Association Between Asymptomatic Hyperuricemia and Kidney Stones

In 31,198 Korean adults, men with the highest uric acid levels had 3.5× the odds of kidney stones — women showed no such link

Journal: PloS One | Published: 2026-05-26 | Type: Cross-sectional study | PMID: 42189825 Authors: Lee Jun Ho et al. (Eulji University / Inje University, Seoul; Cardiff University) Funding/COI: Funding not reported; authors declare no competing interests

Summary

Men with serum uric acid in the top quartile (≥6.9 mg/dL) had 3.5× the odds of kidney stones compared to men in the lowest quartile, even after adjustment for age, BMI, blood pressure, lipids, renal function, calcium, and vitamin D. The association held across three levels of hierarchical regression. Women showed no statistically significant association, which the authors attribute to estrogen-driven urate clearance — plausible but not directly tested here.

Claims

• Overall nephrolithiasis prevalence: 1.4% (427/31,198); 1.8% in men, 0.8% in women
• Men in the highest uric acid quartile (≥6.9 mg/dL): adjusted OR 3.546 (95% CI: 1.240–10.144) vs. lowest quartile
• Uric acid as a continuous variable in men: adjusted OR 1.364 per mg/dL increase (95% CI: 1.068–1.743)
• In fully adjusted models for men, only age (aOR 1.053) and serum uric acid remained independently significant
• No significant association between uric acid and nephrolithiasis in women across any model
• Median uric acid: 6.0 mg/dL in men, 4.4 mg/dL in women

Study Quality

This is a large (n = 31,198) retrospective cross-sectional analysis from a single academic hospital health screening program in Seoul, covering 2010–2020. The methodological highlight is the covariate adjustment: previous Korean and Chinese studies on this question did not account for calcium or vitamin D, both of which are established lithogenic risk factors. Including them here strengthens the independence of the uric acid signal. Gout patients (n = 230) were excluded, which is the right call — the question is specifically about asymptomatic hyperuricemia, not gout-associated stone disease.

The stone definition (≥5 mm on ultrasound or KUB radiography) is reasonable but will miss smaller stones, likely underestimating true prevalence. Critically, urine pH was not measured. Uric acid stones form almost exclusively in acidic urine (pH < 5.5), and two people with identical serum uric acid can have very different lithogenic risk depending on urine pH. Without it, the mechanistic claim is incomplete and the population-level OR is mixing stone types with different etiologies.

Red Flags

• Cross-sectional design: direction of causality is unestablished — prevalent stones could theoretically influence uric acid metabolism rather than the reverse
• The quartile OR's confidence interval is wide (1.240–10.144), reflecting low event counts in subgroups; 427 total stone cases across 31,198 participants limits power for stratified analyses
• Only 99 women had stones — the null finding in women may be a power issue, not a true biological null
• Stone composition not determined: unable to distinguish uric acid stones from calcium oxalate stones promoted by urate crystals, which have different clinical implications
• Single academic hospital, voluntary health screening cohort — healthier and wealthier than the general population (selection bias)
• Funding source not reported — minor but notable for a PloS One publication

Strengths

• Largest Korean dataset on this specific question (asymptomatic hyperuricemia, not gout)
• Extensive covariate adjustment, including calcium and vitamin D — an improvement over prior comparable studies
• Hierarchical regression showing the uric acid-stone association is robust to progressive confounding control
• Appropriate exclusion of gout patients, diuretic users, and those with congenital renal anomalies
• Sex-stratified analysis rather than pooled, which is the right analytical choice given known metabolic sex differences

Verdict

A solid epidemiological observation that adds to a thin literature. The association between uric acid and kidney stones in asymptomatic men is real, replicates prior Korean cohort work, and survives more rigorous adjustment than previous studies. But this paper cannot establish causality, cannot tell you what type of stone is involved, and is silent on urine pH — the variable that matters most mechanistically. The wide confidence interval on the headline OR (1.240–10.144) is a flag; the signal is present but imprecisely estimated. Read this as hypothesis-strengthening epidemiology that justifies a prospective cohort study, not as evidence of a modifiable causal pathway.