Annual Meeting Abstracts: View

<< Back to Search Results

(Abst. 2.294), 2017

A Temporal Evaluation of Antiepileptic Drugs and Suicidality: Is it the Medication or the Comorbidity
Authors: Mary Jo Pugh, South Texas Veterans Health Care System and UT Health San Antonio; Hari Sagiraju, UT Health San Antonio; Chen-Pin Wang, UT Health San Antonio; Hamada Hamid. Altalib, Yale University; and Anne Van Cott, University of Pittsburgh School of Medicine
Content: Rationale: Although studies have found an association of suicide related behavior (SRB) and antiepileptic drugs (AED) inconsistent findings lead to uncertainty in this relationship. Increasing use of AEDs in non-epileptic conditions such as psychiatric disorders and chronic pain complicate the understanding of the relationship between AED exposure and SRB. Similar to studies examining SRB and depression treatment, studies examining temporal patterns of SRB and AED treatment find that the peak of SRB occurs BEFORE initiation of AEDs. We compared trends of SRB before and after initiation of AED therapy among AED users with and without epilepsy to individuals without AED use controlling for sociodemographic characteristics and mental health comorbidity. SRB trends were further evaluated by AED type (anticonvulsant mood stabilizers vs other AED’s). Methods: We obtained Veterans Health Administration (VA) health system data for a retrospective cohort study of Veterans deployed in support of Iraq and Afghanistan wars (IAV) and who received VA care during fiscal years 2013 and 2014 and who did not receive AEDs prior to study period. The index date was the date of first AED prescription for AED users and the first health care encounter for individuals without AED exposure. We identified SRB 12 months before and after the index date based on ICD-9 codes. We conducted generalized estimation equation (GEE) analyses to assess the trend of SRB prevalence the year prior to and after the index date controlling for sociodemographic factors and indicators of mental health comorbidity (ICD9 codes), psychiatric medication and hospitalization prior to index date. To further evaluate differential SRB trend among AED users we conducted similar GEE analyses stratified by (i) epilepsy status (adjusting for inverse propensity scores weights of being diagnosed for epilepsy) and (ii) type of AED received. Results: The GEE analysis of the full cohort showed a significant curvilinear trend of SRB prevalence over the 24-month study period for AED users indicating that the probability of SRB diagnoses increased over time before the index month and then decreased after the index month. Similar curvilinear SRB probability trajectory was observed among non-AED users, but the likelihood for SRB was significantly lower (Odds Ratio: 0.16; 95% CI 0.14-0.19) throughout compared to AED users. GEE analysis among AED users suggested no differential trend in SRB prevalence trajectories when stratified by epilepsy status. Overall SRB was greater for those receiving anticonvulsant mood stabilizers.  Conclusions: While individuals with AED exposure had higher likelihood of SRB, the trend started before the AED exposure and was not differentially associated with epilepsy status or AED type received. Given the limited information available on depression severity in this administrative data it is possible that the significant effect for AED exposure is due to residual confounding of mental health comorbidity. Patients with risk factors for SRB, whether taking AED or not, should be carefully monitored for mental health symptoms and SRB. Funding: VA Health Services Research and Development Service I01 HX000329-05
Figure 1
Figure 2