Exploring the psychoneurobiology of the placebo response in gambling, depressive, and anxiety disorders

The discovery of new therapeutic agents for psychiatric disorders is potentially hindered by the large placebo responses seen in psychotropic drug trials. Through improved understanding of the placebo response, we might understand how to better design clinical trials and identify novel therapeutic targets. However, the placebo response in psychiatric disorders has been scarcely explored. Trial-level variables, such as duration of the trial or number of centres, appear to influence placebo response rate. Within-patient factors, such as previous experience of treatment, expectations, and neurobiology, are also likely to influence placebo response. Which factors matter, and how they influence placebo response in psychiatric disorders, remains unknown.There are large placebo response rates in clinical trials of gambling disorder (sometimes exceeding 70%). I aimed to identify predictors of placebo and medication response in gambling disorder through a pooled analysis of individual patient data from six treatment studies. Multiple linear regression models demonstrated that baseline severity and number of weeks completed in the trial were predictors of medication response (p’s < 0.01). By contrast, predictors of placebo response included increased baseline depressive symptoms, reduced baseline anxiety symptoms, and non-Caucasian ethnicity (p’s < 0.05). These results were robust to choices made in the analysis.It was noteworthy that symptoms of anxiety and depression were predictors of placebo response in gambling disorder. It is possible that biomarkers associated with placebo effects on anxiety or depressive symptoms might be relevant in other psychiatric disorders. I carried out a systematic review of the functional neuroimaging literature to identify neural correlates of placebo response in patients with anxiety or depression. Due to a small number of included studies and significant heterogeneity in study design, I was not able to carry out a formal meta-analysis. The rostral anterior cingulate cortex and default mode network, the ventral striatum, orbitofrontal cortex, and dorsolateral frontal cortex appeared to be key anatomical nodes in placebo antidepressant or anxiolytic effects. Important neurotransmitters might include endogenous opioids, dopamine, and serotonin.The review highlighted a relative lack of research into placebo anxiolysis. This could be due to a lack of convenient experimental paradigms free of confounders. I developed and tested two experimental paradigms designed to induce placebo anxiolysis in healthy volunteers. The first combined verbal suggestions of improvement and a conditioning procedure with administration of a placebo nasal spray in the 7.5% CO2 inhalational model of anxiety. The conditioning procedure induced significant expectations of improvement (p < 0.001) but this did not translate into reduced anxiety compared with a control group who did not undergo conditioning (p’s > 0.350). In the second, volunteers with symptoms of social anxiety disorder were shown a placebo abstract video accompanied by suggestions that it would reduce their anxiety before completing a novel online social interaction task (InterneT-based Stress test for Social Anxiety Disorder, ITSSAD). The placebo again did not reduce anxiety compared with a control group who did not receive verbal suggestions of improvement (p’s > 0.236). An exploratory analysis suggested that expectations did not influence outcomes. Acute anxiety is associated with nocebo effects, and it is possible that such experimental models are not conducive to inducing placebo effects.My findings highlight several gaps in the literature for future study, as well as a few challenges. Further research is needed to explore whether application of knowledge regarding predictors of placebo response improves detection of medication efficacy in clinical trials, to understand how to optimise prospective neuroimaging study design into placebo effects in mental disorders, and to explore the interaction between autonomic arousal and placebo conditioning