Table 3. Stepwise ordered logistic regression analysis^* with oxytocin tertiles as the dependent variable.

Characteristics*	Odds Ratio	95% CI	P-value
Hypertension (Y/N)	2.42	0.84 to 6.95	0.100
Natural log AUC C-Peptide	0.24	0.12 to 0.50	0.001
Fasting insulin	1.03	1.00 to 1.05	0.016
Glomerular filtration rate	1.04	1.01 to 1.07	0.014
Smoking (Y/N)	4.09	1.53 to 10.90	0.005
Psychiatric medications (Y/N)	4.98	1.50 to 16.54	0.009

*Only variables that had a p <0.15 were included in the final model. Model R² = 0.21. Odds Ratio (OR) compares highest to lowest tertile of oxyocin. The following independent variables were considered for the model: waist circumference, fasting insulin, natural log AUC C-Peptide, creatinine, eGFR, IGF-1, diabetes status (Yes/No), smoking status (Yes/No), psychiatric Meds (Yes/No), total number of medications (any). Post-test estimation: Brant test of parallel regression assumption indicated that the parallel regression had not been violated.

Abbreviations: AUC, area under the curve.