Table 7.
Diabetes medication classes and pancreatic cancer risk.
| RR 95% CI | P value for effect | P value for Egger's test | I2 (%) | P value for difference between subgroups | ||
|---|---|---|---|---|---|---|
| Biguanides | ||||||
| All studies (n = 8) | 1.08 (0.77–1.52) | 0.65 | 0.53 | 96.10 | – | |
| Study design | Cohort (n = 5) | 1.01 (0.62–1.64) | 0.98 | – | 97.18 | 0.43 |
| Case–control (n = 3) | 1.25 (1.03–1.50) | 0.02 | – | 43.74 | ||
| Risk of bias | Low risk (n = 5) | 1.05 (0.69–1.59) | 0.83 | – | 97.74 | 0.97 |
| Moderate risk (n = 3) | 1.04 (0.78–1.39) | 0.80 | – | < 0.001 | ||
| Study population | Asian (n = 3) | 0.97 (0.51–1.83) | 0.92 | – | 98.56 | 0.62 |
| Western (n = 5) | 1.15 (0.91–1.45) | 0.23 | – | 62.28 | ||
| Incretin-based medicines | ||||||
| All studies (n = 13) | 0.73 (0.47–1.11) | 0.14 | 0.56 | 96.69 | – | |
| Study design | RCT (n = 2) | 0.76 (0.20–2.95) | 0.69 | – | 28.67 | 0.53 |
| Cohort (n = 10) | 0.70 (0.42–1.19) | 0.19 | – | 97.33 | ||
| Case–control (n = 1) | 0.96 (0.81–1.12) | 0.58 | – | – | ||
| Risk of bias | Low risk (n = 10) | 0.68 (0.42–1.09) | 0.11 | – | 97.10 | 0.02 |
| Moderate risk (n = 1) | 1.41 (1.13–1.75) | < 0.001 | – | – | ||
| High risk (n = 2) | 0.76 (0.20–2.95) | 0.69 | – | 28.67 | ||
| Study population | Asian (n = 4) | 0.84 (0.41–1.69) | 0.62 | – | 96.77 | 0.67 |
| Western (n = 6) | 0.63 (0.24–1.65) | 0.34 | – | 97.79 | ||
| Multiple continents (n = 3) | 0.95 (0.81–1.12) | 0.55 | – | < 0.001 | ||
| Alpha-glucosidase inhibitors | ||||||
| All studies (n = 1) | 1.15 (0.99–1.34) | 0.06 | – | – | – | |
| Study design | Cohort (n = 1) | 1.15 (0.99–1.34) | 0.06 | – | – | – |
| Risk of bias | Low risk (n = 1) | 1.15 (0.99–1.34) | 0.06 | – | – | – |
| Study population | Asian (n = 1) | 1.15 (0.99–1.34) | 0.06 | – | – | – |
| Insulin secretagogues | ||||||
| All studies (n = 5) | 1.26 (1.01–1.57) | 0.04 | 0.79 | 86.08 | – | |
| Study design | Cohort (n = 2) | 1.09 (1.01–1.17) | 0.02 | – | < 0.001 | 0.04 |
| Case–control (n = 3) | 1.46 (1.11–1.91) | 0.01 | – | 74.71 | ||
| Risk of bias | Low risk (n = 4) | 1.31 (1.01–1.69) | 0.04 | – | 89.37 | 0.31 |
| Moderate risk (n = 1) | 1.05 (0.77–1.45) | 0.74 | – | – | ||
| Study population | Asian (n = 2) | 1.09 (1.01–1.17) | 0.02 | – | < 0.001 | 0.04 |
| Western (n = 3) | 1.46 (1.11–1.91) | 0.01 | – | 74.71 | ||
| Thiazolidinediones | ||||||
| All studies (n = 7) | 0.73 (0.47–1.14) | 0.17 | 0.66 | 86.70 | – | |
| Study design | RCT (n = 1) | 0.31 (0.12–0.85) | 0.02 | – | – | 0.01 |
| Cohort (n = 4) | 0.59 (0.39–0.89) | 0.01 | – | 69.18 | ||
| Case–control (n = 2) | 1.26 (0.79–2.02) | 0.33 | – | 67.60 | ||
| Risk of bias | Low risk (n = 5) | 0.77 (0.44–1.34) | 0.35 | – | 89.99 | 0.12 |
| Moderate risk (n = 1) | 0.98 (0.64–1.48) | 0.91 | – | – | ||
| High risk (n = 1) | 0.31 (0.12–0.85) | 0.02 | – | – | ||
| Study population | Asian (n = 2) | 0.80 (0.37–1.75) | 0.58 | – | 76.96 | 0.75 |
| Western (n = 5) | 0.68 (0.35–1.34) | 0.27 | – | 88.92 | ||
| Insulins | ||||||
| All studies (n = 8) | 2.41 (1.08–5.36) | 0.03 | 0.61 | 99.01 | – | |
| Study design | Cohort (n = 5) | 2.92 (0.98–8.70) | 0.06 | – | 99.25 | 0.43 |
| Case–control (n = 3) | 1.77 (0.95–3.30) | 0.07 | – | 92.94 | ||
| Risk of bias | Low risk (n = 6) | 2.06 (1.55–2.73) | < 0.001 | – | 77.20 | 0.62 |
| Moderate risk (n = 2) | 3.60 (0.40–32.23) | 0.25 | – | 99.38 | ||
| Study population | Asian (n = 3) | 2.16 (1.52–3.09) | < 0.001 | – | < 0.001 | 0.76 |
| Western (n = 5) | 2.56 (0.92–7.14) | 0.07 | – | 99.42 | ||