Ravi K Narang1, Greg G Gamble1, Ruth Topless2, Murray Cadzow2, Lisa K Stamp3, Tony R Merriman2, Nicola Dalbeth4. 1. Department of Medicine, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand. 2. Department of Biochemistry, University of Otago, Dunedin, New Zealand. 3. Department of Medicine, University of Otago, Christchurch, New Zealand. 4. Department of Medicine, Faculty of Medical and Health Sciences, University of Auckland, Auckland, New Zealand. Electronic address: n.dalbeth@auckland.ac.nz.
Abstract
RATIONALE & OBJECTIVE: The association between hyperuricemia and urolithiasis has been previously reported. However, this association is based on observational data, which are prone to residual confounding. The aim of this work was to use Mendelian randomization (MR) to evaluate if this relationship represents a causal effect of hyperuricemia. STUDY DESIGN: MR analysis using 2 approaches: 2-stage MR and 2-sample MR. SETTING & PARTICIPANTS: Participants aged 40-69 years from the UK Biobank Resource. EXPOSURE: Serum urate. OUTCOME: Urolithiasis. ANALYTICAL APPROACH: An observational analysis testing for an association between serum urate level and urolithiasis was performed using logistic regression. For MR analyses, serum urate-associated single-nucleotide polymorphisms, identified from genome-wide association data, were used as instrumental variables for serum urate. In the 2-stage MR analysis, a weighted genetic urate score was calculated from the instrumental variables, and a control function estimation model was fit. In the 2-sample MR analysis, multiple-instrument MR via the inverse-variance weighted method was performed. RESULTS: Individual-level data were available for 359,827 participants, of whom 6,398 (1.8%) reported urolithiasis. In the observational analysis, serum urate was positively associated with urolithiasis in an unadjusted analysis (odds ratio [OR], 1.47 [95% CI, 1.42-1.51]); however, after adjustment for relevant confounders, no association was observed (OR, 1.03 [95% CI, 0.99-1.08]). In the 2-stage MR analysis, no significant causal effect of serum urate level on urolithiasis was observed in the unadjusted (OR, 0.93 [95% CI, 0.81-1.08]) or adjusted (OR, 0.94 [95% CI, 0.80-1.09]) models. In the 2-sample MR analysis, multiple-instrument MR did not indicate a causal effect of serum urate on urolithiasis. LIMITATIONS: Stone composition and urinalysis data, including urine pH, were not available for this study. CONCLUSIONS: Our analyses do not support a causal effect of serum urate level on urolithiasis. The association between serum urate level and urolithiasis reported in observational studies is likely due to residual confounding.
RATIONALE & OBJECTIVE: The association between hyperuricemia and urolithiasis has been previously reported. However, this association is based on observational data, which are prone to residual confounding. The aim of this work was to use Mendelian randomization (MR) to evaluate if this relationship represents a causal effect of hyperuricemia. STUDY DESIGN: MR analysis using 2 approaches: 2-stage MR and 2-sample MR. SETTING & PARTICIPANTS: Participants aged 40-69 years from the UK Biobank Resource. EXPOSURE: Serum urate. OUTCOME: Urolithiasis. ANALYTICAL APPROACH: An observational analysis testing for an association between serum urate level and urolithiasis was performed using logistic regression. For MR analyses, serum urate-associated single-nucleotide polymorphisms, identified from genome-wide association data, were used as instrumental variables for serum urate. In the 2-stage MR analysis, a weighted genetic urate score was calculated from the instrumental variables, and a control function estimation model was fit. In the 2-sample MR analysis, multiple-instrument MR via the inverse-variance weighted method was performed. RESULTS: Individual-level data were available for 359,827 participants, of whom 6,398 (1.8%) reported urolithiasis. In the observational analysis, serum urate was positively associated with urolithiasis in an unadjusted analysis (odds ratio [OR], 1.47 [95% CI, 1.42-1.51]); however, after adjustment for relevant confounders, no association was observed (OR, 1.03 [95% CI, 0.99-1.08]). In the 2-stage MR analysis, no significant causal effect of serum urate level on urolithiasis was observed in the unadjusted (OR, 0.93 [95% CI, 0.81-1.08]) or adjusted (OR, 0.94 [95% CI, 0.80-1.09]) models. In the 2-sample MR analysis, multiple-instrument MR did not indicate a causal effect of serum urate on urolithiasis. LIMITATIONS: Stone composition and urinalysis data, including urine pH, were not available for this study. CONCLUSIONS: Our analyses do not support a causal effect of serum urate level on urolithiasis. The association between serum urate level and urolithiasis reported in observational studies is likely due to residual confounding.