how to account for the case-control pair in mixed-effect model--proc mixed

[Jie111]

I want to compare the BMI change between people with and without disease. I have cases and their matched controls, matching by age and sex.

How could I take the matching pair into account in the proc mixed?

               proc mixed data = data;
                   class  t ;
                  model edss_score = follow_up_year follow_up_year*edu_self_report_3g edu_self_report_3g ;                                        
                      repeated t/ type= vc  subject=study_id r rcorr;
                      random intercept follow_up_year/type= vc  subject=study_id ;   
                 run;