Clinical evidence for the influence of uric acid on hypertension, cardiovascular disease, and kidney disease: a statistical modeling perspective.

This article critically evaluates the clinical evidence regarding the influence of uric acid on hypertension, cardiovascular disease, and kidney disease. Data on these relationships are largely observational and exceedingly complex. The complexity is owing to indirect and direct relations, and bidirectional influences, simultaneously operating on multiple outcomes. Limitations of previous analyses include inadequate statistical methods using only bivariate correlations or poorly specified multiple regression models. As a result, great controversy developed as to whether uric acid is an independent predictor of important outcomes. An example of such analytic limitations is including hypertension as an independent variable, together with uric acid, in a multivariate model for predicting cardiovascular disease. Hypertension may predict significant variance in cardiovascular disease, but the contribution of uric acid may not be recognized if uric acid exerts its influence indirectly through hypertension. Path analysis, which can model direct and indirect influences on outcomes simultaneously, would address this substantive question. Studies of uric acid in relation to hypertension, cardiovascular disease, and kidney disease using a path-analytic approach would help specify such conditions as well as optimize design of clinical trials to determine if decreasing uric acid levels improves outcomes.