Intranasal oxytocin attenuates attentional bias for eating and fat shape stimuli in patients with anorexia nervosa

Psychoneuroendocrinology (2014) 44, 133—142

Available online at www.sciencedirect.com

ScienceDirect j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / p s y n e u e n

Intranasal oxytocin attenuates attentional bias for eating and fat shape stimuli in patients with anorexia nervosa Youl-Ri Kim a,*, Chan-Hyung Kim b, Valentina Cardi c, Jin-Sup Eom d, Yoori Seong a, Janet Treasure c a

Eating Disorders Clinic, Department of Neuropsychiatry, Seoul Paik Hospital, Inje University, Seoul, Republic of Korea b Department of Psychiatry, Severance Mental Health Hospital, Yonsei University College of Medicine, Gyeonggi Do, South Korea c Section of Eating Disorders, Department of Psychological Medicine, King’s College London, Institute of Psychiatry, London, United Kingdom d Department of Psychology, Chungbuk National University, Cheongju, Republic of Korea Received 31 October 2013; received in revised form 18 February 2014; accepted 24 February 2014

KEYWORDS Anorexia nervosa; Oxytocin; Attention bias; Dot probe task

Summary We examined the impact of oxytocin on attentional processes for eating, shape, and weight stimuli in patients with anorexia nervosa (AN). A double-blind, placebo-controlled withinsubject crossover design was used. Intranasal oxytocin or placebo followed by a visual probe detection task with food, weight, and shape images was administered to 64 female subjects: 31 patients with AN and 33 control students. The AN group showed significant reductions in the attentional biases toward eating-related stimuli ( p = 0.030, d = 0.516) and toward negative shape stimuli ( p = 0.015, d = 0.498) under the influence of intranasal oxytocin. The effect of oxytocin was correlated with autistic spectrum traits in the AN group. Oxytocin had no effect on the amount of juice consumed in either group. The results of this study suggest that oxytocin attenuates the attentional vigilance to eating and fat shape stimuli in patients with AN. Further studies using oxytocin as a form of intervention for patients with AN are needed. # 2014 Elsevier Ltd. All rights reserved.

* Corresponding author at: Department of Neuropsychiatry, Seoul-Paik Hospital, Inje University, 85 Jeo-dong 2 Ga, Jung-gu, Seoul 100-032, Republic of Korea. Tel.: +82 2 2270 0063; fax: +82 2 2270 0344. E-mail addresses: [email protected], [email protected] (Y.-R. Kim). http://dx.doi.org/10.1016/j.psyneuen.2014.02.019 0306-4530/# 2014 Elsevier Ltd. All rights reserved.

134 Anorexia nervosa (AN) is characterized by avoidance of eating, which causes weight loss and is associated with concerns about weight and shape. The National Institute of Health suggested using a method of classification that characterizes disorders according to their underlying neurobiology to facilitate translational work and an understanding of the basic mechanisms underpinning psychiatric disorders (Cuthbert and Insel, 2013). The psychopathology of AN is not thought to be a lack of appetite, but rather a fear of appetite associated with possible anomalies in systems related to fear and defense of appetitive behaviors, particularly in relationship to food, shape and weight. Anomalies in these systems can impact attention. Attention is part of the process of selectively processing information. The allocation of attention is influenced by topdown information (representations of goals and their relevance). These goals are merged with bottom-up priority signals. A balance and synthesis of these inputs determines the allocation of attention. Certain bottom-up signals such as threats have high salience. Additionally, motivational drives such as hunger can also change attentional processing (Mogg et al., 1994). Anxiety disorders are thought to be underpinned by maladaptive attentional biases toward threats. In a meta-analysis of 172 studies, patients with a wide variety of anxiety disorders had a small to moderate attentional bias to threat cues [effect size (ES) = 0.45] (Bar-Haim et al., 2007). ‘‘Fear of fat’’ is central to the diagnosis of AN as it relates to body shape or weight with a fear of ingesting fats. Patients with AN usually report anxiety in anticipation of a meal, concern about the content of the foods consumed, and fear of the effects of food on shape and weight (Sunday et al., 1995; Steinglass et al., 2011). Thus, both food and fatness are feared in AN. Cognitive models of eating disorders suggest that selective attention to shape-, weight- and food-related information may lead to the maintenance of eating disorder symptoms (Vitousek and Hollon, 1990). Attentional biases to food, weight, and shape cues have been documented in systematic reviews of the eating disorder literature (Faunce, 2002; Dobson and Dozois, 2004; Johansson et al., 2005; Brooks et al., 2011; Aspen et al., 2013). Most studies measuring this construct have used the Stroop paradigm and found small to moderate interference effects in AN and bulimia nervosa with an effect sizes of 0.38 (n = 13) and 0.43 (n = 11), respectively (Brooks et al., 2011). However, the Stroop task is not a simple measure of attention. The dot probe task is more specific, but it has been less frequently used in this population. A meta-analysis of a small number of studies (n = 3, some with overlapping populations) measuring attention using the dot-probe task in women with eating disorders (predominantly bulimia nervosa) found vigilance to binge food (ES = 0.8) and a small vigilance bias toward shapes depicting fatness (ES = 0.24) (Aspen et al., 2013). The interpretation of studies using the dot probe task is difficult, as this task does not distinguish emotional valence. For example, increased attention may be due to threat processing, as has been found using angry faces, for example in people with anxiety disorders (Mogg and Bradley, 1998), or it may be due to reward processing, as has been found in the case of addictions (Friederich et al., 2006). The startle paradigm is sensitive to valence and has shown that the mechanism of an increased attentional bias may differ between AN and bulimia nervosa. For example, food stimuli

