Table 3.
Meta regression analyses of excess relative risk for each major cardiovascular disease endpoint group, with restriction by dose and dose rate
| Cardiovascular disease endpoint | No. of study endpoints | Meta excess relative risk/Gy (95% CI) | Residual heterogeneity P value | I2 (%) (Knapp-Hartung 95% CI) | Egger selection test P value | Duval-Tweedie trim-fill selection bias corrected meta excess relative risk/Gy (95% CI) |
|---|---|---|---|---|---|---|
| Full data | ||||||
| Ischaemic heart disease | 38 | 0.073 (0.047 to 0.099) | 0.01 | 17.80 (3.52 to 95.64) | 0.002 | 0.073 (0.052 to 0.095) |
| Other heart disease* | 27 | 0.034 (0.020 to 0.049) | 0.49 | 0.00 (0.00 to 84.81) | 0.13 | 0.034 (0.021 to 0.048) |
| Cerebrovascular disease | 32 | 0.188 (0.093 to 0.283) | <0.001 | 83.56 (68.52 to 99.22) | 0.05 | 0.184 (0.099 to 0.269) |
| Other cardiovascular disease† | 12 | 0.172 (–0.029 to 0.373) | <0.001 | 90.27 (68.02 to 98.43) | 0.49 | 0.172 (–0.029 to 0.373) |
| All cardiovascular disease (using maximal cardiovascular disease data per study) | 105 | 0.106 (0.076 to 0.135) | <0.001 | 88.61 (85.58 to 95.99) | <0.001 | 0.102 (0.075 to 0.129) |
| Maximum dose ≤0.5 Gy only | ||||||
| Ischaemic heart disease | 6 | 0.438 (–0.131 to 1.007) | 0.17 | 21.89 (0.00 to 99.74) | 0.68 | 0.438 (–0.131 to 1.007) |
| Other heart disease* | 2 | –0.108 (–0.528 to 0.313) | 0.76 | 0.00 (0.00 to 68.13) | 0.39 | –0.188 (–0.617 to 0.242) |
| Cerebrovascular disease | 7 | 0.542 (–0.281 to 1.366) | 0.24 | 0.00 (0.00 to >97.54) | 0.79 | 0.542 (–0.281 to 1.366) |
| Other cardiovascular disease† | 1 | –1.80 (–11.95 to 8.35) | NA | NA | NA | NA |
| All cardiovascular disease (using maximal cardiovascular disease data per study) | 9 | 0.452 (0.064 to 0.840) | 0.36 | 17.88 (0.00 to 98.38) | 0.26 | 0.452 (0.092 to 0.811) |
| Low dose rate data only | ||||||
| Ischaemic heart disease | 22 | 0.202 (0.085 to 0.319) | 0.13 | 45.34 (0.00 to 86.89) | 0.86 | 0.202 (0.085 to 0.319) |
| Other heart disease* | 6 | –0.207 (–0.456 to 0.042) | 0.98 | 0.00 (0 to >0) | 0.55 | –0.241 (–0.727 to 0.246) |
| Cerebrovascular disease | 21 | 0.298 (0.101 to 0.495) | <0.001 | 61.07 (29.38 to 99.94) | 0.18 | 0.294 (0.130 to 0.458) |
| Other cardiovascular disease† | 6 | 0.166 (–0.069 to 0.401) | 0.17 | 39.95 (0.00 to 99.58) | 0.79 | 0.166 (–0.069 to 0.401) |
| All cardiovascular disease (using maximal cardiovascular disease data per study) | 41 | 0.229 (0.136 to 0.322) | <0.001 | 68.23 (31.40 to 92.81) | 0.09 | 0.224 (0.134 to 0.315) |
| Maximum dose ≤0.5 Gy or low dose rate only | ||||||
| Ischaemic heart disease | 24 | 0.205 (0.092 to 0.318) | 0.07 | 39.39 (0.00 to 93.39) | 0.31 | 0.205 (0.092 to 0.318) |
| Other heart disease* | 8 | –0.168 (–0.429 to 0.094) | 0.90 | 0.00 (0.00 to 55.12) | 0.29 | –0.233 (–0.567 to 0.101) |
| Cerebrovascular disease | 23 | 0.306 (0.127 to 0.485) | <0.001 | 57.05 (22.37 to 99.87) | 0.17 | 0.303 (0.148 to 0.458) |
| Other cardiovascular disease† | 6 | 0.166 (–0.069 to 0.401) | 0.17 | 39.95 (0.00 to 99.58) | 0.79 | 0.166 (–0.069 to 0.401) |
| All cardiovascular disease (using maximal cardiovascular disease data per study) | 44 | 0.231 (0.141 to 0.320) | <0.001 | 65.03 (31.27 to 94.25) | 0.04 | 0.226 (0.141 to 0.311) |
All four main endpoints and all cardiovascular disease are analysed separately. Values used in the analysis are from supplement S3 tables S3.4-S3.5. Random effects models are fitted via restricted maximum likelihood. CI=confidence interval; NA=not available.
*
Heart disease other than ischaemic heart disease.
†
Cardiovascular disease other than heart disease and cerebrovascular disease.