Best Evidence Topics
  • Send this BET as an Email
  • Make a Comment on this BET

Acupuncture for neck pain

Three Part Question

[In adults with chronic musculoskeletal neck pain] is [acupuncture better than standard treatment] at [improving pain relief and speeding return to normal activities]

Clinical Scenario

A 34 year old man complains of persisting neck pain following a road traffic accident 3 months previously. Clinical examination is normal apart from tenderness in the neck muscles. There are no neurological symptoms. You suggest a series of self-administered exercises under the supervision of a physiotherapist with NSAIDs for symptomatic relief. He seems unwilling and volunteers that he has been told by a work colleague that acupuncture would cure his symptoms. As you have seen a number of people fail to improve on physiotherapy you wonder if he might be right.

Search Strategy

Medline on OVID interface on the world wide web 1966-10/04
EMBASE 1986-10/04 on OVID interface
CinAHL 1982-11/04
AMED 1985-10/04 on OVID interface
Bandolier on line 11/04
Cochrane Edition 4, 2004
[exp ACUPUNCTURE/ or acupuncture] AND [exp NECK PAIN/ or exp NECK INJURIES/ or NECK/ or exp NECK MUSCLES/ or exp WHIPLASH INJURIES/ or whiplash]
"Acupuncture" as key word
"Acupuncture" as key word

Search Outcome

66 papers were found on the MEDLINE search. No additional relevant papers were found on any of the other searches.
Two high quality systematic reviews were published in 1999 and 2000. Individual papers prior to 1999 are not shown. The table includes the two systematic reviews as well as 4 randomized controlled trials subsequently published.
One additional recent paper identified through news source (not yet indexed on Medline).

Relevant Paper(s)

Author, date and country Patient group Study type (level of evidence) Outcomes Key results Study Weaknesses
He D
24 female office workers in sedentary occupations with chronic neck pain. Subjects received either acupuncture at anti-pain points or "sham" points. Patients had 10 sessions of therapyPRCT with blinded assessment and patients.Self reported pain intensity after treatment70% reduction for acupuncture vs. 29% reduction for sham p<0.001Small numbers. Very specific population. Not all post-traumatic. Only women aged 20-50. No control group of "non-acupuncture therapy". Difficult to understand how effects are lost at 6 months then re-appear at 3 years.
Self reported pain intensity at 6 months24% reduction for acupuncture vs. 36% for sham p=0.15
Self reported pain intensity at 3 years77% reduction for acupuncture vs. 8% reduction for sham p<0.04
Self reported pain frequency after treatment65% reduction for acupuncture vs. 40% for sham p=0.04
Self reported pain frequency at 6 months44% reduction for acupuncture vs. 34% for sham p=0.18
Self reported pain frequency at 3 years66% reduction for acupuncture vs. 2% for sham p=0.003
Nabeta T
34 volunteers from an acupuncture school. Subjects had once weekly acupuncture for 3 weeks. All subjects were "confident acupuncturists" themselves. Subjects randomized to either real or sham acupuncture.PRCTPain as measured on a visual analogue scaleNo difference between groupsLikely to be a highly biased study group. Some authors would suggest treatments too infrequent. Basic data reporting inadequate. No control group of non-acupuncture treatment.
Pain pressure threshold as measured by algometryMore pain following acupuncture, units not described so difficult to interpret the clinical significance
Zhu XM
29 volunteers with chronic neck pain.Age range 31-70 years. 14 given real acupuncture first, 15 received sham acupuncture first. Followed by washout period of 3 weeks and then crossover to other therapyPRCTFive subjective measures (pain intensity, pain duration, pain pill count, pain on VAS, neck disability index)No significant difference between sham and real acupuncture. Both treatments significantly reduced all variables.No "no acupuncture group" as control. Short washout period. The authors state that both practioners and subjects had positive views of acupuncture which may lead to bias. A retrospective sample size done by the authors show this study to be underpowered.
Length of treatment effectChanges were noted for 16 weeks post treatment. This suggests the washout period between therapies may be too short.
Pain threshold at 9 specified sitesOnly one acupoint showed a significant difference between sham and real acupuncture. The authors state that this may be a type 1 error.
Range of movement using a Plutrimeter-V deviceNo difference between sham and treatment groups.
Irnich D
177 patients with chronic neck pain. Aged 18-85 years. Patients received either acupuncture (56), massage (60) or sham laser acupuncture (61). All patients were offered 5 treatments.PRCTMaximum pain on motion measured on 100mm VAS at one weekReduction in pain score on 100mm VAS. Acupuncture 24.22mm vs. massage 7.89mm vs. sham 17.28. No statistical difference between sham and acupuncture p=0.327. Acupuncture vs. massage significantly different p=0.0052Early secondary outcomes were slightly better in the sham and acupuncture groups. Those differences were not present at 6 months. At 6 months there were no significant differences between the groups. Although a difference is demonstrated it is a short one in a chronic condition. The authors suggest that 5 sessions is too short a period of treatment.
Range of motion (total degree difference) at 6 monthsAcupuncture 8.9 vs massage 5.5 vs. sham 3.5. p=NS
Pressure pain threshold as measured by algometry at 6 monthsAcupuncture 0.19 vs massage 0.05 vs. sham 0.03. p=NS. (Kg/cm2)
Role physical measured on SF-36 at 6 monthsAcupuncture 0.83 vs massage 4.95 vs. sham 5.83. p=NS
Smith LA
Extensive electronic search. Authors not contacted. Small studies excluded. Smith et al re-analysed much of the original data and changed the original conclusions in light if their interpretationSystematic ReviewNumber of papers identified23 of which 13 were included after appraisal. Three excluded because unobtainable.Smith LA(5) Extensive electronic search. Authors not contacted. Small studies excluded. Smith et al re-analysed much of the original data and changed the original conclusions in light if their interpretation Systematic review Number of papers identified 23 of which 13 were included after appraisal. Three excluded because unobtainable. Only full published reports included. Only RCTs included. Heterogeneous methodologies made meta-analysis impossible.
Acupuncture in acute pain2 papers on back pain. Smith et al concluded that there was no difference between acupuncture and sham acupuncture
Chronic pain – single session of acupuncture2 papers. No benefit beyond 24 hours after therapy.
Chronic pain, multiple sessions9 papers found. 3 showed benefit with acupuncture 6 did not.
Adjustment for high quality studiesPapers with high validity scores (e.g. those with blinding) were more likely to show a negative effect p=0.023
White AR
Extensive electronic search. Authors were not contactedSystematic reviewNumber of papers identified32 papers found. 18 excluded. 14 included in analysisOnly RCTs included. Acupuncture encompassed needle, laser and electro-acupuncture. Heterogeneous methodologies made meta-analysis impossible. All trials had small numbers of patients range 13-172 (trial with 172 had 4 treatment arms).
Authors view of balance of resultsIncluded papers were split between benefit and no effect.
Quality of studies assessed using a modified Jadad score (max 5)Quality of studies assessed using a modified Jadad score (max 5) One study scored 4/5. 7 studies scored 3/5. 5 scored 2/5 and one scored 1/5.
Adjustment for higher quality studies (3 or more on the quality scale)3 showed benefit to acupuncture. 5 showed no benefit
White P.
135 patients with chronic neck pain. 18-80 years. 124 patients completed study. Randomly assigned to either acupuncture or mock transcutaneous electrical stimulation using a decomissioned electroacupuncture unit.Single blind RCTPain 1 week after treatmentless with acupuncture (difference of 6.3mm CI 1.4-11.3mm, on 100mm VASNo control group of no therapy, would they have improved anyway? Difference is statistically but not clinicall significant. One week follow up for a chronic condition is probably not clinically important. (report currently from abstract - full paper awaited)