Y.-R. Kim et al. accentuated the startle response in AN which contrasted with the attenuated startle in people with bulimia nervosa, suggesting that food is associated with fear in AN but reward in bulimia nervosa. In AN, changes in attention to both food and fatness may be caused by increased threat sensitivity. Many of the neural circuits that are relevant to AN, including problems with social processing, fear, anxiety, and stress, are moderated by oxytocin (Kirsch et al., 2005). One of the mechanisms by which oxytocin produces its effect is through its influence on attentional processing. For example, intranasal oxytocin given to rhesus monkeys specifically decreases the attention paid to threatening facial expressions (Parr et al., 2013). Thus, oxytocin might be expected to reduce attention vigilance to the specific food, weight and shape stimuli that are threatening to people with AN. A recent review synthesized evidence for abnormal oxytocin functioning in patients with AN (Maguire et al., 2013). The cerebrospinal fluid levels levels of oxytocin decrease during the starvation phase of AN (Demitrack et al., 1990; Chiodera et al., 1991). The nocturnal serum oxytocin levels in patients with AN are also reduced (Lawson et al., 2011). Lawson et al. (2012) reported that oxytocin release increases in response to a meal in the acute state of AN and decreases after recovery. This observation suggests that there are both state and trait anomalies in oxytocin function in patients with AN. Anomalies in oxytocin secretion are correlated with the severity of the eating disorder psychopathology and with the level of activation in the brain circuitry in response to images of food (Lawson et al., 2012). Finally, in a proof-of-concept study, intranasal oxytocin given to patients with AN over 6 weeks during nutritional rehabilitation reduced their eating and weight concerns (Russell et al., 2012). A possible mechanism to explain this finding is the effect of oxytocin on attentional responses and the threat of eating disorderrelated stimuli. Over the past decade, oxytocin has been diversely tested as a form of treatment for many psychiatric disorders. Clinical trials of patients with autism, social anxiety, postnatal depression, obsessive-compulsive problems, schizophrenia, borderline personality disorder, and post-traumatic stress (n = 19) were included in a recent comprehensive review by Bakermans-Kranenburg and van Ijzendoom (BakermansKranenburg and van Ijzendoorn, 2013). Overall, a moderate effect was seen (d = 0.32) and, in particular, a large ES was observed in studies on autism spectrum disorder (d = 0.57). This result is of interest for research into the effect of oxytocin on patients with eating disorders, as Gillberg (1992) suggested a strong association between eating disorders and autism, connected by low levels of empathy. Several lines of evidence summarized in recent reviews support this hypothesis. First, there is frequently co-morbidity within cases and families. Second, there are similar patterns of neuropsychological functioning (Zucker et al., 2007; Oldershaw et al., 2011; Treasure et al., 2012). Third, there are shared traits such as those measured by the Autism Spectrum Questionnaire (Hambrook et al., 2008; BaronCohen et al., 2013). It is possible that oxytocin may have a greater effect in the subgroup of patients with eating disorders with higher levels of autistic spectrum traits. In addition to its effects on fear and reward circuits, oxytocin has an effect on appetite in animals. For example, oxytocin selectively suppresses sugar intake, perhaps by

Oxytocin attenuates attentional bias in AN reducing hedonic eating (Sabatier et al., 2013). Thus, it is possible that oxytocin might reduce the reward derived from food. In principle, this effect might make the use of oxytocin in AN a cause of concern, as it might further reduce appetite through this mechanism. However, as discussed above, the dominant mechanism underpinning food avoidance in people with AN is because there is a learned response to food, as if it is a threat rather than a reward. It is therefore important for the impact of oxytocin on consummatory behavior to be examined. This is a proof of concept study that aims to broadly examine whether oxytocin might be beneficial in the short term. Further, it will lay the ground work for further studies which can provide answers to the indication, doses and longterm effect. The aim of this study was to examine whether oxytocin has an impact on attentional processes specific to the psychopathology of people with AN (food, weight and shape concerns) and also on consummatory behavior in patients with AN in comparison to healthy controls. The hypotheses that we tested were: first, patients with AN would have an increased attentional bias to food stimuli; second, patients with AN would have an increased attentional bias to fat shape stimuli; third, patients with AN would have an increased attentional bias to weight stimuli; fourth, oxytocin would reduce the attentional bias to these stimuli; fifth, oxytocin would decrease the amount of a fruit drink consumed; sixth, patients with AN and with a high level of autistic spectrum traits would have the greatest response.

