Does the placebo effect on hot flashes depend on the placebo dose?

purposeTo investigate the presence of a placebo dose-response effect in four randomized, double-blind, placebo-controlled, multi-dose hot flash clinical trials conducted at Mayo Clinic.

methodsHot flash score, frequency, and hot flash-related distress for each placebo dose level were summarized at each time point by mean and standard deviation and changes from baseline were plotted to visualize a possible placebo dose-effect response. Furthermore, a meta-analysis was conducted for each endpoint in the highest and lowest dosage arms across the four trials.

resultsLongitudinal plots of mean hot flash scores, frequencies, and hot flash-related distress scores in patients taking placebo in each study showed a decline in hot flash scores over time without any clinically meaningful differences between the lowest and highest dosage arms in each study. The meta-analysis for each endpoint in the highest and lowest dosage arms across the four trials revealed no clinically important differences either.

conclusionWhile the current study cannot rule out the existence of a placebo dose-response effect in multi-dose placebo-controlled trials in patients with hot flashes or other conditions, it suggests, along with the available data in the placebo literature, that, at least in well-conducted multi-dose clinical trials in which the placebo was used as control, such an effect, if it exists at all, should be very small. Therefore, pooling data from different placebo subgroups is unlikely to compromise the validity of comparisons between the combined placebo arms and each treatment arm.