Seizures occurring both at night and during the day can affect sleep architecture and produce daytime sleepiness. In a study comparing the effects of generalized and focal seizures on sleep, nocturnal generalized seizures produced a decrease in sleep time and reduced REM sleep percentage, prolongation of REM latency, and more sleep fragmentation (Touchon et al., 1991). Stages 1 and 2 sleep were increased while the percentage of slow-wave sleep (SWS) was unchanged. REM rebound [an increase in REM percentage after REM-suppressing situations (i.e., sleep deprivation)] was not observed later in a night after a seizure or during seizure-free nights. Focal seizures occurred during REM and nonrapid eye movement (NREM) sleep. When occurring in isolation, focal seizures produced little or no sleep disruption. However, multiple focal seizures in a night produced a significant reduction of REM sleep and duration of REM periods. As compared with control subjects, patients with epilepsy had significantly more disrupted sleep and lower sleep efficiency (total sleep time/time in bed). These findings held true even on seizure-free nights.
The effects of daytime and nocturnal seizures on sleep and vigilance were studied in 11 patients admitted to an epilepsy monitoring unit (Bazil et al., 1997; Castro et al., 1997). The percentage of time spent in REM sleep was significantly reduced and REM latency prolonged in nights after daytime seizures (Bazil et al., 1997). However, daytime seizures did not affect sleep efficiency. After daytime seizures, patients fell asleep more quickly on the MSLT, suggesting that daytime seizures are a significant cause of EDS. As compared with seizure-free nights, nights with seizures were characterized by a reduction of REM and stage 3 sleep (Castro et al., 1997). In addition, REM latency was prolonged and sleep efficiency was reduced. Patients had more severe daytime sleepiness on days after nocturnal seizures as compared with days after seizurefree nights, as measured by the MSLT. These findings were independent of seizure duration.
antiepileptic drug effects on sleep and vigilance
AEDs produce a variety of alterations in sleep architecture and varying degrees of daytime sleepiness. In seven healthy subjects taking 700 mg daily CBZ of for 10 days (mean serum concentration of 11.8 ± 1.1 /¿g/ml), SWS percentage increased significantly while REM sleep decreased (Yang et al., 1989). Treatment produced a shortened REM latency, although not statistically significant. The SWS-enhancing effects were thought to reflect the effect of CBZ on 5-HT levels or its effect on adenosine receptors that modulate the release of 5-hydroxytryptamine (5-HT) and catecholamines. The acute and long-term effects of controlled release CBZ (CBZ CR) were studied in seven adults with newly diagnosed temporal lobe epilepsy (TLE) (Gigli et al., 1997). The findings were compared to nine control subjects who underwent PSG after a single 400 mg dose of CBZ CR. The single dose produced a significant reduction of REM sleep and increase in REM fragmentation in the TLE group, and an increase in stage shifts in normal subjects. In patients with TLE, these findings were almost completely reversed after 1 month of treatment with 400 mg twice per day. The first dose produced a significant shortening of the mean sleep latency on the MSLT only in normal subjects. After a single dose, daytime sleepiness was reported more often by controls than TLE patients. Other investigators found no change in sleep macroarchitecture with CBZ therapy in patients with epilepsy (Declerck and Wauquier, 1991). However, administration of the drug to healthy volunteers produced shorter sleep cycle duration.
The effect of therapeutic concentrations of phenobarbital (PB) and PHT was studied in 40 patients with epilepsy using a random, crossover design (Wolf et al., 1984). Treatment with PB shortened sleep latency and produced a reduction of body movements and arousals. An increase in stage 2 sleep, reduction of REM sleep, and higher sleep efficiency were observed, while SWS was not affected. Compared to the baseline, each REM period started later, with the most prolonged latency occurring in the first REM period of the night. Treatment with PHT also shortened sleep latency, but produced a reduction of stage 1 and 2 sleep and an increase in SWS. The frequency of arousals and REM percentage were unchanged. No correlation between sleep architecture and serum concentrations of either drug was found.
