Skip to main content
. 2017 Jan 30;7:41681. doi: 10.1038/srep41681

Table 4. The pooled ORs on the association between bilirubin concentration and the risk of diabetic complications.

  No. of study data Model for meta-analysis OR (95%CI) I2 (%) P for heterogeneity
Overall 38 R 0.77 (0.73, 0.81) 87.7 <0.001
Subgroup Analysis
Country
China 19 R 0.73 (0.68, 0.79) 92.4 <0.001
Korea 8 R 0.65 (0.49, 0.84) 72.0 0.001
Japan 9 R 0.90 (0.85, 0.95) 62.1 0.007
Else 2 R 0.71 (0.44, 1.14) 59.1 0.118
Number
<2000 30 R 0.82 (0.79, 0.86) 84.4 <0.001
≥2000 8 R 0.57 (0.45, 0.71) 78.4 <0.001
Design
Case-control 5 F 0.90 (0.87, 0.93) 0.0 0.453
Cross-sectional 26 R 0.73 (0.68, 0.78) 90.7 <0.001
Cohort 7 R 0.78 (0.65, 0.93) 70.8 0.002
Source of subjects
Hospital-based 31 R 0.78 (0.74, 0.82) 89.4 <0.001
Population-based 7 F 0.63 (0.49, 0.81) 41.7 0.113
FBG
<9.00 13 R 0.53 (0.43, 0.64) 90.5 <0.001
≥9.00 6 R 0.86 (0.81, 0.92) 74.3 0.002
HbA1C
<9.00 16 R 0.70 (0.63, 0.79) 90.6 <0.001
≥9.00 7 F 0.91 (0.89, 0.93) 0.0 0.622
NOS
Moderate quality 8 F 0.90 (0.87, 0.93) 21.8 0.257
High quality 30 R 0.72 (0.67, 0.77) 89.7 <0.001

R, random-effects model, F: fix-effects model.

HHS Vulnerability Disclosure