Incidence of mental health conditions by seizure control among adults with epilepsy in the United States

Abstract

BackgroundMental health conditions (MHCs) are frequent comorbidities among people with epilepsy; however, the influence of seizure control on the incidence of MHCs is not well reported. This retrospective observational cohort study based on claims data evaluated the effects of indicators of poor seizure control on the incidence of MHCs among MHC-naïve people with epilepsy. We hypothesized that poor seizure control is associated with new-onset MHC diagnoses and/or new prescription drugs for MHCs. MethodsThis study utilized a sample of patients from HealthVerity Marketplace, which includes more than 150 US commercial, Medicare, and Medicaid payers, to identify a cohort of adults (age ≥18 years) with prevalent epilepsy. Follow-up started on day 1 (January 1) after a 1-year eligibility assessment period occurring in calendar year 2017 or 2018. Patients were followed up until the occurrence of an incident MHC event (primary outcome), defined as a mental health diagnosis or psychotropic drug prescription. Time from follow-up to incident MHC diagnosis or to a drug prescription specific to depression or anxiety disorder was analyzed as a secondary outcome. Multivariate Cox proportional hazards regressions were estimated with time-varying covariates, measured in 6-month intervals during follow-up. Time-varying covariates were based on the occurrence of 4 variables used as indicators of poor seizure control in the prior period: epilepsy-related emergent care admissions, epilepsy-related inpatient admissions, epilepsy electroencephalography referrals, and exposure to one or more new antiseizure medications (ASMs). ResultsFrom a random sample of 40,000 people with epilepsy, 2563 (mean age 46.1 years; 50.6% male) were included in the analysis. Incident MHC events were observed in 27.7% (incidence rate 24.4 events per 100 person-years over 2,915.7 total person-years of follow-up). Mean (standard deviation [SD]) time to event was 232.7 (186.3) days. Among the 4 variables, epilepsy-related emergent care admissions were associated with an increased risk of incident MHC events in the following 6-month period (hazard ratio [HR] = 1.676, 95% confidence interval [CI]: 1.386, 2.026, p < 0.001) as were prescriptions for new ASMs in the previous period (HR = 1.702, 95% CI: 1.359, 2.132, p < 0.001). Previous epilepsy-related emergent care admissions (HR = 1.650, 95% CI: 1.347, 2.021, p < 0.001) and new ASMs (HR = 1.632, 95% CI: 1.280, 2.081, p < 0.001) also predicted an increased risk of incident depression or anxiety in the following 6-month period. ConclusionsPrevious indicators of poor seizure control, including epilepsy-related emergent care admissions and new ASMs, predicted increased risk of new MHC events, including depression and anxiety, during the following 6-month interval in MHC-naïve patients with prevalent epilepsy. These data suggest that poor seizure control can increase the subsequent risk of new mental health diagnoses and treatment among people with epilepsy.