Choice and the nocebo effect: If a little is good, more is better?

Lack of choice over treatment may increase the nocebo effect, whereby unpleasant side effects can be triggered by the treatment context, beyond any inherent physiological effects of the treatment itself. Excessive choice may also increase the nocebo effect. The current studies tested these possibilities.

Participants took part in studies ostensibly investigating the influence of beta-blockers (Study 1, n = 71) and benzodiazepines (Study 2, n = 120) on anxiety. All treatments were placebos. In Study 1, participants were randomly allocated to three groups: no-treatment control, no-choice, and choice between two treatments. In Study 2, a ten-choice group was added. Participants were warned about possible treatment side effects. These warned symptoms were assessed, and scores summed. Nocebo effects were evidenced by significantly higher warned symptoms scores in any placebo-treated group compared to the control group.

In both studies, the no-choice groups experienced a nocebo effect (S1: p = .003, η_p²= 0.121; S2: p = .022, η_p²= 0.045). A significant nocebo effect was not present in groups who chose between two treatments (S1: p = .424, η_p²= 0.009; S2: p = .49, η_p²= 0.004). In Study 2, choosing between ten treatments resulted in a nocebo effect (p = .006, η_p²= 0.065).

Lack of choice resulted in the development of nocebo effects, while having a limited choice between two placebos did not generate significant nocebo effects. However, a larger choice between ten placebos generated a nocebo effect of similar magnitude to lack of choice. Facilitating (some) choice in medical care may reduce the development of nocebo effects, but more extensive choice options may not offer similar benefits.