Sleep improvement for restless legs syndrome patients. Part III: effect of treatment assignment belief on sleep improvement in restless legs syndrome patients. A mediation analysis

Correspondence: Fred Burbank Salt Creek International Women’s Health Foundation, 1235 Puerta del Sol, Suite 700, San Clemente, CA, 92673, USA Tel +1 949 496 0026 Fax +1 949 492 0650 Email [email protected] Purpose: Two parallel-design, randomized, sham-controlled clinical trials were conducted to study the safety and efficacy of vibratory stimulation (VS) on restless legs syndrome (RLS) patients (Part I of this series of articles). Pooled data from the two studies was retroactively analyzed to compare the relative effects of actual pad assignment with therapeutic pad assignment belief on sleep improvement for patients with RLS. Patients and methods: One hundred fifty-eight patients with at least moderately severe RLS, as measured by a score of 15 points or greater on the International Restless Legs Syndrome Study Group rating scale (IRLS), were enrolled in the study. Patients were randomly assigned to treatment (patient-controlled vibration) or sham (patient-controlled sound or light-emitting) pads. Patients and clinicians were blinded to pad assignment. The pad was placed under the patient’s legs while in bed at night and activated during an RLS episode. Improvements in Medical Outcomes Study Sleep Problems Index II (MOS-II) scores from baseline to week 4 were examined as a function of pad assignment (independent variable) and therapeutic pad assignment belief held by each patient (mediator variable) through mediation analysis. Results: Therapeutic pad assignment belief influenced change in MOS-II scores more than actual pad assignment. Patients who believed they had been assigned a therapeutic pad had substantially greater sleep improvement than those who concluded the opposite. When a patient believed that a therapeutic pad had been assigned, sleep improvement was comparable in magnitude, independent of the type of pad assigned (vibrating or sham). Patients assigned vibrating pads believed that they had been assigned a therapeutic pad 2.6 times more frequently than patients assigned sham pads. Consequently, vibrating pads were more efficient at improving sleep than sham pads. Similarity of sleep improvement for those who believed that they had been assigned a therapeutic pad among vibrating, sound, and light pad patients suggests a common counterstimulation therapeutic mechanism of action within the brain. Conclusion: Therapeutic pad assignment belief influenced improvement in MOS-II scores more strongly than actual pad assignment. Therapeutic pad assignment belief was more commonly associated with vibrating pads than sham pads. These results may have implications for the type of shams used in future device studies.