The primary finding in this study was that endothelial dysfunction, as determined by measuring FMD, was independently associated with both reduced eGFR and with elevations in serum uric acid levels. Specifically, we evaluated endothelial function in 263 nondiabetic subjects with CKD and a mean eGFR of 26 ml/min/1.73 m2
. In this group, the median serum uric acid was 7.5 mg/dl. FMD varied markedly, with a range from 4 to 8.2% (mean 6.4%). Importantly, endothelial function correlated inversely with serum uric acid levels (fig. ). Multivariate analysis confirmed that uric acid was an important predictor of endothelial function, independent of numerous risk factors including eGFR and the Framingham risk factors. These studies suggest that uric acid could have a role in driving endothelial dysfunction in subjects with nondiabetic CKD. Since endothelial dysfunction appears to be a major risk factor for progression of CKD [1
], our data provide a rationale for lowering uric acid in CKD, as a means to slow renal disease progression.
There is already consistent evidence that uric acid may have a role in the progression of CKD. For example, several studies have shown that an elevated uric acid predicts the development of CKD in the general population [28
]. An elevated uric acid also predicts the development of diabetic nephropathy in subjects with type 1 diabetes [31
]. Experimental studies have reported that lowering uric acid can slow renal disease progression in both diabetic and nondiabetic renal disease [33
]. Furthermore, several pilot studies suggest that lowering uric acid with a xanthine oxidase inhibitor can improve renal function in subjects without CKD, and may slow renal progression in subjects with CKD [35
]. In this regard, Goicoechea et al. [37
] recently reported a randomized trial in which allopurinol (100 mg/day) slowed renal progression and reduced cardiovascular events in subjects with stage 3 CKD.
In this analysis, we excluded subjects receiving ACE inhibitors or statins as they are known to influence endothelial function and hence could be confounders.
We were most interested in the role of reduced eGFR itself, as this may lead to retention of substances such as uric acid and asymmetric dimethylarginine, the latter considered a circulating uremic toxin that competes with L
-arginine for endothelial NO synthase [19
]. Importantly, both eGFR and uric acid emerged as independent risk factors for endothelial dysfunction. This suggests that uric acid might affect endothelial function distinct from the effects of reduced GFR, and hence may provide a new therapeutic target to reduce renal disease progression.
It is important to recognize that the independent association of uric acid with endothelial function in this study does not imply that uric acid is directly responsible for the endothelial dysfunction. Indeed, uric acid could reflect underling xanthine oxidase activity, since xanthine oxidase generates uric acid from xanthine. In addition, xanthine oxidase is also known to generate oxidants. In this way, an elevated uric acid could simply reflect the presence of xanthine oxidase-associated oxidants, which may be ultimately responsible for the endothelial dysfunction [38
]. Indeed, George et al. [38
] reported that endothelial function in subjects with congestive heart failure improved only when uric acid was lowered by a xanthine oxidase inhibitor as opposed to when a uricosuric was utilized. Waring et al. [39
] also suggested that the acute infusion of uric acid actually improves endothelial function rather than impairs it, in subjects with diabetes. However, while uric acid in the extracellular environment is an antioxidant, there is increasing evidence that uric acid, when entering the cells, is prooxidative and directly induces endothelial dysfunction [10
]. Allopurinol may improve endothelial function better than uricosurics by reducing intracellular uric acid more effectively.
In conclusion, subjects with nondiabetic CKD display a wide range of endothelial function, as measured by FMD. Subjects with a higher uric acid (>7.5 mg/dl) have a significantly worse FMD compared to subjects with uric acid levels <7.5 mg/dl. This relationship persists after controlling for all Framingham risk factors and eGFR. Since endothelial function is an important risk factor for renal progression, this study suggests the importance of determining if lowering uric acid can improve endothelial function in CKD and whether this can translate into significant slowing or improvement in renal function.