Abstracts

RATIONALE: Obstructive sleep apnea (OSA) is a common condition affecting up to 9% of women and 24% of men (Young et al., NEJM 1993;328:1230-1235). A higher prevalence, approaching 33%, has been reported in adults with epilepsy (Malow et al., Neurology 2000;55:1002-1007). Treatment of obstructive sleep apnea may reduce seizure frequency and improve daytime sleepiness in epilepsy patients (Malow et. al., Neurology 1997; 48:1389-1394). Therefore, identifying factors associated with OSA may contribute to improved management of these patients. We explored the relationship between OSA, antiepileptic drugs (AEDs), and other factors in epilepsy patients.
METHODS: We reviewed the records of 156 adult epilepsy patients (ages 18-73, 56% men) who underwent polysomography in our clinical Sleep Laboratory or in our General Clinical Research Center between 1995-2001. We determined the presence or absence of OSA, defined by an apnea-hypopnea index of 5 or greater on polysomnography. Age, gender, body-mass index (BMI), and AED type and number at the time of polysomnography were recorded. To analyze the association between OSA and these variables, we performed chi-square (specific AEDs and gender) and independent sample t-tests (age, number of AEDs, body mass index). A general linear model was also performed to include significant variables in the model. We limited our AED analysis to medications taken by 10 or more patients. Level of significance was set at p [lt] 0.05.
RESULTS: The following AEDs were included in the analysis: phenytoin, carbamazepine, valproic acid, lamotrigine, topiramate, gabapentin, and barbiturates/benzodiazepines (phenobarbital, clonazapam, and primidone were combined into one category). A statistically significant association was seen between phenytoin use and OSA (p = 0.03); no other AEDs were associated with OSA. Other variables associated with OSA were older age (p [lt] 0.0001), male gender (p = 0.009) and higher BMI (p = 0.004). In the general linear model, including age, gender, BMI, and phenytoin use, the only significant variables associated with OSAwere age (p [lt] 0.001) and BMI (p = 0.002). Patients taking phenytoin as compared to those not taking phenytoin were older (40.6 ? 12.5 years vs. 36.5 ? 11.6 years, mean ? standard deviation; p = 0.04); this association between age and AED use was not seen with any other AED studied. The BMI was not significantly different in the phenytoin patients (29.9 ? 8.0 vs. 28.7 ? 6.3).
CONCLUSIONS: In our sample, older age and BMI were independent risk factors for OSA. After adjusting for age and body-mass index, phenytoin use and male gender were not independent risk factors for OSA. Limitations were that sample size of individual AEDs was small so that the effects of specific AEDs on OSA may have been missed; in addition, several of the newer AEDs were not included. Additional study with larger numbers of patients will be needed to determine the relative contributions of specific AEDs and other variables to obstructive sleep apnea.
Objective: Participants should be able to understand factors associated with OSA in epilepsy patients.
[Supported by: NINDS KO2 NS02099 (BAM) and University of Michigan General Clinical Research Center grant M01-RR00042.]