Abstracts

Rationale: Co-morbid obstructive sleep apnea (OSA) is frequent in epilepsy. Treatment with nasal continuous positive airway pressure (nCPAP) reduces seizure frequency. However, predictive features enabling triage for polysomnography (PSG) and nCPAP efficacy remain unclear. We aimed to describe clinical characteristics of co-morbid OSA in epilepsy and to analyze the effect of nCPAP therapy on seizure frequency. Methods: We retrospectively analyzed epilepsy patients with OSA (E OSA, n=111, PSG apnea-hypopnea index (AHI)>5/hour) and without OSA (E-OSA, n=74) between 1/1/00-5/30/09. Demographic, seizure frequency, antiepileptic drug (AED) load, and PSG variables were recorded. Seizure frequency was compared between E OSA patients compliant with nCPAP and those who either deferred nCPAP or were non-compliant. Group comparisons were analyzed with 2-tailed T-tests or Wilcoxon signed rank tests in JMP or SAS (Chicago, IL). Results: PSG yielded OSA diagnosis in 111 (60%). Average AHI in E OSA was 22.4 20.8. Hypersomnia was more frequent in E-OSA v. E OSA (95.9% v. 84.3%, p=0.014), although objective sleepiness was no different (mean sleep latencies 12.8 v. 12.2 minutes, p=0.66). Clinical variables that were more frequent in E OSA than E-OSA were male sex (59.5 v. 37.8%, p=0.004), older age (mean 45.9 v 39.6 years, p=0.0018), weight over ideal (BMI>25, 90 v. 75.7%, p=0.005), enlarged neck (42.6 v. 39.6 cm, p=0.0007), and sleep hypoventilation (36.9 v. 9.5%, p<0.001). Snoring, witnessed apneas, nocturnal seizures, and AED load were similar between groups. 43 patients received nCPAP. Of 33 patients having pre- and post-nCPAP data, seizure frequency decreased significantly (t=2.138, p=0.04). 79% reported seizure reduction and 61% were responders (50% or greater seizure reduction), and effect held when AEDs were unchanged or reduced (n=18, p=0.01). In untreated patients, there was no difference in seizure frequency (n=28, t=-1.411, p=0.17). Conclusions: PSG has a diagnostic yield of 60% for suspected OSA in epilepsy patients. Clinical factors associated with OSA include male sex, older age, enlarged neck circumference, and overweight body habitus. The frequency of nocturnal hypoventilation in OSA patients with epilepsy was an unexpected finding requiring further study. Interestingly, epilepsy patients without OSA more frequently reported subjective sleepiness, but were not objectively sleepier, suggesting that seizures, antiepileptic drug load, and other sleep co-morbidities (i.e., periodic limb movement disorder, insufficient sleep quantity, or primary CNS hypersomnia) may also cause sleepiness in epilepsy patients. nCPAP treatment significantly reduces seizure frequency, comparable to or exceeding the effect of adjunctive AEDs. Further prospective studies of the impact of co-morbid OSA on seizure burden and interictal state factors crucial to quality of life (i.e., mood state, vulnerability to AED adverse effects, and sleepiness/vigilance) are warranted.