Incidence of diabetes mellitus and gout: a study of competitive risk
DONG Xiao-mei1, ZHANG Jia-yi1, ZHANG Si-heng1, WU Xue-ji2, PAN Bing-ying2, PANG Min-hui1, YANG Jian-wei1, ZHANG Jie1, CHEN Xiong-fei1
1. Department of Epidemiology, Medical College of Jinan University, Guangzhou, Guangdong 510632, China; 2. Department of Basic Public Health, Guangzhou Municipal Center for Disease Control and Prevention, Guangzhou, Guangdong 510440, China
Abstract:Objective To survey the changing tendency of risk of diabetes mellitus and gout incidence with the increasing age, and to explore how the two diseases mutually influence each other in their incidence. Methods A stratified cluster sampling method was used to select 14,752 permanent residents aged 35 years and above in Guangzhou City. The historical cohort about diabetes mellitus and gout onset was established based on a face-to-face interview and a retrospective survey of the residents’ health records. Multistate Markov model with competitive risk was introduced to estimate the risk of empirical onset of diabetes mellitus and gout at different ages and in different state. The risk curves of the diseases were fitted by parametric method. Using six common risk-of-risk functions, the distribution curve was fitted to the empirical risk data, and the most appropriate distribution curve was selected. Results Logistic distribution was the most suitable distribution curve. The cumulative incidence rates of populations from health to suffering from diabetes mellitus, from health to suffering from gout, from suffering from diabetes mellitus to suffering from both diabetes mellitus and gout, and from suffering from gout to suffering from both diabetes mellitus and gout were 0.314,8, 0.163,9, 0.172,8 and 0.420,4 respectively. The instantaneous incidence rates of populations from health to suffering from diabetes mellitus, from health to suffering from gout, from suffering from diabetes mellitus to suffering from both diabetes mellitus and gout, and from suffering from gout to suffering from both diabetes mellitus and gout were found to reach the maximum values at the ages of 70.8, 79.0, 60.4 and 68.1 years respectively. Conclusions Diabetes mellitus and gout are risk factors for each other. Gout and diabetes mellitus make the peak incidence of the counterpart happen 10 years earlier, and the incidence curve is higher and steeper. Differences in sample age range and age composition may be the important reasons for the inconsistencies in the present studies. The effect of sample age on the outcome should be further explored when we study the interaction of chronic diseases.
[1] Bo S, Cavallo-Perin P, Gentile L, et al. Hypouricemia and hyperuricemia in type 2 diabetes:two different phenotypes[J].Eur J Clin Invest, 2001, 31(4):318-321. [2] 李仕明. 糖尿病足的临床诊断与治疗进展[J]. 内科急危重症杂志,2002,8(1):1-4. [3] Cook DG, Shaper AG, Thelle DS, et al. Serum uric acid, serum glucose and diabetes:relationships in a population study[J].Postgrad Med J , 1986, 62(733):1001-1006. [4] Kodama S, Saito K, Yachi Y, et al. Association between serum uric acid and development of type 2 diabetes[J].Diabetes Care, 2009, 32(9):1737-1742. [5] Rodríguez G, Soriano LC, Choi HK. Impact of diabetes against the future risk of developing gout[J].Ann Rheum Dis, 2010, 69(12):2090-2094. [6] 杜琳,潘冰莹. 2009广州市社区卫生诊断报告[M]. 广州:广州出版社,2012:28,52. [7] 杨文英. 中国2型糖尿病防治指南(2010年版)[J]. 中国实用乡村医生杂志,2012,19(1):5-14. [8] 曾学军. 《2010年中国痛风临床诊治指南》解读[J]. 中国实用内科杂志,2012,32(6):438-441. [9] Meister R, Schaefer C. Statistical methods for estimating the probability of spontaneous abortion in observational studies—analyzing pregnancies exposed to coumarin derivatives[J]. Reprod Toxicol, 2008, 26(1):31-35. [10] Ruan PK, Gray RJ. Analyses of cumulative incidence functions via non‐parametric multiple imputation[J]. Stat Med, 2008, 27(27):5709-5724. [11] Allignol A, Beyersmann J. Software for fitting nonstandard proportional subdistribution hazards models[J]. Biostatistics, 2010, 11(4):674. [12] Choi HK, Ford ES. Haemoglobin A1c, fasting glucose, serum C-peptide and insulin resistance in relation to serum uric acid levels—the Third National Health and Nutrition Examination Survey[J]. Rheumatology (Oxford), 2008, 47(5):713-717. [13] Puig JG, Ruilope LM. Uric acid as a cardiovascular risk factor in arterial hypertension[J].J Hypertens, 1999, 17(7):869-872. [14] Tung YC, Lee SS, Tsai WC, et al. Association between gout and incident type 2 diabetes mellitus: a retrospective cohort study[J]. Am J Med, 2016, 129(11):1219. [15] Rho YH, Lu N, Peloquin CE, et al. Independent impact of gout on the risk of diabetes mellitus among women and men:a population-based, BMI-matched cohort study[J]. Ann Rheum Dis, 2016, 75(1):91-95. [16] Pan A, Teng GG, Yuan JM, et al. Bidirectional association between diabetes and gout: the Singapore Chinese Health Study[J]. Sci Rep, 2016, 6:25766. [17] Wijnands JMA, van Durme CMPG, Driessen JHM, et al. Individuals with type 2 diabetes mellitus are at an increased risk of gout but this is not due to diabetes:a population-based cohort study[J]. Medicine, 2015, 94(32):e1358. [18] Rodríguez G, Soriano LC, Choi HK. Impact of diabetes against the future risk of developing gout[J].Ann Rheum Dis, 2010, 69(12):2090-2094.