Francis Mawanda, Robert B. Wallace, Kimberly McCoy, Thad E. Abrams

To determine the associations between PTSD, psychotropic medication use, and the risk for dementia. Retrospective cohort. Nationwide sample of US veterans (N = 417,172) aged ≥56 years during fiscal year (FY) 2003 without a diagnosis of dementia or mild cognitive impairment at baseline (FY02-03) and ≥1 clinical encounter every 2 years during follow-up (FY04-12). Demographic characteristics; diagnosis of PTSD, dementia, and medical and psychiatric comorbidity (defined by ICD-9 codes); and psychotropic medication use including selective serotonin reuptake inhibitors (SSRI), serotonin-norepinephrine reuptake inhibitors (SNRI), novel antidepressants (NA), benzodiazepines (BZA), and atypical antipsychotics (AA). Cox proportional hazard models examined for associations between PTSD diagnosis, psychotropic medication use, and risk for a dementia diagnosis. PTSD diagnosis significantly increased the risk for dementia diagnosis (HR = 1.35; [95% CI = 1.27-1.43]). However, there were significant interactions between PTSD diagnosis and use of SSRIs (P < .001), NAs (P = .014), and AAs (P < .001) on the risk for dementia diagnosis. HR for dementia diagnosis among veterans diagnosed with PTSD and not using psychotropic medications was 1.55 [1.45-1.67]. Among veterans diagnosed with PTSD prescribed SSRI, SNRI, or AA, HR for dementia diagnosis varied by drug class use ranging from 1.99 for SSRI to 4.21 for AA, relative to veterans without a PTSD diagnosis and no psychotropic medication receipt. BZAs or SNRIs use at baseline was associated with a significantly increased risk for dementia diagnosis independent of a PTSD diagnosis. PTSD diagnosis is associated with an increased risk for dementia diagnosis that varied with receipt of psychotropic medications. Further research would help to delineate if these findings are due to differences in PTSD severity, psychiatric comorbidity, or independent effects of psychotropic medications on cognitive decline.