Multinomial regression

#1
Hi,

I have carried out a 'multinomial regression analysis' to investigate the relationship between a dependent variable (which has six categories) and five independent variables. All of the variables (both dependent and independent) are categorical. The results I obtain are pretty weird to me but I have not found any theory regarding the required number of dependent and independent variables. Does anybody know if there are any assumptions regarding the number of variables?

Thank you very much in advance.