Hi, you can use linear mixed models (also called mixed effect models) which allows you to include also a proper treatment for the random effects. You can find some useful info here : http://courses.education.illinois.edu/edpsy587/lectures/longitudinal_beamer_online.pdf

If you are familiar with R, a good package is lme4 ( https://cran.r-project.org/web/packages/lme4/lme4.pdf ). You can find some examples using R with the lme4 package here : https://cran.r-project.org/web/packages/lme4/vignettes/lmer.pdf. 

thanks Paolo, mixed model is an extension of GLM which accounts for subject random effect. I'm hoping of getting a new method entirely different from GLM