Acupuncture for frequent tension headaches: what counts as a good result?

A headache diary isn’t glamorous. But before you book acupuncture for headaches that keep returning, it can settle a more useful question: what kind of headache are you trying to prevent? Tension-type headache and migraine aren’t interchangeable, and frequent use of pain medication can complicate the picture. A clinician can help with the diagnosis; recording headache and medication days gives that conversation something solid to work with. NICE’s headache guideline recommends a diary for diagnosis and monitoring.

Then there’s the question of what success would look like. In acupuncture studies of frequent or chronic tension-type headache, a common yardstick is at least 50% fewer headache days after a course of treatment. That’s prevention measured over weeks or months - not relief from the headache you have this afternoon. The Cochrane review found a possible preventive benefit, but its answer changes depending on what acupuncture was compared with. A 2023 review also warned that much of the evidence was low-certainty.

First, which headache is it?

#

Chronic tension-type headache means headache on at least 15 days a month for more than three months under NICE’s classification. Someone with headaches on fewer days can still have frequent episodic tension-type headache, but the categories aren’t a technicality: the treatment guidance and the more recent trial discussed below don’t automatically apply to both. Migraine can overlap with a person’s description of a “tension headache,” too. If the episodes feature migraine-like symptoms, our separate acupuncture and episodic migraine explainer addresses that distinct evidence base. Don’t borrow its response numbers for tension-type headache.

And keep track of the tablets. NICE flags medication-overuse headache when certain acute headache drugs are used on 10 or 15 days a month, depending on the drug, for at least three months. That’s a reason to discuss the pattern with a clinician, not to abruptly stop prescribed medication on your own.

A useful result isn’t necessarily a cure

#

The Cochrane review, which searched studies through 2016, included 12 randomized trials involving 2,349 adults with frequent episodic or chronic tension-type headache. The reader summary describes about 48 in 100 people receiving acupuncture alongside routine care or acute-only treatment reaching at least a halving of headache frequency after roughly three to four months, versus about 17 in 100 without added acupuncture. But those figures summarize two different unblinded trials, not a single pooled experiment. In the review’s detailed abstract, one trial reports 302 of 629 versus 121 of 636 responders; the other reports 60 of 132 versus 3 of 75. Different starting conditions and care in the comparison groups make a tidy “extra people helped per 100” claim misleading.

What happens when the comparison is sham acupuncture instead? In four Cochrane trials, 205 of 391 receiving acupuncture and 133 of 312 receiving sham met the at-least-50% response threshold - roughly 52% versus 43%. That’s a smaller gap than the routine-care comparisons. It doesn’t mean the whole difference between those two sets of trials measures a “placebo effect.” The participants and controls weren’t the same, sham procedures can involve touch or superficial needling, and NCCIH cautions that sham acupuncture needn’t be physiologically inert. Appointment time and expectations might also matter. We can’t cleanly assign percentages to each ingredient of the experience.

Here’s the distinction worth taking to an appointment: compared with having no added acupuncture, a course might look more attractive; compared with a needling-like procedure, the extra effect is harder to pin down. Neither comparison tells you exactly how your next headache will feel.

The newer chronic-headache trial offers a clearer yardstick

#

A 2022 randomized trial enrolled 218 people with chronic tension-type headache, averaging 21.5 headache days per month at baseline. Participants had 20 sessions over eight weeks of either acupuncture intended to elicit the characteristic needling sensation or superficial needling at the same sites without that sensation. So the control group got appointments and needles, not “nothing.”

By week 16, 68.2% of the acupuncture group had at least halved their monthly headache days, against 48.1% in the superficial-needling group. Their average reductions from baseline were 13.1 and 8.8 headache days per month, respectively - a between-group difference of 4.3 days (95% confidence interval, 2.0 to 6.5). The trial still reported a response gap at week 32: 68.2% versus 50.0%, with a 4.5-day difference in mean monthly headache-day reductions. Substantial improvement in the superficial-needling arm is part of the story, not a footnote.

But it’s one protocol in one chronic-headache study. Twenty research visits over eight weeks may be a real cost in time and money, and follow-up to week 32 doesn’t show that benefit lasts for years. Nor does that trial establish the same result for occasional tension headaches, migraine, children or pregnant patients.

Why the reviews still disagree about certainty

#

A 2023 systematic review assessed 14 studies involving 2,795 people. It found lower headache frequency with acupuncture than with sham after treatment and at follow-up, yet rated the evidence generally very low to low quality. Its trial-sequential analysis said the sham comparison hadn’t reached the required information size. In plainer language: a favorable pooled estimate isn’t the same as a settled answer. The review also found a difference versus no acupuncture, for which its required information size was met. Those are separate comparisons.

A 2025 review selected six randomized trials involving 927 participants and reported a modest standardized difference in headache frequency six weeks after treatment. Its study selection, outcomes and timing differ from the 2023 analysis; primary trials may overlap. So the newer paper isn’t a second, independent set of 927 people confirming one fixed-sized benefit. And a standardized effect isn’t a number of headache days you can enter into a calendar.

This is why quoting only Cochrane’s moderate-quality responder comparison, or only the 2023 low-certainty warning, would leave out something important. Study design, the comparison group and the way results are combined all change how confident we should be. There isn’t good evidence here that acupuncture outperforms physiotherapy, massage, relaxation or exercise; Cochrane found the reporting on other active approaches inadequate.

Where it fits - and where it doesn’t

#

The UK’s NICE guideline says to consider a course of up to 10 acupuncture sessions over five to eight weeks for prevention of chronic tension-type headache. That’s a conditional clinical option, not a rule that everyone needs 10 visits. And it shouldn’t be confused with the 20-visit 2022 trial. If you’re weighing it, ask your clinician how you’ll track headache and medication days, what other preventive options fit your diagnosis, and when to decide whether the time and expense are worth continuing. There’s no evidence here to replace prescribed treatment or delay medical assessment.

Needles also bring practical safety questions. Temporary soreness, small bruises or bleeding can occur; Mayo Clinic and NCCIH describe the risk of more serious complications with improper or nonsterile needling, including infection or organ injury. Look for a qualified practitioner using sterile, single-use needles. Tell both your clinician and the practitioner if you have a bleeding disorder, take anticoagulants, have immune suppression or a skin infection, are pregnant or breastfeeding, or have significant chronic illness. If electroacupuncture is proposed, mention an implanted electronic device. These are topics for individual screening - not a claim that each circumstance makes acupuncture universally unsafe or that needles interact with a drug in the way a supplement might. Adult headache trials don’t establish safety or effectiveness for pregnancy, breastfeeding or children.

And don’t schedule a needling visit to investigate an alarming new headache. A sudden severe headache, new weakness or other neurologic change, impaired consciousness, fever with a worsening headache, or a headache after injury needs prompt medical assessment; NICE lists these and other warning signs. A marked change in your usual pattern deserves evaluation as well.

Bottom line

#

For someone with clinician-assessed frequent tension-type headache, acupuncture can be a reasonable prevention conversation, especially if the diary shows chronic headache days. In studies, more people reached a meaningful reduction than with routine care alone; the advantage over sham was smaller and its certainty remains contested. I’d start with the diary and the diagnosis, then ask what outcome would make a course worth the visits - rather than treating a needle appointment as an answer to every recurring headache.