Table 2.
Risk of adenoma with randomized aspirin treatment and subsequent NSAID use*
| Study phase | No. of subjects with adenoma/Total no. of subjects (%) | Adjusted RR (95% CI) | Ptrend† | |
| Randomized trial‡ | ||||
| Placebo | 143/285 (50.2) | 1.00 (reference) | ||
| Aspirin | 256/565 (45.3) | 0.90 (0.77 to 1.05) | ||
| Low-dose aspirin | 118/284 (41.6) | 0.84 (0.70 to 1.00) | ||
| High-dose aspirin | 138/281 (49.1) | 0.96 (0.81 to 1.14) | NA | |
| Posttreatment observational follow-up§ | ||||
| Aspirin treatment group | Posttreatment NSAID use‖ | |||
| Placebo | <2 days per week | 81/203 (39.9) | 1.00 (reference) | |
| Placebo | 2–4 days per week | 12/32 (37.5) | 1.02 (0.61 to 1.69) | |
| Placebo | ≥4 days per week | 18/43 (41.9) | 0.91 (0.59 to 1.39) | .46 |
| Aspirin | <2 days per week | 153/386 (39.6) | 1.00 (0.80 to 1.24) | |
| Aspirin | 2–4 days per week | 21/64 (32.8) | 0.84 (0.57 to 1.25) | |
| Aspirin | ≥4 days per week | 30/105 (28.6) | 0.67 (0.47 to 0.94) | .02 |
| Low-dose aspirin | <2 days per week | 76/194 (39.2) | 1.01 (0.78 to 1.30) | |
| Low-dose aspirin | 2–4 days per week | 8/28 (28.6) | 0.78 (0.43 to 1.41) | |
| Low-dose aspirin | ≥4 days per week | 15/56 (26.8) | 0.62 (0.39 to 0.98) | .03 |
| High-dose aspirin | <2 days per week | 77/192 (40.1) | 0.98 (0.76 to 1.27) | |
| High-dose aspirin | 2–4 days per week | 13/36 (36.1) | 0.89 (0.55 to 1.43) | |
| High-dose aspirin | ≥4 days per week | 15/49 (30.6) | 0.72 (0.46 to 1.12) | .08 |
Included 833 participants who completed study questionnaires and had at least one colonoscopy during the posttreatment follow-up interval. “Aspirin” refers to both doses combined. RR = relative risk; CI = confidence interval; low-dose aspirin = 81 mg; high-dose aspirin = 325 mg; NA = not applicable; NSAIDs = nonsteroidal anti-inflammatory drug.
Relative risks adjusted for age, sex, clinical center, number of lifetime adenomas before randomization, and duration of follow-up until the end of randomized treatment.
P value for the response trend with NSAID frequency of use among each randomized treatment group from a two-sided Wald test.
Relative risks adjusted for aspirin dose, age, sex, center, body mass index, time to year 3 study colonoscopy (overall aspirin treatment effects) or time to first colonoscopy after the end of randomized treatment (posttreatment NSAID effects), number of lifetime adenomas before randomization, and presence of adenomas during the randomized period.
NSAID use defined as the average number of days per week subjects took NSAIDs between the end of randomized aspirin treatment and the subsequent posttreatment colonoscopy.