WARNING message in proc genmod

[Kunko]

I am trying to model the rish of diet score in tertiles (i.e. three leveles:T1, T2 T3) on risk of developing a disease (binary outcome) using modified poisson regression. I am gettng the outputs with estimates (Relative risks and 95% CIs), but receiving a warning "More coefficients than levels specified for effect WDP_Terttile2e. Some coefficients will be ignored." How can address this problem? Will it have any effect on my estimates (Relative risks) and confidence interval if I use the current outputs that I am getting?
607 Proc genmod descending data = bloodprofile30_darsy;
608 class codea WDP_Terttile2e muspgen reduc21_c income21cat rcig21_c rrrchalc21_c TVHrPerW
608! TPA_scorecat M211 ;
609 model TCHDLR30_Q=WDP_Terttile2e muspgen age21ffq reduc21_c income21cat rcig21_c
609! rrrchalc21_c TVHrPerW TPA_scorecat M211 LEnergy_kC /dist = /*binomial */ poisson link=log
609! lrci type3;
610 repeated subject = codea/ type = unstr;
611 estimate 'T2 vs T1' WDP_Terttile2e -1 1 0 0/exp;
612 estimate 'T3 vs T1' WDP_Terttile2e -1 0 1 0/exp;
613 format WDP_Terttile2e WDP_Terttile2e.;
614 run;

NOTE: The DESCENDING option applies only to the multinomial distribution with ordinal response.
NOTE: Class levels for some variables were not printed due to excessive size.
NOTE: Algorithm converged.
NOTE: Algorithm converged.
NOTE: The empirical covariance matrix estimate is used in the ESTIMATE statement.
WARNING: More coefficients than levels specified for effect WDP_Terttile2e. Some coefficients
will be ignored.
WARNING: More coefficients than levels specified for effect WDP_Terttile2e. Some coefficients
will be ignored.
NOTE: PROCEDURE GENMOD used (Total process time):
real time 1.01 seconds

[PaigeMiller · Accepted solution]

Your tertile variable has 3 levels.

Your ESTIMATE statement for comparing tertile levels has 4 levels.

[Kunko]

Fixed! Thank so much Paige!