UCLA

Anorexia Nervosa (AN) and Body Dysmorphic Disorder (BDD) are disorders of body image that share phenomenological and psychological patterns of flawed perception of appearance. Our working model of visual processing dysfunction in AN and BDD describes a primary deficit in configural/holistic visual processing. The consequence of this is a secondary, “inappropriate,” reliance on detailed processing of image features, which is abnormally deployed in situations in which healthy controls deploy configural processing. We hypothesize that this results in a diminished holistic template that is less able to aid in integration, producing a conscious perception dominated by details. We used EEG and fMRI to investigate possible biomarkers of aberrant early visual system activity, that may inform diagnoses, treatments, and therapies in the future.

The aim of my thesis is to understand the neural dynamics underlying visual processing abnormalities in AN and BDD. I used neuroimaging techniques, specifically functional magnetic resonance imaging (fMRI) and electroencephalography (EEG), to address this research question. Several fMRI studies have shown abnormal brain activation patterns during global encoding of faces and objects for BDD (1–3) and AN (4; 5). In addition to fMRI, I used EEG, which provides superior time resolution and thereby an opportunity to discover biomarkers complementary to the underlying pathophysiology uncovered in fMRI studies. To date, there have been no studies investigating early visual processing in electrophysiological components in either AN or BDD groups, one key area of focus for my dissertation.

In order to determine if individuals with AN and BDD have similar abnormalities in early holistic processing and slower detail processing relative to controls, I investigated amplitude and latency differences in early visual event related potentials (ERP), namely the P100 and N170.

These components are thought to reflect configural and detailed processing, respectively, and are abnormal in those with schizophrenia (6) and William’s Syndrome (7), two disorders in which individuals may suffer from similar abnormalities in global processing. The results of reduced P100 amplitudes and delayed N170 latency suggests early visual processing deficiencies in AN. Moreover, there was evidence of a possible brain-behavior relationship in the BDD group, as worse insight correlated with reduced N170 amplitude.

Next, I performed a joint analysis of EEG and fMRI signals using a technique called Fusion independent components analysis (ICA) to generate a joint spatiotemporal profile that leverages spatial and temporal resolution advantages of the respective modalities. This allowed us to localize early electrophysiological processes using the high spatial resolution of fMRI, and had yet to be performed in these populations. We found that AN and BDD showed hypoactivity in early and dorsal visual stream systems for low spatial frequencies, suggesting that a common deficiency in holistic processing is operating in primary visual structures as early as 100 ms post exposure and extends later in time (170 ms) into dorsal stream higher order regions. However, the patterns of hypoactivity are not identical; BDD but not AN additionally demonstrated hyperactivity compared to controls in ventral visual stream systems for high spatial frequency houses, and BDD showed hypoactivity in dorsal visual regions for low spatial frequency faces in comparison to participants with AN. Thus, an imbalance in detailed versus configural/holistic processing for non appearance-related stimuli may characterize both disorders, but the defect appears to be more pronounced in BDD. This dissertation contributes novel findings to the understanding of the pathophysiology underlying visual processing in these disorders, using methods that have not yet been performed in these populations. We found evidence of similar abnormalities between AN and BDD. There were also important differences in electrophysiological and hemodynamic signatures as well, which warrants further investigation. Further experiments that test other ERP components, frequency analyses, and simultaneous EEG-fMRI studies, and replication of the current findings in larger samples, will allow further characterization of neural signatures that can be used as possible biomarkers for more accurate diagnoses and treatment. Deeper understanding of the early visual processing mechanisms in AN and BDD could better inform perceptual treatments, and ERP or fMRI components may have the potential to be to used to track and monitor patient symptoms or disease severity in the future.