Hyperuricemia and the risk of gout are commonly seen in patients with type 2 diabetes. It has also been reported that chronically elevated circulating uric acid concentrations are associated with an increased risk of hypertension, cardiovascular disease and chronic kidney disease, all of which are known diabetes‐associated complications1, 2.

By analyzing the USA nationwide commercial insurance database, Fralick et al.3 reported a lower risk of gout in new users of sodium–glucose cotransporter 2 (SGLT2) inhibitors, as compared with new users of glucagon like peptide‐1 (GLP‐1) receptor agonists (with 1:1 propensity score matched), in patients with type 2 diabetes. Several features of this study should be taken into consideration. It involved real‐time data and enrolled nearly 300,000 adults with type 2 diabetes who had received SGLT2 inhibitors or GLP‐1 receptor agonists. Additionally, the study collected information regarding the diagnosis of gout, as well as the prescribed medication used for treatment from March 2013 to December 2017. What the authors discovered was that new users of SGLT2 inhibitors experienced a 64% lower risk of gout, as compared with new users of GLTP‐1 agonists (4.9 vs 7.8 events per 1,000 person years) in patients with type 2 diabetes. Data obtained from this study are particularly welcome for two reasons. First, there are a limited number of drugs that have been shown to safely lower uric acid levels. As a matter of fact, there have been concerns raised in a recent report that the use of febuxostat, a relatively new uric acid‐lowering agent, caused a higher risk for cardiovascular death and all‐cause mortality as compared with allopurinol, a traditional uric acid lowering drug. Second, the US Food and Drug Administration is encouraging the use of applications of real‐world data to evaluate supplemental indications for already approved medications.

Presumably, the glycosuria caused by the use of SGLT2 inhibitors helps uric acid to be secreted into the urine. The most likely mechanism postulated was that, through SGLT2 inhibitors, an increase in uricosuria occurred by suppressing the activity of glucose transporter 9b, a hexose/urate transporter located at the proximal tubular cells across the basolateral membrane1. However, the exact mechanisms have not been fully illustrated. It is estimated that the increased uric acid elimination through the use of SGLT2 inhibitors usually lowers circulating uric acid concentrations by approximately 35–45 μmol/L (0.60–0.75 mg/dL) in individuals with a baseline uric acid value in the normal concentration range of approximately 200–400 μmol/L (~3.3–6.7 mg/dL)1. In the Empagliflozin Reduced Mortality and Hospitalization for Heart Failure Across the Spectrum of Cardiovascular Risk (EMPA‐REG OUTCOME) trial, patients who received empagliflozin showed a lower serum uric acid level of approximately 30 μmol/L compared with those who had been assigned a placebo4. Alternatively, it was shown that the use of GLP‐1 agonists in patients with type 2 diabetes did not show any effect on serum uric acid levels, when compared with a placebo or comparator medication. To confirm their observations, the authors also examined those who had undergone up to 1 year of index medication exposure, and found that adults with type 2 diabetes who had received SGLT2 inhibitors had a 73% lower risk of gout, as compared with those who had received GLP‐1 agonists (5.6 vs 7.7 events per 1,000 person years). The authors also clearly showed that the users of SGLT2 inhibitors had a reduced risk of gout with a hazard ratio of 0.66 (95% confidence interval 0.58–0.75) compared with a group of new users who were prescribed dipeptidyl peptidase‐4 inhibitors, which is one of the most commonly prescribed second‐line medications for adults with diabetes, without lowering serum uric acid levels.

Previous observational and interventional studies have shown that uric acid acts as a modifiable risk factor in the progression of chronic kidney disease for type 2 diabetes1, 2. This is also true for those diabetes patients with lesser degrees of hyperuricemia. Several potential benefits are listed from the use of SGLT2 inhibitors in Table 1. One item of note that should be pointed out is that the authors particularly excluded patients with a history and diagnosis of gout at baseline. Those with prevalent gout and those with a higher baseline risk for gout (such as the elderly, and those diagnosed with chronic renal diseases and established cardiovascular disease) were not enrolled. Thus, the benefits from the use of SGLT2 inhibitors might be underestimated.

The magnitude in the reduction of serum uric acid levels in those users of SGLT2 inhibitors was generally modest, and believed to be less potent than xanthine oxidase inhibitors, such as allopurinol or febuxostat. However, given that the modes of action are different and might be potentially complementary, it is likely that the use of SGLT2 inhibitors for diabetes with hyperuricemia might offer an additive effect if combined for use with a xanthine oxidase inhibitor. Alternatively, it has been postulated that increased uric acid elimination in the urine through the use of an SGLT2 inhibitor might lead to an increased risk of renal calculi; however, until now, studies have yet to observe this possible long‐term risk. It is also hoped that non‐diabetic adults with hyperuricemia and/or a history of gout might gain benefits from the use of SGLT2 inhibitors. If proven, this will be useful for adults with hyperuricemia who do not have diabetes.

