Hamner et al. [61] retrospectively reviewed the files of 30 patients with PTSD (and 67% MDD) who received gabapentin as adjunctive medication. Gabapentin was mainly prescribed first to facilitate sleep; with a dose of 300–3600 mg/day, the majority (77%) showed moderate or greater improvement in the duration of sleep and a decrease in frequency of nightmares. The improvement appeared to be dose dependent; the group with moderate or marked improvement received 1344 ± 701 mg, and the group with mild or no improvement received 685 ± 227 mg.

There is some evidence that lamotrigine is helpful for PTSD; however, posttraumatic sleep disturbance or sleep difficulties caused by nightmares were not measured in studies such as that performed by Hertzberg et al. [67]. Their study included 15 patients with PTSD enrolled in a 12-week double-blind study of lamotrigine (start 25 mg and titration up to 500 mg if possible) and placebo. One patient dropped out, ten were on lamotrigine, and four on placebo; more patients on lamotrigine responded (50–25%), and patients on lamotrigine improved more on reexperiencing and avoidance/numbing symptoms.

Schneider et al. [144] studied the effect of di-n-propylacetic acid (DPA) (valproic acid) on sleep in 11 healthy volunteers. After short-term application (2 days), a shortening of the time to fall asleep and of the waking time was found, whereas under long-term administration (2 weeks), a decrease in deep synchronous sleep was observed. No marked influence on REM sleep was observed. Subjective sleep experiences did not change. Hamner et al. [62] performed a placebo-controlled study with divalproex in 29 chronic PTSD patients of whom most used other medication (antidepressants, anxiolytics); the authors found no difference compared with placebo concerning the sleep-related measures and even found a decrease in avoidance/numbing scores and improvement in the Clinical Global Impression Scale favoring placebo.

Tiagabine is a GABAergic anticonvulsant. Mathias et al. [98] conducted a double-blind, placebo-controlled study of a single oral dose of 5 mg tiagabine on nocturnal sleep in ten healthy elderly volunteers (mean age 68 years) and found tiagabine to increase sleep efficiency, tendentially decreased wakefulness, and prominently increased both SWS and low-frequency activity in the EEG within non-REM sleep. Except for self-rated sleep intensity, there was no significant change in subjectively assessed sleep parameters nor perceived state upon awakening. Taylor [160] did a case series on seven patients with PTSD and reported positive findings on sleep disturbance, especially nightmares; the mean dose was 8 mg tiagabine . Connor et al. [34] performed an open-label tiagabine for 12 weeks with tiagabine (initiated at 2 mg bid and up to 16 mg daily max, mean dose 10.8 mg) in 29 outpatients with PTSD; those who showed significant improvement (n = 19) continued with placebo or tiagabine , and there was a greater trend toward a likelihood of further remission but no significant differences. Distressing dreams, nightmares, and PSQI improved significantly during the open-label part, but, in the placebo-controlled phase, the improvements stayed the same in both the placebo and tiagabine group. Walsh et al. [168], in a placebo-controlled, randomized, double-blind, parallel-group study in insomnia patients (n = 232) with different dosages of tiagabine, found a significant dose-dependent increase of slow-wave sleep and a decrease in stage 1 sleep but no change in WASO, sleep latency, or total sleep compared to placebo.

Davidson et al. [43] performed a 12-week, double-blind, randomized, multicenter study and included 232 patients with PTSD (tiagabine and placebo; both n = 116). There was no difference in change by tiagabine (mean dosage 11.2 mg/day) and placebo concerning PTSD symptoms (CAPS) (p = 0.85) baseline vs. final visit. There were no differences in sleep ratings between the tiagabine and placebo groups. Krystal et al. [83] performed an open-label 3-week study on 20 adults with PTSD who took 2–12 mg tiagabine daily (two times) with polysomnography measures. Contrary to the findings of Davidson et al. [58], there appeared to be an improvement in different sleep parameters such as the WASO (effect size 0.49; p0.033) and nightmares (e.g., PSQI item 5 h gave an effect size of 0.44; p = 0.008).

They also concluded that the first treatment night predicted PTSD response at 3 weeks. A decrease in self-reported and objective time awake after onset of sleep and an increase of SWS accounted for 94% of the week 3 PTSD score. More important, they found positive and significant correlations between changes in sleep parameters and total PTSD scores, such as WASO with PSG (r = 0.6; p < 0.001) or self-reported (r = 0.82; p < 0.001). This means that improvements in sleep can lead to improvements in PTSD symptoms.

In several case and open-label studies on topiramate [13–15], some effect of topiramate on PTSD was found, especially on sleep problems . The study of Berlant [14] was an open-label study with topiramate as monotherapy or augmentation (median dosage 50 mg/day); this led to a decline of PTSD symptoms in median 9 days, and 94% of the patients with nightmares reported full cessation after 4 weeks. Tucker et al. [162] conducted a double-blind, placebo-controlled study on topiramate monotherapy (median final dose 150 mg/day) in 38 patients with PTSD; the authors found no significant reduction in the CAPS score, but there was a significant decrease in reexperiencing symptoms and the treatment outcome PTSD scale. No separate sleep measures were reported.

Alderman et al. [5] performed an 8-week open-label pilot study of topiramate in 43 patients with PTSD; they found reductions in several scales and a significant reduction of PTSD symptoms (CAPS). The Stanford sleepiness scale tendentially improved (p = 0.08). There was a significant reduction in nightmares and a reduction in the number of patients who were anxious to fall asleep and the number of patients with high-risk drinking patterns.

Kinrys et al. [78] performed a retrospective chart review on 40 patients who have taken 9.3 weeks (sd 5.1) adjunctive levetiracetam . There was improvement, but change of severity was measured by use of clinical global inventory (CGI) only, and no information about sleep or nightmares was given.

