The need to investigate nocebo effects in more detail

While much discussion has focused on placebo effects over the past years, less attention has been paid to nocebo phenomena, in both clinical trials and medical practice1, 2, 3. Unfortunately, psychiatry is one of the disciplines in which we know the least about nocebo effects, although their involvement across a variety of mental disorders is likely to be very significant. Indeed, mental disorders are difficult to investigate in this respect, from both a clinical and a neuroscientific perspective, compared to conditions such as pain and motor disorders.

The nocebo effect represents the evil twin of the placebo effect, whereby the patient's negative expectations may lead to clinical worsening1, 2. Therefore, it can provide important information on the psychological factors involved in the generation and time course of a disease.

Although there are a number of clinical trials in psychiatry in which nocebo effects have been assessed and described, little information can be derived from these studies, as the possible psychological and neurobiological underpinnings are difficult to extrapolate. For example, in a meta‐analysis of antidepressant clinical trials, high nocebo effects were found, yet the possible sources of these effects could not be identified4. Indeed, in a clinical trial without a no‐treatment control arm, psychological factors cannot be disentangled from the natural history of the disease and from regression to the mean.

What we need today in the field of psychiatry is to approach the nocebo effect in the same way as done for other medical conditions, where we have understood some of the underlying mechanisms. The task is not easy, and certainly it represents a challenge for future psychiatric research, but it is worthwhile, considering that the neuroscientific approach to nocebo phenomena is paying dividends in other conditions1, 2.

For example, pain is the condition where nocebo effects have been analyzed in most detail. Many mechanisms are at work in nocebo hyperalgesia, including patient‐related factors, the psychosocial context, and neurobiological factors. Recent research has identified many biological underpinnings, such as cholecystokinergic and cyclooxygenase hyperactivity2. Likewise, brain imaging techniques, including functional magnetic resonance and positron emission tomography, have documented the involvement of several brain regions, and even the spinal cord, in the nocebo hyperalgesic response5.

This mechanistic approach to the nocebo phenomenon is important for at least two reasons. First, it demonstrates that nocebo effects are associated with changes in the patient's brain. Second, it suggests that the understanding of these effects may lead to better medical practice and clinical trials: in fact, what we want to do in routine clinical practice is to maximize placebo effects while minimizing nocebo effects, whereas in clinical trials we want to minimize both placebo and nocebo effects.

Nocebo phenomena are also important to better address some issues related to the biopsychosocial model. For example, in a recent study, we investigated the role of negative expectations, so important in nocebo phenomena, in hypoxia headache, in order to understand their relative contribution to the generation of headache pain6. We found that biological, psychological and social factors are additive not only in the generation of headache, but also in inducing the biochemical changes related to hypoxia, such as the increased activity of cyclooxygenase. This is a straightforward example of how negative psychological factors may interact with biological factors in the generation of illness.

In the setting of clinical trials, nocebo effects represent an important source of confusion and misinterpretations. For example, the rates of adverse events reported in the placebo arms of clinical trials for three different classes of anti‐migraine drugs (non‐steroid anti‐inflammatory drugs, triptans and anticonvulsants) were very high and, most interestingly, the adverse events in the placebo arms corresponded to those of the anti‐migraine medication with which the placebo was compared7. The most likely explanation for these effects is that the list of possible adverse events in the informed consent forms generates negative expectations.

Depression shows the same effects. In a comparison of the rates of adverse effects reported in the placebo arms of tricyclic antidepressant and selective serotonin reuptake inhibitor (SSRI) trials, the way in which adverse events were recorded influenced the rate of these effects substantially8. A total of 143 placebo‐controlled randomized trials and data from 12,742 patients were analyzed. More systematic assessment led to higher rates than less systematic assessment. Far more adverse effects were reported in tricyclic antidepressant compared to SSRI placebo groups, e.g. dry mouth, drowsiness, constipation, sexual problems. In general, the adverse effect profiles were strongly influenced by the expectations of investigators and patients, with the adverse effect pattern of the placebo group closely resembling the adverse effects of the drug group.

A better understanding of nocebo effects in psychiatry could be crucial both in the setting of clinical trials and in routine clinical practice. By controlling patients’ negative expectations, we could be able to reduce to some extent poor compliance and dropouts.

For example, the way in which informed consent is formulated should probably be revised in order to pay more attention to sentences that could lead to negative expectations. Likewise, doctor‐patient interaction should be aimed at avoiding negative communication. In a study on influenza vaccination, people who were informed of the proportion of individuals who do not develop side effects (positive communication) showed less adverse effects than those who were informed of the proportion of individuals developing side effects (negative communication)9.

In conclusion, we believe that future psychiatric research should try to better understand nocebo phenomena from different perspectives. The neuroscientific approach could give us information on the biology of nocebo effects in mental disorders, while the methodological perspective could help us design better clinical trials. Overall, both medical practice and doctor‐patient relationship could benefit from this.