Date of Award

Spring 1-1-2016

Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Tor D. Wager

Second Advisor

Yuko Munakata

Third Advisor

Leaf Van Boven

Fourth Advisor

Christopher Lowry

Fifth Advisor

McKell Carston


Placebos do not possess any direct pharmacological method of action, however placebo treatment has been shown to markedly reduce pain across multiple studies. This effect, termed placebo analgesia, is typically mediated by expectations, such that greater reductions in pain are explicitly related to greater expectations for pain relief. This dissertation examines the psychological processes and neurobiological mechanisms involved in placebo analgesia across three different studies, and the results are interpreted within a dual process framework.

The first study introduces a novel procedure used to induce expectation-independent placebo analgesia without pharmacologically active drugs. In this study, subjects who experienced multiple conditioning sessions across several days continued to report placebo analgesia both before (Pre-Reveal) and after (Post-Reveal) a complete and convincing disclosure of the placebo manipulation. The second study explores the neurobiological mechanisms involved in placebo analgesia using a similar paradigm. Pre-Reveal, analgesia was associated with reduced activity within brain regions associated with pain experience, including both cortical (dorsolateral prefrontal cortex (dlPFC), dorsal anterior cingulate cortex, insular cortex) and subcortical (putamen, thalamus, rostral ventromedial medulla (RVM)) regions. Analgesia was correlated with decreased expression of the Neurologic Pain Signature (NPS), a multivariate pattern sensitive and specific to pain, and was captured by changes in a classifier trained to predict analgesia based on brain activity (5-fold CV, r = 0.32, t37 = 2.36, p = 0.023). Post-Reveal, analgesia was correlated with reduced activity within dlPFC and RVM, but was not captured by changes in NPS or other trained classifiers. The third and final study evaluates whether expectations are important during learning by informing half of the subjects that the intensity of the painful stimuli was reduced during conditioning. These subjects experienced placebo analgesia Pre-Reveal, but not Post-Reveal.

These studies extend the current model of placebo effects to include separable underlying processes that either respond rapidly to new information or slowly accumulate information over time. This work highlights the importance of managing expectations during treatment of clinical disorders, as initial expectations for treatment may be critical for inducing a therapeutic placebo response that could persist even when expectations about that treatment are reduced.

Available for download on Friday, July 31, 2020