Table 2.
Subgroup | No. of studies | OR (95% CI) | I2 (%) | P a | P b |
---|---|---|---|---|---|
Study design | |||||
Cohort | 4 | 1.33 (1.18, 1.49) | 43.8 | 0.148 | 0.518 |
Cross-sectional | 5 | 1.47 (1.28, 1.70) | 90.7 | <0.001 | |
Region | |||||
USA | 5 | 1.33 (1.26, 1.40) | 37.8 | 0.154 | 0.106 |
Non-USA | 3 | 1.64 (1.26, 2.14) | 86.3 | 0.001 | |
Sample size | |||||
< 35,000 | 5 | 1.59 (1.32, 1.92) | 73.4 | 0.005 | 0.079 |
≥ 35,000 | 4 | 1.32 (1.25, 1.40) | 54.3 | 0.087 | |
Gender | |||||
Men | 4 | 1.31 (1.25, 1.37) | 13.4 | 0.326 | 0.547 |
Women | 3 | 1.43 (1.21, 1.69) | 78.2 | 0.010 | |
Adjustment for confounding factors | |||||
< 5 | 5 | 1.51 (1.30, 1.76) | 91.2 | <0.001 | 0.275 |
≥ 5 | 4 | 1.32 (1.14, 1.52) | 0 | 0.497 |
OR odds ratio, CI confidence interval
a P value for heterogeneity among studies assessed with Cochran’s Q test
b P value for interaction evaluated by meta-regression models