A regression model has an outcome and one or more predictors. You only mention outcomes, so you need to clarify what is the response variable you want the model to predict, and what are the predictor variables that you want to use in that model. Then, regarding the response variable, is it categorical or continuous, and if it is continuous, what sort of distribution does it have in any one setting of the predictor variables? Since you use the word "longitudinal", there are presumably clusters of repeated response measurements, such as multiple responses from subjects in a study. Describe this.
