May 2009  << Back  

  Can't open the PDF? Click here for help.

 

 RESEARCH ARTICLE

A preliminary controlled trial of divalproex in posttraumatic stress disorder

Mark B. Hamner, MD

Mental Health Service, Ralph H. Johnson VA Medical Center, Department of Psychiatry, Medical University of South Carolina, Charleston, SC, USA

Richard A. Faldowski, PhD

Department of Psychiatry, Medical University of South Carolina, Charleston, SC, USA

Sophie Robert, PharmD

Mental Health Service, Ralph H. Johnson VA Medical Center, Department of Psychiatry, Medical University of South Carolina, Charleston, SC, USA

Helen G. Ulmer, MSN

Mental Health Service, Ralph H. Johnson VA Medical Center, Department of Psychiatry, Medical University of South Carolina, Charleston, SC, USA

Michael David Horner, PhD

Mental Health Service, Ralph H. Johnson VA Medical Center, Department of Psychiatry, Medical University of South Carolina, Charleston, SC, USA

Jeffrey P. Lorberbaum, MD

Department of Psychiatry, Penn State University, Hershey Medical Center, Hershey, PA, USA

BACKGROUND: Case reports and open trials have reported beneficial effects of divalproex in the treatment of posttraumatic stress disorder (PTSD). The objective of this study was to conduct a placebo-controlled study of the efficacy and tolerability of divalproex in chronic PTSD patients.

METHODS: Patients were randomized to receive placebo or divalproex. The primary outcome measure was the Clinician Administered PTSD Scale (CAPS).

RESULTS: Of 29 patients randomized, 16 received divalproex and 13 placebo. There were no significant differences between groups in mean change from baseline to end point (last observation carried forward) on the CAPS total score or subscales except for a significant decrease in avoidance/numbing scores with placebo. The only significant difference in secondary outcomes was a greater improvement in Clinical Global Impression Scale–Severity favoring placebo.

CONCLUSIONS: Divalproex was not superior to placebo in this study. This could be due to lack of efficacy of divalproex in this population, inadequate sample size to detect differences, or other factors. Further study of divalproex is needed to better clarify the role of this agent in PTSD.

KEYWORDS: posttraumatic stress disorder, divalproex, anticonvulsant, mood stabilizer, psychopharmacology

ANNALS OF CLINICAL PSYCHIATRY 2009;21(2):89–94

  INTRODUCTION

The mainstay of medication treatment for posttraumatic stress disorder (PTSD) is antidepressant medication. Two antidepressants, sertraline and paroxetine, are now FDA-indicated for PTSD.1-3 However, many patients do not have an adequate response to antidepressants alone. Evidence suggests that combat veterans with PTSD are not as likely to benefit from antidepressants as are civilian trauma victims. Combat PTSD is also associated with a high rate of comorbidity and a chronic course.3 Consequently, many clinicians consider a variety of augmenting agents, including anticonvulsant mood stabilizers, atypical antipsychotics, anxiolytic or hypnotic agents, antiadrenergic agents, and others.4,5 Good clinical practice also dictates that psychotherapy be optimized for PTSD patients.

Although anticonvulsant medications, most with mood stabilizing properties, are frequently used in this population, few data support this practice. There has been one small, positive controlled trial of lamotrigine in PTSD.6 Recent controlled trials with topiramate and tiagabine failed to confirm benefits observed with these agents in open-label studies.7-14 To our knowledge, reports on other agents such as carbamazepine, phenytoin, gabapentin, and levetiracetam have involved case reports or open trials.15-19

Both gamma-aminobutyric acid (GABA) and glutamate/N-methyl-D aspartate (NMDA) alterations have been implicated in preclinical and clinical studies of PTSD.14,20-23 Divalproex may affect several neurotransmitter systems including facilitation of GABA neuro-transmission and inhibition of glutaminergic neuro-transmission.24,25 The agent also induces neurogenesis of GABAergic neurons in preclinical studies.26

