Table 6.

Results of two-piecewise logistic regression model

OBS Index	Elevated ePWV, OR (95%CI)#, P value
Inflection point (K)	Unadjusted models	Adjusted models#
≤ 5	1.11 (1.06, 1.17) P < 0.0001	0.70 (0.51, 0.94) P = 0.019
> 5	0.98 (0.98, 0.99) P < 0.0001	1.00 (0.96, 1.04) P = 0.945
P for log likelihood ratio test	< 0.001	0.028

Two-piecewise logistic regression model was used to calculate the threshold effect of the OBS. If the log likelihood ratio test > 0.05, it means the two-piecewise logistic regression model is not superior to the single-line logistic regression model

^#Adjusted for age, sex, BMI, race, poverty income ratio, SBP, DBP, smoking status, HbA1c, CRP, TC, HDL, eGFR, DM, CVD, Hypertension, antihypertensive drugs, Lipoprotein-lowering drugs, hypoglycemic drugs