Clinical Research: Phase 1 - Phase 4

The Placebo Problem, Part 12: The Nocebo Response

This is the twelfth installment of our look at the increasingly high placebo response that is plaguing clinical trials in analgesia and psychiatry. Catch up on the rest of the posts in the series here.

After focusing exclusively on the placebo effect in this series, today we’re going to take a quick detour, turning to the placebo’s opposite, the nocebo response.

In a placebo response, trial participants improve when administered an inert substance; in a nocebo response, their symptoms get worse. More broadly, just as the placebo effect is the brain’s response the positive psychosocial context in which treatment occurs, the nocebo effect is the brain’s response to negative psychosocial context. A classic example of the nocebo effect, first described in the 1960s, is when trial participants who are told the drug they will receive has side effects later report stronger and/or more frequent adverse effects than participants not given any information about potential side effects.

Unlike the placebo response, the nocebo response has been much less well studied, in part due to ethical concerns. While the term “placebo” originates from the Latin phrase for “I shall please,” “nocebo” arises from the Latin for “I shall harm.” A nocebo response is, by definition, one that induces a negative experience, so detailed examinations of the nocebo response are few and far between.

Nevertheless, the nocebo response does seem to have a big impact on clinical trials. A meta-analysis of fibromyalgia and painful diabetic neuropathy trials calculated that the nocebo effect accounted for 72 percent of patient drop-outs in the drug groups, while a meta-analysis of the nocebo effect in pain trials found overall nocebo effect sizes to be moderate to large, highly variable, and similar in magnitude to placebo responses. This study also noted that, again similar to the placebo response, larger nocebo effects sizes occurred in trials that used a combination of verbal suggestion and conditioning compared to trial that used verbal suggestion alone.

Believing bad
As in the placebo response, expectation seems to play a large role in driving the nocebo effect. A trial participant is told there is a possibility of adverse effects, causing him or her to expect these effects, which then leads to the belief coming true. In a meta-analysis of clinical trials of an antidepressant versus placebo, adverse events were most often reported in phase II clinical trials, followed by phase III trials. Phase IV trials had the fewest adverse events, suggesting that the nocebo response decreases as a treatment comes to be viewed as less experimental and uncertain.

Additional evidence that expectation plays a role in the nocebo response comes from a systematic review that found that the factors which predicted a greater nocebo response included higher perceived doses, verbal suggestion that the treatment would produce symptoms, and greater expectations about symptoms.

Interestingly, a sense of control over treatment also appears to contribute to the nocebo response. A study in which healthy volunteers could either choose between two equivalent beta blockers or be randomly assigned to one of the two found that the nocebo response was greater in the group that did not have a choice in their treatment.

Bad on the brain
The neurochemical contributors to the nocebo response seem to, at least partially, overlap with those that play a role in the placebo response, although they exert opposite effects. As we discussed in our fourth post in this series, placebo analgesia has been associated with increases in opioid and dopamine neurotransmission. One study of placebo analgesia found that the small number of nocebo responders showed decreased dopamine and opioid neurotransmission in the same brain areas that showed increases in these neurotransmitters in placebo analgesia.

The neuropeptide cholecystokinin (CCK) has also been linked to the nocebo response in pain. Verbal suggestions of worsening pain seem to increase anxiety about the impending increase in pain, which in turn activates the CCK system, leading to increased pain neurotransmission. In support of this theory, both anxiolytics (benzodiazepines) and a CCK antagonist can block nocebo-induced hyperalgesia.

Join us next week as we return our focus to the placebo response, examining how it affects a unique population: pediatric patients. We’ll take a look at the magnitude of the placebo response in kids, discuss what we know about underlying mechanisms, and examine whether the factors that can modify the placebo response are similar in adult and pediatric populations.

Remember to take a look back at this series from the beginning, starting with our overview.  And sign up for more industry insights here.