Text version
Let’s look at a trial of zolpidem for insomnia in patients with suicidal depression. The idea here is when starting an antidepressant in a patient with insomnia, should we also start a hypnotic? Moreover, if we do, can that reduce suicidal ideation more quickly? The authors recruited 1,300 candidates with severe depression in this multicenter study. Then, after a structured interview to exclude patients with bipolar, schizophrenia, substance use problems, and after a sleep study to rule out sleep apnea and then thorough informed consent, they were down to 100 subjects because everyone in this study had suicidal ideation. Everyone was started on an SSRI, mostly fluoxetine. Half of them also received zolpidem, and the other half received a placebo in addition to their antidepressant.
Insomnia improved more with zolpidem than with the placebo, but the target in this study was suicidal ideation. It was measured by self-report with the scale for suicidal ideation. It was also measured by an observer using the Columbia-Suicide Severity Rating Scale, in which patients are asked detailed questions about their thoughts regarding suicide. The difference between these 2 measures matters because, by self-report, zolpidem had no impact on suicidal ideation. However, by the Columbia-Suicide Severity Rating Scale, there was a slightly higher statistically significant reduction in suicidal ideation compared to placebo.
Some other secondary measures were also examined, including hopelessness, quality of life, dysfunctional attitudes towards sleep, and nightmares. Furthermore, for none of these was there a difference between zolpidem and placebo when added to an SSRI. Overall, it looks like the study began with high hopes, and none of them panned out, except for a slight statistical advantage in terms of reduction of suicidal ideation on 1 scale, the more objective scale.
There were some interesting side discoveries. Clinically, it is impressive what happened when they stopped the zolpidem, which was done with no taper. For the next 2 weeks, they followed up by phone and found no worsening of sleep or suicidal ideation in either the placebo or the zolpidem group. Both had those pills stopped abruptly, and then patients were followed for 2 weeks. The zolpidem dose was 6.25 mg to start, but 75% of the subjects were increased to the 12.5 mg CR dose. The insomnia index improved during follow-up despite the discontinuation of zolpidem. The improvement was even more in patients who received zolpidem than those who received the placebo. This result does not suggest that stopping it suddenly left people with sleep problems.
Also, noteworthy is the simple fact of the study’s existence in the first place. All of these patients had suicidal ideation. That was an inclusion criterion, whereas, as the authors point out, usually patients with suicidal ideation are excluded. Based on their experience, no deaths or suicide attempts during this study. The authors conclude, “there are no reasons to systematically exclude suicidal outpatients from psychotropic randomized clinical trials.” That is a forceful statement right there in the American Journal of Psychiatry. One could conclude that this massive study was negative, as 1 of the main outcome measures showed no difference between zolpidem and placebo as add-ons to an antidepressant. Yes, insomnia improved more with zolpidem, but suicidal ideation did not in the main measure. The second measure did show a statistically significant edge for zolpidem, but the effect was small. The results do not spur the addition of zolpidem when starting an SSRI, although the authors point out that the effect of zolpidem on suicidal ideation was greater for those with severe insomnia. Indeed, their data do not entirely rule out the strategy.
As the authors point out, CBT-I—cognitive behavioral therapy for insomnia—is also an effective treatment that might be considered in this context. They chose zolpidem because they hoped for a much more rapid effect. However, given the results, it is hard to argue for zolpidem over CBT-I, if you can get CBT-I for patients. It would be fascinating to see a replication using CBT-I instead, but the massive scale of this multicenter study suggests that it is unlikely.
In summary, adding zolpidem for patients with suicidal ideation and insomnia did not gain a lot relative to adding a placebo, except perhaps for some patients with severe insomnia.
Abstract
Reducing Suicidal Ideation Through Insomnia Treatment (REST-IT): A Randomized Clinical Trial
William V McCall, Ruth M Benca, Peter B Rosenquist, Nagy A Youssef, Laryssa McCloud, Jill C Newman, Doug Case, Meredith E Rumble, Steven T Szabo, Marjorie Phillips, Andrew D Krystal
Objective: The authors sought to determine whether targeted treatment of insomnia with controlled-release zolpidem (zolpidem-CR) in suicidal adults with insomnia would provide a reduction in suicidal ideation superior to placebo.
Methods: Reducing Suicidal Ideation Through Insomnia Treatment was an 8-week three-site double-blind placebo-controlled parallel-group randomized controlled trial of zolpidem-CR hypnotic therapy compared with placebo, in conjunction with an open-label selective serotonin reuptake inhibitor. Participants were medication-free 18- to 65-year-olds with major depressive disorder, insomnia, and suicidal ideation. Suicidal ideation was the main outcome, measured first by the Scale for Suicide Ideation and second by the Columbia-Suicide Severity Rating Scale (C-SSRS).
Results: A total of 103 participants were randomly assigned to receive zolpidem-CR (N=51) or placebo (N=52) (64 women and 39 men; mean age=40.5 years). Zolpidem-CR had a robust anti-insomnia effect, especially in patients with the most severe insomnia symptoms. No significant treatment effect was observed on the Scale for Suicide Ideation (least squares mean estimate=-0.56, SE=0.83, 95% CI=-2.19, 1.08), but the reduction in scores was significantly positively related to improvement in insomnia after accounting for the effect of other depression symptoms. The C-SSRS indicated that zolpidem-CR had a significant treatment effect (least squares mean estimate=-0.26, SE=0.12, 95% CI=-0.50, -0.02). The advantage for zolpidem-CR in reducing suicidal ideation on the C-SSRS was greater in patients with more severe insomnia. No deaths or suicide attempts occurred.
Conclusions: Although the results do not support the routine prescription of hypnotic medication for mitigating suicidal ideation in all depressed outpatients with insomnia, they suggest that coprescription of a hypnotic during initiation of an antidepressant may be beneficial in suicidal outpatients, especially in patients with severe insomnia.
Download PDF and other files
Reference
McCall, W. V., Benca, R. M., Rosenquist, P. B., Youssef, N. A., McCloud, L., Newman, J. C., Case, D., Rumble, M. E., Szabo, S. T., Phillips, M., & Krystal, A. D. (2019). Reducing suicidal ideation through insomnia treatment (REST-IT): A randomized clinical trial. American Journal of Psychiatry, 176(11), 957-965
