Multilevel regression or Conjoint analysis

Hi. We have made a survey to find out if men have a more positive thinking of protein than woman with a rating of some products which ranges from 1 to 7 as a health perception. We have made some levels of "High, Medium and low" of all the three macronutriens (Protein, Carbs and fat).
Furthermore we have people age, so we can make the analysis form age, if the gender does not give us anything significant.

We are now very troubled with how to do the actual research and which of the two models are best to use. We have both JMP and SAS, but we have only worked in JMP.
Our questions are:
Why do JMP remove high level of protein and age? (See picture)

Does anybody have some code in SAS that we could make our research from? We are very bad at programming ourselves, so we hope that some of you guys could help us out.

And does anybody have a step-by-step guide to making a Conjoint Analysis in JMP?