Association between the uric acid to high-density lipoprotein cholesterol ratio and estimated glomerular filtration rate decline in community-managed patients with type 2 diabetes: a retrospective cohort study
Background The uric acid to high-density lipoprotein cholesterol ratio (UHR), a composite biomarker integrating pro-inflammatory stress and vascular protection loss, has been linked to diabetic kidney disease (DKD) in cross-sectional studies, but longitudinal evidence on its association with continuous eGFR decline remains limited. Methods This retrospective cohort study included 8,827 community-managed type 2 diabetes patients from 54 health centers in Shenzhen, China, followed for 3 years (2023–2025). Baseline UHR was calculated as serum uric acid (μmol/L) divided by HDL-C (mmol/L). The primary outcome was annual eGFR slope, derived from individual linear regressions. Multivariable linear regression, restricted cubic spline analysis, model comparison, and pre-specified subgroup and sensitivity analyses were performed. Results In fully adjusted models, each 1-unit increase in baseline UHR was associated with greater annual eGFR decline (β = −0.002, 95% CI − 0.003 to −0.001, p = 0.002). In model comparison, UHR demonstrated a larger standardized effect (β = −0.052) than UA alone (β = −0.045) or HDL-C alone (β = 0.028) and achieved explanatory power comparable to the joint UA + HDL-C model with a lower AIC, indicating superior parsimony. Comparing extreme tertiles, the highest UHR tertile (T3) exhibited an additional 0.56 mL/min/1.73 m2/year decline versus the lowest (T1) (β = −0.558, p = 0.001; P for trend = 0.001), cumulating to 5.6 mL/min/1.73 m2 over 10 years. Truly negative eGFR slopes were more frequent in T3 (44.3%) than T1 (21.4%), and T3 had 34% higher odds of rapid eGFR decline (OR 1.34, 95% CI 1.11–1.61, p = 0.002). RCS analysis confirmed linearity (P for nonlinearity = 0.301). The association was stronger in patients aged >60 years and those with HbA1c ≥ 7% (P for interaction = 0.005 and <0.001) and remained robust in sensitivity analyses, including three-year average UHR (β = −0.009, p < 0.001). Conclusion Higher baseline UHR independently predicted faster eGFR decline in community-managed type 2 diabetes patients. As an inexpensive, readily available index, UHR may serve as a biomarker for identifying patients at elevated risk of progressive renal function loss.