Table 3. Association between oral bisphosphonate use and colorectal cancer sites, Kaiser Permanente, Northern California, 1997–2011.
| N cases | OR | (95% CI) | |
|---|---|---|---|
| Site-specific associations | |||
| Cecum | 2,195 | 0.86 | (0.72, 1.03) |
| Appendix | 131 | 0.19 | (0.03, 1.42) |
| Ascending colon | 1,648 | 0.79 | (0.63, 0.99) |
| Hepatic flexure of colon | 541 | 0.84 | (0.57, 1.26) |
| Transverse colon | 830 | 1.15 | (0.86, 1.54) |
| Splenic flexure of colon | 348 | 0.63 | (0.33, 1.21) |
| Descending colon | 547 | 1.01 | (0.64, 1.60) |
| Sigmoid colon | 2,803 | 0.77 | (0.61, 0.96) |
| Overlapping lesion of colon | 61 | 0.55 | (0.15, 2.10) |
| Colon, NOS | 149 | 0.40 | (0.14, 1.15) |
| Rectosigmoid junction | 752 | 0.66 | (0.42, 1.02) |
| Rectum | 2,500 | 0.73 | (0.56, 0.94) |
Cases and controls were matched on sex, age at time of index date (+/− 2 years), duration of membership prior to index date (+/− 1 year), race, and region of residence.
The adjusted model additionally included age, smoking, alcohol use, Charlson comorbidity index, use of NSAIDs, and previous lower endoscopy.