Skip to main content
. 2020 May 13;8:e9090. doi: 10.7717/peerj.9090

Table 2. Logistic regression models for NAFLD as a predictor for non-specific ST-T segment changes.

Logistic regression model OR (95% CI) P-value
NAFLD (yes vs no)
Model 1 1.925 (1.729–2.143) <0.001
Model 2 1.935 (1.725–2.170) <0.001
Model 3 1.596 (1.415–1.800) <0.001
Model 4 1.289 (1.122–1.480) <0.001
Other independent predictors of non-specific ST-T segment changes in Model 4
Sex
Female Reference
Male 0.359 (0.317∼0.405) <0.001
Age 1.058 (1.052∼1.064) <0.001
BMI 1.072 (1.050∼1.093) <0.001
Hypertension
No Reference
Yes 1.837 (1.622∼2.081) <0.001
FPG
Normal Reference
High 1.258 (1.104∼1.434) 0.001
TC
Normal Reference
High 1.195 (1.021–1.397) 0.026
HDL-C
Normal Reference
Low 0.839 (0.711–0.989) 0.037

Notes.

Sample size, n = 32922. Data are expressed as OR (95% CI).

OR
odds ratio
CI
confidence interval
BMI
body mass index
FPG
Fasting plasma glucose
TC
total cholesterol
HDL-C
high density lipoprotein cholesterol

Covariates included in multivariable regression models were as follows: model 1: unadjusted; model 2: age, sex; model 3: adjustment for heart rate, hypertension (blood pressure ≥140/90 mmHg) and fasting plasma glucose and plus the same variables included in model 2; model 4: adjustment for the same variables included in model 3 plus BMI, triglyceride, total cholesterol, HDL-C.