Does Acupuncture Work? Evidence, Risks & What Really Helps

Does acupuncture work? Yes – for certain conditions, the research is now clear enough that Federal health agencies and leading medical guidelines recommend it, and for others the research evidence is weak or nonexistent. An honest response balances sorting through what tens of thousands of patients in research trials actually tells us, including proven indications, still-contentious ones, and what science really can’t tell us about the benefits of acupuncture.

The Short Answer: Where Acupuncture Works and Where It Doesn’t

Most authoritative evidence on the effect of acupuncture for chronic pain comes from a 2018 individual-patient-data meta-analysis of 39 randomized controlled trials involving 20,827 patients. Relative to no acupuncture, this treatment shows a moderate clinical trial effect size (roughly 0.5 SDs) in pain scores. Relative to sham acupuncture, where needles may be superficially and at non-specific points inserted, the effect is smaller (roughly 0.2 SDs). Both results are statistically significant, and the trial follow-up was for a full year with a modest 15% reduction in impact (Vickers et al., Journal of Pain, 2018). An interconnected series of 39 individual trials utilizing the most rigorous methodology confers uncommon weight to the conclusion for any used healthcare intervention.

Understanding that nuance is important. Acupuncture is not a placebo, but the low difference between real (vs sham) acupuncture is less of a difference than acupuncture and no intervention at all, indicating the role of the patient/clinician interaction, the ritual of treatment, and stimulus via the acupuncture needle itself. For particular conditions such as acupuncture for IVF success, acupuncture for smoking cessation, acupuncture for hot flashes, that low real-vs-sham gap isn’t even there. Studies on the efficacy of acupuncture for these specific indications consistently fail to show specific benefit.

How Acupuncture Works: Two Frameworks That Don’t Always Agree

Tell a Chinese medicine practitioner and a Harvard neurologist how acupuncture works and both explanations will be internally coherent, yet neither has been shown to fully explain why using acupuncture produces measurable effects.

How does acupuncture work scientifically?

Mechanistically, the most defensible answer is that acupuncture has at least three overlapping effects- nervous-system effects (detailed in the animal and imaging literature cited above), tissue effects (on the connective tissue at the site of the needle), and large nonspecific effects from expectation, practitioner desire, and the ring of the ritual.

National Center for Complementary and Integrative Health health information clearly states that “how acupuncture works is not fully understood,” a very honest, if not exactly reassuring statement from a federal health agency, and one that Western medicine still does not want to acknowledge.

“Decreases in pain after acupuncture cannot be explained solely in terms of placebo effects. Variations in effect size are driven predominantly by differences in treatments received by the control group rather than by differences in the characteristics of acupuncture treatment itself.”

— Andrew J. Vickers and the Acupuncture Trialists’ Collaboration, Journal of Pain 2018

Conditions Acupuncture Helps (Sorted by Evidence Strength)

The following Evidence-Tier Framework classifies indications based on the strength of the systematic-review evidence rather than how frequently the conditions are marketed as “treatable with acupuncture.” References are cited adjacent to each row, so you can check the source of each tier.

Strongest effects concentrate around musculoskeletal complaints — chronic low-back pain, knee OA, neck pain, and headache. Acupuncture appears to have reasonable evidence for a number of non-pain indications- allergic rhinitis, urinary incontinence, treatment-related nausea- and weak or no evidence for fertility, smoking cessation, weight loss, and many cosmetic claims.

What the Best Studies Actually Show (and Why They’re Hard to Run)

What makes acupuncture research difficult is creating a convincing placebo. You can give one group a sugar pill that appears identical to the real medication. In an acupuncture trial — particularly trials of acupuncture for chronic pain — you can’t mimic the needle very easily. Researchers have used retractable “sham” needles that touch the skin without puncturing, or inserting real needles at a nonmeridian point. Both of these options still results in some sensation, some practitioner attention, and some expectation in the patient-meaning the placebo condition is often a treatment in its own right.

This methodological issue is why the bulk of meta-analyses still show a real-vs-sham difference, just a lot smaller than the no-treatment-vs-acupuncture difference. American Pain Society’s conclusion based on the Vickers data is that acupuncture is “a reasonable option for a patient with chronic pain” – wary language, not a sales pitch. Acupuncture may not be a cure, but where evidence is strong, it is worth its place in pain management.

Is acupuncture better than placebo?

Yes, modestly. In the chronic pain conditions examined by Vickers, real acupuncture outperformed sham by an average of 0.2 standard deviations in pain scores- small but statistically significant. In select indications (hot flashes, IVF, carpal tunnel) the difference disappears. This honest interpretation: acupuncture has specific effects over placebo for chronic musculoskeletal pain and headache, but the difference varies by condition.

What Happens at an Acupuncture Session (And What It Feels Like)

Most first visits take about an hour. The acupuncturist will ask about your presenting symptoms, habits and use of medication. The practitioner will examine your tongue, its shape and coating, your face color and the pulse felt at three points on each wrist. Because TCM diagnosis is based upon patterns rather than specific diseases, the intake process is somewhat more detailed than for most practitioners.

The therapy requires 5 to 20 disposable sterile acupuncture needles at a time, at precise locations. Insertion is usually painless – the needles are extremely fine. These thin needles are about the diameter of a human hair. Once the practitioner feels the needle has reached the appropriate depth, the patient may feel what’s known as a “de qi” sensation: tingling, numbness, heaviness, soreness, pressure. In traditional Chinese medicine, that sensation indicates that the physician has accessed the correct location, where thin needles into the skin are believed to engage the meridian most fully.