Hindmarch et al. [71] performed a randomized, double-blind, placebo- and active-controlled, 3-way crossover study with 24 volunteers (23 completers) who took pregabalin 150 mg t.i.d., alprazolam 1 mg t.i.d., and placebo t.i.d. for 3 days. Pregabalin increased slow-wave sleep and reduced sleep latency. REM sleep was reduced, but REM sleep latency appeared to be equal to placebo. Pregabalin reduced the number of awakenings. Subjective sleep improved, but ratings of behavior after awakening showed impairments.

Strawn et al. [159] wrote a case report about a patient with PTSD who took 75 mg b.i.d. in addition to her other medication; her nightmares stopped and insomnia improved within 2 weeks. Pae et al. [120] conducted an open-label study with nine patients with PTSD who were on stable doses of antidepressants. They were treated with flexibly dosed pregabalin [mean dose 200 mg/day (range 150–300 mg/day)] for 6 weeks and improved on PTSD complaints; however, but no sleep measures were made. Paslakis et al. [122] published a case report about a patient with PTSD who previously used different kinds of medication but improved on pregabalin with later addition of quetiapine because of an additional bipolar disorder. The patient reported improvement in sleep and nightmares within the first week.

Risperidone was used in a study which consisted of two parts [148]. The first was a placebo-controlled, double-blind, crossover study on eight volunteers who took risperidone 1 mg for one night; this resulted in a significantly decreased REM sleep. The second part was an open-label study on eight patients with MDD who did not respond to a therapeutic dose of antidepressants and received 2 weeks of risperidone 0.5–1 mg (final mean 0.7 mg) daily. The depressed patients had significantly less wake and REM sleep as well as a significant decline in depressive symptoms.

There are a few case reports about the positive outcome when using risperidone for PTSD, e.g., the study of Leyba and Wampler [95]. Stanovic et al. [155] performed a retrospective chart study in acutely burned hospitalized patients with distressing acute stress symptoms that probably would not respond to brief psychotherapeutic interventions. Patients received 0.5–2 mg risperidone (mean 1 mg) at bedtime and had less sleep disturbances, nightmares/flashbacks, and hyperarousal. David et al. [36] conducted an open-label study of flexible dose adjunctive risperidone (1–3 mg) in patients who partially responded to medication (different use of AD, AED, and anxiolytics). A total of 17 Vietnam veterans completed at least 6 weeks and showed improvement of sleep disturbance measured by self-report sleep measures. Less awakenings and reduction in trauma-related dreams (CAPS item) were found, but improvement was especially found in sleep log data and not in retrospective scales, such as the PSQI .

Rothbaum et al. [138] performed a randomized augmentation study with risperidone and sertraline . The patients who did not remit during 8 weeks treatment with open-label sertraline received risperidone or placebo for 16 weeks. Of the 45 patients that started, 34 completed the sertraline part and 25 went on to the second part, of whom 20 completed. All patients improved, with no group differences. Post hoc analyses showed that the group that received risperidone improved more on the sleep item of the Davidson Trauma Scale (p = 0.03) and the sleep item of the CAPS Scale (trend p = 0.09). Krystal et al. [85] conducted a 6-month, randomized, double-blind, placebo-controlled multicenter trial with 367 screened patients, of which 296 were diagnosed with military-related PTSD and ongoing symptoms after at least two SRI treatments. Risperidone (up to 4 mg) or placebo was given to 247, and there was no significant change in total CAPS score. Post hoc analyses [86] showed a significant but small reduction in reexperiencing and hyperarousal symptoms, perhaps clinically not detectable. Risperidone use resulted in more adverse events than placebo, such as weight gain, fatigue, and somnolence. Sleep was measured by means of the PSQI, and risperidone provided some improvement (p = 0.034) as well as less severe nightmares (measured by means of CAPS item and p = 0.033). The improvements in sleep correlated with PTSD symptom reductions as a whole (measured by the CAPS; r-0.28, p = 0.001) and improvement in general mental health measured by means of the SF-36 V subscale (r = 0.26, p = 0.003). The scores became significantly different from placebo in week 24; before that time, the differences were not significant, and the positive effect was not seen in measures for global clinical status and general measures of quality of life. This gave the authors reason to conclude that the results have some, but limited, clinical significance.

Sharpley et al. [149] studied olanzapine and its effects on sleep when used as augmentation to ineffective treatment with SSRIs in depression . In an open trial, 12 patients with SSRI-resistant depressive disorder took 2.5–10 mg olanzapine (mean 4.8 mg) for 3 weeks; the depressive symptoms decreased, while sleep efficiency increased, as did the subjective sleep quality and SWS. In general, the improvements occurred after the first week.

Labbate and Douglas [88] and Jakovljević et al. [75] reported on patients with PTSD given olanzapine 5–20 mg as augmentation; they described improvements in symptoms (e.g., nightmares and sleep disturbance) starting after 1–4 days. Stein et al. [156] performed a double-blind, placebo-controlled study with augmentation of olanzapine (15 mg/day) or placebo in 19 patients with PTSD who hardly responded to an SSRI; the authors found a significant clinical improvement of PTSD and depressive and sleep measures. The enhanced sleep accounted for much of the reported improvement. They concluded that the overall clinical magnitude of effects was modest for most of the patients but clinically meaningful for some. Carey et al. [28] performed a randomized 8-week placebo-controlled study with olanzapine (flexible dosage; ending at mean 9.3 mg/day) as monotherapy in 34 patients with PTSD (ten dropped out, and four of them were included in the analysis) and found more improvement on the CAPS compared with placebo. There were no special measurements on insomnia or nightmares.