Participants were recruited from the Omaha Veterans Affairs Medical Center following local IRB approval. We obtained written consent from all the subjects who were recruited in the study. Veterans between the ages of 19 and 65 years with chronic PTSD (diagnosis for \>6 months) attributable to military combat exposure were included in the study. Patients were required to be clinically stable on their psychotropic medication regimen for at least three months prior to study entry. In addition to meeting DSM-IV criteria for PTSD and endorsing subjective complaints of memory difficulties, patients had to score least one standard deviation below the mean performance of a standardized, age- and sex-matched population on the Spatial Span, Logical Memory I, and Letter-Number Sequencing subtests of the Wechsler Memory Scale III (Third edition) for study entry. Patients with a history of dementia, schizophrenia, bipolar disorder, traumatic brain injury, and seizure were excluded. Patients with any history of alcohol or illicit drug abuse or dependence within the past one month were excluded. Patients requiring concomitant treatment with drugs with potential effects on the glutamergic system, such as amantadine,dextromethorphan, or carbonic anhydrase inhibitors, were excluded.
Subjects initially received memantine 5 mg once daily, which was increased weekly by 5 mg/day in divided doses to a dose of 20 mg/day. Memory was assessed using the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) \[Randolph, 1998\] using both forms A and B at baseline, end of week 8, and end of week 16. The RBANS is composed of 10 subtests that yield a total score and five index scores: immediate memory, visuospatial/constructional, language, attention, and delayed memory. Each index score has a normal mean of 100 and standard deviation of 15 based on the performance of a standardization sample matched to the U.S. Census on sex, ethnicity, and level of education. Alternative forms of the RBANS (Forms A and B) were used to avoid bias due to practice effects. We administered Clinician Administered PTSD Scale (CAPS), Hamilton Depression Scale (HAM-D), Hamilton Anxiety Scale (HAM-A), Quality of Life Enjoyment and Satisfaction Questionnaire (Q-LES-Q), and Sheehan Disability Scale (SDS) to assess the secondary measures. Changes in the scores over time with repeated measures were estimated with mixed-effects models. The primary outcome measures of interest were index and percentile scores in RBANS total and subscale scores. The repeated measures model included visit (as a categorical variable) as a fixed effect. An unstructured covariance matrix was used to fit the within patient repeated measures effect. Tukey's method was used to compare pair-wise means.
Secondary outcome measures, CAPS, HAM-A, HAM-D, Q-LES-Q, and SDS, were analyzed similarly to the RBANS with repeated measures models. P-values less than 0.05 are considered to be statistically significant. SAS software version 9.1 (SAS Institute, Cary, NC) was used for the analysis
