TCPR: A lot of patients with PTSD have trouble sleeping. How does this differ from the insomnia we see in other disorders?
Dr. McCarthy: Insomnia is different in PTSD. Some of that is explained by the symptoms of PTSD, such as nightmares, hyperarousal, anxiety, and physical restlessness. But there’s another thing we’re seeing that isn’t as obvious: sleep apnea. This is not the typical sleep apnea where obesity is driving the problem; many of these patients are young and thin.

TCPR: What’s causing this sleep apnea?
Dr. McCarthy: We’re not sure, but it probably has something to do with the autonomic nervous system, which controls the smooth muscles in the airway. PTSD causes hyperarousal, which could affect smooth muscle tone through the autonomic nervous system. There are other signs of autonomic hyperactivity in PTSD, like high blood pressure, high pulse rate, and digestive disturbances.

TCPR: How common is sleep apnea in PTSD?
Dr. McCarthy: The rates range from 40% to 75%, depending on the cutoffs that are used to define it. The cutoffs are important because sleep apnea is part of a spectrum of breathing problems in sleep, from mild upper airway resistance (ie, snoring) to full obstructive sleep apnea (Williams SG et al, Sleep and Breathing 2015;19(1):175–182). So your index of suspicion should be high, even if the patient is young and thin.

TCPR: PTSD has a lot of sleep symptoms: insomnia, nightmares, and sleep apnea. How do these sleep disturbances affect the PTSD itself?
Dr. McCarthy: Many PTSD symptoms are made worse by insomnia. These patients are more irritable, more anxious, and less able to manage daily stressors when they haven’t slept well. And when PTSD is worse, there’s more hyperarousal, which in turn makes the insomnia worse. It’s a vicious cycle.

TCPR: Can PTSD improve by treating sleep?
Dr. McCarthy: Yes, many PTSD symptoms lessen just by restoring sleep. Those vicious cycles are turned down, and good-quality sleep helps patients process trauma. For example, it’s thought that REM sleep helps reshape traumatic memories from disorganized, limbic memories into a more cognitive type of memory.

TCPR: How do you treat insomnia in PTSD?
Dr. McCarthy: You could start by treating PTSD itself, such as with a trauma-focused therapy, and that may improve sleep indirectly. But a new direction is to address sleep first. The thinking is that patients are better able to process, learn, and make use of therapy when they are sleeping better. The treatment team might add in cognitive behavioral therapy for insomnia (CBT-I), address sleep apnea, or consider a hypnotic. A handful of studies have examined these strategies, and so far the results look good (Colvonen PJ et al, Curr Psychiatry Rep 2018;20(7):48).

TCPR: Which sleep medications do you prefer in PTSD?
Dr. McCarthy: My first-line is prazosin, mainly because its mechanism addresses one of the core ­problems in PTSD: autonomic hyperarousal. It’s usually thought of for nightmares, but it tends to be relatively sedating, so it can help sleep as well. Unlike the z-hypnotics and benzodiazepines, prazosin lacks muscle relaxant effects, so it’s safe in sleep apnea. There’s also some indication from the data that prazosin helps with daytime PTSD symptoms. On the other hand, prazosin is not for everyone. It failed to work in a recent controlled trial, which was the largest to date, but that trial had a few flaws. The placebo rate was unusually high, and the investigators may have enriched their sample with patients who were less likely to respond to prazosin (Raskind MA et al, N Engl J Med 2018;378:507–517).

TCPR: How was it enriched that way?
Dr. McCarthy: This was a VA study, which limits the pool of potential prazosin responders because the medication is already widely used there for PTSD. They also excluded patients who responded to trazodone, which has adrenergic effects that are similar to ­prazosin’s. I don’t think this study refutes prazosin’s benefits for nightmares, but it does remind us that PTSD is a complex illness that affects many types of patients.

TCPR: Do benzodiazepines have a role for sleep in PTSD?
Dr. McCarthy: Short-acting benzodiazepines like temazepam are recommended by the Academy of Sleep Medicine for general insomnia, but they are less useful in patients with PTSD. When you’re undergoing psychotherapy, you really need to be able to pull up the anxiety-provoking aspects of the memory in order to work through them, and if you’re taking a benzo, that exposure therapy will be less effective (Rothbaum BO et al, Am J Psych 2014;71(6):640–648). This population is also at risk for alcohol use disorders, which is another reason to avoid benzos, as is sleep apnea. Benzos can worsen apnea by relaxing the smooth muscle; prazosin lacks this risk. On the other hand, benzos aren’t contraindicated in PTSD, and they did have modest benefits for sleep in PTSD according to a meta-analysis.

TCPR: Can you describe the ideal case where you’d consider a benzo in PTSD?
Dr. McCarthy: Maybe someone who has successfully completed cognitive therapy and has a good handle on the primary PTSD symptoms, without an alcohol or substance use disorder, where you’re just targeting residual insomnia and you’ve ruled out sleep apnea.

TCPR: What about the z-hypnotics?
Dr. McCarthy: We don’t have much research on that class in PTSD. Eszopiclone (Lunesta) has one study, and it found improved insomnia and daytime functioning in PTSD without any tolerance or dependence (Pollack MH et al, J Clin Psychiatry 2011;72:892–897). I feel comfortable extrapolating from eszopiclone to zolpidem (Ambien) or zaleplon (Sonata), but there may be differences that haven’t been brought out yet in clinical trials.

TCPR: How do the atypical antipsychotics affect sleep in PTSD?
Dr. McCarthy: Risperidone and olanzapine improved sleep in PTSD in small to medium studies, and the sedating antipsychotic quetiapine (Seroquel) has improved daytime symptoms of PTSD. However, the side effects with this class—particularly metabolic—often outweigh their benefits. Mirtazapine is another one that’s usually best avoided for sleep unless it’s used for comorbid depression. Its sedative effects tend to wear off, and weight gain is a risk with it.

TCPR: Any other hypnotics you’d consider in PTSD?
Dr. McCarthy: Hydroxyzine actually has some data in PTSD, both for sleep and nightmares. In a head-to-head study with placebo and prazosin, it brought mild improvements over placebo but was not as good as prazosin. Trazodone has small studies in PTSD, and while the studies are limited, it may be helpful specifically for reducing nightmares (Warner MD el al, Pharmacopsychiatry 2001;34(4):128–133). However, tolerance can develop with it, and it has pretty well-documented cognitive side effects that can persist into the daytime (Roth AJ et al, J Sleep Res 2011;20:552–558). Other hypnotics have promise, like suvorexant and ramelteon, but haven’t been studied in PTSD.

TCPR: Any thoughts on gabapentin (Neurontin)?
Dr. McCarthy: As with many of these drugs, there’s not a lot of data. A small, retrospective case series suggested it might help PTSD-related nightmares, but the Academy of Sleep Medicine does not generally recommend it for insomnia, except perhaps when a pain disorder is present (Hamner MB et al, Ann Clin Psychiatry 2001;13(3):141–146).

TCPR: Thank you for your time, Dr. McCarthy.