Hyperventilation and 6-hour EEG recording in evaluation of absence seizures.

Seven patients with poorly controlled absence seizures were evaluated clinically and electroencephalographically (EEG) in control and placebo periods and then evaluated after 10 weeks on valproic acid. For analysis of epileptiform activity, 5 minutes of controlled hyperventilation recording was more reliable than 6-hour recording as a predictor of clinical seizure frequency. This apparently paradoxic result can be explained by considering that the number of epileptiform discharges in the EEG is affected by many variables (such as level of consciousness, anxiety, diurnal seizure variation, or blood glucose level) that are difficult to control for prolonged periods.