RESUMO
INTRODUCTION: We aimed to investigate the effects of changes in sleep architecture on long-term clinical outcome in patients with Parkinson's disease (PD) who underwent deep brain stimulation of subthalamic nuclei (STN DBS). METHODS: We followed up eight PD patients before and three years after STN DBS surgery. In addition to clinical assessments, polysomnography (PSG) followed by multiple sleep latency tests was performed before and after STN DBS, while stimulator was ON and OFF. RESULTS: Subjective sleep latency was significantly decreased (P=0.033) and sleep duration was increased (P=0.041), as measured by Pittsburgh sleep quality index. Latency to REM sleep stage was shortened after surgery with STN DBS ON (P=0.002). Index of central type of abnormal respiratory events was significantly increased while stimulator was ON (P=0.034). Total number of major body movements was found to be increased when stimulator was turned OFF (P=0.012). Among PSG data obtained during STN DBS ON, it was observed that duration of N3 sleep was negatively correlated with UPDRS scores at 1st (P=0.038) and 3rd (P=0.045) post-operative years. Among PSG variables during STN DBS OFF, durations of N3 sleep (P=0.017) and REM sleep (P=0.041) were negatively correlated with UPDRS scores at post-operative 1st year. CONCLUSION: Disturbances in sleep architecture are associated with higher UPDRS scores and worse prognosis at 1st and 3rd post-operative years. Similar results obtained while stimulator was OFF at the end of 1st year support the presence of microlesion effect after STN DBS, which is probably not long lasting.