I would like to conduct an analysis using mri_glmfit but I'm getting this error:

ERROR: matrix is ill-conditioned or badly scaled, condno = 1e+08

My command was: 

mri_glmfit --y lh.area.fwhm10.fsaverage.mgh --fsgd finalgroupanalysis.fsgd dods --C group_effect.mtx --surf fsaverage lh --cortex --glmdir lh.surfacearea.glmdir 

My FSGD file is attached. It contains 3 classes and 1 variable (age). 

And my design matrix is 

--

Wanda Truong, MSc