The Vanderbilt article -- is that the one you mean? -- says:
One genetic mutation (rs4728029) was found to relate to both ventricle dilation and cognition and is a marker of neuroinflammation.
It doesn't say how
it relates, but if the mutation is a "marker " of inflammation, it's probably a marker in the "plus" direction (i.e., more
inflammation). This would mean the ancestral allele, G, is the better allele.
But it would be good to get the article, or contact the researchers. These university pop summaries are sometimes wrong.