Conversely, we want to underline that a covariate known as highly significant in regression modeling and therefore well-known in the literature to be related with the outcome of interest, may not result in large improvements in prediction and therefore may not be a good predictor candidate [ 7 ].
← all excerpts
The role of baseline serum 25(OH)D concentration for a potential personalized vitamin D supplementation.
1
—
—