PSG data from five active and six sham group patients were excluded from the analyses because of poor data quality. The final PSG data used for analysis were as follows: 39 (GSC group), 23 (active treatment group of ID patients), and 24 (sham treatment group of ID patients).