Several case reports, open trials, and a retrospective chart review have evaluated the efficacy of valproate in PTSD.27-33 Fesler29 studied 16 combat veterans with PTSD treated with valproate and noted significant improvement in 10 of the 16, particularly in hyperarousal and hyperreactivity symptoms. Clark and colleagues30 also studied 16 combat veterans treated with divalproex in an 8-week open trial. They used the Clinician Administered PTSD Scale (CAPS) as the primary outcome measure and noted significant improvement in intrusion and hyper-arousal symptoms but not in avoidance/numbing symptoms. Interestingly, Hamilton Anxiety scale (HAM-A) and Hamilton Depression scale (HAM-D) scores also demonstrated significant improvement. Petty and colleagues31 conducted an 8-week open trial of divalproex in 21 combat veterans with PTSD. They noted improvement in total CAPS ratings as well as in HAM-D and HAM-A scores. Davis and colleagues33 conducted a retrospective chart review to evaluate the effects of divalproex in a larger naturalistic sample of veterans with PTSD. A rater blinded to the order of study visits, medication doses, and serum valproate levels rated the progress notes of 50 patients with the Clinical Global Impression–Severity (CGI-S) and Clinical Global Impression–Improvement (CGI-I) scales. Twenty-five (50%) were rated as very much or much improved on the CGI-I.

One study of valproate monotherapy has been conducted in a civilian population. Otte et al32 studied 10 patients (6 male, 4 female) in an open trial of 8 weeks. They noted no significant change in the PTSD Diagnostic Scale (PDS), Impact of Events Scale (IES), or Beck Depression Inventory (BDI) at the end of the study. They noted the use of self-rating instruments, small sample size, high attrition rate, and low valproate concentrations at the end of the study (62 mcg/mL vs 69 to 80 mcg/mL in earlier studies) as limitations. Attrition in this and other studies merits comment: although only 5 patients completed the trial, this is not out of line with attrition rates of 22% to 53% observed in earlier studies.30,31

Two controlled studies of valproate in PTSD have been published. Hollander and colleagues34 evaluated the efficacy of valproate in a 12-week study in the treatment of impulsive aggression in patients with cluster B personality disorder, intermittent explosive disorder, and PTSD. The majority of patients were male. A small proportion of them were receiving concomitant psychiatric medications (antidepressants and/or hypnotics). Most traumas were related to physical or sexual abuse. No significant treatment differences between divalproex and placebo were noted on the Overt Aggression Scale–Modified (OAS-M) in the subset of patients with PTSD. The large improvement (a 69% decrease in OAS-M scores) noted with placebo may have offset any potential differences. Additionally, average dose and plasma levels were not provided for that subset, so it difficult to evaluate the adequacy of the therapeutic trial.

Davis and colleagues35 evaluated the efficacy of divalproex in 85 veterans with PTSD with a special focus on hyperarousal symptoms. Eighty-two patients were included in the efficacy evaluations in this randomized, placebo-controlled 8-week trial. There were no significant differences in primary or secondary outcome measures between patients randomized to placebo vs divalproex. The final mean dose of divalproex was 2309 ± 507 mg/d, with mean levels of 82 ± 30 mg/L. The authors commented that further studies were warranted, in particular those investigating combination therapies with divalproex.

To our knowledge, the following study is the first controlled study of divalproex as an adjunctive therapy in combat veterans with PTSD.

  METHODS

Study design and subjects

A 10-week randomized, double-blind, placebo-controlled trial was conducted with patients from the Ralph H. Johnson VA Medical Center in Charleston, SC, who met DSM-IV36 criteria for PTSD based on the Structured Clinical Interview for DSM-IV37 and the 1-month diagnostic version of the Clinician Administered PTSD Scale (CAPS-1).38 Patients were eligible for the study if they were between age 18 and 65, were competent to give informed consent, and had a score ≥50 on the CAPS-1. Exclusion criteria included: history of hypersensitivity to divalproex; unstable medical conditions or medical contraindications to the safe administration of divalproex; alcohol or substance use disorders within 1 month of study entry; diagnoses of schizophrenia, schizoaffective disorder, or bipolar disorder; or change in psychotropic medication within 4 weeks of randomization. The protocol was approved by the Institutional Review Board, and all study-related procedures were performed after patients signed informed consent.

