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.