wait for the professional

proc genmod data=resp descend;
   class id treatment(ref="P") center(ref="1") sex(ref="M")
      baseline(ref="0") / param=ref;
   model outcome=treatment center sex age baseline / dist=bin;
   repeated  subject=id(center) / logor=fullclust;
run;

LOGOR=log-odds-ratio-structure-keyword specifies the regression structure of the log odds ratio used to model the association of the responses from subjects for binary data. The response syntax must be of the single variable type, the distribution must be binomial, and the data must be binary.

   data resp;
      keep id active center center2 female age baseline visit outcome;
      input center id treatmnt $ sex $ age baseline visit1-visit4;
      active=(treatmnt='A');
      center2=(center=2);
      female=(sex='F');
      visit=1;  outcome=visit1;  output;
      visit=2;  outcome=visit2;  output;
      visit=3;  outcome=visit3;  output;
      visit=4;  outcome=visit4;  output;
      cards;
   1  1 P M 46 0 0 0 0 0
   1  2 P M 28 0 0 0 0 0
   1  3 A M 23 1 1 1 1 1
   1  4 P M 44 1 1 1 1 0
   1  5 P F 13 1 1 1 1 1
   1  6 A M 34 0 0 0 0 0
   1  7 P M 43 0 1 0 1 1
   1  8 A M 28 0 0 0 0 0
   1  9 A M 31 1 1 1 1 1
   1 10 P M 37 1 0 1 1 0
   1 11 A M 30 1 1 1 1 1
   1 12 A M 14 0 1 1 1 0
   1 13 P M 23 1 1 0 0 0
   1 14 P M 30 0 0 0 0 0
   1 15 P M 20 1 1 1 1 1
   1 16 A M 22 0 0 0 0 1
   1 17 P M 25 0 0 0 0 0
   1 18 A F 47 0 0 1 1 1
   1 19 P F 31 0 0 0 0 0
   1 20 A M 20 1 1 0 1 0
   1 21 A M 26 0 1 0 1 0
   1 22 A M 46 1 1 1 1 1
   1 23 A M 32 1 1 1 1 1
   1 24 A M 48 0 1 0 0 0
   1 25 P F 35 0 0 0 0 0
   1 26 A M 26 0 0 0 0 0
   1 27 P M 23 1 1 0 1 1
   1 28 P F 36 0 1 1 0 0
   1 29 P M 19 0 1 1 0 0
   1 30 A M 28 0 0 0 0 0
   1 31 P M 37 0 0 0 0 0
   1 32 A M 23 0 1 1 1 1
   1 33 A M 30 1 1 1 1 0
   1 34 P M 15 0 0 1 1 0
   1 35 A M 26 0 0 0 1 0
   1 36 P F 45 0 0 0 0 0
   1 37 A M 31 0 0 1 0 0
   1 38 A M 50 0 0 0 0 0
   1 39 P M 28 0 0 0 0 0
   1 40 P M 26 0 0 0 0 0
   1 41 P M 14 0 0 0 0 1
   1 42 A M 31 0 0 1 0 0
   1 43 P M 13 1 1 1 1 1
   1 44 P M 27 0 0 0 0 0
   1 45 P M 26 0 1 0 1 1
   1 46 P M 49 0 0 0 0 0
   1 47 P M 63 0 0 0 0 0
   1 48 A M 57 1 1 1 1 1
   1 49 P M 27 1 1 1 1 1
   1 50 A M 22 0 0 1 1 1
   1 51 A M 15 0 0 1 1 1
   1 52 P M 43 0 0 0 1 0
   1 53 A F 32 0 0 0 1 0
   1 54 A M 11 1 1 1 1 0
   1 55 P M 24 1 1 1 1 1
   1 56 A M 25 0 1 1 0 1
   2  1 P F 39 0 0 0 0 0
   2  2 A M 25 0 0 1 1 1
   2  3 A M 58 1 1 1 1 1
   2  4 P F 51 1 1 0 1 1
   2  5 P F 32 1 0 0 1 1
   2  6 P M 45 1 1 0 0 0
   2  7 P F 44 1 1 1 1 1
   2  8 P F 48 0 0 0 0 0
   2  9 A M 26 0 1 1 1 1
   2 10 A M 14 0 1 1 1 1
   2 11 P F 48 0 0 0 0 0
   2 12 A M 13 1 1 1 1 1
   2 13 P M 20 0 1 1 1 1
   2 14 A M 37 1 1 0 0 1
   2 15 A M 25 1 1 1 1 1
   2 16 A M 20 0 0 0 0 0
   2 17 P F 58 0 1 0 0 0
   2 18 P M 38 1 1 0 0 0
   2 19 A M 55 1 1 1 1 1
   2 20 A M 24 1 1 1 1 1
   2 21 P F 36 1 1 0 0 1
   2 22 P M 36 0 1 1 1 1
   2 23 A F 60 1 1 1 1 1
   2 24 P M 15 1 0 0 1 1
   2 25 A M 25 1 1 1 1 0
   2 26 A M 35 1 1 1 1 1
   2 27 A M 19 1 1 0 1 1
   2 28 P F 31 1 1 1 1 1
   2 29 A M 21 1 1 1 1 1
   2 30 A F 37 0 1 1 1 1
   2 31 P M 52 0 1 1 1 1
   2 32 A M 55 0 0 1 1 0
   2 33 P M 19 1 0 0 1 1
   2 34 P M 20 1 0 1 1 1
   2 35 P M 42 1 0 0 0 0
   2 36 A M 41 1 1 1 1 1
   2 37 A M 52 0 0 0 0 0
   2 38 P F 47 0 1 1 0 1
   2 39 P M 11 1 1 1 1 1
   2 40 P M 14 0 0 0 1 0
   2 41 P M 15 1 1 1 1 1
   2 42 P M 66 1 1 1 1 1
   2 43 A M 34 0 1 1 0 1
   2 44 P M 43 0 0 0 0 0
   2 45 P M 33 1 1 1 0 1
   2 46 P M 48 1 1 0 0 0
   2 47 A M 20 0 1 1 1 1
   2 48 P F 39 1 0 1 0 0
   2 49 A M 28 0 1 0 0 0
   2 50 P F 38 0 0 0 0 0
   2 51 A M 43 1 1 1 1 0
   2 52 A F 39 0 1 1 1 1
   2 53 A M 68 0 1 1 1 1
   2 54 A F 63 1 1 1 1 1
   2 55 A M 31 1 1 1 1 1
    ;

proc genmod data=resp descend;
class id center;
model outcome=center2 active female age baseline / dist=bin;
repeated  subject=id(center) / logor=fullclust;
run;

logor= gives the Alternative Logistic Regression Algorithm for repeated measurements, which is irrelevant to the question.
In genmod, there is no way to directly give OR in parameter estimates output, as far as I know. what we normally do is taking the Parameter Estimates first and then doing the exponential calculation.
another way to work around the problem is through using statement ESTIMATE/EXP syntax but it needs you to write out the correct estimate formula, which often intimidates some beginners.
JingJu

这个回答是靠谱的。
proc genmod data=fmpr descend;
class fno isb4(ref="3") /param=ref;
model glucose=isb4 /dist=bin;
repeated subject=fno/corr=un;
estimate "OR：0 vs 3" isb4 1 0 0 -1/exp;/**第一组与第三组（对照）相比的OR**/;
run;