The table below is the output of polytomous (multinomial) logistic regression of clinical classification on insulin response (IR), steady-state plasma glucose (SSPG), and relative weight (RW). The clinical classification has three categories: 1 = overt diabetes, 2 = chemical diabetes, and 3= normal (which has been used as reference category).

Write the polytomous logistic regression model in terms of logit functions:

Odds 95% C.I. Variable Coefficient Z-Test p-value Ratio Lower Upper s.e. Logit 1: (2/3) 2.336 Constant -7.615 -3.26 0.001 RW 3.473 2.446 1.42 32.23 0.156 0.27 3894.2 SSPG 0.016 0.005 3.29 0.001 1.02 1.01 1.03 IR 0.004 0.002 1.53 0.127 1.00 1.00 1.01 Logit 2: (1/3) 3.463 Constant -1.845 -0.53 0.594 3.867 RW -5.868 -1.52 0.129 5.53 0.00 0.00 SSPG 0.046 0.009 0.000 4.92 1.05 1.03 1.07 IR -0.0134 0.005 -2.66 0.008 0.99 0.98 1.00 Log-Likelihood = -68.415 G = 159.369 df = 6 p-value < 0.000