Table 1.
* Slope estimates for association between log salivary cortisol concentration and Δ carbohydrate.
| Sample Time (Sample Number) | Diet | Intervention Week | Slope | S.E. | P-Value |
|---|---|---|---|---|---|
| Pre-TSST (9) | DGA | Baseline | 0.001463 | 0.001487 | 0.3266 |
| DGA | 9 | −0.00312 | 0.001487 | 0.0372 | |
| 30 min post-TSST (10) | DGA | Baseline | 0.001142 | 0.001487 | 0.4434 |
| DGA | 9 | −0.0039 | 0.001487 | 0.0096 ** | |
| 60 min post-TSST (11) | DGA | Baseline | 0.001814 | 0.001507 | 0.2306 |
| DGA | 9 | −0.00312 | 0.001517 | 0.041 | |
| 90 min post-TSST (12) | DGA | Baseline | 0.001689 | 0.001493 | 0.2596 |
| DGA | 9 | −0.00298 | 0.00152 | 0.0516 | |
| 120 min post-TSST (13) | DGA | Baseline | 0.002341 | 0.001492 | 0.1186 |
| DGA | 9 | −0.00193 | 0.001488 | 0.1955 | |
| Pre-TSST (9) | TAD | Baseline | 0.000565 | 0.001402 | 0.687 |
| TAD | 9 | −0.00013 | 0.001438 | 0.93 | |
| 30 min post-TSST (10) | TAD | Baseline | 0.000172 | 0.001402 | 0.9023 |
| TAD | 9 | 0.000992 | 0.001438 | 0.4908 | |
| 60 min post-TSST (11) | TAD | Baseline | −0.00139 | 0.001402 | 0.3228 |
| TAD | 9 | 0.000077 | 0.001439 | 0.9571 | |
| 90 min post-TSST (12) | TAD | Baseline | −0.00083 | 0.001405 | 0.5551 |
| TAD | 9 | −0.00017 | 0.001442 | 0.9073 | |
| 120 min post-TSST (13) | TAD | Baseline | −0.00106 | 0.001406 | 0.4501 |
| TAD | 9 | −0.00041 | 0.001439 | 0.7742 |
* The significant diet group–week–saliva sample–Δ carbohydrate interaction suggested the nature of the association between cortisol and Δ carbohydrate differed among diet groups (DGA, N = 22; TAD, N = 21), intervention week, and saliva samples. Therefore, we also used the mixed model procedure, accounting for repeated measurements, to estimate coefficients (slopes) for the relationship between log salivary cortisol (nmol/L) and magnitude of change (Δ) in carbohydrate consumption from the pre-intervention (habitual) to the diet intervention period as a continuous variable at each combination of diet group, intervention week, and saliva sample. The P-value in bold indicates that the slope was significantly different from 0, while the double asterisk (**) indicates that the slopes for that week and sample number significantly differed between the DGA and TAD groups. Baseline age, education, BMI, and hunger score from the Three-Factor Eating Questionnaire were included as independent variables in the statistical model. TSST: Trier Social Stress Test; Pre-TSST: Saliva sample taken immediately prior to the stress test; DGA: Dietary Guidelines for Americans diet intervention group; TAD: typical American diet intervention group.