Table 4.

The Association of log-morning serum cortisol with Prevalent Diabetes in the Jackson Heart Study

Logistic Regression Model – Odds Ratio (95% CI) for Prevalent Diabetes
	Unadjusted	Model 1	Model 2
Log-Cortisol	1.86 (1.53 – 2.24), p<0.001	2.13 (1.71 – 2.65), p<0.001	2.07 (1.66 – 2.59), p<0.001
Morning Serum Cortisol in Quartiles
Quartile 1	Referent	Referent	Referent
Quartile 2	1.46 (1.16 – 1.85), p=0.001	1.45 (1.13 – 1.86), P=0.004	1.44 (1.12 – 1.85), p=0.005
Quartile 3	1.55 (1.23 – 1.96), p<0.001	1.50 (1.17 – 1.94), p=0.002	1.47 (1.14 – 1.90), p=0.003
Quartile 4	2.05 (1.64 – 2.57), p<0.001	2.32 (1.80 – 2.99), P<0.001	2.26 (1.75 – 2.91), p<0.001

Model 1 – Adjusted for age, sex, education, occupation, systolic blood pressure, waist circumference, current smoking, physical activity, hormone replacement therapy, beta-blocker medications Model 2 – Model 1 + time of cortisol collection

For the logistic regression, the odds ratios are expressed as a percentage of higher prevalence per log unit increase (continuous) and a percentage of higher prevalence of diabetes per quartile compared to Quartile 1.

Interpretation: In the continuous analysis (unadjusted) a 1-unit increase in log-cortisol is associated with an 86% higher odds of prevalent diabetes and Quartile 4 compared with Quartile 1 was associated with a 105% higher odds of prevalent diabetes