We appreciate the concerns raised by Sharman et al 1 regarding methodological aspects of our work. We would therefore like to take the opportunity to present additional details on polysomnographic recordings and scoring, as well as a supplementary analysis accounting for independence of all individual recordings and potential effects of outliers.Polysomnographs were performed with a MEPAL system (MAP, Martinsried, Germany) between January 2012 and December 2016. Electroencephalogram (EEG; C3 and C4 with linked M1 and M2 as reference electrodes until February 2016; F3, C3, and O1 with M2 as reference electrode thereafter) and electrooculogram (EOG; 2 channels) were analyzed with a low-pass filter at 35 Hz and high-pass filter at 0.5 Hz, with a sampling rate of 100 Hz. Chin electromyography (EMG; 2 channels) was analyzed with a low-pass filter at 35 Hz, high-pass filter at 10 Hz, and a sampling rate of 100 Hz.Since our equipment did not fulfill all American Academy of Sleep Medicine (AASM) recommendations 2 (EEG/EMG sampling rates of 100 Hz instead of ≥ 200 Hz, 2-channel instead of 3-channel chin EMG), sleep was scored according to R&K. 3 All recordings were scored by a single polysomnographic technologist (M.B.) with > 20 years of experience. We acknowledge that recordings prior to 2016 did not include occipital electrodes; however, we want to emphasize that the lack of occipital derivations does not violate Rechtschaffen and Kales recommendations. 3 We want to clarify that the arousal index is not the key result of our study. Compared to Rechtschaffen and Kales, AASM criteria affect the distribution of non-rapid eye movement stages, 4 which may relate to differences in arousal scoring. However, differences in total sleep time, sleep efficiency, or rapid eye movement