The needles will need to stay in for 10-20 minutes while the person remains, ideally, very still. Some practitioners move the needles about, and in some cases the needles will have small electrical impulses applied (electroacupuncture). When removed the person will typically feel either very relaxed, energized, or nothing at all.

How often should you get acupuncture?

Part of a typical chronic pain course would be one or two visits per week for 6-8 weeks. If a person has not shown improvement after roughly 6 visits the chances of improvement decline rapidly and most practitioners would suggest change or cessation of treatment. For an acute condition 1-3 visits are likely to suffice. For some conditions, like post stroke or for fertility, ongoing visits 4-8 weekly are often used.

Safety, Side Effects, and Choosing a Qualified Acupuncturist

Acupuncture is a safe therapy when used by trained and accredited practitioners with sterile needles – which are actively monitored as Class II devices by the FDA in the United States. Its side effects tend to be minimal and brief lasting: mild bruising, pin point bleeding at needle site, brief dizziness; there have been no serious side effects noted for nearly 200 000 patients observed in prospective audits. The most common adverse event of significence may be the pneumothorax (partially collapsed lung due to overdeep needling) that has been recorded.

• ✔
  Check that acupuncturist is licensed by your state – the majority of states now require non medical doctor practitioners to have NCCAOM certification
• ✔
  Make sure that electrical as well as regular needles are cleared for single use, having been cleared by the Food and Drug Administration
• ✔
  Find out how many patients with your problem they see – generalists are OK for many types of pain; specialists matter for fertility, post stroke
• ✔
  Disregard the popular (and unsubstantiated) idea of five banned points: there is no real classical list of universally deadly points, only increased risks associated with needling too deeply near the Bowels and other structures

Acupuncture vs. Dry Needling: Why the Difference Matters

Another common question people have is whether acupuncture is the same as dry needling, which, despite use of similar needles, is based upon very different principles and tissue targets.

Bottom line: if there is a specific muscle problem a physical therapist can help you with, dry needling might be an option. For more general chronic pain pictures, headache, nausea, allergic rhinitis – or for someone you want to have studied the full TCM diagnostic system – acupuncture is probably the better-trained pathway.

Types of Acupuncture: From Manual to Battlefield Protocols

“Acupuncture” is a blanket term for several distinct styles with different needling techniques and points.

• Manual acupuncture, the classical technique: a practitioner inserts and manipulates tiny needles at various body points. Most evidence-based published studies use this style.
• Electroacupuncture is manual needles with electrical current running between them. It’s indicated for pain, nausea, stress urinary incontinence; the JAMA 2017 incontinence trial used 18 sessions of electroacupuncture.
• Auricular acupuncture (also called ear acupuncture) involves needle insertion or placement of beads at points on the outer ear. Has decent evidence for ongoing back pain and cancer pain in 2019-2020 reviews.
• Battlefield Acupuncture (BFA) is a 5-point ear protocol developed by the U.S. Department of Defense and adopted by Veterans Affairs facilities for rapid pain relief.
• Scalp acupuncture direct to points on the scalp, sometimes used in stroke rehab and neurological conditions.
• Cosmetic and facial acupuncture is marketed for skin tone and fine lines. Evidence base remains very weak and mostly in promotional forms rather than RCT reporting.

Cost, Insurance, and Whether Medicare Covers It

A typical cash-only acupuncture visit in the US costs about $75 to $200, with the initial visit costing more because it includes a longer intake. Community acupuncture clinics who see several patients in one room will offer sliding-scale rates as low as $20 to $50 a visit.

Most people aren’t aware just how long a course of treatment adds up: 6-8 visits added together takes several hundred dollars, even without insurance reimbursements.

Insurance coverage has expanded over time, but is patchy. Among acupuncturist visits with some form of insurance coverage, the share rose from 41.1% in 2010-11 to 50.2% in 2018-19 (data from the Medical Expenditure Panel Survey). Medicare started to cover acupuncture on for chronic low-back pain in 2020- up to 12 visits, with another 8 covered if the first 12 produce noticeable improvement (CMS CAG-00452N). Medicaid coverage varies by state, and private coverage often only covers certain diagnoses like for chronic pain or chemo-related nausea. Always double check before making an appointment.

Where Acupuncture Research Is Heading

Several lines of work are reshaping the conversation:

• Research is converging around fascia as a mechanistic basis of some points of acupuncture. Director of the NCCIH Helene Langevin’s 2021 article in Life about fascia flexibility and use in myofascial pain has brought renewed interest in this area, beside the older neural-gate-control mechanism.
• Coverage continues to grow. Medicare’s 2020 decision on chronic low-back pain proved to be the most significant coverage shift in decades of U.S. coverage; even more expansion is probable if cost-effectiveness data are favorable.
• Federal coverage is being scrutinized. Systematic reviews can — and should — reevaluate frequently-used protocols, just as the Battlefield Acupuncture account provides an example of. Expect more such reviews, as the VA system accumulates outcome data.
• Use of acupuncture in the US is climbing. Adult U.S. acupuncture use went from 1.0% in 2002 to 2.2% in 2022, according to National Health Interview Survey information, published in JAMA.

If in 2026 you are contemplating acupuncture for chronic pain, consider whether your practitioner’s protocol is appropriate to provide evidence-strength (which is highest) and whether the practitioner can explain how six sessions with no progress should lead to termination.

Frequently Asked Questions