Endogenous opiates and the placebo effect

Abstract

Objective

A meta-analysis was performed to investigate the ability of placebo administration to reduce self-report of pain and to examine whether placebo-induced pain reduction might have physiological and psychological underpinnings.

Method

Forty-five effect sizes and 1183 participants from 12 studies were meta-analyzed for the effects of placebo and the opioid antagonist, naloxone, on self-report of pain.

Results

Analyses showed that placebo administration was associated with a decrease in self-report of pain, and a hidden or blind injection of naloxone reversed placebo-induced analgesia. Furthermore, there were significant between-group differences for type of pain (experimental vs. postoperative/clinical) for placebo studies.

Conclusions

The results support the literature illustrating that the belief and expectation of analgesia induces discrete physiological changes, leading to relief from pain, and this response may be mediated by endogenous opioids. The implications of these findings are discussed in terms of the symbolic aspect of health care and mental health providers' words and context, and their potential impact on the course of illness and well-being.

Introduction

The placebo response has received attention in many different areas of medicine and science. Its effects have been studied in the context of whether it is truly an appropriate control condition for clinical trials of drugs and other medical or surgical treatments. While some reports estimate that between 25% and 60% of patients report improvement with placebo treatment across various clinical conditions, such as pain, asthma, cardiovascular diseases, and depression [1], [2], other reports have found little evidence supporting significant clinical effects [3].

The fact that physiological changes occur in response to an “inert” treatment, or a symbolic event, has important implications for many different therapeutic modalities. For example, it has been shown that medical context (i.e., all factors that comprise the atmosphere surrounding a patient, including the specific wording of statements made by doctors) has profound effects on the outcome of medical treatment, and this effect is modulated by specific neurochemical messengers [4]. Others have suggested that placebo-like factors may play a role in determining treatment outcome for general medical outcomes [5], as well as specific conditions, such as Parkinsonism [6] and clinical depression [7], [8]. For depression, these placebo-induced improvements can be mapped to changes in brain glucose metabolism [9], [10] and electrical activity [11].

The physiological mechanisms underlying the placebo response have been studied extensively but are not well understood. The most well-studied phenomenon with respect to biochemical changes comes from studies examining placebo-induced analgesia. Mediators implicated include endogenous opioids (e.g., β-endorphin; [12], [13], [14]), cholecystokinin [13], and dopamine [15].

Because studies in this literature are difficult to compare due to the variability of conditions, populations studied, and methodologies, there was a need to examine the significance and consistency of effect sizes across the literature. Past reviews in this area have either been qualitative reviews [16], meta-analyses examining placebo effects in clinical trials collapsed across a variety of clinical conditions [3], or are somewhat outdated [17].

The main objective of this meta-analysis was to provide a focused quantitative review of the effects of placebo analgesia and the plausible mechanisms underlying this phenomenon (i.e., opiate related). For the primary analyses, it was hypothesized that placebo administration (i.e., administration of an inert agent along with verbal communication, that the individual would be receiving an analgesic agent) would decrease the subjective rating of both nonexperimental (postoperative/clinical) and experimentally induced pain. Second, it was hypothesized that the antagonism of opiate receptors (via the administration of naloxone) would attenuate placebo-induced analgesia. Third, as an exploratory analysis, it was hypothesized that there would be a difference in placebo-mediated analgesia among the different types of pain because the literature in this area is variable.