BACKGROUND: Acupuncture is widely used for chronic non-specific low back pain (CNLBP), yet the optimal dosing parameters remain unclear. This systematic review and meta-analysis evaluated the overall efficacy of acupuncture for CNLBP and explored dose-response relationships between acupuncture parameters and pain reduction.
METHODS: PubMed, Embase, the Cochrane Library, and Web of Science were searched from inception to December 2025 for randomized controlled trials (RCTs) comparing acupuncture with sham, no-intervention, or active controls. The primary outcome was end-of-treatment pain intensity (VAS/NRS). Effect sizes were expressed as Hedges' g and pooled using a random-effects model (REML, Knapp-Hartung). Dose-response relationships for treatment sessions, duration, and needle retention time were examined by weighted meta-regression with restricted cubic splines (RCS).
RESULTS: Twelve randomized controlled trials involving 1,155 participants were included. Acupuncture significantly reduced pain intensity compared with controls (pooled Hedges' g = 0.57, 95% CI: 0.19-0.95), with substantial between-study heterogeneity (I2 = 85.3%). Exploratory dose-response and subgroup analyses indicated a potential nonlinear association for treatment duration - notably with a possible plateau or turning range around 6-7 weeks - though it was sensitive to the most influential trial and must be viewed as hypothesis-generating. Total sessions, needle retention time, and treatment frequency showed no significant dose-response relationships. No publication bias was detected, though statistical power was limited.
CONCLUSION: Acupuncture provides moderate pain relief for CNLBP. The observed nonlinear duration signal-including a potential turning range around 6-7 weeks-is preliminary and hypothesis-generating, given the high heterogeneity and sensitivity to the most influential trial. Session number and needle retention time showed no detectable dose-response association within the available evidence. Dedicated dose-finding trials are required.
| Discipline Area | Score |
|---|---|
| Physician |  |
| Rehab Clinician (OT/PT) | |
Physician rater
Over 80% of trials were at high risk for participant blinding bias. The control groups included sham treatments, usual care, physiotherapy, and placebo TENS, which likely overestimates the specific effect. Outcomes assessed only end-of-treatment pain. Acupuncture might be a viable addition to multimodal rehabilitation, but it is not established as an effective standalone treatment.
Physician rater
Acupuncture offered moderate pain relief overall, but tests for publication bias lacked sufficient power, and the evidence certainty was not formally assessed.
Rehab Clinician (OT/PT) rater
The study adds support to published evidence reporting the positive effects acupuncture in reducing short-term pain intensity. The study was unable to demonstrate significant dose-related acupuncture parameters in reducing pain. I was not surprised the authors reported the evidence was limited by substantial heterogeneity of studies. CLBP is a heterogeneous syndrome involving complex biopsychosocial factors such as pain-related fear or patient expectations and treatment preferences. The authors examined outcomes through the lens of a single treatment modality. Without controlling for complex modifying biopsychosocial interactions, the results should be interpreted cautiously by clinicians.
, McMaster University