A comprehensive medical and psychiatric history as well as urine drug screen were performed at screening, as well as a physical and neurological examination if indicated by medical history or if none had been performed in the last year. Laboratory tests and ECG were also performed if none had been performed within 6 months of study entry.

Following a 1-week single-blind placebo lead-in, eligible patients were randomized to divalproex or placebo. Clinical evaluation and tolerability assessment were performed weekly for the first 2 weeks, then biweekly until the end of the study.

Divalproex was initiated at 750 mg/d in divided doses and increased by 250 mg every 3 to 4 days, if tolerated, until positive clinical response was achieved or desired trough levels were reached. Blood draws for valproate levels were obtained at the end of weeks 2, 6, and 10 from patients in both groups. A pharmacist not involved in clinical ratings monitored the valproate levels and instructed the study physician to adjust the dose to achieve trough levels between 50 and 125 mcg/mL unless the patient had a positive clinical response at a lower dose. Concomitant psychiatric medications (except for mood stabilizers) were permitted if they were maintained at a stable dose for at least 1 month prior to enrollment.

Efficacy measures

The primary outcome variable was the change from baseline to end point in global CAPS scores, 1-week symptom version (CAPS-2).38 Secondary efficacy variables included the IES,39 the CGI-I,40 the HAM-D,41 the Leibowitz Social Anxiety Scale (LSAS),42 the Pittsburgh Sleep Quality Index (PSQI),43 and the PSQI Addendum for PTSD (PSQI-A).44

Safety and tolerability

Adverse events and vital signs were evaluated at each visit. Blood samples were drawn 12 hours after the evening dose of divalproex at weeks 2, 6, and 10 to measure trough valproate levels. Liver function tests were obtained at screening and weeks 2, 6, and 10, and complete blood cell counts were done at screening and week 10.

Statistical analyses

Because of the small sample sizes and likelihood of non-normal error distributions on primary outcome variables, exact nonparametric statistical methods were used for all analyses. These included exact Wilcoxon and Kruskal-Wallis tests for continuous measures, and Fisher exact and Fisher-Freeman-Halton exact tests for categorical variables. All statistical tests were 2-tailed, and P values of .05 were considered statistically significant.

  RESULTS

This study enrolled 29 patients; 13 were randomized to placebo and 16 to divalproex. One patient in the latter group did not have at least one postrandomization visit for efficacy assessments, thus the analysis sample consisted of 15 in the divalproex and 13 in the placebo groups.

The baseline characteristics of subject participants were similar in both groups. All were male except for one female, and the mean age was 52.6 ± 7.9 and 52.1 ± 5.7 years in the divalproex and placebo groups, respectively. Most patients had a comorbid Axis I diagnosis, the most common being major depression (TABLE 1). The traumatic events included combat (n = 28) and sexual assault (n = 1). Most patients were receiving an antidepressant medication, namely a selective serotonin reuptake inhibitor. Several patients were receiving low doses of tricyclic antidepressants and/or trazodone (TABLE 1). Two patients in the divalproex group and one in the placebo group received no other psychiatric medicines.

There were no statistically significant differences between groups in change from baseline to end point in any of the efficacy measures, except for the re-experiencing symptoms (TABLE 2). Mean change scores on the CAPS-C improved significantly for placebo (-6.31 ± 9.32) vs divalproex (1.07 ± 4.88), exact P = .035. The CGI severity of illness also improved significantly for placebo (-0.38 ± 0.65) vs divalproex (0.13 ± 0.35), exact P = .01.


