[Objective assessment and therapeutic efficacy of an improved mandibular advancement device for snoring and sleep apnea syndromes with polysomnography].

In the treatment of snoring (SN) and sleep-related breathing disorders (SRBD), mandibular advancement devices (MAD) are of increasing importance. Their mode of action is based on the advancement of the mandible, thereby increasing various upper airway dimensions and thus airway patency and airflow during sleep. The aim of the present study was to investigate efficacy and tolerability of an individually fitted MAD on 11 patients (10 males, 1 female), mean age 57 years, using sleep laboratory methods in 3 subsequent nights (adaptation-, baseline-, treatment night). The MAD consists of 2 separate parts that attach to both dental arches. On occlusion the upper maxillary part with a protruding cone meets an inclined plane of the lower mandibulary part, thereby forcing the mandible to advance. 10 patients (6 with obstructive sleep apnea, 3 with obstructive hypopnea and 1 primary snorer) tolerated the MAD well; one patient (primary snorer) removed the MAD after 1 hour. Regarding the target variable, the snoring index (SI), confirmatory statistics demonstrated a significant improvement from 108 to 53/h sleep, though normalisation could not be achieved. Descriptive data analysis showed significant improvement of the apnea-hypopnea index (AHI) from 15 to 5.5/h and of the oxygen desaturation index (O2-DI) from 21 to 13/h sleep. Arousal variables and periodic leg movement index (PLMI) improved as well. Objective sleep efficiency and subjective sleep- and awakening quality remained unchanged. Thus, besides the good therapeutic efficacy (the medians of improvement of the SI, AI, AHI, O2-DI and PLMI were 37, 48, 53, 51 and 29%, respectively), acute acceptance of the MAD was also satisfactory. Last but not least our present study showed once more the necessity of an adaptation night, as from the first to the second sleep laboratory night respiratory indices deteriorated significantly.