Table 2.

Result of Bayesian binary/ordinal logistic regression with random subject effect and handling missingness

Independent variables
Dependent variables	Sex‡	Age at baseline (year)	Obesity§	Remnant cholesterol (mg/dL)	Physical activity (TMET h/day)
	OR (95% Credible intervals)
Univariate model
Pre-T2DM	0.844 (0.630,1.072)	1.027 (1.010,1.045)*	1.140 (1.033,1.390)*	1.009 (1.005,1.015)*	0.997 (0.972,1.019)
HBP (pre-HTN or HTN)	1.076 (1.056,1.104)	1.312 (1.062,1.776)*	2.251 (1.803,2.969)*	1.0129 (1.010,1.020)*	0.999 (0.972,1.027)
Bivariate model
Pre-T2DM	0.956 (0.743,1.238)	1.055 (1.040,1.068)*	1.399 (1.129,1.730)*,†	1.086 (1.043,1.119)*,†	0.990 (0.970,1.012)
HBP (pre-HTN or HTN)	1.084 (1.067,1.102)	1.419 (1.077,1.877)*	2.443 (1.978,3.031)*,†	1.093 (1.067,1.121)*,†	0.999 (0.972,1.027)

*Significant results based on 95% credible intervals, ^†Significant difference between univariate and bivariate logistic regressions, ^‡Female is the reference group, ^¦Nonobese is the reference group. Pre-T2DM=Pretype 2 diabetes mellitus; HBP=High blood pressure; HTN=Hypertension; Pre-HTN=Prehypertension; OR=Odds ratio; TMET=Total metabolic equivalent