I have a balanced repeated measurements data-set containing continuous
measurements on four occasions, for 81 subjects. The first occasion is prior
to surgery, the second is after the surgery and before the initiation of a
treatment, the third is 6 months after treatment initiation and the last 12
months after treatment. At the same occasions I have measurements for all
subjects of another continuous marker which is related to the marker of
interest. I want to estimate the effect of treatment while accounting for
the second marker�s levels. My first thought was to use :
