Acupuncture massage vs Swedish massage and individual exercises vs group exercises in low back pain sufferers - a randomised controlled clinical trial in a 2x2 factorial design

Objective: Rehabilitation programs for low back pain (LBP) almost always contain massage and exercise therapy in one form or another. Aim: This study aimed to quantify the effectiveness of therapeutic 'Acupuncture' massage (APM; i.e. tonic stimulation of entire meridians) according to Penzel versus Swedish massage (SM) and individual medical exercises (IE) versus group exercises (GE) in LBP sufferers. Patients and Methods: 109 patients participating in a complex in-patient rehabilitation program were randomised to four groups in a 2x2 factorial design. Main outcome measures were functional ability/disability (Functional Questionnaire Hanover, FFbH) and pain intensity (10 cm visual analogue scale, VAS). Pre/post changes were evaluated by means of 2-way analysis of variance (ANOVA). Additionally, lumbar motility was measured by a 2-inclinometer technique. Results: Baseline mean FFbH score was 66 (SD =18) %, mean pain intensity on VAS was 4.5 (SD = 2.4) cm. Lumbar flexion and extension were 49 (13) and 13 (7). Because of some differences between groups at baseline, group-standardized outcomes were used for analysis. APM showed beneficial effects for both disability and pain compared with SM (group differences: Delta FFbH 7.0% [95% confidence interval (CI) 2.5-11.6], p = 0.003; Delta VAS 0.8 cm [95% CI: 2-15], p = 0.024). Standardized response means were SRMFFbH = 0.5 and SRMVAS = 0.8 for APM, as opposed to SRMFFbH = -0.01 and SRMVAS = 0.4 for SM. Neither significant group differences between both exercise groups [Delta FFbH -0.5% (95% CI -5.2 to 4.2); Delta Vas 0.4 cm (95% CI 0.3 to 1.1)] nor significant interactions between medical exercise and massage were found. Conclusions: Given the fact that even the treatments considered to be the best available achieve at best moderate effects, the observed effect sizes with APM are promising and warrant further investigation in replication studies. In contrast to common view, no superiority of individual versus group exercises could be found in the present study.