As with all observational studies, the findings by Fralick et al.3 might be subject to unmeasured confounding, despite many confounders being well balanced by propensity score matched analysis. There are very limited laboratory data regarding hemoglobin A1c and serum creatinine levels available (these values were available for approximately 5% of the patients with a baseline hemoglobin A1c level of 8.6%, and a baseline creatinine level of 78.7 μmol/L (0.89 mg/dL). This greatly devalues the opportunity to examine the effects of the changes in glycemic control and the alterations of renal function on the risk of gout. Additionally, there were no serum uric acid levels available that might enable a comparison with the uric acid‐lowering effects of these two glucose‐lowering agents. Information with regard to the bodyweight of the participants was not reported in this article. Therefore, it is not known whether the benefits of less frequent gout favored those with a higher or lower body mass index given the known effects of SGLT2 inhibitors on bodyweight5. One might argue that the use of a diagnosis code for gout, plus a prescription claim for medication used to treat gout for up to 14 days, would be sufficient to capture all incidences of gout and flare up.

There was 41 days difference in the mean follow‐up time (302 days for SGLT2 inhibitor vs 261 days for GLP‐1 agonist)3. In addition, a sizable number of patients did not refill their medications during the follow‐up period. Specifically, 45% of the patients discontinued their SGLT2 inhibitor medication, with 49% discontinuing their GLP‐1 agonist. There were also mixed effects in that 9% of the patients who had started GLP‐1 agonist treatment and subsequently filled a prescription for an SGLT2 inhibitor, whereas 8% of the patients who had started treatment with an SGLT2 inhibitor subsequently filled a prescription for a GLP‐1 agonist. As these phenomena did occur in real‐world practice, the data presented should be interpreted cautiously.

In summary, the authors of this real‐world observational cohort study reported a relative risk reduction in gout of nearly 40%, and an absolute risk reduction of approximately three fewer adults who experienced gout per 1,000 person‐years by new users of SGLT2 inhibitors, as compared with new users of GLP‐1 agonists. Given the potential limitations, such as unmeasured confounding, missing data, incomplete laboratory data, clinical information (use of alcohol, dietary patterns, body mass index) and a low baseline risk for gout, the authors’ findings are clinically relevant and demand further investigations.

高尿酸血症和痛风风险在 2 型糖尿病患者中很常见。另据报道，循环尿酸浓度长期升高与高血压、心血管疾病和慢性肾脏疾病的风险增加有关，所有这些都是已知的糖尿病相关并发症1, 2 。

Fralick等人通过分析美国全国商业保险数据库。 3报告称，与胰高血糖素样肽 1 (GLP-1) 受体激动剂的新使用者相比（1:1 倾向评分匹配），钠-葡萄糖协同转运蛋白 2 (SGLT2) 抑制剂的新使用者患痛风的风险较低。 2型糖尿病患者。应考虑本研究的几个特点。它涉及实时数据，并招募了近 300,000 名接受过 SGLT2 抑制剂或 GLP-1 受体激动剂治疗的 2 型糖尿病成人患者。此外，该研究还收集了 2013 年 3 月至 2017 年 12 月期间有关痛风诊断以及用于治疗的处方药物的信息。作者发现，SGLT2 抑制剂的新使用者的痛风风险比服用 SGLT2 抑制剂的新使用者降低了 64%。 2 型糖尿病患者新使用 GLTP-1 激动剂（每 1000 人年 4.9 次事件 vs 7.8 次事件）。从这项研究中获得的数据特别受欢迎，原因有两个。首先，已被证明可以安全降低尿酸水平的药物数量有限。事实上，最近的一份报告引起了人们的担忧，即与传统降尿酸药物别嘌呤醇相比，使用相对较新的降尿酸药物非布索坦会导致更高的心血管死亡和全因死亡率风险。降酸药。其次，美国食品和药物管理局鼓励使用真实世界数据来评估已批准药物的补充适应症。

