Epilepsy and TBI: Unexpected Findings

Traumatic brain injury (TBI) raises the risk of subsequent epilepsy (post-traumatic epilepsy: PTE). The risk also correlates with TBI severity, and is higher in the presence of a preexisting seizure disorder. These are fairly well accepted principles. However, much about the relationship between TBI and epilepsy remains unclear. A material portion of the larger and longer studies are derived from military cohorts, and some of the conclusions do not generalize well to a civilian population. Also lacking are rigorously obtained data on how demographic and social factors influence risk. For example, the relationship between age and subsequent seizure risk is murky.

Investigators at the Medical University of South Carolina sought to address this. They identified a sample cohort from a comprehensive state TBI database, which documented persons hospitalized with TBI. Patients were then interviewed annually, up to three times. The cohort was large: 2,118 patients were identified and completed the first year interview. By the third year, however, only 1,173 interviews could be completed: This is a 45% attrition rate. The questionnaire was changed part way through the data collection period, to increase the sensitivity for PTE.

Of the initial cohort, 145 had epilepsy pre-TBI and 115 developed epilepsy post-TBI. PTE incidence was 9.1/100 persons over the three years of the study, with the expected increment with TBI severity: Mild - 4.4; moderate - 7.6; Severe - 13.6. There was some variation by year post-injury: The incidence of PTE was higher in the second year post-TBI than the first.

The most interesting findings, however, were related to the social and demographic variables. PTE patients, compared with the non-epilepsy group, had an increased probability of being middle-aged, receiving Medicaid, and having sustained TBI by violence. Also of note, PTE patients were more likely to have at least three comorbid medical conditions, as well as a history of previous TBI, stroke, or depression. Race, gender, income, and substance abuse history did not influence PTE risk. Data on depression and substance abuse as factors in PTE have not been well studied in the past. Multivariate analysis showed the strongest factors to be early PTE, severe TBI, at least three comorbidities, and preexisting depression.

Study pitfalls are, as always, present. Whether patients followed for the full three years differ from those lost to follow up is a concern. The investigators addressed this: An analysis of participation at 3 years v. early attrition showed that insurance status and education correlated with persistence in the study, but none of the other studies factors changed with longer study participation. A potential material weakness was the redesign of the questionnaire part of the way through the study. However, I reviewed the changes, and they don't appear to materially alter the conclusions. Another concern is that the cohort was derived from people who were hospitalized, so the mild TBI group may have been skewed. The usual caveats about self-reporting bias and comparing dissimilar cohorts also obtain.

I think there are two lessons to be leared nere. Depression, a common comorbidity in epilepsy, doubled PTE risk - that is a new finding. Also, TBI causes a large increase in PTE risk. The incidence of epilepsy in the general population has been estimated at 0.05 per 100 person-years. The second year incidence in this study was 2.2, which is significantly higher. This is in line with previous studies, but is a good reminder.