Background/Purpose: While a few studies have examined colchicine’s ability to prevent gout attacks, NSAIDs have not been formally studied in this regard, despite use of naproxen and its apparent efficacy in the febuxostat trials. Little data are available on effectiveness of prophylaxis with anti-gout medications on risk of gout attacks as used in the community. Further, dosing and choice of NSAID type for prophylactic purposes may vary widely in the community. We evaluated the association of colchicine and NSAID use with risk of recurrent gout attacks in an internet-based cohort of persons with gout.

Methods: We conducted an internet-based case-crossover study (each person acts as his/her own control, eliminating the effect of time-invariant confounders) to assess risk factors for gout attacks among persons with pre-existing gout. Subjects with gout who had ≥1 attack in the prior year were recruited online from across the US, with their gout diagnosis verified through medical records review.  Participants logged onto the study website when they had a gout attack and provided exposure information (including medication use) over the 14-day period prior to the gout attack (case-period) using an online questionnaire. The same questionnaire was collected for a 14-d period during an intercritical period (control-period).  Medication use between 3-14 days was not collected in an earlier study period (’03-’07) and was imputed assuming the data was missing at random. We examined the relation of colchicine and NSAID use, respectively, over specific time spans (prior 1-2 days, 1-7 days, and 1-14 days) to the risk of gout attacks using conditional logistic regression.

Results:   Of the 724 participants (from 49 US states and D.C.) who experienced ≥1 gout attack during the study period, 78.5% were male, 89% were White, and 58% had a college education. 38.4% were on some form of ULT on study entry (majority on allopurinol 300mg/d). Colchicine was taken primarily as 0.6mg/d and the most common NSAID used was ibuprofen. Colchicine protected against gout attacks when taken consistently over a 14 day period, but not over shorter periods (Table). The increased risk of gout attacks when colchicine was used only in the prior 2 days (i.e., without having taken it in the prior 3-14 days) likely indicates its use in anticipation of an impending gout attack (confounding by indication). However, very few only took colchicine in the prior 1 or 2 days only, limiting its interpretability. NSAID use did not appear to protect against gout attacks, regardless of consistent use (Table).

Conclusion: Colchicine was effective in protecting against gout attacks when taken consistently; intermittent or short-term use did not confer protection. We cannot exclude the possibility that particular formulations and/or doses of NSAIDs are effective; nonetheless, as used in the community, NSAIDs were not associated with lower risk of gout attacks.