Fronto-limbic Neural Activity in Response to Basic Emotion Cues in Adults and Adolescents with Anorexia Nervosa: An fMRI Study

Abstract

BACKGROUND: Anorexia Nervosa (AN) is a severe mental illness associated with chronicity, poor treatment outcome, high rates of morbidity and mortality, and significant individual and societal costs. Behavioral research studies show that AN is associated with deficits in several aspects of emotional processing, including emotional learning/memory, awareness, recognition, and regulation. It is possible that emotional processing abnormalities, driven by underlying genetic and neurobiological factors, contribute to AN development and persistence. Therefore, better understanding disruptions in basic emotional functioning in AN and the neural correlates that underlie them is critically important. METHOD: 19 adolescent and adult females with restricting-type AN (10 medicated, 9 unmedicated) and 19 healthy age and gender matched controls underwent functional MRI in conjunction with an emotional faces task where BOLD response was examined when viewing images of happy and fearful faces (presented in a block design format). Scanning was completed using a 3.0 Tesla scanner. Data preprocessing and analysis was carried out using FEAT in FSL. Whole brain group level analyses were conducted using a mixed-effects model (FLAME) with cluster-wise significant testing (min Z=2.32; cluster significance = p<0.05, corrected). Between group differences in eight regions of interest (ROIs: bilateral amygdala, insula, pregenual anterior cingulate, and subgenual anterior cingulate) were also examined (MANCOVA). Correlational analyses then investigated the relationships between brain response and self-reported eating disorder symptoms as well as between brain response and participant age. RESULTS: Whole brain comparisons showed that, in response to viewing fearful versus happy faces, participants with AN had lower activation in areas of the pregenual anterior cingulate and ventral prefrontal cortex. When examining ROIs, participants also showed reduced activity in pregenual anterior cingulate regions, which remained when controlling for body mass index and self-reported depression and anxiety. Findings did not appear to be a consequence of psychotropic medication use. Greater severity of eating disorder, depression, and anxiety symptoms was associated with lower activation in left ventral prefrontal regions. In addition, brain response in the left insula correlated with participant age and measures of eating disorder severity. No significant correlations were observed between clinical data and other ROIs, including the amygdala. CONCLUSION: Findings from the current study suggest that, during the ill state of restricting-type AN, fearful facial expressions elicit decreased functional activity in anterior cingulate and ventral prefrontal regions, areas of the brain that are central to emotional processing. Disruptions in neural activity may contribute to deficits in emotional awareness, recognition, and unconscious emotional regulation which have been widely observed within behavioral studies in AN. Future research is needed to further examine how fronto-limbic deficits in this disorder may be related to disease vulnerability and symptom maintenance. Such studies may guide identification of treatment targets and development of novel interventions, which are severely limited in this disorder.