据推测，使用SGLT2抑制剂引起的糖尿有助于尿酸分泌到尿液中。最可能的假设机制是，通过 SGLT2 抑制剂，通过抑制葡萄糖转运蛋白 9b 的活性而导致尿酸尿增加，葡萄糖转运蛋白 9b 是一种己糖/尿酸盐转运蛋白，位于跨基底外侧膜的近端肾小管细胞处1 。然而，确切的机制尚未完全阐明。据估计，对于基线尿酸值在正常浓度范围内的个体，通过使用 SGLT2 抑制剂增加尿酸消除通常可将循环尿酸浓度降低约 35–45 μmol/L (0.60–0.75 mg/dL)大约 200–400 μmol/L (~3.3–6.7 mg/dL) 1 。在恩格列净降低心血管风险范围内的心力衰竭死亡率和住院率 (EMPA-REG OUTCOME) 试验中，与接受安慰剂的患者相比，接受恩格列净的患者血清尿酸水平降低约 30 μmol/L 4 .另外，研究表明，与安慰剂或对照药物相比，2 型糖尿病患者使用 GLP-1 激动剂并未对血清尿酸水平产生任何影响。为了证实他们的观察结果，作者还检查了那些接受过长达 1 年指标药物暴露的患者，发现接受 SGLT2 抑制剂治疗的 2 型糖尿病成人患者，与接受过 SGLT2 抑制剂治疗的患者相比，患痛风的风险降低了 73%。接受 GLP-1 激动剂治疗（每 1000 人年 5.6 次与 7.7 次事件）。作者还清楚地表明，SGLT2 抑制剂的使用者患痛风的风险降低，风险比为 0.66（95% 置信区间 0.58-0.58）。75）与一组服用二肽基肽酶 4 抑制剂的新用户进行比较，二肽基肽酶 4 抑制剂是成人糖尿病患者最常用的二线药物之一，但不会降低血清尿酸水平。

先前的观察性和介入性研究表明，尿酸是 2 型糖尿病慢性肾病进展中的一个可改变的危险因素1, 2 。对于那些高尿酸血症程度较轻的糖尿病患者也是如此。表 1 列出了使用 SGLT2 抑制剂的几个潜在益处。应该指出的一项值得注意的是，作者特别排除了基线时有痛风病史和诊断的患者。痛风流行者和痛风基线风险较高的人（例如老年人、被诊断患有慢性肾脏疾病和已确诊的心血管疾病的人）没有被纳入研究。因此，使用 SGLT2 抑制剂的益处可能被低估。

SGLT2抑制剂使用者的血清尿酸水平降低幅度通常不大，并且被认为不如黄嘌呤氧化酶抑制剂（例如别嘌呤醇或非布索坦）有效。然而，鉴于作用方式不同并且可能具有潜在的互补性，如果与黄嘌呤氧化酶抑制剂联合使用，SGLT2 抑制剂治疗高尿酸血症糖尿病可能会产生附加效果。另外，据推测，通过使用 SGLT2 抑制剂增加尿液中的尿酸消除可能会导致肾结石的风险增加；然而，到目前为止，研究尚未观察到这种可能的长期风险。还希望患有高尿酸血症和/或痛风病史的非糖尿病成年人可能从使用 SGLT2 抑制剂中获益。如果得到证实，这对于患有高尿酸血症但未患有糖尿病的成年人来说将很有用。

与所有观察性研究一样，Fralick等人的研究结果。尽管许多混杂因素通过倾向得分匹配分析得到了很好的平衡，但3可能会受到无法测量的混杂因素的影响。关于血红蛋白 A1c 和血清肌酐水平的可用实验室数据非常有限（这些值适用于大约 5% 的患者，其基线血红蛋白 A1c 水平为 8.6%，基线肌酐水平为 78.7 μmol/L（0.89 mg/L）。 dL）。这大大降低了检查血糖控制变化和肾功能改变对痛风风险的影响的机会。此外，没有可用的血清尿酸水平可以与尿酸进行比较。本文没有报告有关参与者体重的信息，因此尚不清楚较少发生痛风的益处是否有利于体重指数较高或较低的人。 SGLT2 抑制剂对体重的已知影响5人们可能会争辩说，使用痛风诊断代码，加上用于治疗痛风长达 14 天的药物处方声明，足以捕获所有痛风和发作的发生率。向上。

平均随访时间有 41 天的差异（SGLT2 抑制剂为 302 天，GLP-1 激动剂为 261 天） 3 。此外，相当多的患者在随访期间没有补充药物。具体而言，45% 的患者停止使用 SGLT2 抑制剂药物，49% 的患者停止使用 GLP-1 激动剂。还存在混合效应，即 9% 开始 GLP-1 激动剂治疗并随后开出 SGLT2 抑制剂处方的患者，而 8% 开始使用 SGLT2 抑制剂治疗的患者随后开出 SGLT2 抑制剂的处方。 GLP-1 激动剂。由于这些现象在现实世界的实践中确实发生过，因此应谨慎解释所提供的数据。

总之，这项真实世界观察性队列研究的作者报告称，新使用者使用 SGLT2 抑制剂后，痛风的相对风险降低了近 40%，并且每 1,000 人年中患痛风的成年人的绝对风险降低了约 3 名。与 GLP-1 激动剂的新使用者相比。考虑到潜在的局限性，例如未测量的混杂因素、缺失数据、不完整的实验室数据、临床信息（饮酒、饮食模式、体重指数）和痛风基线风险较低，作者的研究结果具有临床相关性，需要进一步研究。