Longitudinal serum uric acid transition patterns and their clinical, genetic, and dietary determinants: prospective prediction of incident hyperuricemia and gout in a Korean population-based cohort.

Longitudinal serum uric acid (SUA) transition patterns and their clinical, genetic, and dietary determinants remain poorly characterized. This prospective cohort study - a secondary analysis of the Korea Genome and Epidemiology Study Health Examinee (KoGES-HEXA) cohort - aimed to characterize SUA transition patterns, identify their determinants, and examine their associations with incident hyperuricemia and gout. Genomic DNA was extracted from peripheral blood leukocytes and genotyped by the KoGES infrastructure; 58,701 participants with successfully genotyped data meeting KoGES quality control criteria were initially available. After applying study-specific exclusion criteria, 54,000 participants were included in the final analysis, yielding four SUA transition groups across two examination visits: Normal SUA (81.5%), Newly Developed Hyperuricemia (NDH; 12.4%), Recovered Hyperuricemia (RH; 5.6%), and Persistently Elevated Hyperuricemia (PH; 0.6%). Cox proportional hazards models evaluated incident hyperuricemia (n = 50,676) as the primary outcome and incident gout (n = 45,333) as the secondary outcome. Genetic risk scores (GRS) were constructed from genome-wide association analysis, and machine learning models identified key predictors of incident hyperuricemia. Lower eGFR and higher fatty liver index showed the strongest associations with adverse SUA transition groups (OR: 4.60-7.55). A 4-SNP GRS linked to urate transporter genes was strongly associated with hyperuricemia risk (OR: 3.64, 95% CI: 3.17-4.18), with nominally significant gene-lifestyle interactions for smoking, alcohol, and plant-based diet. Plant-based diet adherence was associated with reduced incident hyperuricemia risk (HR: 0.902) and attenuated genetic susceptibility across risk strata. Baseline hyperuricemia showed the strongest association with incident gout (HR: 9.34, 95% CI: 5.74-14.2). Machine learning models achieved good discrimination (AUROC: 0.862-0.885), with eGFR, GRS, sex, and BMI as top predictors. Longitudinal SUA transition patterns were strongly associated with incident hyperuricemia and gout risk, with reduced renal function, hepatic steatosis, and genetic susceptibility as the strongest determinants. Plant-based dietary patterns were associated with attenuated hyperuricemia risk across genetic risk strata, supporting integrated clinical, genetic, and lifestyle approaches to hyperuricemia prevention. Key Points • Longitudinal analysis identified four distinct serum uric acid trajectories, with 18.5% experiencing dynamic hyperuricemia patterns. • Impaired renal function and hepatic steatosis are the strongest predictors of adverse trajectories. • Plant-based dietary adherence significantly attenuates genetic susceptibility to hyperuricemia. • Machine learning integration of clinical, genetic, and lifestyle factors enables accurate hyperuricemia prediction.