Mixed model of repeated measures versus slope models in Alzheimer’s disease clinical trials