1. Methods 1.1. Participants Sixty-four women (31 patients with AN and 33 healthy university students) between the age of 16 and 45 took part in this double-blind, placebo-controlled, cross-over study. The first participant entered the study on August 16, 2012, and the last participant was examined on November 30, 2013. The patients with AN were recruited from the Eating Disorders Clinic of Seoul Paik Hospital, Seoul, South Korea. As the study aimed to examine broadly whether oxytocin might be of benefit in the short term, patients from both the outpatient clinic (n = 13) and the inpatient ward (n = 18) were recruited. All patients who volunteered to participate were screened during the early phase of treatment. None of them were in the weight-recovered stage. The diagnosis of AN was confirmed by the Structured Clinical Interview from the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition (First et al., 2002). The exclusion criteria for patients were: active substance use disorder, diagnosis of a psychotic disorder (schizophrenia, schizoaffective, psychosis not otherwise specified), diagnosis of autism or Asperger’s syndrome, and presence of an animal phobia. All other comorbid diagnoses were allowed. Patients taking psychiatric medications other than fluoxetine were excluded [three patients who had been taking a stable dose of fluoxetine (20 mg/d) for several weeks were included]. The participants in the comparative group were undergraduate or graduate students who responded to an advertisement posted in the psychology department at a women’s university in Seoul, South Korea. The inclusion criteria were

135 healthy females without a history of medical or psychiatric illnesses and a minimum of 17 years of age. All subjects were nonsmokers, heterosexual, nulliparous, and were not taking any medications (including the contraceptive pill). Exclusion criteria included a self-reported history of major depression; bipolar, panic, or psychotic disorders; substance dependence, epilepsy, eating disorders, autism spectrum disorders, animal phobia, or traumatic brain injury. The Korean version of the Wechsler Adult Intelligence Scale was used to measure IQ. Healthy participants were tested during the follicular phase of their menstrual cycle and all the patients with AN were amenorrheic. Compensation was provided for traveling expenses and time. This study protocol was approved by both the Korean Food and Drug Association Institutional Review Board (12061) and the Institutional Review Board of Seoul Paik Hospital (IIT-2012-096). All participants provided written informed consent prior to participation in the study. This study was registered at the Clinical Research Information Service (http://cris.nih.go.kr) (registration number: KCT0000716).

1.2. Drug preparation Intranasal oxytocin sprays were formulated by JW Pharmaceuticals (Seoul, South Korea) from oxytocin powder (Hemmo Pharmaceuticals, Mumbai, India). 35.2 mg oxytocin (568 U) was mixed into 300 mL 0.9% sodium chloride solution and the pH was adjusted to 4.01 with 10 diluted acetic acid. The placebo spray (pH 4.01) was also formulated with 0.9% sodium chloride solution and acetic acid, but without the peptide. The filtered and sterilized solutions were sealed in 1.5 mL vials and stored frozen. On the day of use, the vials were thawed and kept in a refrigerator (4 8C) until required. A clinician prepared the nasal spray by transferring oxytocin or the placebo from the vial into a nebulizer. The nebulizer was primed and given to the participants, who self-administered the nasal spray while being monitored by the clinician. It was designed that each oxytocin spray would deliver 40 IU of oxytocin, based on a recent review which reported that the short-term use of intranasal OTof up to 40 IU (per dose) in humans, both males and females, resulted in no detectable subjective changes in controlled research settings (MacDonald et al., 2011).

1.3. Drug administration Each participant visited the laboratory twice for testing; once in the placebo condition and once in the oxytocin condition. The order of placebo or oxytocin administration for each participant was determined randomly by the project coordinator who was not involved in the experiment, using Microsoft Excel as follows: (1) the RAND function was used to generate a random number between 0 and 1, (2) If the generated number was smaller than 0.5, the participant was given the placebo first and (3) If the generated number was greater than or equal to 0.5, the participant was given oxytocin first. Both patients and researchers involved in the experimental process were unaware of the drug assignments. When the participant arrived, she was asked to self-spray the full 1.5 mL dose of oxytocin or the placebo nebulizer into her two nostrils under the supervision of a clinician. Each inhalation of the test

136 solution into a single nostril contained approximately 4 IU of OT. Subsequently, the participants received oxytocin or placebo in ten puffs delivered to alternating nostrils at 30 s intervals. Neither the participant nor the clinician was informed about the drug assignment of the day. The neuropsychological tasks began 45 min after the administration of the drug. The second appointment was scheduled at least four days and at most seven days after the first appointment.

