Testing the boundaries for expectation effects in health and disease

Expectations affect our physiology and clinical outcomes, however the boundaries for this modulation are poorly understood. The purpose of this thesis is to investigate minimal requirements to elicit expectation effects on health-related outcomes, using experiments and an independent patient data meta-analysis. More specifically, this thesis will build on three aspects that traditionally are considered important for the formation of expectation effects: conscious awareness (Study I), endogenous opioids (Study II) and face-to-face interaction with a health-care representative (Study III, Study IV). I will thus investigate expectation effects from neurobiology up to our interaction with the context around us. In study I the role of conscious awareness in pain processing was investigated in a non-clinical population (N=114) to see if expectations can shape pain even when the participant is not aware of getting noxious stimuli. This was done by assessing whether noxious heat given while asleep would lead to changes in pain ratings in a subsequent test-phase when awake. Two control experiments consisted of only the test-phase. The results showed that participants who had been getting noxious heat while they were sleeping, displayed the same pattern of heightened pain ratings (i.e., pain alarm response) as participants in the control conditions who had not been exposed to the noxious stimuli during sleep. In comparison, the awake condition rated all test-phase stimuli the same. The results emphasize how important expectations are for shaping pain perception. Study II investigated if endogenous opioids are necessary for placebo-like effects/conditional responding. Healthy participants (N=30) were randomized to naltrexone/placebo before a pain-cue conditioning, using pressure pain and functional magnetic resonance imaging (fMRI). Results show comparable conditioned analgesic (pain relieving) and hyperalgesic (pain enhancing) responses in participants with naltrexone or placebo. These findings indicate that full function of the endogenous opioid system during pain conditioning is not necessary for conditional responding. Study III investigated if placebo effects can be created through online communication. Healthy participants (N=30) were randomized to empathetic/neutral communication online where they learnt about a sham analgesic TENS machine (fake pain-relieving machine). After this, a placebo experiment face-to-face was performed, in which the communication was held to a minimum. Results showed that placebo effects were induced during online communication, both in the empathetic and the neutral condition. In Study IV expectation ratings and how they relate to treatment outcome in online and face-to-face psychological treatment were investigated in an individual patient data meta-analysis. Individual participant data from studies that randomized patients to online versus face-to-face psychological intervention and who administered the Credibility and Expectancy Questionnaire (CEQ) were analyzed. Results shows comparable effects of how expectation ratings predicted clinical outcomes post treatment between online and face-to-face treatments. These results suggest that pain processing such as pain alarm response is affected by conscious awareness, endogenous opioids are not necessary in all situations to create pain cue conditioning, placebo effects can be created through online communication and expectations seem to be just as important for online treatments as it is for treatments delivered face-to-face. In sum, these results challenge formerly known boundaries for expectation effects