TABLE 1

Baseline characteristics of patients randomized to divalproex or placebo

  Divalproex (n = 16) Placebo (n = 13)
Age, years (SD) 52.6 (7.9) 52.1 (5.7)
Sex, male 15 13
Race    
  Caucasian 15 11
  African-American 1 1
Marital status    
  Married 9 6
  Divorced 4 5
  Never married 0 2
  Widowed 1 0
  Unknown 2 0
Comorbid Axis I    
  Major depression 12 8
  Other anxiety disorder 5 2
History of substance use disorder 3 2
Concomitant psychiatric medications    
  SSRI 10 7
  Nefazodone 2 0
  Venlafaxine 1 0
  Mirtazapine 1 1
  Bupropion 2 3
  Trazodone 5 2
  Tricyclic antidepressants 2 4
  Benzodiazepines 3 4
  Zolpidem 2 1
  Other 0 2
SSRI: selective serotonin reuptake inhibitor.

The mean daily dose was 1196 ± 246 mg (range, 500 to 1500 mg) for divalproex. The mean final daily dose was 1250 mg ± 327 mg. The mean trough valproate plasma level was 64.8 ± 36.2 mcg/mL (range, 0.23 to 144.0 mcg/mL). There were no significant correlations between valproate plasma levels and improvement on any of the rating scales.

The proportion of patients who completed the study was slightly higher in the divalproex group (n = 9 [56.3%]) vs placebo (n = 6 [46.2%]), although the difference was not statistically significant. Early protocol terminations occurred for 7 patients taking divalproex (2 for lack of efficacy; 1 for increased blood pressure; 1 for unsteady gait; 1 for nausea, diarrhea, and weakness; 1 for noncompliance with protocol; and 1 was lost to follow-up) and for 7 patients taking placebo (3 for lack of efficacy; 1 for lethargy; 1 due to an unstable medical condition; 1 for lack of time; and 1 was lost to follow-up). Study completers and study dropouts did not statistically differ on any background characteristic, suggesting an absence of bias among completers compared to those with attrition.

Divalproex was generally well tolerated. The most common side effects were mild and included dizziness (n = 10, 62.5%), fatigue (n = 6, 37.5%), tremors (n = 5, 31.3%), somnolence (n = 4, 25%), dyspepsia (n = 4, 25%), and diarrhea (n = 6, 37.5%). The most common side effects observed with placebo included fatigue (n = 4, 30.8%), insomnia (n = 4, 30.8%), headache (n = 4, 30.8%), and dry mouth (n = 3, 23.1%). There were no unexpected or serious adverse events and no clinically significant changes in vital signs or weight.


TABLE 2

Efficacy outcomes for patients randomized to divalproex or placebo

Variable Divalproex (n = 15) Placebo (n = 13) Exact P value
CAPS total      
  Baseline 76.67 (23.80) 77.62 (21.91)  
  Change 1.67 (15.96) -10.15 (23.88) .20
CAPS-B      
  Baseline 16.27 (8.02) 14.23 (7.92)  
  Change 0.80 (8.72) -2.46 (8.66) .39
CAPS-C      
  Baseline 32.87 (10.20) 36.46 (8.40)  
  Change 1.07 (4.88) -6.31 (9.32) .035
CAPS-D      
  Baseline 27.53 (11.48) 26.92 (11.20)  
  Change -0.20 (6.89) -1.38 (10.98) .67
CGI-I      
  Endpoint 3.47 (1.25) 3.23 (1.30) .75
CGI-S      
  Baseline 4.00 (0.53) 4.00 (0.41)  
  Change 0.13 (0.35) -0.38 (0.65) .01
HAM-D      
  Baseline 20.91 (8.95) 19.69 (8.37)  
  Change -2.19 (8.34) -0.08 (7.10) .56
IES      
  Baseline 44.33 (19.49) 42.42 (14.48)  
  Change -2.26 (13.46) -10.08 (16.27) .17
LSAS      
  Baseline 48.69 (33.62) 82.92 (11.57)  
  Change 15.40 (34.18) -0.77 (20.07) .78
PSQI      
  Baseline 13.20 (4.61) 10.30 (4.50)  
  Change 0.20 (3.05) -0.50 (2.88) .72
PSQI-A      
  Baseline 9.92 (4.65) 7.67 (4.64)  
  Change 0.31 (3.43) -0.67 (4.38) .70
CAPS: Clinician Administered PTSD Scale; CAPS-B: CAPS re-experiencing symptoms; CAPS-C: CAPS avoidance/numbing symptoms; CAPS-D, CAPS hyperarousal symptoms; CGI-I: Clinical Global Impression-Improvement scale; CGI-S: Clinical Global Impression-Severity scale; HAM-D: Hamilton Depression scale; IES: Impact of Events Scale; LSAS: Leibowitz Social Anxiety Scale; PSQI: Pittsburgh Sleep Quality Index: PSQI-A, Pittsburgh Sleep Quality Index Addendum for PTSD; PTSD: posttraumatic stress disorder.
Baseline and change scores are mean (SD).