1.4. Modified dot probe task (Shafran et al., 2007) 1.4.1. Stimuli The photographic stimuli used for the task consisted of eating, shape, and weight-related images and had previously been used on people with eating disorders (Shafran et al., 2007). The eating images included 3 types of valences which were developed for use in bulimia nervosa (Shafran et al., 2008). These were rated as ‘‘Positive’’ (low calorie food), ‘‘Neutral’’, or ‘‘Negative’’ (high calorie/palatable food). The shape images included 3 types of valences. ‘‘Negative’’ shape images depicted fat body parts of Western young women that are typically associated with body shape (e.g. thighs, stomach; head excluded) and ‘‘positive’’ shape images depicted slimmer counterparts of such body parts. ‘‘Neutral’’ shape images included pictures of body parts infrequently associated with body shape (e.g., eyes, elbows). The weight images included pictures of weighing scales or Western young women weighing themselves. The target stimuli consisted of 18 eating pictures, 18 shape pictures (six thin, six fat, and six neutral), and 11 weight pictures. Each target stimulus was paired with an animal picture of the same emotional valence. For example, positive images included kittens and puppies, negative images included snakes and insects, and neutral items included birds. 1.4.2. Design and procedure Attentional biases were assessed using a visual probe detection task (Posner et al., 1980) in which pairs of pictures [target (i.e., eating/shape/weight) and non-target (animal)] were presented (Macleod et al., 1986). A fixation cross was displayed in the center of the screen for 750 ms, and then the target and non-target stimuli pair appeared for 1000 ms, following the paradigm of Mansell et al. (Mansell et al., 1999). As the two images disappeared from the screen, one of two probes (‘:’ or ‘

’) appeared where the center of one of the two pictures used to be. The participants were required to press one of two buttons on a keyboard (marked with white stickers) corresponding to the probe (Q for ‘:’ and Z for ‘

’). When a response was made, the probe disappeared and the next trial started. The participants were advised to identify the probe as quickly and accurately as possible. The target stimuli positions and the probe positions were balanced across trials, so each appeared in either location (right or left; probe behind the target or non-target) with equal frequency, following the paradigm of Mogg and Bradley (Mogg and Bradley, 1998). Attention was measured as the time taken to correctly respond to the probe. It was assumed that the subject would more readily respond to the probe if her attention was already allocated to the place where the probe

Y.-R. Kim et al. appears. Responding quicker when the probes appeared behind target stimuli, as compared to when they appeared behind non-target stimuli, was interpreted as vigilance for a threat. A session consisted of eight practice trials, followed by 112 experimental trials with a break in the middle. Test trials were presented in a randomized order and recorded for each participant using E-Prime software (Psychology Software Tools, Sharpsburg, USA).

1.5. Assessment measures 1.5.1. The Eating Disorder Examination Interview (EDE) (Fairburn and Cooper, 1993) The Korean version of the clinician-rated EDE was conducted for the patients with AN to establish the degree of eating psychopathology. The Korean version of the EDE has demonstrated a good inter-rater reliability (Pearson’s correlation coefficients = 0.98—1.0) and internal consistency (Cronbach’s alpha = 0.72—0.89) for variables (Heo et al., 2004).

1.6. Self-report measures 1.6.1. Eating Disorder Examination Questionnaire (EDE-Q) (Fairburn and Beglin, 1994) The EDE-Q assesses the main behavioral features of an eating disorder over the past 28 days based on self-report. The questionnaire consists of 36 items on a 7-point forced choice rating scale and measures weight, shape, and eating related concerns, and dietary restraint. We used the standardized Korean version of the 12th edition of the EDE-Q, which has a high internal consistency and good 2-week reliability. 1.6.2. Autism-Spectrum Quotient (AQ) (Baron-Cohen et al., 2001) The AQ is a 50-item, self-administered questionnaire that assesses the degree to which adults with normal intelligence have autism spectrum traits. The questionnaire measures social skill, attention switching, attention to detail, communication, and imagination as subscales of autism spectrum traits. 1.6.3. Other measurements Depression and anxiety levels were assessed for each subject using the standardized Korean versions of the Beck Depression Inventory (BDI) (Beck et al., 1961) and the Spielberger State and Trait Anxiety Inventory (STAI) (Spielberger et al., 1983), respectively.

1.7. Procedures The participants were tested in a private room at Seoul Paik Hospital. The experiments were carried out at 1430 h, corresponding to 2 h after the participants had eaten lunch. The participants were instructed to eat lunch at 2 h before drug administration and were also refrained from further eating and drinking (other than water). Alcohol and caffeine were abstained on the day of drug administration. Upon arrival, the participants completed baseline measures of physical symptoms including abdominal, neurological, dermatological, and cardiac symptoms.

Oxytocin attenuates attentional bias in AN After the intranasal administration of the oxytocin or the placebo, the participants completed self-report measures to assess their psychological state for 45 min, after which the neuropsychological tasks were performed. At the end of the day, the participants were asked to drink as much as they could of a carton of apple juice (Delmont, 190 mL per carton), and then each participant carefully completed follow-up measures for adverse physical symptoms.

