Abstract
Epidemiologic evidence is conflicting regarding the potential for nonsteroidal antiinflammatory drugs (NSAIDs) to lower the risk of lung cancer. We thus determined the incidence of lung cancer among a cohort of over 242,000 adults in Denmark who had been issued NSAID prescriptions during 1991–2002 and calculated the relative risks of this cancer according to the numbers of prescriptions received. Similar analyses were conducted among a subset with chronic obstructive pulmonary disease, of whom the large majority was assumed to have been smokers, as an indirect control for smoking. Utilization of Danish registries ensured essentially complete and unbiased ascertainment of prescriptions and cancer outcomes. The relative risks of lung cancer among NSAID users redeeming 10 or more prescriptions vs. nonusers in the general population and among chronic obstructive pulmonary disease patients were 1.51 (95% CI = 1.33–1.71) and 1.55 (95% CI = 0.93–2.60) and tended to decline with decreasing numbers of NSAIDs prescribed. Separate analyses for adenocarcinoma and squamous cell carcinoma also failed to show reduced risks among NSAID users. Confounding by indication may have contributed to the increased risks, but the study provides no evidence to suggest that recent NSAID intake may be protective for lung cancer. © 2005 Wiley‐Liss, Inc.