. 2019 Jul 5;9:9737. doi: 10.1038/s41598-019-46166-z

Table 3.

Univariate and multivariate logistic regression of the occurrence of obstructive sleep apnea (OSA) versus preoperative and demographic variables. ^aStatistical analysis was carried out using univariate logistic regression (the occurrence of OSA vs. preoperative and demographic variables).

Preoperative Variables	Univariate Analysis			Multivariate Logistic Regression
Preoperative Variables	OR	95% CI	P Value^a	OR	95% CI	P Value^b
Relative snoring time (%)	1.062	1.012–1.115	0.015^*	1.096	1.017–1.182	0.017^*
Lowest oxygen saturation (%)	0.486	0.299–0.788	0.003^*	0.427	0.235–0.775	0.005^*
BMI (kg/m2)^†	1.456	1.084–1.957	0.013^*	none	none	0.617
Female gender^†	0.119	0.014–1.042	0.054	none	none	0.855
The amount of mandibular setback (mm)^†	1.126	0.923–1.375	0.242	none	none	0.438
Total sleep time (minutes)^†	1.010	0.995–1.025	0.202	none	none	0.640
Sleep efficiency (%)^†	1.067	0.980–1.162	0.133	none	none	0.929
Wake after sleep onset (minutes)^†	0.985	0.965–1.006	0.164	none	none	0.983
Total RERA index (number/h)^†	1.121	0.934–1.346	0.221	none	none	0.981

^bStatistical analysis was carried out using multivariate logistic regression (the occurrence of OSA vs. preoperative and demographic variables). ^*P < 0.05. ^†Variables not significant in multivariate analysis (those exempted from the equation).