Table 4.
Association between hepatic steatosis with or without coexisting significant fibrosis and the 10-year estimated CVD risk (using the Steno type 1 risk engine or the ASCVD risk calculator)
| Logistic Regression Analyses | Odds Ratios (95% confidence intervals) | P-value |
|---|---|---|
| Y = High or moderate risk vs. low Steno type 1 risk score | ||
| Unadjusted model | ||
| Patients with HSI ≤ 36 (n = 654) | Ref. | - |
| Patients with HSI > 36 and FIB4 < 1.3 (n = 509) | 1.16 (0.92–1.47) | 0.202 |
| Patients with HSI > 36 and FIB4 ≥ 1.3 (n = 91) | 20.2 (7.33–55.7) | < 0.001 |
| Adjusted model 1 | ||
| Patients with HSI ≤ 36 (n = 654) | Ref. | - |
| Patients with HSI > 36 and FIB4 < 1.3 (n = 509) | 0.94 (0.65–1.36) | 0.753 |
| Patients with HSI > 36 and FIB4 ≥ 1.3 (n = 91) | 11.4 (3.54–36.9) | < 0.001 |
| Adjusted model 2 (n = 1,043) | ||
| Patients with HSI ≤ 36 (n = 532) | Ref. | - |
| Patients with HSI > 36 and FIB4 < 1.3 (n = 432) | 0.79 (0.52–1.18) | 0.244 |
| Patients with HSI > 36 and FIB4 ≥ 1.3 (n = 79) | 10.9 (2.99–39.9) | < 0.001 |
| Y = High or intermediate riskvs.low ASCVD risk score | ||
| Unadjusted model | ||
| Patients with HSI ≤ 36 (n = 654) | Ref. | - |
| Patients with HSI > 36 and FIB4 < 1.3 (n = 509) | 1.12 (0.88–1.41) | 0.364 |
| Patients with HSI > 36 and FIB4 ≥ 1.3 (n = 91) | 8.38 (4.64–15.1) | < 0.001 |
| Adjusted model 1 | ||
| Patients with HSI ≤ 36 (n = 654) | Ref. | - |
| Patients with HSI > 36 and FIB4 < 1.3 (n = 509) | 0.99 (0.71–1.39) | 0.982 |
| Patients with HSI > 36 and FIB4 ≥ 1.3 (n = 91) | 4.83 (2.39–9.78) | < 0.001 |
| Adjusted model 2 (n = 1,043) | ||
| Patients with HSI ≤ 36 (n = 532) | Ref. | - |
| Patients with HSI > 36 and FIB4 < 1.3 (n = 432) | 0.93 (0.64–1.36) | 0.697 |
| Patients with HSI > 36 and FIB4 ≥ 1.3 (n = 79) | 6.93 (3.00-15.9) | < 0.001 |
Cohort size, n = 1,254, except where indicated. Data are expressed as odds ratio (OR) and 95% confidence interval, assessed by univariable and multivariable logistic regression analyses. The dependent variable of logistic regression models was: (a) the high or moderate Steno type 1 risk groups combined vs. the low Steno type 1 risk group, or (b) the high or Intermediate ASCVD risk groups combined vs. the low ASCVD risk group. Regression model 1 was adjusted for sex, BMI, diabetes duration, HbA1c, presence of CKD (defined as e-GFR < 60 mL/min/1.73 m2 or abnormal albuminuria), and lipid-lowering medication use. Regression model 2 was adjusted for the same model’s 1 covariates after excluding those with significant alcohol intake (n = 211)