They overexpressed this protein in transgenic mice, then did a comparison of the genes expressed in that overexpressed group vs. wild-type. They took the genes, classified them into known pathways, and what you're seeing in the table are the p-values associated with up-regulation (these pathways are more expressed in the transgenic mice than in WT).
You should interpret this as an implication that the overexpressed protein (PGC-1a) could be related to the gene expression/modulation of known Alzheimer's, Huntington's, and Parkinson's disease genes/gene pathways. It's an association, not relative risk; to get odds or relative risk, they would have to do a study in clinical cases.