Table 3.
Subgroup and sensitivity analysis- vitamin D and atrial fibrillation incident, per 10 ng/ml increase
| Items | Number of studies | RR (95%CI) | P | P* h (%) | P# | |
|---|---|---|---|---|---|---|
| Result of primary analysis | 10 | 0.88 (0.78–0.98) | 0.03 | 85 | – | |
| Mean age | < 65 years | 3 | 0.87 (0.72–1.06) | 0.17 | 83 | 0.15 |
| > 65 years | 7 | 0.68 (0.52–0.89) | 0.005 | 95 | – | |
| Study design | Case-control | 4 | 0.13 (0.04–0.44) | < 0.001 | 74 | 0.001 |
| Cohort | 6 | 0.96 (0.93–0.99) | 0.007 | 0 | – | |
| Sample size | < 2000 | 5 | 0.21 (0.06–0.69) | < 0.001 | 96 | 0.01 |
| ≥2000 | 5 | 0.95 (0.92–0.99) | 0.005 | 0 | – | |
| Case | < 200 | 4 | 0.21 (0.06–0.69) | 0.007 | 74 | 0.001 |
| ≥200 | 6 | 0.97 (0.96–0.98) | 0.001 | 0 | – | |
| Region | European | 1 | 0.96 (0.91–1.01) | 0.14 | – | 0.006 |
| America | 5 | 0.96 (0.92–0.99) | 0.02 | 0 | – | |
| Asia | 4 | 0.13 (0.04–0.44) | 0.001 | 74 | – | |
| Study quality | < 6 | 2 | 0.49 (0.12–2.06) | 0.33 | 92 | 0.32 |
| ≥6 | 8 | 0.74 (0.61–0.91) | 0.004 | 94 | – | |
| Repeated with fixed model | 10 | 0.97 (0.96–0.98) | < 0.001 | 93 | – | |
| Exclusion of subjects | – | |||||
| Case-control omitted | 6 | 0.96 (0.93–0.99) | 0.007 | 0 | – | |
| Low quality excluded | 8 | 0.74 (0.61–0.91) | 0.004 | 94 | – | |
| Not adjusted age excluded | 9 | 0.75 (0.64–0.88) | < 0.001 | 94 | – | |
| Not adjusted BMI excluded | 8 | 0.77 (0.66–0.89) | < 0.001 | 94 | – | |
| Not adjusted sex excluded | 9 | 0.75 (0.64–0.88) | < 0.001 | 94 | – | |
| Not adjusted BMI excluded | 6 | 0.74 (0.60–0.92) | 88 | 0.006 | – | |
*P value of heterogeneity. #P for subgroup of subgroup. BMI body mass index; RR Ratio risk, CI confidence intervals