BACKGROUND

Mandibular advancement device (MAD) is a good alternative for patients with obstructive sleep apnea (OSA). However, the treatment response varies amongst individuals.

OBJECTIVE

This study aimed to explore the role of craniofacial features in the response to MADs to improve prognostication and patient selection.

METHODS

The retrospective trial contained 42 males aged 41.5 ± 9.0 years, and with an apnea-hypopnea index (AHI) of 21.5 ± 13.8 events/h. According to the mandibular plane angle, participants were divided into three groups: low angle (n = 13), average angle (n = 14) and high angle (n = 15). Under the monitoring of home sleep testing, adjustable MADs were used to titrate the mandible forward from 0 mm with an increment of 0.5 mm every day. The polysomnography outcomes, mandibular protrusion amounts, changes in upper airway MRI measurements and nasal resistance were compared amongst the three groups.

RESULTS

The normalisation rate (AHI <5 /h) was 92.3%, 57.1% and 46.7%, respectively, in the low-, average- and high-angle groups (p = .027). The effective protrusion where AHI was reduced by half was 20 (11.3 ~ 37.5) %, 31.3 (23.6 ~ 50) % and 50 (36.9 ~ 64.9) % of the maximal mandibular protrusion, in the low-, average- and high-angle groups (p = .004). Multivariate logistic regression revealed that increased gonion angle (OR = 0.878) and baseline AHI(OR = 0.868) can reduce the probability of normalisation.

CONCLUSION

The high mandibular plane angle might be an unfavourable factor to MAD treatment and more protrusion was needed to achieve a 50% reduction in AHI. Vertical craniofacial pattern (gonion angle) and baseline AHI constituted the model for predicting the effect of MADs.