  DISCUSSION

There were no significant differences in composite CAPS ratings (the primary efficacy variable) or in subscale scores between divalproex or placebo-treated patients, except for the difference observed on the CAPS C subscale (avoidance/numbing) favoring placebo. This may have been due to lack of efficacy of divalproex in PTSD, or at least in this more treatment-refractory population. Other factors that limit interpretation of the study include small sample size (low power to detect differences); relatively high attrition rates, although comparable between groups and similar in magnitude to that observed in other studies30,31; inadequate plasma levels of divalproex; or other factors such as nonadherence.

The patients enrolled in this study had chronic, severe, and relatively treatment-refractory PTSD symptoms as reflected by their high mean baseline CAPS scores, despite current pharmacologic treatment. This needs special attention in future studies because, although this is a population that is more likely to receive adjunctive medications in current clinical practice, it also is more difficult to ascertain treatment effects. Future controlled studies should consider, in addition to a much larger sample size, inclusion of new treatment-seeking patients and patients who have responded historically to other agents. In light of the low therapeutic levels seen in some patients in this study, future work should also consider attempting to achieve higher average divalproex levels. Furthermore, since there is a higher risk for bipolar disorder in PTSD patients,45 studies investigating the response of PTSD comorbid with bipolar disorder would be of interest. Future studies should also consider including primary and secondary outcome measures, eg, measures of aggression or irritability, that could help better define symptom characteristics of PTSD patients who may respond to divalproex or other anticonvulsant agents.

DISCLOSURE: Dr. Hamner receives grant support from Abbott Laboratories, AstraZeneca, and Otsuka Pharmaceutical and owns stock in Pfizer Inc and Merck. Drs. Faldowski, Robert, Horner, and Lorberbaum and Ms. Ulmer report no financial relationship with any company whose products are mentioned in this article or with manufacturers of competing products.

ACKNOWLEDGMENTS: This study was funded by Abbott Laboratories (investigator-initiated protocol) and was conducted at the Ralph H. Johnson VA Medical Center, Charleston, SC. The results were presented in part at the 44th Annual New Clinical Drug Evaluation Unit, June 1-4, 2004; Phoenix, AZ. We are grateful to Charlotte Teneback, MD, Jennifer Mixson, RPh, and Deborah Agbor-Tabi, MPH, for their technical support, and to the veterans who participated in this study.

