Table 2.
Use of cellular phones and risk of tumors in subgroup meta-analysis of case-control studies.
| Factor | All | Hardell et al. Studies | INTERPHONE-Related Studies | Studies by Other Groups | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| No. | OR (95% CI) | I2 (%) | No. | OR (95% CI) | I2 (%) | No. | OR (95% CI) | I2 (%) | No. | OR (95% CI) | I2 (%) | |||
| 36 | 0.99 (0.91 to 1.07) | 47.4 | 10 | 1.15 (1.00 to 1.33) * | 40.1 | 9 | 0.81 (0.75 to 0.88) | 1.3 | 17 | 1.02 (0.92 to 1.13) | 8.1 | |||
| Difference in response rates a | Smaller (<14.5%) | 16 | 1.07 (0.94 to 1.21) | 54.2 | 10 | 1.15 (1.00 to 1.33) * | 40.1 | 1 | 0.81 (0.70 to 0.94) | n.a. | 5 | 0.99 (0.81 to 1.2) | 21.1 | |
| Larger (>14.5%) | 17 | 0.91 (0.82 to 1.02) | 23.8 | n.a. | 8 | 0.81 (0.73 to 0.91) | 13.7 | 9 | 1.02 (0.90 to 1.17) | 0.0 | ||||
| Use of blinding at interview | Used | 10 | 1.16 (1.01to 1.34) * | 39.4 | 9 | 1.16 (1.00 to 1.35) * | 45.4 | n.a. | 1 | 1.19 (0.54 to 2.59) | n.a. | |||
| Not used | 26 | 0.91 (0.84 to 0.99) | 32.1 | 1 | 0.90 (0.44 to 1.70) | n.a. | 9 | 0.81 (0.75 to 0.88) | 1.3 | 16 | 1.02 (0.91 to 1.13) | 13.0 | ||
| Methodolog-ical quality b | High | NOS | 17 | 1.11 (1.00 to 1.22) * | 20.1 | 9 | 1.16 (1.00 to 1.35) * | 45.4 | 1 | 0.90 (0.66 to 1.23) | n.a. | 7 | 1.08 (0.92 to 1.27) | 0.0 |
| NHLBI | 20 | 1.09 (0.99 to 1.20) | 29.3 | 8 | 1.18 (1.00 to 1.40) | 50.7 | 2 | 0.80 (0.54 to 1.20) | 0.0 | 10 | 1.03 (0.91 to 1.15) | 0.0 | ||
| Low | NOS | 19 | 0.88 (0.80 to 0.97) | 33.9 | 1 | 0.90 (0.44 to 1.70) | n.a. | 8 | 0.81 (0.74 to 0.88) | 8.5 | 10 | 0.99 (0.85 to 1.16) | 30.5 | |
| NHLBI | 16 | 0.86 (0.78 to 0.95) | 27.2 | 2 | 0.95 (0.64 to 1.41) | 0.0 | 7 | 0.81 (0.74 to 0.90) | 22.4 | 7 | 0.99 (0.79 to 1.24) | 31.2 | ||
| Funding by cellular phone industry | Not funded | 28 | 1.07 (0.98 to 1.17) | 21.9 | 10 | 1.15 (1.00 to 1.33) * | 40.1 | 1 | 0.95 (0.53 to 1.71) | n.a. | 17 | 1.02 (0.92 to 1.13) | 8.1 | |
| Funded | 8 | 0.81 (0.74 to 0.89) | 10.6 | n.a. | 8 | 0.81 (0.74 to 0.89) | 10.6 | n.a. | ||||||
| Type of case-control study | HCC | 9 | 0.95 (0.80 to 1.12) | 22.4 | n.a. | n.a. | 9 | 0.95 (0.80 to 1.12) | 22.4 | |||||
| PCC | 27 | 1.00 (0.91 to 1.09) | 53.7 | 10 | 1.15 (1.00 to 1.33) * | 40.1 | 9 | 0.81 (0.75 to 0.88) | 1.3 | 8 | 1.10 (0.96 to 1.26) | 0.0 | ||
| Malignancy | Malignant | 21 | 1.08 (0.97 to 1.20) | 31.4 | 9 | 1.18 (1.02 to 1.37) | 38.5 | 2 | 0.84 (0.54 to 1.31) | 0.0 | 10 | 0.97 (0.84 to 1.12) | 8.8 | |
| Benign | 14 | 0.86 (0.77 to 0.95) | 21.9 | 3 | 0.92 (0.74 to 1.14) | 38.6 | 8 | 0.81 (0.72 to 0.90) | 14.6 | 3 | 1.07 (0.83 to 1.39) | 4.3 | ||
a A difference in response rates between cases and controls was measured based on the average difference in response rates of 14.5% points between cases and controls when combining all the studies. Three studies [51,52,54] did not report response rates; b The methodological quality of each study was assessed by the Newcastle-Ottawa Scale (NOS) and the National Heart, Lung, and Blood Institute (NHLBI) quality assessment tool of case-control studies. The NOS score of ≥7 stars or the NHLBI score of ≥9 were considered as having high quality, and that of <7 stars and that of <9 were considered as having low quality; No.,number of studies; n.a., not available; HCC, hospital-based case-control study; PCC, population-based case-control study; ‘*’ indicates that cellular phone use statistically significantly increases the risk of tumor.