1.8. Statistical analysis We followed the standard analysis procedure used for this task (Bradley et al., 1999). Only the reaction times (RT) for the trials in which the probes were correctly identified were included. The correct response rates were 99.9% for the AN group and 99.7% for the healthy control (HC) group. Mean RTs were calculated for each participant. Outliers were removed by excluding detection latencies that were beyond two standard deviations from their mean (i.e., from each individual’s mean RTs across all stimuli). Unlike Shafran et al. (2007), the eating stimuli were not divided into emotional valences based on the caloric content of the food in our analysis, because patients with AN tend to rate all food stimuli, not just high caloric food, as aversive (Strober, 2004; Steinglass et al., 2011). Therefore, in this analysis, the different valences of eating stimuli were regarded altogether as negative valences and merged as overall eating stimuli to increase the statistical power. Attentional bias scores were calculated for each matched trial type (eating-animal, shape-animal, and weight-animal) as PN—PT, where PN is the mean RT for probes replacing non-target stimuli, and PT is the mean RT for probes replacing target stimuli. Therefore, positive scores suggest increased attention toward the target stimuli, whereas negative scores indicated attention away from the target stimuli. A score of 0 suggest a lack of bias. The baseline attentional bias of the AN and the HC were compared using the t-test. The attentional responses to the shape stimuli were investigated via a series of initial 2 (group: AN and HC)  2 (drug: oxytocin and placebo)  3 (valence: positive, negative, and neutral) repeated-measures analyses of variance (ANOVAs) followed by post hoc analyses. Attentional responses to the eating and weight stimuli were examined via 2 (group: AN and HC)  2 (drug: oxytocin and placebo) repeated-measures ANOVAs. Post hoc analyses were conducted through paired t-test comparisons for each group. Greenhouse—Geisser corrections were applied on the assumption that sphericity was violated. The effect size was calculated and shown if appropriate [Cohen’s d for t-tests; partial eta square (Dh2) for ANOVA]. The estimation of Cohen’s d is as follows: negligible (=0 and <0.15), small (0.15 and <0.40), medium (0.40 and <0.75), large (0.75 and <1.10), very large (1.10 and <1.45), and huge (>1.45) (Cohen, 1992). Pearson’s correlations were used to investigate the relationship between the effect of the oxytocin on attentional bias (drug effect = RTplacebo RToxytocin) and the psychological measures. P values < 0.05 were considered significant, and two-tailed tests were used. Bonferroni corrections were applied if multiple testing was used. Analyses were performed using SPSS 19.0 (SPSS Inc., Chicago, IL, USA).

137

2. Results 2.1. Demographic and group characteristics The demographic and group characteristics of the participants are presented in Table 1. The AN and HC groups were similar in age and intelligence, but significantly different in body mass index and all four subscales of the EDE-Q. The AN group also scored higher in the BDI, the STAI, and the Social Skill and Communication subscales of AQ compared to the HC group. The mean EDE interview scores for the AN were 5.07  6.04 for the Restraint subscale, 3.73  4.41 for the Eating Concern subscale, 4.35  8.33 for the Weight Concern subscale, and 3.85  3.10 for the Shape Concern subscale.

2.2. Attentional bias scores Regarding the overall eating stimuli, in the placebo condition, both the AN and HC participants showed baseline attentional biases toward the stimuli (AN: 19.04  72.26; HC: 7.26  49.30) and the difference between the baseline vigilance of the two groups was insignificant (t(62) = 0.765, p = 0.062, d = 0.192). The two-way group (AN vs. HC)  drug (oxytocin vs. placebo) repeated-measures ANOVA identified a main effect of drug (F(1,62) = 8.09, p = 0.006, Dh2 = 0.115), but no main effect of group (F(1,62) = 0.187, p = 0.667, Dh2 = 0.003) or group  drug interaction effect (F(1,62) = 0.506, p = 0.479, Dh2 = 0.008) on the attentional bias to the eating stimuli. Post hoc tests showed that the oxytocin significantly decreased the attentional bias only in the AN group (t(30) = 2.281, p = 0.030, d = 0.808). The three-way group (AN vs. HC)  drug (oxytocin vs. placebo)  valence (positive; low calorie: neutral: negative; high calorie/binge) found no main effect of valence on the attentional bias to the eating stimuli (F(2,124) = 1.173, p = 0.313, Dh2 = 0.020). The attentional bias scores for the two groups in each condition for all three types of stimuli are shown in Table 2 and Fig. 1. Regarding the overall weight stimuli, in the placebo condition, the AN and HC groups did not differ in their baseline attentional bias (AN: 0.30  92.14; HC: 1.48  96.36; t(62) = 0.075, p = 0.707, d = 0.190). The group (AN vs. HC)  drug (oxytocin vs. placebo) repeated-measures ANOVA identified no main effect of group (F(1,62) = 0.074, p = 0.787, Dh2 = 0.001) or of drug (F(1,62) = 0.062, p = 0.805, Dh2 = 0.001) and no group  drug interaction effect (F(1,62) = 0.148, p = 0.702, Dh2 = 0.002) on the attentional bias to the weight stimuli (Table 2) (Fig. 1). Regarding the valenced shape stimuli, in the placebo condition, no differences between the attentional biases of the AN group and those of the HC group were observed for the positive (thin) shape stimuli (t(62) = 0.192, p = 0.849, d = 0.048) or for the neutral shape stimuli (t(62) = 0.764, p = 0.448, d = 0.195). There was, however, a small vigilance bias to the negative (fat) shape in the AN group compared to the HC group in the placebo condition, which was not significant due to the large variance (t(62) = 1.507, p = 0.137, d = 0.380). The three-way group (AN vs. HC)  drug (oxytocin vs. placebo)  valence (negative, neutral, positive) repeated-measures ANOVA identified no main effect of drug (F(1,62) = 1.372, p = 0.246, Dh2 = 0.024) or of valence (F(2,124) = 1.590, p = 0.209, Dh2 = 0.027) and no group  drug  valence