    REFERENCES

  1.  American Psychiatric Association. Practice guidelines for the treatment of patients with acute and chronic posttraumatic stress disorder. Arlington, VA: American Psychiatric Association; 2004.
  2.  PTSD Treatment Guidelines Task Force. Guidelines for treatment of PTSD. J Trauma Stress. 2000;13:539–588.
  3.  Veterans Administration/Department of Defense Clinical Practice Guideline Working Group. Management of post-traumatic stress. Washington, DC: Veterans Health Administration, Department of Veterans Affairs and Health Affairs, Department of Defense; 2003. Office of Quality and Performance publication 10Q-CPG/PTSD-04.
  4. Ballenger JC, Davidson JR, Lecrubier Y, et al. Consensus statement update on posttraumatic stress disorder from the International Consensus Group on Depression and Anxiety. J Clin Psychiatry. 2004;65(suppl 1):55–62.
  5.  Treatment of posttraumatic stress disorder. Expert Consensus Panels for PTSD. J Clin Psychiatry. 1999;60(suppl 16):3–76.
  6. Hertzberg MA, Butterfield MI, Feldman ME, et al. A preliminary study of lamotrigine for the treatment of posttraumatic stress disorder. Biol Psychiatry. 1999;45:1226–1229.
  7. Tucker P, Trautman RP, Wyatt DB, et al. Efficacy and safety of topiramate monotherapy in civilian post-traumatic stress disorder: a randomized, double-blind, placebo-controlled study. J Clin Psychiatry. 2007;68:201–206.
  8. Lindley SE, Carlson EB, Hill K. A randomized, double-blind, placebo-controlled trial of augmentation topiramate for chronic combat-related posttraumatic stress disorder. J Clin Psychopharmacol. 2007;27:677–681.
  9. Davidson JR, Brady K, Mellman TA, et al. The efficacy and tolerability of tiagabine in adult patients with post-traumatic stress disorder. J Clin Psychopharmacol. 2007;27:85–88.
  10. Berlant JL. Topiramate in posttraumatic stress disorder: preliminary clinical observations. J Clin Psychiatry. 2001;62(suppl 17):60–63.
  11. Berlant J, van Kammen DP. Open-label topiramate as primary or adjunctive therapy in chronic civilian posttraumatic stress disorder: a preliminary report. J Clin Psychiatry. 2002;63:15–20.
  12. Berlant JL. Prospective open-label study of add-on and monotherapy topiramate in civilians with chronic nonhallucinatory posttraumatic stress disorder. BMC Psychiatry. 2004;4:24.
  13. Taylor FB. Tiagabine for posttraumatic stress disorder: a case series of 7 women. J Clin Psychiatry. 2003;64:1421–1425.
  14. Connor KM, Davidson JR, Weisler RH, et al. Tiagabine for posttraumatic stress disorder: effects of open-label and double-blind discontinuation treatment. Psychopharmacology (Berl). 2006;184:21–25.
  15. Lipper S, Davidson JR, Grady TA, et al. Preliminary study of carbamazepine in posttraumatic stress disorder. Psychosomatics. 1986;27:849–854.
  16. Wolf ME, Alavi A, Mosnaim AD. Posttraumatic stress disorder in Vietnam veterans: clinical and EEG findings, possible therapeutic effects of carbamazepine. Biol Psychiatry. 1988;23:642–644.
  17. Bremner JD, Mletzko T, Welter S, et al. Treatment of posttraumatic stress disorder with phenytoin: an open-label pilot study. J Clin Psychiatry. 2004;65:1559–1564.
  18. Hamner MB, Brodrick PS, Labbate LA. Gabapentin in PTSD: a retrospective clinical series of adjunctive therapy. Ann Clin Psychiatry. 2001;113:141–146.
  19. Kinrys G, Wygant LE, Pardo TB, et al. Levetiracetam for treatment-refractory posttraumatic stress disorder. J Clin Psychiatry. 2006;67:211–214.
  20. Harvey BH, Oosthuizen F, Brand L, et al. Stress-restress evokes sustained iNOS activity and altered GABA levels and NMDA receptors in rat hippocampus. Psychopharmacology (Berl). 2004;175:494–502.
  21. Harvey BH, Bothma T, Nel A, et al. Involvement of the NMDA receptor, NO-cyclic GMP, and nuclear factor K-beta in an animal model of repeated trauma. Human Psychopharmacology. 2005;20:367–373.
