شواهد الکتروفیزیولوژیک برای یک جانبداری توجه در محرک های بدن در پردازش بولیمیا

Empirical evidence suggests abnormalities in the processing of body stimuli in bulimia nervosa (BN). This study investigated central markers of processing body stimuli by means of event-related potentials in BN. EEG was recorded from 20 women with BN and 20 matched healthy controls while watching and evaluating underweight, normal and overweight female body pictures. Bulimics evaluated underweight bodies as less unpleasant and overweight bodies as bigger and more arousing. A higher P2 to overweight stimuli occurred in BN only. In contrast to controls, no N2 increase to underweight bodies was observed in BN. P3 was modulated by stimulus category only in healthy controls; late slow waves to underweight bodies were more pronounced in both groups. P2 amplitudes to overweight stimuli were correlated with drive for thinness and body dissatisfaction. We present novel support for altered perceptual and cognitive-affective processing of body images in BN on the subjective and electrophysiological level.

Human bodies provide important social cues, which contribute to the identification of other people, their age and gender as well as their intentions and affective states (Minnebusch & Daum, 2009). Abnormalities in perception and evaluation of body shape are a hallmark of eating disorders (ED) (Birtchnell et al., 1985, Lindholm and Wilson, 1988, Miyake et al., 2010, Stice and Shaw, 2002, Uher et al., 2005 and Williamson et al., 1989), most notably Anorexia Nervosa (AN) and Bulimia Nervosa (BN). Bruch (1962) is acknowledged as being the first to describe dysfunctional body image experiences as a main feature of eating disorders. The current diagnostic criteria for AN and BN outlined in DSM-V (American Psychiatric Association, 2013) demand the “undue influence of body weight or shape on self-evaluation” as well as a “disturbance in the way in which one's body weight or shape is experienced” (American Psychiatric Association, 2013). Previous research has demonstrated that body image disturbances are one factor preceding the onset of ED (Jacobi, Hayward, de Zwaan, Kraemer, & Agras, 2004), and, in addition, predict the maintenance (Stice, 2002) and relapse process of ED (Keel et al., 2002). Usually, disturbances of body image can be separated into a perceptual component and a cognitive-affective component (Cash and Brown, 1987, Cash and Deagle, 1997 and Slade and Brodie, 1994). The perceptual disturbance reflects problems to assess one's own body size exactly (body size distortion) (Cash and Brown, 1987, Cash and Deagle, 1997 and Slade and Brodie, 1994). Previous research has demonstrated that women with AN and BN seem to overestimate their own body dimensions to a comparable degree, whereas healthy women show a tendency to underestimate them (Cash and Deagle, 1997, Mohr et al., 2011 and Vocks et al., 2007). The cognitive-affective component of body image disturbance comprises negative body-related attitudes and emotions (body dissatisfaction) (Cash and Brown, 1987, Cash and Deagle, 1997 and Slade and Brodie, 1994). Eating-disordered individuals are highly dissatisfied with their size, shape and body appearance (Cash & Deagle, 1997). A negative evaluation of one's own body is often associated with body-related avoidance (e.g., not looking in the mirror or hiding one's body under baggy clothes) (Trautmann, Worthy, & Lokken, 2007). Body weight/shape dissatisfaction is the highest in bulimics compared with AN and controls, whereas perceptual distortion parameters do not differentiate between AN and BN and yielded significantly smaller effects than attitudinal indices comparing ED with controls (Cash & Deagle, 1997). Findings on AN imply that body image disturbance is a result of cognitive-evaluative dissatisfaction rather than of perceptual deficits (Skrzypek, Wehmeier, & Remschmidt, 2001). Cognitive biases are assumed to play a fundamental role in the formation and maintenance of eating disorders (Lee and Shafran, 2004, Williamson et al., 1999 and Williamson et al., 2004). According to cognitive-behavioral theory, dysfunctional schemata in individuals with an ED lead to a biased processing of disorder-relevant information (e.g. body-related stimuli), being expressed in the form of an attention, memory or interpretation bias (Williamson et al., 1999 and Williamson et al., 2004). An attention bias refers to the fact that individuals with an ED selectively attend to weight/shape-related stimuli (Williamson et al., 1999 and Williamson et al., 2004). Studies using the dot-probe paradigm (Blechert et al., 2010, Rieger et al., 1998, Shafran et al., 2007 and Shafran et al., 2008), the visual search task (Smeets, Roefs, van Furth, & Jansen, 2008) and the emotional stroop task (Dobson and Dozois, 2004 and Lee and Shafran, 2004) assessed an attentional bias for shape and weight stimuli in individuals with an ED. Eye-tracking studies indicate avoidance behaviors as well as increased direction of attention to unliked body regions and a