Risk for early onset Dementia among Veterans: The contributions of TBI and Epilepsy

Abstract

<h3>Research Objectives</h3> To identify the impact of TBI and epilepsy on subsequent diagnoses of early onset dementia (EOD) in a cohort of post-9/11 era Veterans as both are associated with dementia in older individuals and are more common in Post-9/11 Veterans than the general population. <h3>Design</h3> Retrospective observational study of Post-9/11 Veterans in VA care FY02-FY18. <h3>Setting</h3> National longitudinal data from Departments of Defense and Veterans Affairs. <h3>Participants</h3> Post-9/11 Veterans 3 years of care. <h3>Interventions</h3> N/A. <h3>Main Outcome Measures</h3> EOD was operationalized using ICD-9-CM/10 diagnoses for diagnoses previously identified as reliable for EOD (Alzheimer's and frontotemporal dementia). TBI severity was identified using self-reports from the VA Comprehensive TBI Evaluation and ICD-9-CM/10 codes. Epilepsy was identified using a previously validated algorithm requiring ICD-9-CM/10 codes and anticonvulsant medications. Covariates included socio-/military demographics, prior deployment and comorbid conditions associated with dementia. We conducted logistic regression analysis predicting EOD to identify associations for epilepsy and TBI severity controlling for potential confounders. <h3>Results</h3> Among the 1,055,873 Veterans who met inclusion criteria, 923 had EOD (7.4/1000 epilepsy; 0.7/1000 no epilepsy). Epilepsy (aOR 2.41 [1.98-2.93]) and TBI of all severity (aORs: mTBI 1.67 [1.40-1.99]; moderate/severe TBI 2.31 [1.80-2.98]; penetrating TBI 2.87 [2.08-3.97]) were significantly associated with EOD. Other significant predictors (aORs >1.5; p < .001) included age 50-64 and 40-49 (vs. 30-39), other neurological conditions, stroke, schizophrenia, depression, and bipolar disorder. <h3>Conclusions</h3> Epilepsy and TBI (of all severities) were associated with EOD. While there was a near linear association of TBI severity and EOD which is consistent with prior research, there was no significant interaction between TBI and epilepsy. This suggests that each condition contributes to EOD. However, those with more TBIs have higher risk for epilepsy and EOD suggesting the mechanism of added impact of TBI severity may work through other neurological/neurodegenerative conditions. Indeed, other strong predictors of EOD (e.g., stroke, other neurological conditions) supports this hypothesis and the idea that multimorbidity associated with TBI may reveal phenotypes of neurodegenerative outcomes that require further evaluation. <h3>Author(s) Disclosures</h3> None.