Effects of Uric Acid-Lowering Treatment on Glycemia: A Systematic Review and Meta-Analysis

Background: Serum uric acid levels have been shown to be associated with increased risk of diabetes. However, it remains unclear whether uric acid-lowering therapy (ULT) is associated with improved glycemic status. This study aimed to summarize evidence from randomized controlled trials (RCTs) to investigate whether ULT reduces fasting blood glucose (FBG) and glycated hemoglobin A1c (HbA1c) levels. Methods: PubMed, Embase, and the Cochrane Library were searched from inception until April 10, 2019. Moreover, in order to maximize the search for articles on the same topic, the reference lists of included studies, relevant review articles and systematic reviews were reviewed. Parallel RCTs investigating the effect of ULT on FBG or HbA1c levels were considered for inclusion. An English language restriction was applied. Data were screened and extracted independently by two researchers. Meta-analyses were performed using random-effects models to calculate the weighted mean differences (WMDs) and 95% confidence intervals (CIs). Results: Four trials with 314 patients reported the effect of ULT with allopurinol on FBG and 2 trials with 141 patients reported the effect of ULT with allopurinol on HbA1c. Treatment with allopurinol resulted in a significant decrease in FBG (WMD: −0.61 mmol/L, 95% CI: −0.93 to −0.28), but only a trend of reduction in HbA1c (WMD: −0.47%, 95% CI: −1.16 to 0.22). Notably, the subgroup analyses showed that treatment with allopurinol was associated with reduced FBG levels in patients without diabetes (WMD: −0.60 mmol/L, 95% CI: −0.99 to −0.20), but not in patients with diabetes. In addition, the dose of allopurinol treatment ≥200 mg daily resulted in a reduction of FBG levels (WMD: −0.59 mmol/L, 95% CI: −0.95 to −0.23), whereas low-dose allopurinol (<200 mg daily) had no effect on FBG levels. Conclusions: The findings suggest that ULT with allopurinol may be effective at reducing glycemia, but such an improvement does not appear to be observed in patients with diabetes. The findings require confirmation in additional trials with larger sample sizes.