Immediate, short- and long-term effects of PHT on sleep architecture were studied in 40 untreated subjects with epilepsy (Roder-Wanner et al., 1987). A single 100-mg dose at bedtime shortened sleep latency and produced a reduction of wake time and an increase in movement arousals and body movements. A decrease in REM sleep was observed in patients with focal epilepsy. During short-term treatment (adjustment to steady state), sleep latency and stage 1 sleep were reduced, SWS was increased, and arousals were more frequent in both REM and NREM sleep. In 12 subjects treated with PHT for a minimum of 4.5 months, most of the short-term effects were reversed. An increase in stages 1 and 2, and a decrease in SWS were observed, while REM percentage was unchanged. There was no correlation between seizure frequency or serum drug concentration and long-term alterations in sleep architecture. An increase in stage 1 and 2 and a minor reduction of REM sleep after treatment with PHT have also been reported (Declerck and Wauquier, 1991).
In 30 healthy subjects, a single dose of 250-mg primidone (PRM) produced an increase in SWS and a reduction of REM percentage and density (Maxion et al., 1975). No significant changes were found after one 200-mg dose of PHT, although SWS percentage was reduced. A shortened sleep latency and reduction of REM density, but not percentage, were observed in patients with epilepsy on 750 mg daily of PRM for three months. Treatment with 600 mg of PHT per day for the same duration produced no notable effects.
Both ethosuximide (ES) and VPA were found to increase the percentage of stage 1 sleep in 11 patients with absence epilepsy (Wolf et al., 1984). Treatment with VPA resulted in a reduction of stage 2 sleep but no change in SWS. Treatment with ES produced a significant reduction in SWS percentage and a trend for increased stage 2 sleep. In 13 patients with primary generalized epilepsy, doses of ES effective in suppressing absence seizures produced a significant increase in stage 1 sleep and a decrease in SWS, without altering stage 2 or REM sleep percentages (Roder and Wolf, 1981). There was no change in sleep latency, number of stage shifts, body movements, or number of awakenings. Interrupted or light sleep, longer sleep latency, and daytime sleepiness were reported by patients taking ES. Treatment with VPA produced an increase in stage 1 and a trend for reduced stage 2 sleep without affecting SWS, REM, sleep latency, or the number of movements or arousals. Patients treated with VPA complained of difficulty initiating sleep and daytime sleepiness.
The effects of high- (1000 mg/day) and low-dose (500 mg/day) VPA and placebo on sleep were compared in 10 healthy subjects (Harding et al., 1985). As compared with low-dose and placebo, high-dose VPA produced a reduction of REM sleep percentage and an increase in SWS. A reduction of REM sleep was also observed with low-dose treatment when compared with the withdrawal phase of the study. Others have reported only a shorter sleep cycle duration in patients and normal subjects treated with VPA (Declerck and Wauquier, 1991).
With use of the MSLT, daytime vigilance was compared in 20 patients with generalized epilepsy taking PB (10) or VPA (10) monotherapy, 10 subjects with focal epilepsy on CBZ, and 10 healthy controls (Manni et al., 1993a,b). All patients were seizure free for 1 year. Sleep time the previous night was similar between groups. The mean sleep latency was 9, 12.5, 12.5, and 12.9 min for PB, VPA, CBZ, and controls, respectively. No sleep onset REM periods were recorded. Mean serum concentrations were PB 19.3 (6-20), VPA 85.7 (69-106), and CBZ 8.2 (5.5-12) ¿¿g/ml. Only two subjects had mean sleep latencies less than 5 min; both were in the PB group. Patients on PB were sleepier based on the MSLT, but this was not accompanied by complaints of EDS.
Of the newer AEDs, only gabapentin has been studied for its effect on sleep. In eight patients treated with 1800 mg per day of gabapentin as add-on therapy, the drug produced a significant increase in REM sleep and mean REM period duration and fewer awakenings (Placidi et al., 1997). Although no PSG analyses have been published on patients taking felbatol, lamotrigine, topiramate, or tiagabine, sleep complaints were commonly reported in premarket trials. The incidence of somnolence and insomnia in patients treated with felbatol was 19 and 18%, respectively (PDR, 1998). Insomnia was a common reason for discontinuation of the drug. Treatment with lamotrigine produced somnolence in 14% and insomnia in 6% of patients (PDR, 1998). Topiramate produced somnolence in approximately 30% of patients treated (PDR, 1998). Somnolence and insomnia occurred in 18 and 6%, respectively of patients receiving tiagabine (Gabatril package insert, Abbott Laboratories, 1997).
Was this article helpful?