It is clearly difficult to do quantitive clinical trials into acupuncture. Blinding of the operator is clearly impossible and may be difficult to achieve for subjects as well. Many trials involve subjects who themselves may have pre-existing views about the efficacy of treatment. Both the systematic reviews conclude that there is insufficient evidence to recommend acupuncture for neck (or back) pain. However, their conclusions are hampered by the overall low quality of trials that they found. Both groups found meta-analysis to be impossible due to poor trial design and heterogenecity. Since the publication of the systematic reviews, the recent papers are also a mixed bag of design and findings. Both the paper be We and that by Imrich found differences in the early periods following treatment, but both found that these changes are not maintained in the long term. When managing a chronic problem then the utility of treatments which are only effective in the short term remains questionable. Sadly, the current evidence is of insufficient quality to either recommend or refute whether acupuncture works. However, it does appear that there is little or no evidence for its benefit in the long term.

Editor Comment

A recent paper has been published in Annals of Internal Medicine (White et al) this has been reported from abstract only and will be updated when this becomes fully available.

Clinical Bottom Line

Acupuncture may be of short term benefit in neck and back pain. There is little or no evidence that short term benefits are maintained.

Level of Evidence

Level 1 - Recent well-done systematic review was considered or a study of high quality is available.


  1. He D, Veiersted KB, Hostmark AT, Medbo JI Effect of acupuncture treatment on chronic neck and shoulder pain in sedentary female workers: a 6-month and 3-year follow-up study Pain 2004;109(3):299-307
  2. Nabeta T, Kawakita K. Relief of chronic neck and shoulder pain by manual acupuncture to tender points--a sham-controlled randomized trial Complementary Therapies in Medicine 2002;10(4):217-222
  3. Zhu XM, Polus B A controlled trial on acupuncture for chronic neck pain. American Journal of Chinese Medicine 2002;30(1):13-28
  4. Irnich D, Behrens N, Molzen H, Konig A, Gleditsch J, Krauss M et al Randomised trial of acupuncture compared with conventional massage and "sham" laser acupuncture for treatment of chronic neck pain BMJ 2001;322(7302):1574-1578
  5. Smith LA, Oldman AD, McQuay HJ, Moore RA Teasing apart quality and validity in systematic reviews: an example from acupuncture trials in chronic neck and back pain Pain 2000;86:119-132
  6. White AR, Ernst E A systematic review of randomized controlled trials of acupuncture for neck pain. Rheumatology 1999;38(2):143-147
  7. White P, Lewith G, Prescott P, Conway J. Acupuncture versus placebo for the treatment of chronic mechanical neck pain. Annals of Internal Medicine 2004;141:911-919