Fig. 2

Influence of covariables on eGFR in the elderly and the general adults. Univariable linear regression was performed to estimate the association of each covariable (rows) with eGFRcrea and eGFRcys in AugUR (A, n = 2,272) and KORA S4 (B, n = 2,900). Shown are beta-estimates and 95% CIs (left). Bar plots show the proportion of explained eGFR variance (R2) by each covariable (right)