difficulty in removing attention from body stimuli in general among females with high levels of drive for thinness and body dissatisfaction (Gao et al., 2013, Janelle et al., 2009 and Janelle et al., 2003). In a social comparison setting with self- and other-photos, patients with BN and women high in body dissatisfaction turned their attention away from their own bodies (Blechert et al., 2010 and Janelle et al., 2009) and toward bodies with lower BMIs compared to bodies with higher BMIs (Blechert, Nickert, Caffier, & Tuschen-Caffier, 2009). Findings from brain imaging also indicate processing abnormalities for body shape stimuli in patients with an ED. The main finding was the activation of a specific neural network in response to own and others body shapes in ED (Friederich et al., 2010 and Uher et al., 2005), including the lateral fusiform gyrus, the inferior parietal cortex and the lateral prefrontal cortex, with regional hypoactivations within this network in women with an ED (Uher et al., 2005). In an imaging study of Fladung et al. (2010), ventral striatal activation was higher during processing of underweight stimuli compared with normal-weight stimuli in AN, while the reverse pattern was observed in healthy women. The authors interpreted their results as support for theories of starvation dependence in maintenance of AN stressing the pivotal role of the human reward system in AN. One recent fMRI study by Mohr et al. (2011) in BN showed that the patients were less sensitive to body size distortions as reflected by the absence of neural modulation when watching distorted body images. Moreover, bulimics overestimated their own body size, suggesting a deficit in body size perception (Mohr et al., 2011). Abnormalities in the processing of body stimuli in ED have also been assessed by self-reported emotions and ratings of body shape and weight. Women with BN and AN rated underweight, normal weight and overweight female body stimuli as more aversive than controls; with higher levels of self-reported aversion to overweight compared with underweight and normal bodies (Uher et al., 2005). Being confronted with self-images, bulimics reported higher levels of insecurity and sadness (Tuschen-Caffier, Vögele, Bracht, & Hilbert, 2003); in a social comparison task with a slim body ideal patients with BN reported increased levels of anxiety (Van den Eynde et al., 2013). Studies which investigated ratings of body weight or shape in BN illustrate that the patients do not only overestimate their own body size (Blechert et al., 2009, Blechert et al., 2010 and Mohr et al., 2011), but also rated bodies with a comparable BMI to their own and bodies of others with higher BMIs larger as compared with AN and controls (Blechert et al., 2009 and Blechert et al., 2010). Bodies of others with lower BMIs, on the contrary, achieved similar ratings by bulimics and controls (Blechert et al., 2009). While imaging approaches (e.g., fMRI, PET) allow for identifying brain areas involved in processing of body stimuli with high spatial accuracy, they lack the fine temporal resolution to investigate the time course of brain dynamics accompanying body-selective processing. Event-related potentials (ERPs) on the contrary allow for recording of neuronal activity underlying attentional processing of stimuli with high temporal resolution. The early deflections in the ERPs of body perception are characterized by a P1 and a prominent N1, better known as the ‘N170′ in the case of face processing, a negative deflection at occipitotemporal electrodes peaking between 140 and 220 ms post-stimulus onset. The N170 is thought to reflect a late stage in the structural encoding of the visual stimulus (de Gelder et al., 2010). Underlying neural generators for body perception in the N170 time window were found in the right lateral occipitotemporal cortex, a location corresponding to the extrastriate body area (de Gelder et al., 2010). In an ERP study, Gao et al. (2011) registered higher amplitudes of the N170 toward fatness-related words compared with thinness-related words in body weight dissatisfied women and interpreted their findings in terms of an early negativity bias. The P2, which is described as a positive deflection (Carretié et al., 2001 and Huang and Luo, 2006) with an onset between 180 and 200 ms after stimulus presentation (Carretié et al., 2004 and Carretié et al., 2001) is attributed to reflect automatic attention processes (Carretié et al., 2004) and is associated with an early attentional negativity bias in emotional perception (Carretié et al., 2001, Delplanque et al., 2004 and Huang and Luo, 2006). Several studies showed higher amplitudes of the P2 for unpleasant relative to pleasant stimuli at frontal, central and parieto-occipital sites (Carretié et al., 2001, Delplanque et al., 2004 and Huang and Luo, 2006). Suggested neural generators of the P2 are the sensory cortex and the anterior cingulate cortex (ACC) (Carretié et al., 2004). One of the first ERP studies with fatness- and thinness-related stimuli assessed higher P2 