Figure 4.

Correlations between variables in the 1-year cohort. (A) In terms of rate of change of a variable over time. For each variable, the relationship between that variable and time was determined within each patient, and the resulting regression coefficients were correlated across variables. (B) The intra-patient stability. For each variable, intra-patient stability was calculated as the average pairwise distance between samples within a patient. These average distances were then correlated across variables. Microbiota are given in green, clinical in blue, pro-inflammatory in orange, anti-protease in yellow, and general parameters in gray. The size of the Spearman’s rank correlation coefficient (Rho) is represented by the width and color of the edges, and only correlations with a Rho greater than 0.3 or smaller than −0.3 are shown. Values of p were adjusted using the Benjamini-Hochberg method. Values of p indicated as: ^*<0.05, ^**≤0.01, and ^***≤0.001.