138 Table 1

Y.-R. Kim et al. Demographic and clinical characteristics of the patients with anorexia nervosa and healthy control subjects.

Characteristics

AN (n = 31)

Controls (n = 33)

Age Intelligence BMI EDE-Q Restraint Eating Concern Weight Concern Shape Concern Global AQ Social Skill Attention Switching Attention to Detail Communication Imagination Total BDI STAI State Trait

23.10 (9.35) 109.27 (14.01) 15.15 (2.51)

22.18 (2.14) 114.62 (9.05) 20.91 (2.22)

t

df

p

0.536 1.788 9.819 ***

32.87 48.51 62

.596 .080 <.001

2.50 2.10 2.63 3.01 2.56

(1.89) (1.80) (1.61) (1.56) (1.56)

0.79 0.55 1.46 2.14 1.23

(0.86) (0.76) (1.13) (1.31) (0.89)

4.558 *** 4.388 *** 3.330 ** 2.435 * 4.120 ***

39.31 38.14 51.06 62 44.91

<.001 <.001 .002 .018 <.001

3.94 4.52 5.03 2.73 2.29 18.60 24.53

(1.93) (1.93) (2.23) (1.82) (1.88) (5.33) (14.01)

2.67 4.21 3.97 1.58 2.12 14.55 7.38

(1.81) (1.36) (2.08) (1.35) (1.50) (4.28) (6.89)

2.710 ** 0.724 1.971 2.848 ** 0.399 3.344 ** 6.057 ***

62 53.67 62 53.17 62 62 41.64

.009 .472 .053 .006 .691 .001 <.001

4.899 *** 4.371 ***

62 62

<.001 <.001

58.47 (13.13) 57.53 (13.46)

43.28 (11.25) 43.75 (11.33)

AN, anorexia nervosa; BMI, body mass index; EDE-Q, Eating Disorder Examination Questionnaire; AQ, autism-spectrum quotient; STAI-State, Spielberger State and the Trait Anxiety Inventory State score; STAI-Trait, Spielberger State and Trait Anxiety Inventory Trait score. * p < 0.05. ** p < 0.01. *** p < 0.001.

interaction effect (F(2,124) = 1.194, p = 0.307, Dh2 = 0.021) on the attentional bias to the shape stimuli (Table 2). In post hoc tests, the decrement in the attentional bias under the influence of oxytocin in the AN group was significant only for the negative shape stimuli (t(30) = 2.592, p = 0.015, d = 0.498).

2.3. Correlations between attentional biases and psychopathological measures in AN The drug effect was defined as the difference in the reaction time in the placebo condition and in the oxytocin

condition for a certain type of stimuli (drug effect = RTplaRToxytocin). A correlation analysis was performed in cebo the AN group between the drug effect on the various types of stimuli and the psychopathological measures (AQ subscales, EDE-Q subscales, BDI, STAI). After a Bonferroni correction for multiple comparisons of the AQ subscales (alpha = 0.05/6), the only significant association observed was between the effect of oxytocin on fat shape stimuli and the Communication subscales of AQ (r = 0.49, p = 0.009) in the AN group. The correlation between the total AQ scale and the oxytocin effect was not statistically significant

Table 2 Attentional bias to eating, weight, and shape visual image stimuli after intranasal administration of oxytocin or placebo to patients with anorexia nervosa and healthy controls. AN (n = 31) Placebo Eating stimuli Overall 19.04 Weight stimuli Overall 0.30 Shape stimuli 1.57 Positive Neutral 21.51 Negative 42.03

HC (n = 33) Oxytocin

t(df = 30) p

Cohen’s d Placebo

Oxytocin

t(df = 32) p

Cohen’s d

(72.26)

14.63 (57.52)

2.281 *

.030

0.516

7.26 (49.40)

