Wang Liya, Sawhney Monika, Zhao Yingnan, Carpio Gandahari Rosa, Fonseca Vivian, Shi Lizheng
Department of Global Health Systems and Development, School of Public Health and Tropic Medicine, Tulane University, New Orleans, Louisiana.
Department of Public Health, College of Health Professions, Marshall University, Huntington, West Virginia.
Clin Ther. 2015 Jun 1;37(6):1206-15. doi: 10.1016/j.clinthera.2015.03.010. Epub 2015 Apr 7.
This study aimed to determine the association between colchicine use and the incidence of diabetes in a cohort of patients with gout.
This is a retrospective study of 27,876 adults with gout identified via the Veterans Integrated Services Network 16 data warehouse. Patients had up to 11 years of follow-up (January 1999 through December 2010). The final study sample consisted of 1046 pairs of 1:1 propensity score-matched patients from the colchicine treated and control cohorts. Time to first diabetes development since the first gout diagnosis was modeled.
After the propensity score matching, the 12-month baseline variables (eg, age, sex, race, index year, body mass index, serum uric acid, antigout drug use, and health care use) were comparable between the matched cohorts (P > 0.05 for all). Among the 1046 matched pairs, 234 patients who had taken colchicine and 224 patients who had never taken colchicine developed diabetes; the incidence rates were 38.95 and 39.02 per 1000 patient-years, respectively. In Poisson and Cox proportional hazards regression, the risk of incident diabetes was reduced with increased duration of colchicine use, but the difference was not statistically significant (P > 0.05). In a time-varying Cox proportional hazards model, the hazard ratio for incident diabetes among patients who had taken colchicine was 0.877 (95% CI, 0.662-1.163; P = 0.362) compared with those who had not taken colchicine.
This study suggests a possible duration- or dose-related association between colchicine use and reduced risk of diabetes in adults with gout even though the risk reduction was not significant. Further studies are needed to confirm findings from this study.
本研究旨在确定痛风患者队列中使用秋水仙碱与糖尿病发病率之间的关联。
这是一项通过退伍军人综合服务网络16数据仓库识别出的27876名成年痛风患者的回顾性研究。患者的随访时间长达11年（1999年1月至2010年12月）。最终研究样本包括来自秋水仙碱治疗组和对照组的1046对1:1倾向评分匹配患者。对自首次痛风诊断以来首次发生糖尿病的时间进行建模。
倾向评分匹配后，匹配队列之间的12个月基线变量（如年龄、性别、种族、索引年份、体重指数、血清尿酸、抗痛风药物使用和医疗保健使用）具有可比性（所有P>0.05）。在1046对匹配对中，服用秋水仙碱的234名患者和从未服用秋水仙碱的224名患者发生了糖尿病；发病率分别为每1000患者年38.95和39.02。在泊松和Cox比例风险回归中，糖尿病发病风险随着秋水仙碱使用时间的增加而降低，但差异无统计学意义（P>0.05）。在时变Cox比例风险模型中，与未服用秋水仙碱的患者相比，服用秋水仙碱的患者发生糖尿病的风险比为0.877（95%CI，0.662-1.163；P=0.362）。
本研究表明，痛风成年患者使用秋水仙碱与糖尿病风险降低之间可能存在持续时间或剂量相关的关联，尽管风险降低并不显著。需要进一步研究来证实本研究的结果。
