Table 4.

Annual change in eGFR according to quintiles of plant-based diet scores in the Atherosclerosis Risk in Communities studya

Diet IndexAnnual Change in eGFR (95% CI), ml/min per 1.73 m2 per yrP for Trend
Overall plant-based diet indexQ1 (n=3377)Q2 (n=2430)Q3 (n=3596)Q4 (n=2348)Q5 (n=2792)
 Model 1b−1.68 (−1.73 to −1.64)−1.58 (−1.63 to −1.53)−1.58 (−1.62 to −1.54)−1.55 (−1.60 to −1.49)−1.55 (−1.59 to −1.50)<0.001
 Model 2c−1.68 (−1.72 to −1.63)−1.58 (−1.62 to −1.53)−1.57 (−1.62 to −1.52)−1.55 (−1.60 to −1.50)−1.54 (−1.59 to −1.49)<0.001
 Model 3d−1.64 (−1.67 to −1.60)−1.55 (−1.59 to −1.51)−1.54 (−1.58 to −1.50)−1.53 (−1.57 to −1.48)−1.52 (−1.56 to −1.50)0.002
Healthy plant-based diet indexQ1 (n=2917)Q2 (n=3142)Q3 (n=3445)Q4 (n=2684)Q5 (n=2355)
 Model 1b−1.62 (−1.66 to −1.57)−1.61 (−1.66 to −1.57)−1.57 (−1.62 to −1.53)−1.57 (−1.62 to −1.52)−1.55 (−1.61 to −1.50)0.04
 Model 2c−1.62 (−1.66 to −1.57)−1.62 (−1.66 to −1.57)−1.58 (−1.62 to −1.53)−1.57 (−1.62 to −1.52)−1.55 (−1.60 to −1.49)0.01
 Model 3d−1.58 (−1.62 to −1.54)−1.58 (−1.62 to −1.54)−1.55 (−1.57 to −1.50)−1.54 (−1.60 to −1.51)−1.52 (−1.57 to −1.48)0.04
Less healthy plant-based diet indexQ1 (n=2954)Q2 (n=3105)Q3 (n=3378)Q4 (n=2756)Q5 (n=2350)
 Model 1b−1.58 (−1.63 to −1.54)−1.63 (−1.68 to −1.58)−1.58 (−1.63 to −1.54)−1.59 (−1.64 to −1.54)−1.56 (−1.62 to −1.51)0.32
 Model 2c−1.57 (−1.63 to −1.52)−1.63 (−1.67 to −1.58)−1.58 (−1.62 to −1.54)−1.58 (−1.63 to −1.54)−1.56 (−1.61 to −1.51)0.24
 Model 3d−1.55 (−1.59 to −1.51)−1.58 (−1.62 to −1.55)−1.55 (−1.59 to −1.51)−1.56 (−1.60 to −1.52)−1.53 (−1.57 to −1.48)0.29
Provegetarian diet indexQ1 (n=3590)Q2 (n=3106)Q3 (n=2235)Q4 (n=2851)Q5 (n=2761)
 Model 1b−1.61 (−1.65 to −1.56)−1.61 (−1.65 to −1.56)−1.57 (−1.62 to −1.51)−1.61 (−1.66 to −1.56)−1.55 (−1.59 to −1.50)0.15
 Model 2c−1.60 (−1.65 to −1.56)−1.60 (−1.65 to −1.55)−1.57 (−1.62 to −1.51)−1.61 (−1.66 to −1.56)−1.54 (−1.59 to −1.49)0.15
 Model 3d−1.62 (−1.66 to −1.59)−1.61 (−1.65 to −1.57)−1.59 (−1.63 to −1.54)−1.64 (−1.68 to −1.59)−1.56 (−1.61 to −1.52)0.24
  • P for trend was obtained from an interaction term between follow-up time and quintiles of plant-based diet indices as an ordinal variable. 95% CI, 95% confidence interval; Q, quintile.

  • a Sample sizes for the analysis of annual change in kidney function differ from the main analyses of incident CKD because we used baseline dietary intakes only to construct the scores (and not the cumulative average of dietary intake assessed at baseline and visit 3).

  • b Model 1 was adjusted for age, sex, race-center, and total energy intake.

  • c Model 2 was adjusted for covariates in model 1 plus education, smoking status, physical activity, alcohol consumption, and margarine consumption.

  • d Model 3 was adjusted for covariates in model 2 plus baseline total cholesterol, lipid-lowering medication use, kidney function (two linear spline terms with one knot at 90 ml/min per 1.73 m2), hypertension, diabetes, and body mass index.