Assessment of heterogeneity among studies examining the association between global DNA methylation and breast cancer risk
| Subgroup | Estimates (n) | RR (95% CI); P | I2 | Meta-regression P |
|---|---|---|---|---|
| Study demographics | ||||
| Mean age | ||||
| <50 | 5 | 0.83 (0.58–1.19); P = 0.31 | 0.0% | 0.08 |
| 50–<60 | 5 | 0.47 (0.32–0.69); P = 0.02 | 64.7% | |
| 60+ | 4 | 1.01 (0.54–1.90); P = 0.97 | 92.0% | |
| Country | ||||
| North American | 10 | 0.83 (0.55–1.25); P = 0.37 | 86.6% | 0.14 |
| Other | 4 | 0.47 (0.28–0.79); P = 0.004 | 67.9% | |
| Study design | ||||
| Type | ||||
| Prospectivea | 5 | 0.62 (0.43–0.87); P = 0.007 | 58.1% | 0.49 |
| Case–control | 9 | 0.77 (0.45–1.32); P = 0.34 | 90.1% | |
| Blood draw type | ||||
| >3 years prior to diagnosis | 3 | 0.52 (0.28–0.97); P = 0.04 | 71.5% | 0.37 |
| <3 years prior to diagnosis | 11 | 0.77 (0.49–1.20); P = 0.24 | 88.8% | |
| DNA methylation assessment | ||||
| Genomic context | ||||
| Illumina 450k | 3 | 0.53 (0.27–1.06); P = 0.07 | 69.8% | 0.26 |
| Global | 3 | 0.59 (0.33–1.04); P = 0.07 | 69.9% | |
| Repetitive elements | 6 | 0.84 (0.63–1.13); P = 0.25 | 54.0% | |
| LUMA | 4 | 1.02 (0.28–3.76); P = 0.98 | 96.4% | |
| Assay type | ||||
| Non-LUMA | 12 | 0.64 (0.50–0.83); P < 0.001 | 58.3% | 0.25 |
| LUMA | 3 | 1.02 (0.28–3.76); P = 0.98 | 96.4% | |
↵aIncludes nested case–control studies.