Table 3.

Analyses of uni- and multivariate regression models for factors to improve alanine aminotransferase level in obstructive sleep apnea patients with nonalcoholic fatty liver disease

Factor	n	Partial regression coefficient	Standard error	β	P value
Univariate model
Decreased factors for 6 mo
BW	50	2.051	0.600	0.443	0.001
BMI	50	5.578	1.664	0.436	0.002
AST	50	1.318	0.134	0.818	< 0.001
TG	47	0.030	0.036	0.124	0.407
LDL-C	50	-0.027	0.063	-0.061	0.674
HbA1c	50	11.076	4.064	0.366	0.009
HOMA-IR	47	0.521	1.496	0.052	0.729
M2BPGi	50	4.126	7.657	0.078	0.593
LSM	48	1.871	1.656	0.164	0.264
CAP	48	0.020	0.047	0.064	0.667
m-CI for 6 mo of CPAP therapy	50	-11.878	5.354	-0.305	0.031
Mutivariate regression model
m-CI for 6 mo of CPAP therapy	50	-9.890	4.928	-0.254	0.051

The multivariate regression model was adjusted for quantities of BW change during the 6 mo of continuous positive airway pressure therapy. AST: Aspartate aminotransferase; BMI: Body mass index; BW: Body weight; CAP: Controlled attenuation parameter; HbA1c: Hemoglobin A1c; HOMA-IR: Homeostasis model assessment-insulin resistance; LDL-C: Low-density lipoprotein cholesterol; LSM: Liver stiffness measurement; M2BPGi: Mac-2-binding protein glycan isomer; m-CI: Mean compliance index; TG: Triglycerides; β: Standardized partial regression coefficient.