Skip to main content
. 2019 Mar 19;13:51. doi: 10.3389/fnbeh.2019.00051

Table 5.

Predictor loadings for neonatal clinical factors and genotypes.

PC1
PC2
Variables Loadings p-valuea FDRb Loadings p-valuea FDRb
Neonatal clinical factors
Infection 0.320 0.0126 0.023 0.105 0.4359 0.533
Invasive procedures 0.483 0.0001 0.000 -0.125 0.3465 0.533
Morphine 0.238 0.0620 0.097 0.265 0.0532 0.146
Surgery -0.085 0.5011 0.612 0.419 0.0019 0.010
Mechanical ventilation 0.502 0.0001 0.000 0.152 0.2663 0.488
SNAP-II Day 1 0.333 0.0111 0.023 0.274 0.0367 0.135
GA -0.460 0.0005 0.001 -0.106 0.4243 0.533
Total brain volume -0.771 0.0000 0.000 -0.513 0.0001 0.001
Genotypes
BDNF -0.075 0.5743 0.632 0.051 0.7024 0.702
5HTTLPR 0.137 0.3232 0.444 -0.229 0.0860 0.189
COMT -0.051 0.7057 0.706 -0.069 0.6166 0.678

Both PC1 and PC2 were negatively related to regional volumes of the limbic system (Table 4). Significant positive loadings in this table indicate the relative contribution of each clinical factor and genotype for each component. For example, taking Tables 4 and 5 together, for PC1, more neonatal invasive procedures and more days on mechanical ventilation were related to smaller regional volume of CA4DG. PC1, principal component 1; PC2, principal component 2; Infection, neonatal culture proven infection; Morphine, morphine exposure, Surgery, number of surgeries; Ventilation, days on mechanical ventilation, SNAP-II day 1, score for neonatal acute physiology-II (i.e., illness severity on day 1); GA, gestation age. aRed font indicate p < 0.0045 after Bonferroni adjustment for multiple comparisons. bRed font indicate p < 0.05 after Benjamini Hochberg false discovery rate (FDR) 5% overall error rate adjustment.