The predictive value of mean serum uric acid levels for developing prediabetes. (August 2016)

• Record Type: Journal Article
• Authors: Zhang, Qing; Bao, Xue; Meng, Ge; Liu, Li; Wu, Hongmei; Du, Huanmin; Shi, Hongbin; Xia, Yang; Guo, Xiaoyan; Liu, Xing; Li, Chunlei; Su, Qian; Gu, Yeqing; Fang, Liyun; Yu, Fei; Yang, Huijun; Yu, Bin; Sun, Shaomei; Wang, Xing; Zhou, Ming; Jia, Qiyu; Zhao, Honglin; Huang, Guowei; Song, Kun; Niu, Kaijun
• Is Part Of: Diabetes research and clinical practice. Volume 118(2016)
• Page Start: 79
• Page End: 89
• Publication Date: 2016-08
• Subjects: Uric acid -- Prediabetes -- Cohort studies; Diabetes -- Periodicals; Diabetes Mellitus -- Periodicals; 616.462
• DOI: 10.1016/j.diabres.2016.06.011
• Languages: English
• ISSNs: 0168-8227

Abstract

Highlights: Mean serum uric acid concentration is a strong predictor for incident prediabetes. Baseline serum uric acid value is only weakly related to prediabetes risk. Mean serum uric acid value is a reliable factor to stratify prediabetes risk. Abstract: Aims: We aimed to assess the predictive value of mean serum uric acid (SUA) levels for incident prediabetes. Methods: Normoglycemic adults ( n = 39, 353) were followed for a median of 3.0 years. Prediabetes is defined as impaired fasting glucose (IFG), impaired glucose tolerance (IGT), or impaired HbA1c (IA1c), based on the American Diabetes Association criteria. Serum SUA levels were measured annually. Four diagnostic strategies were used to detect prediabetes in four separate analyses (Analysis 1: IFG. Analysis 2: IFG+IGT. Analysis 3: IFG+IA1c. Analysis 4: IFG+IGT+IA1c). Cox proportional hazards regression models were used to assess the relationship between SUA quintiles and prediabetes. C-statistic was additionally used in the final analysis to assess the accuracy of predictions based upon baseline SUA and mean SUA, respectively. Results: After adjustment for potential confounders, the hazard ratios (95% confidence interval) of prediabetes for the highest versus lowest quintile of mean SUA were 1.22 (1.10, 1.36) in analysis 1; 1.59 (1.23, 2.05) in analysis 2; 1.62 (1.34, 1.95) in analysis 3 and 1.67 (1.31, 2.13) in analysis 4. In contrast, for baseline SUA, significance was only reached in analyses 3 and 4. Moreover, compared with baseline SUA, mean SUA value was associated with a significant increase in the C-statistic ( P < 0.001). Conclusions: Mean SUA value was strongly and positively related to prediabetes risk, and showed better predictive ability for prediabetes than baseline SUA.