Multi-unit relations among neural, self-report, and behavioral correlates of emotion regulation in comorbid depression and obesity

Depression is a leading cause of disability and is commonly comorbid with obesity. Emotion regulation is impaired in both depression and obesity. In this study, we aimed to explicate multi-unit relations among brain connectivity, behavior, and self-reported trait measures related to emotion regulation in a comorbid depressed and obese sample (N = 77). Brain connectivity was quantified as fractional anisotropy (FA) of the uncinate fasciculi, a white matter tract implicated in emotion regulation and in depression. Use of emotion regulation strategies was assessed using the Emotion Regulation Questionnaire (ERQ). We additionally measured reaction times to identifying negative emotions, a behavioral index of depression-related emotion processing biases. We found that greater right uncinate fasciculus FA was related to greater usage of suppression (r = 0.27, p = 0.022), and to faster reaction times to identifying negative emotions, particularly sadness (r = −0.30, p = 0.010) and fear (r = −0.35, p = 0.003). These findings suggest that FA of the right uncinate fasciculus corresponds to maladaptive emotion regulation strategies and emotion processing biases that are relevant to co-occurring depression and obesity. Interventions that consider these multi-unit associations may prove to be useful for subtyping and improving clinical outcomes for comorbid depression and obesity.


RAINBOW-ENGAGE Science of Behavior Change Study (ENGAGE) Exclusion Criteria
Exclusion criteria for the RAINBOW study included active suicidal ideation, any Axis I disorder outside of major or minor depressive disorder as well as comorbidities with any anxiety disorder, active bulimia nervosa within the past three months, active alcohol or substance use disorder, ongoing psychiatric care with a provider outside of Palo Alto Medical Foundation (PAMF), plans to undergo bariatric surgery during the study period or having underwent the surgery in the previous year, preexisting diabetes (other than during pregnancy), pre-existing cardiovascular disease, diagnosis of cancer (other than non-melanoma skin cancer) that is or was active or treated with radiation or chemotherapy in the past year, diagnosis of a terminal illness and/or residence in a long-term care facility, cognitive impairment (measured from the Callahan 6-item screener 60 ), inability to speak, read, or understand English, having no reliable telephone service or no regular Internet access via a computer or mobile device, plans to move out of the study area, current pregnancy or plans to become pregnant during the study period, concurrent enrollment in any other study that could act as a confound, family members of study staff, as well as exclusion at the investigator's discretion for clinical safety or protocol adherence reasons. Exclusion criteria for ENGAGE included a weight limit of 350 lbs, an inability to fit in the MRI system, known neuroanatomical structural abnormalities (e.g., tumor or trauma-induced abnormality), as well as other MRI-related constraints.

Quality Assurance
After completion of Automated Fiber Quantification (AFQ), we performed several quality assurance steps to ensure high quality fiber clustering of our data. This included a visual inspection of all rendered uncinate fasciculi. Visual inspection of rendered fiber tracts revealed an irregular right uncinate in one participant. This rendering depicted many fibers that deviated from the core of the fiber tract. We conducted additional analyses removing this individual to ensure that this participant's data did not influence our results.
For quality assurance, we evaluated the calculated number of fibers for all uncinate fasciculi. The two lowest fiber counts were at least 2.07 standard deviations of the mean (M=1431.3, SD=508.6). To assess their contribution to reported effects we removed these two individuals and conducted additional analyses. After removing the two lowest fiber counts, all correlations we had previously identified remained significant before and after controlling for age (Table S1).
As an additional quality check, we removed one participant from our analysis on the basis of visual inspection of their uncinate rendering (Fig. S3). After adding this step to all previous quality checks (removal of +/-3 standard deviation outliers and bottom 2 fiber counts), all correlations for FA of the right uncinate remained significant before and after controlling for age (Table S2). Table S1: Correlations between Fractional anisotropy (FA) of the right uncinate fasciculus with behavioral and suppression scores after removing the two lowest fiber count individuals and outliers (|Z| > 3). Results are also included for age-corrected partial correlations. All correlations included at least one non-normally distributed variable (as assessed by the Lillifor's Test) and therefore tests were conducted using Spearman correlation.

Neg
RT=averaged negative reaction time, S RT=sad reaction time, D RT=disgust reaction time, A RT=anger reaction time, F RT=fear reaction time, Suppr=ERQ suppression score.