How to get the change rate using proc mixed? how to explain proc mixed output?

[PSZ]

I have 1000 persons data with 5 years' annual bone mineral density (BMD) measurement.  Now I want to get the bone mineral density (BMD) change rate per year. Because this is longitudinal data, I think I should use proc mixed.

My data is like:

id     visit                    group           bmd

1      0(baseline)          a                 1.01

1      1                         a                 1.00

1      2                         a                 1.00

1      3                         a                 0.99

1      4                         a                 0.98

1      5                         a                 0.96

2      0(baseline)          b                 1.07

2      1                         b                 1.06

2      2                         b                 1.04

2      3                         b                 0.99

2      4                         b                 0.95

2      5                         b                 0.93

......

My code is :

PROC MIXED DATA=mydaa noclprint;
CLASS id  visit   group;
MODEL  bmd = visit   group  group*visit / SOLUTION CL DDFM=KR OUTPM=aaa;
REPEATED visit / SUBJECT=id   TYPE=UN R RCORR;
RUN;

Is this code correct? and where can I get the change rate in the output? I read one paper said the estimate is the change rate, but the  numbers are positive (the change rate should negative). 

And I plot the result like this:

PROC SGPLOT DATA=aaa;
SERIES x = visit  y = pred / GROUP = group  MARKERS;
RUN;

The plot looks correct. But how can I get the numbers for the slop?

Thank you very much for your help.

[rbikes]

The estimate is only the estimate of the dependent variable. Here, the dependent variable is bmd, is bmd a change rate? Looking at the structure I would say it isn't. Assuming I am correct I would suggest you create baseline variable and then from that and bmd calculate the change and use the change variable as the dependent variable and also include baseline into the model.

[PSZ]

Thanks for your reply. If I understand correctly, the coefficient of the interaction term (group * visit) (i.e. the slop of the model) is the change rate. 

[rbikes]

If you just want to find the slop I would suggest to follow this tutorial by Rick Wicklin

https://blogs.sas.com/content/iml/2013/02/27/slope-of-a-regression-line.html