Altered polysomnographic profile in juvenile myoclonic epilepsy

• Sleep architecture was altered in JME compared to controls despite seizure freedom.
• Patients >20 years had reduced TST compared to those <20 years (p = 0.01).
• Patients with seizures for >5 years had reduced NREM sleep %age (p = 0.04).
• Those on VPA therapy >1 year had a longer N2 (p = 0.03) and N3 latency (p = 0.03).
• Patients on ≤600 mg/day of VPA had often isolated limb movements (p = 0.01).

SummaryPurposeTo study the spectra of sleep profile using PSG in a cohort of patients with JME attending a University hospital.MethodologyThis prospective cross-sectional case–control study involved 25 patients of JME (age: 22.0 ± 6.3 years; M:F = 13:12) on valproic acid (VPA) and 25 matched healthy controls (age: 23.2 ± 3.04 years; M:F = 16:9) were recruited. All patients underwent clinical assessment, electroencephalogram (EEG), and evaluation with sleep questionnaire and PSG.ResultsPSG analysis revealed significant alterations in sleep architecture in the JME group in the form of reduced mean sleep efficiency (p = <0.035) and number of patients with reduced sleep efficiency (p = 0.001), increased mean sleep onset latency (p = 0.04) and number of patients with increased sleep latency (p = 0.023), reduced mean N2 sleep percentage (p = 0.005) and reduced mean total NREM (non-rapid eye movement) sleep (p = 0.001) and increased mean wake percentage (p = 0.001). The frequency of arousals, involuntary limb movements, and event related arousals in the JME groups was not different from the controls. Patients >20 years had reduced total sleep time compared to those <20 years (p = 0.012). Patients with seizures for >5 years had reduced NREM sleep percentage (p = 0.042) and those on VPA therapy >1 year had a longer stage 2 (p = 0.03) and N3 latency (p = 0.03). Patients on ≤600 mg/day of VPA had a higher prevalence of isolated limb movements (p = 0.01).ConclusionsPSG revealed significant alterations in sleep architecture in JME despite adequate seizure control. There was variable degree of PSG–phenotypic correlation.