attached youŽll find the two y.fsgd file from the analysis run in GLM_FIT. the results are completely different (the z-score converted age produces a believable result, while the result correcting with raw age scores seems completely wrong. 

best regards

Olof