Tension-type headache is the common headache felt on both sides of the head as a pressing or tightening pain. NICE, the body that writes clinical guidance for England, calls it episodic when it occurs on fewer than 15 days a month and chronic when it occurs on 15 or more days a month for more than three months. Trials of acupuncture for it test prevention: a course of sessions, followed by a count of headache days in a diary. Cochrane last pooled them in a 2016 review by Klaus Linde and colleagues at the Technical University of Munich, and its findings differ by comparison: large against routine care, smaller against sham acupuncture, and none in acupuncture’s favour against physiotherapy, massage or exercise. Sham acupuncture is a control procedure made to resemble acupuncture, such as shallow needling away from the points; Sham acupuncture sets out the kinds in use and why none is inert.
The 2016 Cochrane review
The review included 12 randomised trials with 2,349 adults who had frequent episodic or chronic tension-type headache, each trial following participants for at least eight weeks after randomisation. Its main outcome was response: a reduction of at least half in headache frequency, measured three to four months after randomisation, once treatment had finished. The authors rated the evidence with GRADE, the scale Cochrane uses to say how much confidence a result supports, from high to very low.
| Compared with | Evidence | Acupuncture vs control | GRADE |
|---|---|---|---|
| Routine care or treatment of attacks only | Trial of 1,265 people | 48% vs 19% | Moderate |
| Routine care or treatment of attacks only | Trial of 207 people | 45% vs 4% | Moderate |
| Sham acupuncture | 4 trials, 703 people | 51% vs 43% | Moderate |
| Physiotherapy, massage or exercise | 4 trials | No trial found acupuncture significantly better; some outcomes slightly favoured the other therapy | Not pooled; high risk of bias |
The review did not pool the two routine-care trials, because their participants started with quite different headache frequencies and their control groups were managed differently. Neither trial was blinded, so participants knew whether they had acupuncture, but the review judged them otherwise at low risk of bias. Neither looked beyond four months. Seven trials compared acupuncture with sham and the review rated them moderate to high quality; five large ones supplied data for its pooled results. Six months after randomisation the results against sham were similar to those at the end of treatment. The four trials against physiotherapy, massage or exercise were inadequately reported. Across the review, the evidence was downgraded mainly for lack of blinding and for effect sizes that varied between trials.
Harms were rarely reported. In six trials, 1 of 420 people having acupuncture and none of 343 having sham dropped out because of adverse effects. In three trials, 17% of the acupuncture group and 12% of the sham group reported an adverse effect, a difference the review did not find significant; both figures were rated low quality. After a treatment covers what acupuncture’s adverse events are and how often they occur. The authors concluded that the results suggest acupuncture is effective for frequent episodic or chronic tension-type headache, and that trials comparing it with other treatments are needed.
Later reviews
A 2023 meta-analysis from Chengdu University of Traditional Chinese Medicine searched to September 2022 and pooled 14 trials with 2,795 participants. It added a trial sequential analysis, a test of whether the pooled number of participants is large enough for a result to be treated as settled. Against no acupuncture, headache frequency fell more with acupuncture (standardised mean difference 0.52) and the sample passed that test; a standardised mean difference expresses the gap between groups in units of the spread of results. Against sham, frequency also fell more, and more people responded both after treatment and at follow-up, but the sample did not pass the test. The authors rated the evidence low to very low.
A 2024 review by Chen and colleagues asked whether the effect lasts. Pooling seven trials with 3,221 participants, it found a higher response rate with acupuncture than with sham at the end of treatment and one to six months later, no significant reduction in painkiller use, and no advantage for acupuncture over physical training or relaxation training in headache days or intensity. No serious adverse events were reported.
The Acupuncture Trialists’ Collaboration pooled individual participants’ data, rather than published averages, from trials of four chronic pain conditions, chronic headache among them. Its 2018 update covered 39 trials and 20,827 patients and found differences of about 0.5 standard deviations against no acupuncture and about 0.2 against sham, with the effect falling by about 15% over a year. Chronic pain reports that analysis in full.
Two trials against sham
The Acupuncture Randomised Trial for tension-type headache, published in 2005, enrolled 270 people at 28 outpatient centres in Germany. It compared acupuncture with minimal acupuncture, meaning shallow needling at points that are not acupuncture points, and with a waiting list, giving 12 sessions over eight weeks. Headache days fell by 7.2 with acupuncture, 6.6 with minimal acupuncture and 1.5 on the waiting list. The difference between acupuncture and minimal acupuncture, 0.6 days, was not statistically significant; 46%, 35% and 4% of the three groups responded. The authors concluded acupuncture was more effective than no treatment but not significantly more effective than minimal acupuncture.
A 2022 trial at Chengdu University of Traditional Chinese Medicine enrolled 218 people with chronic tension-type headache, who had had headaches for a mean of almost 11 years and on a mean of 21.5 days a month. Both groups had 20 sessions over eight weeks. The acupuncture group was needled to produce deqi, the aching or heavy sensation acupuncturists seek at a point; the control group had the same number of sessions with superficial needling that avoided deqi at each site. At week 16, 68% of the acupuncture group and 48% of the control group had halved their headache days; at week 32 the figures were 68% and 50%. Monthly headache days fell by 4.3 more in the acupuncture group at week 16 and by 4.5 more at week 32. Four mild adverse events were reported. Needling and deqi describes the sensation the two groups were separated by.
The two trials used different controls: one needled non-points shallowly, the other needled superficially and avoided deqi. Reading an acupuncture trial explains why a control’s design shapes the difference a trial can find.
What guidelines say
NICE’s headache guideline, CG150, was published in September 2012 and last updated in June 2025. Recommendation 1.3.9, unchanged since 2012, says a course of up to 10 sessions of acupuncture over 5 to 8 weeks can be considered to prevent chronic tension-type headache. The guideline makes no acupuncture recommendation for the episodic form. Clinical guidelines compares what other guideline bodies say.
What the reviews could not settle
The two trials behind the comparison with routine care stopped at four months. Against sham, the two trials above range from no significant difference in headache days to a gap of 20 percentage points in responders. The 2023 meta-analysis found its sham-controlled sample too small for a settled answer, and its authors call for more high-quality trials against sham. The comparison with physiotherapy, exercise and relaxation rests on a few small, poorly reported trials that did not favour acupuncture. None of the reviews could say which points or how many sessions matter.
AcuiQ’s tension-type headache page and its chronic tension-type headache page list the protocols individual studies prescribed, each with its citation. Migraine covers the other common primary headache, which NICE says commonly overlaps with the chronic tension-type form.