Hi,
I'm trying to run a cortical thickness group comparison (HD v Controls), controlling for site, gender (1 or 2) and age. Without age as a covariate the results are as expected but once age is added on all significant results disappear. I think this is a problem with the method rather than because all results are dependent on age. My fsgd and contrast files for the glm are below. I would really appreciate your input as to what might be going wrong to effect the results so strongly. Am I doing something wrong?
GroupDescriptorFile 1 Title HDvControl
Class HD-london-1 Class HD-leiden-1 Class HD-paris-1 Class HD-ulm-1 Class HD-london-2 Class HD-leiden-2 Class HD-paris-2 Class HD-ulm-2 Class C-london-1 Class C-leiden-1 Class C-paris-1 Class C-ulm-1 Class C-london-2 Class C-leiden-2 Class C-paris-2 Class C-ulm-2
Variables Age Input 30009 HD-paris-2 26 Input 30014 C-ulm-2 51 Input 30016 C-london-1 48 Input 30024 C-leiden-1 38 Input 30029 HD-london-2 66 Input 30030 HD-london-1 54 Input 30037 HD-london-2 42 Input 30041 C-london-1 43 Input 30042 HD-london-2 40 Input 30043 HD-london-2 41 Input 30047 HD-london-2 66 Input 30051 C-london-2 45 Input 30054 HD-london-2 55 Input 30056 C-london-2 66 Input 30060 C-london-1 50 Input 30068 C-london-1 51 Input 30071 HD-london-1 43 Input 30072 HD-london-2 44 Input 30074 C-london-2 58 Input 30075 HD-london-2 48 Input 30078 C-london-2 53 Input 30079 HD-london-1 58 Input 30082 C-london-2 62 Input 30089 HD-london-2 54 Input 30093 HD-leiden-1 44 Input 30098 HD-leiden-2 45 Input 30100 HD-leiden-2 42 Input 30102 C-leiden-2 48 Input 30107 HD-leiden-2 61 Input 30112 HD-leiden-2 50 Input 30114 C-leiden-1 48 Input 30115 HD-leiden-2 47 Input 30117 HD-leiden-2 49 Input 30118 C-leiden-1 51 Input 30119 HD-leiden-2 59 Input 30124 C-leiden-1 57 Input 30125 HD-leiden-2 39 Input 30136 C-leiden-2 51 Input 30138 C-leiden-2 56 Input 30139 HD-leiden-2 33 Input 30142 HD-leiden-2 32 Input 30148 HD-leiden-1 48 Input 30154 C-leiden-1 47 Input 30155 HD-leiden-2 46 Input 30160 HD-leiden-2 38 Input 30162 HD-leiden-2 52 Input 30163 C-leiden-2 42 Input 30168 C-ulm-2 58 Input 30177 C-london-2 49 Input 30185 C-ulm-2 34 Input 30189 HD-ulm-1 50 Input 30191 C-ulm-1 44 Input 30193 HD-ulm-1 46 Input 30195 C-ulm-2 50 Input 30198 C-ulm-1 56 Input 30199 HD-ulm-1 53 Input 30208 HD-ulm-1 59 Input 30209 C-ulm-2 44 Input 30210 HD-ulm-1 30 Input 30214 HD-ulm-2 57 Input 30215 HD-ulm-1 34 Input 30220 HD-ulm-2 54 Input 30222 HD-ulm-1 64 Input 30223 C-ulm-1 44 Input 30229 HD-ulm-1 40 Input 30235 HD-ulm-1 48 Input 30240 HD-ulm-1 36 Input 30241 C-paris-2 65 Input 30247 HD-paris-1 24 Input 30248 HD-paris-1 44 Input 30250 HD-paris-2 62 Input 30254 HD-paris-1 37 Input 30256 HD-paris-1 23 Input 30259 HD-paris-1 64 Input 30265 C-paris-2 50 Input 30269 C-paris-2 58 Input 30273 C-paris-2 49 Input 30275 C-paris-1 59 Input 30280 HD-london-2 42 Input 30281 HD-london-2 65 Input 30288 HD-london-2 51 Input 30311 HD-paris-1 40 Input 30312 HD-paris-2 51 Input 30314 HD-paris-2 46 Input 30315 HD-leiden-2 67 Input 30318 C-paris-2 64 Input 30320 C-paris-1 49 Input 30321 HD-paris-1 61 Input 30326 C-paris-2 45 Input 30334 HD-london-1 49 Input 30347 C-paris-1 62 Input 30349 HD-paris-2 48 Input 30352 C-paris-2 66 Input 30357 C-ulm-1 28 Input 30361 HD-paris-2 63 Input 30363 HD-ulm-1 51 Input 30365 C-ulm-2 46 Input 30369 HD-ulm-2 45 Input 30378 C-leiden-1 53 Input 30385 HD-leiden-2 61 Input 30392 HD-ulm-2 51
Contrast: 1 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0
Thanks, Elin