I am working on longitudinal data with a 4-time point. I want to fit a model exploring how outcome Y(Binary) changes over time (no missing in any time-point) with the exposures X, W, T and Z (these exposures are also time-changing - some are full observed and missing in points - so unbalanced). But on the same model, I want to account for other exposure variables which do not change over time(Constant Baseline Covariates).

I have tried GEE models, and also explored some references that recommend generating another binary variable for each exposure per time point, but these will generate quite many exposures as 4 covariates * 4 point=16 new variables + other 6 constant covariates the model won't converge.

Please can you advise on how best I can analyse this type of data, or provide any reference or analysis program that I can learn more about this?