Massage for neck pain contrasted against standard (non-surgical) treatment: A systematic review update

Objectives: The aim of this systematic review update was to determine the average effect of massage for adults with neck pain (NP) contrasted against another standard treatment. Methods: Randomised controlled trials comparing massage to standard treatments were included; placebo/no treatment comparisons were excluded. Databases were searched (CENTRAL, MEDLINE, EMBASE, CINAHL, ICL, trial registries) from inception to Oct-1-2023. We used the standard Cochrane methodological procedures: rated Risk of Bias 1.0, abstracted mean differences (MD), meta-analysed data, and rated the level of certainty (GRADE). Results: We included 42 studies (2656 participants; 67% high RoB) contrasted against 10 unique treatments. Trials studied ages 18–70, 70% female, and mean pain severity 52 Visual Analogue Scale (VAS). Very-low to moderate-certainty evidence for pain (MD VAS 0–100, 95% CI) at ≤12 weeks follow-up follows. The pre-stated minimal important difference margin was 10 VAS points. Massage was 1. Favoured over an education-booklet (MD -4.80, −9.17 to −0.43), traction (MD -19.10, −31.07 to −7.13), and higher over lower dose (MD -17.40, −25.40 to −9.40). 2. Likely comparable to acupuncture (MD 2.09, −1.13 to 5.30), exercise (MD -2.50, −6.41 to 1.41), and ultrasound (MD -12.80, −26.43 to 0.83). 3. Not favoured (less effective) when compared to kinesiotaping (MD 9.59, 0.68 to 18.51), manipulation-mobilisation (MD 7.93, 0.01 to 15.85), and mobilisation-plus-exercise (not reported). Massage may not increase risks of minor adverse events: RR 0.37 (95% CI 0.22 to 0.61). Conclusion: For subacute-chronic NP, pain reduction varied by comparison. The evidence was limited by imprecision and high RoB. Focused planning for adequately dosed longer-term trials is needed.