Exploring Differences in the Aspirin-Colorectal Cancer Association by Sex and Race/Ethnicity : The Multiethnic Cohort Study
Menée sur une cohorte multi-ethnique incluant 183 199 participants (durée médiane de suivi : 16,1 ans), cette étude évalue, en fonction du sexe et de l'origine ethnique, l'association entre l'utilisation d'aspirine ou d'autres anti-inflammatoires non stéroïdiens et le risque de cancer colorectal (4 882 cas)
Background : Evidence has accumulated that long-term use of aspirin and other non-steroidal anti-inflammatory drugs (NSAIDs) protects against colorectal cancer (CRC). We tested whether the inverse associations between NSAIDs and CRC is similarly observed across sexes and five racial/ethnic groups (Japanese, Latino, African American, Native Hawaiian, and white) in the Multiethnic Cohort (MEC) Study. Methods : During a mean follow-up of 16.1 years, we identified 4,882 invasive incident CRC cases among 183,199 eligible participants. Cox proportional hazards models were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs). Results: Use of aspirin and other NSAIDs was associated with a lower incidence of CRC in men (HR=0.77; 95% CI, 0.69-0.86 for current vs. never users of aspirin) but not in women (Pinteraction=0.005). Among male current users, a reduced risk was observed with {greater than or equal to}6 years of aspirin or total NSAID use. The inverse association with current NSAID use in men was observed in all racial/ethnic groups, except for Native Hawaiians, and was stronger in whites. Conclusion : Our findings suggest that the benefit of NSAIDs for CRC may be strongest for white men and generalizes to African American, Japanese and Latino but not to Native Hawaiian men. The lack of inverse association observed in women and Native Hawaiian men in the MEC should be interpreted with caution. Impact : Since only very few ethnic/racial groups are likely to be represented in trials of NSAIDs and CRC, it is important to conduct prospective observational studies in various populations to test the generalizability of their results.