Table 2.
Subgroup-sensitivity analyses on studies investigating the association between nonalcoholic fatty liver disease and incident hypertension
| Hazard ratios (95% CI) | Test for overall effect | Study number | Between group heterogeneity | |
|---|---|---|---|---|
| Duration of follow-up | ||||
| <6 years | 1.48 (1.16–1.89) | z = 3.182, P = 0.001 | 7 | P = 0.176 |
| ≥6 years | 2.10 (1.35–3.20) | z = 3.325, P = 0.001 | 4 | |
| Adjustment for baseline BP | ||||
| Absent | 1.78 (1.16–2.73) | z = 2.647, P = 0.008 | 5 | P = 0.655 |
| Present | 1.60 (1.33–1.93) | z = 4.881, P < 0.001 | 6 | |
| Adjustment for adipositya | ||||
| Absent | 2.44 (1.84–3.22) | z = 6.229, P < 0.001 | 4 | P < 0.001 |
| Present | 1.36 (1.20–1.54) | z = 4.871, P < 0.001 | 7 | |
| Geographical region | ||||
| Europe/USA | 1.97 (1.23–3.15) | z = 2.830, P = 0.005 | 3 | P = 0.401 |
| Asia | 1.58 (1.27–1.96) | z = 4.078, P < 0.001 | 8 | |
All studies included in Fig. 2 were analyzed in these subgroup analyses.
a
Inclusion of either BMI or waist circumference in the multivariable logistic regression model.
BP, blood pressure; CI, confidence interval.