amplitudes for body words than for neutral words for both weight dissatisfied women and controls (Gao et al., 2011). Subsequent phases of stimulus processing are reflected in the N2 and P3. The N2 is a negative deflection with a peak at 240 ms which is assumed to represent a final stage of automatic attention-related processing (Carretié et al., 2004) and deviation from implicit expectations (Gramann, Toellner, Krummenacher, Eimer, & Muller, 2007). Together with the P3a, the N2 is suggested to be part of an orienting complex, being involved in the detection of novel stimuli (Campanella et al., 2002). The P3, a later component in stimulus processing is described as a positive deflection in the time window from 250 to 500 ms after stimulus onset (Picton, 1992 and Polich, 2007). The P3 is associated with attention and working memory (Linden, 2005 and Polich, 2007), and it is assumed to reflect an attention-driven comparison process to evaluate the representation of stimuli in working memory (Polich, 2007). The P3 component is regarded to mirror the cognitive evaluation of the meaning of a stimulus, which involves a deliberative, conscious and controlled process (Campanella et al., 2002 and Huang and Luo, 2006). In previous studies, the P3 was observed to be sensitive for the emotional content of stimuli, as negative information induced larger P3 amplitudes as compared to positive and neutral stimuli (Huang and Luo, 2006 and Ito et al., 1998). The later slow wave, which commences in the region of the P3 is a positive deflection which can be recorded from a 200–300 ms time window after stimulus onset with a peak at 1000 ms (Cuthbert, Schupp, Bradley, Birbaumer, & Lang, 2000). Late positive waves indicate selective, increased attentive processing of emotional stimuli and are assumed to reflect motivational systems in the brain (Cuthbert et al., 2000). In patients with AN, higher P3 amplitudes for both body images and geometrical shapes were recorded, which reveals a nonspecific hyperarousal (Dodin & Nandrino, 2003). In weight dissatisfied women, Gao et al. (2011) recorded higher P3 amplitudes for fat and thin body words, compared with neutral words, whereby thin body words elicited largest P3 amplitudes. The authors interpreted their results as an attentional bias in terms of high relevance of thin body stimuli for weight esteem in weight dissatisfied women (Gao et al., 2011). Horndasch, Heinrich, Kratz, and Moll (2012) who presented female underweight, normal-weight and overweight body pictures to girls with AN recorded highest amplitudes of the late positive potential for images with underweight women in AN in an early and late time window, whereas typically developing girls showed highest amplitudes for stimuli of overweight women in an early time window. The authors interpreted their findings in terms of motivational significance of underweight body stimuli in AN (Horndasch et al., 2012). To our knowledge, no ERP study on BN currently exists that systematically varied the depicted weight of pictorial body stimuli. One severe shortcoming of recent studies is that using word stimuli and distorted body images provokes an artificial situation and does not reflect everyday confrontation with body-specific information and real-life body images. Therefore, in the present study, we wanted to address the question whether bulimic patients showed altered evaluation and processing of real body images depicting different weight classes, by analyzing subjective picture ratings and electrophysiological correlates of early and late processing of external body stimuli in time course. Our main aim was to draw further conclusions in terms of body image disturbance in BN. Regarding the self-report data, we hypothesized that bulimic patients would evaluate underweight and overweight stimuli as less pleasant and more arousing than controls, which should mirror the cognitive-affective component of body image disturbance. In addition, we expected that individuals with BN would perceive overweight body stimuli as bigger than controls, which should reflect the perceptual component of disturbed body image in terms of distorted weight estimation. Referring to the electrophysiological data, we assumed that the perceptual level of disturbed body image could be related to differences between healthy and bulimic persons in the processing of body stimuli as reflected in the N170 and P2. We expected an early attentional negativity bias toward overweight body images in women with BN, reflected by higher amplitudes of the N170 and the P2. The N2, the P3 and the late slow wave reflect higher cognitive and affective processing during later stages in time course (Campanella et al., 2002, Carretié et al., 2004, Cuthbert et al., 2000 and Polich, 2007) and are therefore good candidates to assess disturbances on the higher order cognitive-affective component of body image. We expected a late attentional/cognitive processing bias toward underweight body images in BN, reflected by larger amplitudes of the N2, the P3 and the late slow wave.