Skip to main content
. 2021 Jun 2;13:121. doi: 10.1186/s13148-021-01082-w

Table 2.

Trans-ethnic meta-analyses of associations between kidney traits and DNAm-based age and lifespan predictors

Clock eGFR CKD
β 95% CI p I2 phet OR 95% CI p I2 phet
HorvathAA  − 0.006  − 0.01, − 0.003 5.15E−04 38.696 0.121 1.019 1.003, 1.034 0.016 13.663 0.323
HannumAA  − 0.007  − 0.011, − 0.004 1.05E−04 0 0.816 1.033 1.016, 1.05 8.55E−05 44.898 0.08
GrimAA  − 0.006  − 0.01, − 0.002 1.94E−03 63.474 0.008 1.027 1.01, 1.044 1.27E−03 77.714 5.2E−05
PhenoAA  − 0.005  − 0.008, − 0.002 2.62E−04 0 0.564 1.031 1.018, 1.044 3.19E−06 33.885 0.158
EEAA  − 0.008  − 0.012, − 0.005 2.09E−06 43.328 0.09 1.038 1.022, 1.055 3.41E−06 49.636 0.053
MRS  − 0.117  − 0.158, − 0.075 4.39E−08 27.67 0.208 1.784 1.474, 2.159 2.71E−09 68.295 0.002a
IEAA  − 0.004  − 0.008, 0 0.051 16.08 0.303 1.007 0.99, 1.025 0.427 2.365 0.411
Clock uACR Microalbuminuria
β 95% CI p I2 phet OR 95% CI p I2 phet
HorvathAA 0.002  − 0.004, 0.008 0.606 0 0.632 1.014 0.992, 1.036 0.223 0 0.703
HannumAA 0.014 0.009, 0.02 2.04E−06 0 0.967 1.054 1.032, 1.076 1.08E−06 0 0.898
GrimAA 0.029 0.021, 0.037 1.07E−12 7.22 0.357 1.106 1.074, 1.138 7.58E−12 0 0.59
PhenoAA 0.01 0.005, 0.015 2.71E−05 0 0.787 1.035 1.017, 1.053 8.96E−05 0 0.612
EEAA 0.013 0.008, 0.017 4.62E−07 0 0.989 1.048 1.03, 1.066 1.42E−07 0 0.86
MRS 0.252 0.179, 0.324 1.01E−11 76.128 0.006a 2.238 1.734, 2.889 6.14E−10 67.72 0.026a
IEAA 0.002  − 0.004, 0.008 0.529 0 0.847 1.015 0.991, 1.04 0.214 0 0.769
Clock Urate
β 95% CI p I2 phet
HorvathAA 0.003  − 0.001, 0.007 0.12 0 0.453
HannumAA 0.005 0.001, 0.009 0.011 0 0.919
GrimAA 0.009 0.004, 0.013 1.17E−04 56.31 0.057
PhenoAA 0.009 0.006, 0.012 4.71E−08 0 0.432
EEAA 0.007 0.004, 0.011 4.37E−05 41.60 0.144
MRS 0.115 0.067, 0.162 2.08E−06 12.16 0.336
IEAA  − 0.001  − 0.005, 0.003 0.675 0 0.797

Results from trans-ethnic meta-analyses of associations between kidney traits and DNAm-based age and lifespan predictors in up to seven population-based studies. Study-level associations were adjusted for chronological age, sex, BMI, blood lipids, hypertension, smoking and diabetes. Beta coefficients are given as changes in one standard deviation (SD) of the continuous kidney trait. Fully adjusted associations of serum-creatinine-based traits (eGFR, CKD) are based on N ≤ 9390 observations, whereas the sample size for urinary albumin-based traits (uACR, microalbuminuria) is N ≤ 4406 and for urate N ≤ 5769. I2 is the heterogeneity statistic, and (Q) phet corresponds to Cochran’s Q heterogeneity statistic

Shown in bold are statistically significant associations (p < 1.43E-03 and consistent direction of effect across studies) with either no evidence of heterogeneity in the fixed-effects model or supporting findings from the random-effects model

aphet of MRS with CKD, uACR and microalbuminuria < 0.05, therefore reported association based on significant random-effects models: MRS-uACR: β = 0.248 [0.1, 0.397], p = 1.061E−03; MRS-CKD: OR = 1.915 [1.316, 2.786], p = 6.89E−04; and MRS-microalbuminuria: OR = 2.197 [1.403, 3.439], p = 5.82E−04 (Additional file 1: Table S3)