Association of Differences in Cystatin C‐ and Creatinine‐Based Estimated Glomerular Filtration Rate With Prevalence and Incidence of Stroke

Background The intraindividual difference between the estimated glomerular filtration rates by cystatin C and creatinine (ie, eGFRdiff) is of clinical importance. This study aimed to investigate the cross‐sectional and longitudinal associations of eGFRdiff with stroke using nationally representative data from the CHARLS (China Health and Retirement Longitudinal Study). Methods This study included 11 869 participants (aged ≥45 years) from the CHARLS 2015 for the cross‐sectional analysis. A total of 11 553 participants free of stroke were recruited in 2015 as baseline and followed up to 2020 for the cohort analysis. The eGFRdiff was calculated by subtracting estimated glomerular filtration rate by serum creatinine from estimated glomerular filtration rate by serum cystatin C at baseline. Logistic regression models and Cox proportional hazards models were used to examine the cross‐sectional and longitudinal associations. Results The mean age of the participants was 60.3±9.6 years, and 5514 (46.5%) were men. According to the cross‐sectional analysis, a negative eGFRdiff was associated with a greater prevalence of stroke (odds ratio per 1 SD, 0.841 [95% CI, 0.741–0.955]). During a median follow‐up of 5.0 years, 838 individuals (7.3%) experienced incident stroke. A lower baseline eGFRdiff was associated with a greater risk of stroke onset (hazard ratio [HR] per 1 SD, 0.884 [95% CI, 0.817–0.955]). Compared with those in the midrange eGFRdiff group (from −15 to 15 mL/min per 1.73 m 2 ), those in the negative eGFRdiff group (<−15 mL/min per 1.73 m 2 ) exhibited a significantly increased risk of incident stroke (HR, 1.202 [95% CI, 1.026–1.407]). Conclusions A large negative eGFRdiff was independently associated with higher prevalence and incidence rates of stroke among middle‐aged and older Chinese adults.