Supplementary Material for: Association of renal function and stroke incidence in general population: A nationwide retrospective longitudinal cohort study

Introduction: Individuals with impaired renal function are known to have an increased risk of stroke. However, studies from general populations that include individuals with various comorbidities remain scarce. This study aimed to investigate the association between renal function and stroke incidence according to entire spectrum of estimated glomerular filtration rate (eGFR) level in the general Korean population using a nationwide longitudinal analysis.
Methods: We included 2,708,874 participants who underwent health examinations between 2010 and 2011 from the Korean NHIS-HEALS database. An eGFR was calculated to assess renal function and then categorized into clinically relevant ranges as well as deciles. Stroke incidence was defined as a new record of the International Classification of Diseases-10 codes I60-I63, requiring at least two claims with a main or secondary diagnosis within a year and confirmation via brain computed tomography or magnetic resonance imaging within one month of diagnosis.
Results: The mean age was 48.63 ± 14.02 years, with 112,913 (4.36%) stroke cases identified over 9.63 years. In the multivariable Cox regression analysis using eGFR ranges and deciles, impaired renal function was significantly associated with an increased stroke incidence. The highest stroke risk occurred in the lowest eGFR range (< 30 ml/min/1.73m2) (hazard ratio [HR], 1.98, 95% confidence interval [CI], 1.83-2.15) and lowest eGFR decile (< 67.57 ml/min/1.73m2) (HR, 1.51, 95% CI, 1.47-1.54), whereas the lowest risk was observed in the highest eGFR range (> 120 ml/min/1.73m2) (HR, 0.32, 95% CI, 0.30-0.34) and eGFR decile (≥ 113.41 ml/min/1.73m2(HR, 0.42, 95% CI, 0.40-0.44). A significant inverse linear trend was confirmed between stroke incidence and both eGFR ranges and deciles. These findings were consistent regardless of age, obesity, and stroke subtypes.
Conclusion: Stroke risk increased linearly with declining eGFR in the general population. Assessment of renal function may be a useful clinical tool for identifying individuals who would benefit most from stroke prevention interventions.