More Relapses for Patients Who Stopped Antidepressants

Adults who continued to take their prescribed antidepressants were less likely to have a relapse of a depressive episode, according to a double-blind, randomized trial.

The U.K. study assigned nearly 500 patients on antidepressants to either placebo (discontinuation) or continued treatment (maintenance). After 52 weeks, the primary outcome — relapse of depression — occurred in 56% of patients in the discontinuation group and 39% of patients in the maintenance group (HR 2.06, 95% CI 1.56-2.70, P<0.001), reported Gemma Lewis, PhD, of University College London, and colleagues.

The percentage of patients who reported feeling worse at 12 weeks than they did at 6 weeks was 44% in the discontinuation group and 21% in the maintenance group (OR 2.88, 95% CI 1.90-4.38), the authors noted in the New England Journal of Medicine.

Secondary outcomes followed a similar trend; researchers used the Patient Health Questionnaire 9-item version (PHQ-9) to assess depressive symptoms, the Generalized Anxiety Disorder Assessment 7-item version (GAD-7) for anxiety symptoms, the modified 14-item Discontinuation-Emergent Signs and Symptoms (DESS) checklist for withdrawal symptoms, and the 12-item Short-Form Health Survey (SF-12) to assess quality of life.

At 12 weeks, the mean score for depressive symptoms was 4.1±3.8 in the maintenance group and 6.3±5.1 in the discontinuation group, for an estimated difference of 2.2 points (95% CI 1.5-2.8). The mean score for anxiety symptoms at the same 12-week point was 3.1±3.3 in the maintenance group and 5.3±4.6 in the discontinuation group, for a difference of 2.4 points (95% CI 1.8-3.0).

The mean score for withdrawal symptoms at 12 weeks was 1.3±2.4 in the maintenance group and 3.1±3.5 in the discontinuation group, for an estimated difference of 1.9 points (95% CI 1.5-2.3), and the mean score for mental health-related quality of life was 46±10 in the maintenance group and 41±11 in the discontinuation group, for an estimated difference of -4.9 points (95% CI -6.4 to -3.3).

For Jeffrey Jackson, MD, MPH, of the Medical College of Wisconsin, the results of this study were significant, disappointing, and unsurprising.

“They confirm what most primary care physicians already knew or intuited,” he wrote in an accompanying editorial. “The frequency of relapse after the discontinuation of treatment is high, particularly among patients with several previous depressive episodes.”

Jackson also highlighted the finding that patients who had three or more previous depressive episodes were more than twice as likely to have a relapse than those who had fewer episodes. This, he said, poses a limitation to the generalizability of these findings for patients who have had only one depressive episode.

After seeking out participants who were being treated at 150 general practices in the U.K., 478 eligible individuals were randomized into two groups: 238 in the maintenance group, and 240 in the discontinuation group.

Eligible participants were adults ages 18 to 74 who had either been on antidepressants for more than 2 years or had reported at least two previous episodes of depression. All patients were on a daily antidepressant regimen — 20 mg of citalopram (Celexa), 100 mg of sertraline (Zoloft), 20 mg of fluoxetine (Prozac), or 30 mg of mirtazapine (Remeron) — for at least 9 months, were recovered from any recent depressive episodes, and reported feeling well enough to consider stopping their medication. At baseline, both groups had similar characteristics: 73% were women, mean age was 54, and 95% were white.

In the maintenance group, 70% of patients adhered to the trial regimen compared with 52% of the discontinuation group.

There were 17 serious adverse events during the study, nine in the maintenance group and eight in the discontinuation group. There were no deaths or suicide attempts.

Lewis and colleagues noted that other significant limitations to their study were the trial’s lack of racial and ethnic diversity, and that only patients who felt ready to discontinue their medication were included in the research.