12.93 (59.36)

1.680

.103

0.292

(92.14)

1.74 (75.69)

0.101

.920

0.024

1.48 (96.36)

7.95 (65.53)

0.430

.670

0.075

(126.33) (105.77) (104.69)

18.78 (104.90) 10.78 (106.52) 7.49 (93.84)

0.612 0.478 2.592 *

.545 .636 .015

0.175 0.101 0.498

6.94 (94.94) 43.87 (122.43) 1.19 (81.13)

19.68 (92.62) 14.77 (87.83) 12.21 (96.11)

0.576 1.094 0.666

.569 .282 .510

0.100 0.190 0.116

Data are shown as mean (standard deviation). AN, anorexia nervosa; HC, healthy control. Statistical values represented by paired t-test between the oxytocin and placebo conditions in each group. * p < 0.05.

Oxytocin attenuates attentional bias in AN

139

Figure 1 Attentional bias to food, shape [classified as positive (thin), neutral (normal), and negative (fat)] and weight stimuli in the AN and HC groups.a A main effect of oxytocin on attentional bias was observed in response to eating stimuli ( p = 0.006). Post hoc tests showed that the attentional bias to food was attenuated under the oxytocin condition in the AN group ( p = 0.030). Attentional bias toward negative shape images was attenuated in the AN group in response to body shape stimuli after oxytocin administration ( p = 0.015). aError bars indicate standard error of the mean. AN, anorexia nervosa; HC, healthy control.

(r = 0.032, p = 0.101). There also was a tendency of an association between the effect of oxytocin on the fat shape stimuli and state anxiety (r = 0.404, p = 0.033). No association was observed between the drug effect on attention and the aspects of psychopathology typically associated with eating disorders (EDE-Q), or depression in the AN group.

2.4. Changes in juice consumption under the influence of oxytocin In the AN group, the mean juice consumption was 89.16  84.77 mL in the placebo condition and 97.84  84.57 mL in the oxytocin condition, which was an increase of 8.9%. In the HC group, the mean juice consumption was 162.91  50.20 mL in the placebo condition and 173.42  43.41 mL in the oxytocin condition, which was an increase of 6.8%. The two-way group (AN vs. HC)  drug (oxytocin vs. placebo) repeated-measures ANOVA identified only a main effect of group (F(1,63) = 23.224, p < 0.01, Dh2 = 0.273), but no effect of the drug (F(1,63) = 1.830, p = 0.181) or drug  group interaction effect (F(1,63) = 0.001, p = 0.976) on the amount of juice consumed. The increments in the juice consumption induced by oxytocin were not statistically significant, as confirmed by the Wilcoxon signed-rank test, in both the AN (z = 0.213, p = 0.831, ES = 0.038) and the HC (z = 1.244, p = 0.214, ES = 0.223) groups. There was no correlation between the amount of juice drunk and the AQ scales.

3. Discussion The aim of this study was to examine whether oxytocin has an impact on attentional processes to food, fatness and eating behavior in patients with AN. Our first hypothesis of increased attentional bias to food in patients with AN was not proved, because although the mean attentional bias was increased, the large variance meant that this did not reach the level of significance (ES = 0.192). In our second hypothesis, there was a tendency for the AN group to be more vigilant to the negative (fat) shape stimuli compared to the HC group with small effect (ES = 0.380), though the evidence in our study was not as strong as that in the case of bulimia nervosa previously reported. We failed to confirm the third hypothesis that patients with AN would have an increased attentional bias to weight stimuli. Our fourth hypothesis was confirmed in that oxytocin decreased selective attention toward images of food and fat body parts in patients with AN. We also found a decrease in attentional bias to food in HC, although this was not significant due to the variance. We found no evidence in either group to support our fifth hypothesis that oxytocin would reduce the amount of drink consumed. Our sixth hypothesis was proved in that the impact of oxytocin was more pronounced in patients with AN with high levels of autistic spectrum trait of communication problem. Although attentional bias to food stimuli using the dot probe paradigm has been widely examined in the healthy, obese and subclinical populations, there has been much less work undertaken in people with eating disorders. Food

