Table 3 Univariate analysis of predictors of head circumference at 1 year in the CHILD cohort study: top 40 features sorted by p-value.

From: Multimodal machine learning for modeling infant head circumference, mothers’ milk composition, and their shared environment

  1. Associations are measured by Kendall’s Tau for numeric features or Wilcoxon rank sum for binary features. Feature colors correspond to modality types as in Fig. 1A. P-values in the last column are corrected using the Benjamini–Hochberg FDR procedure across all considered features. Significant features with and without correction are marked by bold font in the respective p-value column. A blue background in the p-value columns marks positive, and orange mark negative associations. A “*” indicate variables that depend on other variables (e.g., diarrhea is only assessed when the baby has a cold). The column “na” lists the number of missing values for the respective variable and “unique” denotes how many unique values the variable is represented by. For a detailed description of features, see Supplemental Table S1.
Back to article page