Placebo, Nocebo Reactions Confound Antidepressant Trials for Persistent Depression

A new meta-analysis found that 31 percent of patients in antidepressant treatment trials for persistent depressive disorder (PDD) responded to placebo and 57 percent reported adverse (nocebo) effects to them.

A new meta-analysis found that 31 percent of patients in antidepressant treatment trials for persistent depressive disorder (PDD) responded to placebo and 57 percent reported adverse (nocebo) effects to them, in numbers comparable to those reported in trials for episodic illness.

"Both placebo and nocebo reactions account for a considerable proportion of the reactions that can be found in active treatment arms of randomized controlled trials," commented Levente Kriston, DPhil, Department of Psychosomatic Medicine and Psychotherapy, University Medical Center, Hamburg, Germany (pictured) and colleagues, "and are therefore highly relevant for estimating the true medication effect."

Although placebo response rate has been previously characterized in trials of antidepressant treatment for episodic major depression, Kriston and colleagues examined the occurrence and contributing factors in treatment trials for conditions persisting for at least two years. The study cohorts included those with continuing mild depressive mood (dysthymia) and chronic major depression.

Twenty-three trials were identified in publications from 1985 through 2013, with sample sizes ranging from 34 to 410. The patient age ranged from 31 to 75 and the proportion of female patients ranged from 37 to 77 percent. Nine trials compared placebo with a selective serotonin reuptake inhibitor (SSRI), three with a tricyclic antidepressant (TCA) and five with another antidepressant. Six trials compared placebo with different antidepressants, using at least two medication treatment arms.

To compare depression symptom severity across trials, ratings from different Hamilton depression scales or observer ratings such as the Montgomery-Asberg Depression Scale (MADRS) were converted to equivalent Hamilton Rating Scale for Depression (HRSD)-17 scores, following published guideline. The baseline depression severity ranged substantially, from 10.2 to 20 points, with remission defined as a 50 percent reduction from baseline score.

The meta-analysis determined a pooled placebo response rate of 31 percent (95% CI=27-35) and a placebo remission rate of 22 percent (95% CI=19-26). The nocebo response was apparent with an adverse event rate of 57 percent (95% CI=48-66) and a 4 percent discontinuation rate due to adverse events to placebo (95% CI=3-7).

Kriston and colleagues reported several factors were associated with higher placebo and nocebo responses. They described a greater likelihood of placebo and nocebo responses among patients with "early onset" illness, although they did not define the corresponding age of onset. Larger trials with greater numbers of patients in treatment arms had higher placebo and nocebo response rates. Patients with less chance of being randomized to placebo (eg. 33 percent chance), however, were less likely to have a placebo response than those with a 50 percent chance.

"We found placebo and nocebo reactions in trials on PDD comparable to those with episodic depression," Kriston and colleagues indicate, which conflicts with the assumption that less placebo and nocebo reactions occur in PDD and show that such reactions should be considered when doing research on and working with patients suffering from PDD."

The meta-analysis of placebo and nocebo responses in antidepressant trials for persistent depressive disorder was published in the June issue of the Journal of Affective Disorders.

Related Coverage

Placebo Predicts Which Patients Respond Best to Depression Drugs

Redefining Major Depressive Episodes and Treatment Guide

Treatment of Major Depressive Disorder Must Include Focus on Residual Symptoms