The Effect of Choice on Placebo Treatment for Sleep

Abstract

Thesis title: The Effect of Choice on Placebo Treatment for Sleep Abstract: Recent studies suggest that choice may enhance the placebo effect in the treatment of acute conditions. However, in those studies, there is often a failure to detect a placebo effect in the absence of choice, making it unclear as to whether choice provides an advantage when a significant placebo effect exists without choice. To address this, this thesis examined the role of choice in placebo effects for a chronic condition where placebo effects exist independent of choice. First, a review of the literature on the placebo effect and choice effects is presented. Second, a meta-analysis was conducted to confirm that placebo effects are reliably produced for sleep difficulty without choice. The meta-analysis indicated statistically reliable small to moderate placebo effects in perceived sleep onset latency (SOL), total sleep time (TST), and global sleep quality (GSQ). Finally, an empirical study tested the influence of choice on the placebo effect for sleep difficulty. 117 participants with self-reported sleep difficulty were recruited under the guise of a hypnotic trial and randomized to receive: placebo treatment with single choice, daily choice, or no choice, or no treatment. Significant placebo effects were found for perceived insomnia severity, fatigue, and sleep quality, with treatment-induced expectancy fully mediating the placebo effects for insomnia severity and sleep quality, and partially mediating effects in fatigue. However, placebo effects were not found for perceived SOL and TST, nor for objective SOL and TST measured via actigraphy. There was no evidence of enhancement of the placebo effect by choice on any of the outcomes, nor were there any differences between making a single choice versus daily choices. These findings suggest that choice may not significantly strengthen placebo effects over longer treatment periods, or provide a significant advantage when placebo effects already exist without choice.