ACUPUNCTURE FOR CHRONIC STROKE: RANDOMIZED SHAM-CONTROLLED STUDY. David E Krebs*1,6,9; Peter M Wayne2; Eric A Macklin3; Rosa Schnyer4; Ted J Kaptchuk5; Stephen W Parker6,10; Donna Moxley Scarborough7; Chris A McGibbon8; Judith D Schaechter11; Joel Stein11; William B Stason11 1. PT and Clinical Investigation, MGH Institute of Health Professions and Massachusetts General Hospital Biomotion Lab, Boston, MA; 2. New Eng Sch of Acupuncture, Belmont, MA; 3. New Eng Sch of Acupuncture, Belmont, MA; 4. New Eng Sch of Acupuncture, Belmont, MA; 5. Harvard, Boston, MA; 6. Harvard, Boston, MA; 7. Biomotion Lab, Boston, MA; 8. U New Brunswick, Fredricton, NB, Canada; 9. MGH, Boston, MA; 10. Harvard, Boston, MA; 11. Harvard, Boston, MA PURPOSE: We compared the effects of traditional Chinese acupuncture to sham acupuncture on upper extremity (UE) function and quality of life in patients with chronic hemiparesis from stroke. BACKGROUNDS/SIGNIFICANCE: Persistent functional limitations and decreased quality of life often follow stroke. SUBJECTS: We studied 33 subjects who had a stroke 0.8 to 24 years previously, with moderate to severe UE functional impairment. METHODS AND MATERIALS: This study was a two-arm, prospective, sham-controlled, randomized clinical trial with patients and assessors blinded, but not acupuncturists. Patients were randomized to active acupuncture individually tailored to Traditional Chinese Medicine diagnoses, including electroacupuncture; or sham acupuncture using blunt-tipped needles that did not penetrate the skin. Up to 20 treatment sessions (x=16.9 sessions) were performed 1 to 50 weeks (x=10.5 weeks); the per- protocol goal was 20 treatments. We assessed UE motor function, spasticity, grip strength, range of motion, activities of daily living (ADL), quality of life (QOL), and mood at baseline and following conclusion of treatment; ADL, QOL, and mood were also measured at x=37 weeks. ANALYSES: We used multiple regression models controlling for each measure at baseline, subject dexterity, baseline FM score, and log-transformed time since stroke. Results are reported as significant on a comparison-wise basis at alpha = 0.05 for two-tailed tests. RESULTS: Intention-to-treat analyses of efficacy measures found no differences between active and sham acupuncture for any outcome variable. "Per-protocol analyses" had significant improvement in wrist spasticity and ROM, and shoulder frontal plane ROM (p = .01) among active compared to the sham acupuncture. Improvements in upper extremity motor function (p = .09) and digit range of motion (p = .06) among protocol-compliant subjects treated with active versus sham acupuncture were nearly significant. In the intention-to-treat analyses, Barthel ADL, Nottingham QOL and CES-D did not differ significantly between treatment groups at either 12 weeks or 6 months after controlling for baseline FM score, dexterity, and log-transformed time since stroke. In the per-protocol analyses, the overall Nottingham scores improved more after 12 weeks among sham than active subjects, implying greater improvement in average health perspectives among sham subjects. However, this was due largely to lower scores at baseline among sham subjects for the energy level dimension. The median change was zero for all other Nottingham dimensions for both sham and active subjects. Nottingham scores at 6 months differed between sham and active subjects along the energy level dimension due to the same difference at baseline, but none of the other dimensions nor the overall score differed significantly. Subjects in both groups indicated high and equal levels of expectancy that acupuncture might be effective (median: 12 on a 0 to 15 scale for both groups). After 2 treatment sessions, only 28% of subjects correctly guessed their treatment assignment and responses did not differ by actual treatment assignment (Fisher's exact p = 0.71). CONCLUSIONS: Our results suggest that the sham acupuncture protocol was effective as a concealed control, consistent with other recent studies. Controls employed in prior studies evaluating acupuncture for stroke rehabilitation have been criticized because they are unblinded and do not control for non-specific physical or psychological effects of acupuncture. Shallow or minimal needling controls and subliminal transcutaneous nerve stimulation controls are unblinded, may not be inert, and may not control for non-specific effects. The sham device we employed represents a significant improvement over these previous controls and shows promise as a valid and effective control for future blinded trials evaluating acupuncture for post-stroke hemiplegia therapy. These results must be interpreted cautiously given our small sample size and multiple statistical comparisons. While our primary analyses were negative, observed improvements in UE function and mobility among protocol-compliant subjects suggest that acupuncture may offer promise as an intervention for chronic stroke symptoms. A larger, more definitive RCT employing the same design features employed in this study is both feasible and warranted. FUNDING SOURCE: Anonymous donor KEYWORDS: Acupuncture, Stroke, Placebo, Functional Analysis Relative change in UE motor function, spasticity, and range of motion following active or sham acupuncture. *1The treatment effect is the least-squares (LS) mean of the active treatment group minus that of the sham group Copyright 2009 by the American Physical Therapy Association. Requests for reprints should be directed to the corresponding author of the article. Educators, students, and other academic customers may receive permission to reprint copyrighted material from Physical Therapy (ISSN 1538-6724) by contacting the Copyright Clearance Center Inc, 222 Rosewood Dr, Danvers, MA 01923. Other types of customers who want permission to reprint should contact the APTA Editorial Office, Attn: Physical Therapy.