, each with its citation.