Table 4.
Model to predict NASH (elaboration set, n = 100).
| Variable | N | NASH N (%) |
P univariate | Adjusted Odds Ratio (95% confidence interval) |
Multivariate p |
|---|---|---|---|---|---|
| Age† | 0.183* | 1.033 (0.983–1.098) | 0.202 | ||
| < 62 years | 50 | 6 (12) | |||
| ≥ 62 years | 50 | 11 (22) | |||
| Sex | 0.374 | 0.346 | |||
| Female | 61 | 12 (20) | 1.845 (0.501–7.170) | ||
| Male | 39 | 5 (13) | |||
| BMI‡ | 0.055* | 1.002 (0.852–1.179) | 0.978 | ||
| < 30 kg/m2 | 67 | 8 (12) | |||
| ≥ 30 kg/m2 | 33 | 9 (27) | |||
| HOMA§ | < 0.001* | 2.221 (1.501–2.286) | < 0.001 | ||
| < 4 | 25 | 3 (4) | |||
| ≥ 4 | 75 | 14 (56) | |||
| Triglycerides§ | 0.026* | 0.992 (0.979–1.005) | 0.206 | ||
| < 150 mg/dL | 76 | 9 (12) | |||
| ≥ 150 mg/dL | 24 | 8 (33) | |||
| HDL-c§ | 0.002* | 0.997 (0.943–1.055) | 0.924 | ||
| < 45 mg/dL | 33 | 11 (33) | |||
| ≥ 45 mg/dL | 67 | 6 (9) | |||
| Metabolic syndrome | 0.068 | 5.885 (0.788–43.962) | 0.084 | ||
| No | 32 | 2 (6) | |||
| Yes | 68 | 14 (21) | |||
| AST§ | 0.432 | – | – | ||
| < 37 mg/dL | 97 | 16 (17) | |||
| ≥ 37 mg/dL | 3 | 1 (33) | |||
| GGT§ | 0.115* | – | – | ||
| < 66 IU/mL | 76 | 10 (13) | |||
| ≥ 66 IU/mL | 24 | 7 (29) | |||
| Liver stiffness† | 0.062* | 1.087 (0.826–1.432) | 0.550 | ||
| < 5.2 kPa | 50 | 5 (10) | |||
| ≥ 5.2 kPa | 50 | 12 (24) | |||
| CAP¶ | 0.005* | 1.016 (1.002–1.031) | 0.021 | ||
| < 248 dB/m | 36 | 1 (2.8) | |||
| ≥ 248 dB/m | 64 | 16 (25) |
CAP controlled attenuation parameter, BMI Body mass index, HOMA homeostasis model assessment.
*Entered in the logistic regression model as continuous variables.
Categories of continuous variables: †By the median; ‡BMI ≥ 30 kg/ m2, indicative of obesity; §Categorized by upper limit of normal; ¶:CAP ≥ 248 dB/m, cut-off for steatosis (18).