Constrained principal component analysis results to explain variance in cortical thickness in relation to neonatal clinical factors.

Comp = Component; PCA = principal component analyses; eigenvalues for 3 components = 14.77, 2.46, 1.56 respectively.
In Principal component analysis (PCA), a component refers to a vector of weightings that best explain the variance. Each extracted component accounts for a portion of the total variance in the data: the first component accounts for the largest amount of variance, with each successive component accounting for a smaller amount of the total variance.
In Constrained Principal Component Analysis, the external analysis consisted of a multivariate multiple regression of the predictor variables on the dependent measures, which produces predicted and residual scores. In the present study, the matrix of predicted scores reflects the variance in cortical thickness that is predicted from the 7 neonatal clinical variables, and the residual matrix reflects the variance that is not predicted by these variables.
The internal analysis consisted of three different PCAs: one on the unconstrained variance in cortical thickness (Overall), one on the variance in the cortical thickness predictable from the 7 neonatal clinical variables (Predictable), and one on the variance in cortical thickness not predictable from the clinical variables (Residual). The variance accounted for by the external analysis and each component extracted in the internal analysis is listed in regular font. The percentages of variance accounted for by the external analysis and each component extracted in the internal analysis are listed in italic font. All internal analyses were separately rotated using varimax.