In Scanner Self-Report Craving and Stress

Supplemental Material

The following provides a detailed description of in-scanner craving and stress ratings for Studies 1 and 2.

Study 1

In Scanner Self-Report Craving and Stress

To examine between group differences for craving and stress on the day of the fMRI procedure, a series of repeated measures ANOVAs were carried out with group as the between subject factor and time as the within subject factor. A main effect of time indicated that there was a statistically significant change in subjective craving for both groups following consumption of the standardized meal (F(1,19) = 40.92, p<.001). A follow up paired samples t-test indicated that craving significantly decreased (t (1,19) = -6.19, p<.001). There was not a statistically significant change in craving from the post-meal ratings to the baseline assessment of craving in the fMRI environment. There was a statistically significant change in craving following exposure to pre-stress food cues (F(1,19) = 36.51, p<.001). A follow-up paired samples t-test indicated that craving significantly increased (t (1,19) = 5.15, p <.001). A main effect of time indicated that for both groups, subjective craving changed following the stress induction (F(1,19) = 12.24 , p<.01). A follow up paired samples t-test indicated that craving significantly decreased at this point (t (1,19) = -3.58, p <.002). Craving again significantly changed following exposure to post-stress food cues (F(1,19) = 14.58, p<.01). A follow up paired samples t-test indicated that craving significantly increased (t (1,19) = 3.76, p<.001). Results of repeated measures ANOVAs also indicated that women with BN had statistically significantly different levels of craving throughout the paradigm. Examination of means and follow-up paired samples t-tests indicated that women with BN reported higher levels of craving than control women. Craving trajectories for both groups are presented in Supplemental Figure 1a.

The same design was utilized to examine changes in stress over the course of the scan session. A significant group x time interaction showed that there were significant differences in stress from pre to post meal between the BN and HC groups (F(1,19) = 4.88, p<.05). An independent samples t-test indicated that stress was significantly lower for HC women than BN women following the meal (t = 4.03, p<.001). A main effect of group showed that the BN group reported statistically significant differences in subjective stress relative to the HC group across the scan session (F(1,19) = 24.01, p<.001). Follow up t-tests indicated that across all time points, women with BN rated their subjective stress as higher than control women. A main effect of time effect indicates that stress changed significantly for both groups following the stress induction (F(1,19) = 32.91, p<.001). A follow-up paired samples t-test indicated that stress significantly increased from pre to post stress induction (t(1,19) = 5.81, p<.001). Ratings of subjective stress significantly changed for both groups following exposure to post-stress food cues (F(1,19) = 88.20, p<.001). A follow up paired samples t-test indicated that stress significantly decreased following exposure to post stress food cues (t (1,19) = -9.20, p<.001). See Supplemental Figure 1b for stress trajectories during the scan session in both groups.

Study 2

In Scanner Self-Report Craving and Stress

A series of repeated measures ANOVAs were conducted to examine the effect of time and cue type on subjective ratings of craving and stress within the BN subjects. The pattern of results was consistent with Study 1. The significant change in subjective craving from presentation of neutral cues to pre-stress food cues (F (1,16) = 26.47, p <.001) was in the expected direction (t(1,16) = 5.19, p<.001). Subjective craving did not significantly change following the stress induction (F (1,16) = .17, p<.68). There was not a statistically significant change in subjective craving from post stress induction to post-stress food cue presentation(F (1,16) = 3.35, p<.09).

There were no statistically significant changes in subjective stress from baseline to presentation of neutral cues (F (1,16) = 3.13, p<.10), nor from presentation of neutral cues to food cues (F (1,16) = 2.13, p<.16). There was a statistically significant change in subject stress from the first run of food cues to period immediately following the stress induction (F (1,16) = 17.56, p.001). A follow paired sample t-test indicated that the stress induction activity successfully elevated subjective stress (t (1,16) = 4.19, p<.001). There was a significant change in subject stress following presentation of the post-stress food cues (F (1,16) = 9.58, p <.007). A follow up paired sample t-test indicated that stress decreased (t(1,16) = -3.09, p<.007). See Supplemental Figure 2a and 2b for trajectories of craving and stress throughout the scan paradigm.

Supplemental Figure 1a. Trajectory of subjective craving levels during scan session. Y-axis = reported value of craving level on a scale of 0-6. Pre-Meal = craving rating before provision of standard meal, Post-Meal = craving rating 1 hour following meal consumption, Baseline = craving rating upon entering fMRI environment and following structural scan, Neutral = craving rating following neutral cues, Pre-Stress Food = craving rating following pre-stress food cues, Stress = craving rating following stress induction, Post-Stress Food = craving rating following post stress food cues.

Supplemental Figure 1b. Trajectory of subjective stress levels during scan session. Y-axis = reported value of stress level on a scale of 0-6. Pre-Meal = stress rating before provision of standard meal, Post-Meal = stress rating 1 hour following meal consumption, Baseline = stress rating upon entering fMRI environment and following structural scan, Neutral = stress rating following neutral cues, Pre-Stress Food = stress rating following pre-stress food cues, Stress = stress rating following stress induction, Post-Stress Food = stress rating following post-stress food cues.

Supplemental Figure 1a.

Supplemental Figure 1b.

Figure Caption 2a and 2b

Supplemental Figure 2a. Trajectory of subjective craving levels during scan session in Study 2. Y-axis = reported value of craving level on a scale of 0-6. Baseline = craving rating upon entering fMRI environment and following structural scan, Neutral = craving rating following neutral cues, Pre-Stress Food = craving rating following pre-stress food cues, Stress = craving rating following stress induction, Post-Stress Food = craving rating following post stress food cues. * = significant two-way interaction.

Supplemental Figure 2b. Trajectory of subjective stress levels during scan session. Y-axis = reported value of stress level on a scale of 0-6. Baseline = stress rating upon entering fMRI environment and following structural scan, Neutral = stress rating following neutral cues, Pre-Stress Food = stress rating following pre-stress food cues, Stress = stress rating following stress induction, Post-Stress Food = stress rating following post-stress food cues. * = significant two-way interaction.

Supplemental Figure 2a.

Supplemental Figure 2b.