  22. Chambers RA, Bremner JD, Moghaddam B, et al. Glutamate and post-traumatic stress disorder: toward a psychobiology of dissociation. Semin Clin Neuropsychiatry. 1999;4:274–281.
  23. Vaiva G, Thomas P, Ducrocq F, et al. Low posttrauma GABA plasma levels as a predictive factor in the development of acute post-traumatic stress disorder. Biol Psychiatry. 2004;55:250–254.
  24. Cunningham MO, Woodhall GL, Jones RS. Valproate modifies spontaneous excitation and inhibition at cortical synapses in vitro. Neuropharmacology. 2003;45:907–917.
  25. Gobbi G, Janiri L. Sodium- and magnesium-valproate in vivo modulate glutaminergic and GABAergic synapses in the medial prefrontal cortex. Psychopharmacology (Berl). 2006;185:255–262.
  26. Laeng P, Pitts RL, Lemire AL, et al. The mood stabilizer valproic acid stimulates GABA neurogenesis from rat forebrain stem cells. J Neurochem. 2004;91:238–251.
  27. Berigan TR, Holzgang A. Valproate as an alternative in post-traumatic stress disorder: a case report. Mil Med. 1995;160:318.
  28. Szymanski HV, Olympia J. Divalproex in post-traumatic stress disorder. Am J Psychiatry. 1991;148:1086–1087.
  29. Fesler FA. Valproate in combat-related post-traumatic stress disorder. J Clin Psychiatry. 1991;52:361–364.
  30. Clark RD, Canive JM, Calais LA, et al. Divalproex in posttraumatic stress disorder: an open-label clinical trial. J Trauma Stress. 1999;12:395–401.
  31. Petty F, Davis LL, Nugent AL, et al. Valproate therapy for chronic, combat-induced posttraumatic stress disorder. J Clin Psychopharmacol. 2002;22:100–101.
  32. Otte C, Wiedemann K, Yassouridis A, et al. Valproate monotherapy in the treatment of civilian with non-combat-related posttraumatic stress disorder: an open-label study. J Clin Psychopharmacol. 2004;24:106–108.
  33. Davis LL, Ambrose SM, Newell JM, et al. Divalproex for the treatment of posttraumatic stress disorder: a retrospective chart review. International Journal of Psychiatry in Clinical Practice. 2005;9:278–283.
  34. Hollander E, Tracy KA, Swann AC, et al. Divalproex in the treatment of impulsive aggression: efficacy in cluster B personality disorders. Neuropsychopharmacology. 2003;28:1186–1197.
  35. Davis LL, Davidson JR, Ward LC, et al. Divalproex in the treatment of post-traumatic stress disorder: a randomized, double-blind, placebo-controlled trial in a veteran population. J Clin Psychopharmacol. 2008;28:84–88.
  36.  Diagnostic and statistical manual of mental disorders. 4th edition. Washington, DC: American Psychiatric Association; 1994.
  37. Spitzer RL, Williams JBW, Gibbon M, et al. Structured clinical interview for DSM-IV (SCID-IV). New York, NY: Biometric Research, New York State Psychiatric Institute; 1994.
  38. Blake DD, Weathers FW, Nagy LM, et al. A clinician rating scale for assessing current and lifetime PTSD: the CAPS-1. Behavior Therapist. 1990;18:187–188.
  39. Horowitz M, Wilner N, Alvarez W. Impact of Event Scale: a measure of subjective stress. Psychosom Med. 1979;41:209–218.
  40. Guy W. ECDEU Assessment manual for psychopharmacology. Rev ed. Washington, DC: US Department of Health, Education, and Welfare; 1976.
  41. Hamilton M. A rating scale for depression. J Neurol Neurosurg Psychiatry. 1960;23:56–62.
  42. Liebowitz MR. Social phobia. Mod Probl Pharmacopsychiatry. 1987;22:141–173.
  43. Buysse DJ, Reynolds CF 3rd, Monk TH, et al. The Pittsburgh Sleep Quality Index: a new instrument for psychiatric practice and research. Psychiatry Res. 1989;28:193–213.
  44. Germain A, Hall M, Krakow B, et al. A brief sleep scale for posttraumatic stress disorder: Pittsburgh Sleep Quality Index Addendum for PTSD. J Anxiety Disord. 2005;19:233–244.
  45. Kessler RC, Sonnega A, Bromet E, et al. Posttraumatic stress disorder in the National Comorbidity Survey. Arch Gen Psychiatry. 1995;52:1048–1060.

CORRESPONDENCE Mark B. Hamner, MD, Mental Health Service 116, Ralph H. Johnson VA Medical Center, 109 Bee Street, Charleston, SC 29401 USA. E-MAIL: hamnermb@musc.edu