140 naturally captures the attention, but it is more pronounced in individuals who describe craving, eating in response to food cues or emotion and overeating or obesity (Brignell et al., 2009; Smeets et al., 2009; Hepworth et al., 2010; Nijs et al., 2010; Werthmann et al., 2011). People with AN who live free from supervision over their eating exhibited greater vigilance toward food cues than healthy women at baseline (Cardi et al., 2012), whereas those undergoing nutritional rehabilitation show less vigilance to food than healthy women (Cardi et al., 2013). We had a mixed population of inpatients and outpatients which may account for the failure to find a significant difference in patients with AN compared to control women at baseline. One explanation for the reduction in attentional bias to food in the oxtytocin condition is that the oxytocin modulated the threat response to food. However, it is interesting that oxytocin showed a tendency to reduce the attentional bias toward food in the control group and the mechanism for this effect, if it were to be shown to be robust, would not be expected to be via threat reduction. The mechanism for this effect might be by oxytocin reducing the reward from food. This is in line with the animal research summarized in a recent review that suggests that oxytocin selectively suppresses the reward driven aspect of eating (Sabatier et al., 2013) and the recent human research in which oxytocin has been found to reduce the intake of high sugar snacks in man (Ott et al., 2013). It is possible that the mechanism accounting for this is through inhibition of the mesolimbic dopamine pathway (Succu et al., 2008). Shafran et al. (2008) observed a vigilance toward both weight and fat shape stimuli in patients with bulimia nervosa. Also, an attentional bias to fatness related stimuli was found in bulimia nervosa in a study using the dot probe paradigm in which words represented negative body shape cues (e.g. ‘‘fat’’) (Rieger et al., 1998). The failure of cues related to weight to elicit an attentional bias was unexpected, as these produced a bias in bulimia nervosa (Shafran et al., 2008). This suggests that weight fear is not a transdiagnositic phenomenon and is less salient to people with AN. This is perhaps not surprising as they are able to reduce their weight effectively. The reduced attention toward fat shape stimuli under the oxytocin condition in patients with AN suggests that oxytocin may have modulated the salience (possibly due to threat) of these stimuli. Oxytocin reduces attentional bias to threat cues in monkeys (Parr et al., 2013). We also found that the effect of oxytocin on fat shape stimuli was more prominent in cases with communication problems as assessed by the AQ and increased state anxiety. Further studies will be needed to examine whether a higher level of autistic traits and vulnerability might be a marker of responsivity to oxytocin. State anxiety also had a tendency to be correlated with the effect of oxytocin on fat shape stimuli. Given that some animal and human studies suggest that oxytocin can reduce the reward aspect of appetite (Sabatier et al., 2013; Ott et al., 2013), in the patient group, it was important to ensure that the impact of oxytocin would not be harmful by reducing consummatory behavior. In fact, there was a trend for both groups to drink more in the oxytocin condition in our study. This unexpected finding will need to be replicated in further studies measuring eating behavior following intranasal oxytocin. In previous studies for patients with AN, the attentional bias to food increased rather than decreased as it does in healthy individuals after food intake,

Y.-R. Kim et al. particularly in patients undergoing nutritional rehabilitation (Cardi et al., 2013). Therefore, it would have been of interest to measure the possible benefit of oxytocin on the attentional bias to food after the juice consumption. Some limitations of this study should be considered. These may account for some of the individual variation and hence reduce the power of this study to show differences. First, two thirds of the patients with AN were in the inpatient ward and would have a positive food balance, as compared to the others treated in the outpatient clinic, who probably have a negative food balance. In healthy controls, the drive state such as the level of hunger can increase attentional processes to food (Mogg and Bradley, 1998; Placanica et al., 2002). Fasting healthy subjects had higher levels of gaze toward food pictures when eye-tracking was measured (Giel et al., 2011). Though we asked the participants to have a meal 2 h before the study, apart from the inpatient group this was not tightly controlled and may have increased the variance. It will be of interest in the future to study cases who are at different stages of the illness and within different phase of treatment, which may reduce the heterogeneity in the response and increase the power to see differences. Second, the comparison group (student volunteers) had higher levels of shape concern than has been reported in Western samples. So it may be possible that students with subclinical symptoms of eating disorders were more likely to volunteer to participate in the study. This would have decreased the size of any effect in the current study. Third, we used the eating, weight, and shape images developed by Shafran et al. (2007) who conducted a study on bulimia nervosa. The food images were complex, as eating behavior (by Western people) as well as food composition, was depicted. Other groups have used general food stimuli or high and low calorie foods (Veenstra and de Jong, 2012). Furthermore, the stimuli used were pictures of Western foods, meal crockery and people which may have impacted on their salience. For example, some items in the images are unfamiliar to Koreans (e.g. a cake stand). The unfamiliarity of some of the stimuli and lack of salience to a Korean audience might be expected to decrease the effect size. In addition, we used vertical and horizontal dots as the target stimuli in the dot probe paradigm, and there is some evidence that less variation in this paradigm is found if letters are used as targets and if the task only involves responding to the position of the dot (Bradley et al., 1999). Fourth, we used a simple test of juice consumption as an endpoint for consummatory behavior. This test may not be ideal and might be expected to have a ceiling effect, because it was delivered in a standard portion size. We chose this as a relatively ‘‘non threatening’’ food item, as we did not want to have a large ‘‘floor’’ effect due to the people with AN not drinking anything. Therefore, whether the impact of oxytocin would reduce the food intake in the patients group still remains unproved. Fifth, three patients were taking fluoxetine. It is possible that this might affect the oxytocin pathway. However, in an additional analysis with the three patients separately, there was no difference in their response to attentional bias compared with the rest of the patients. Finally, we did not measure other individual factors that are known to impact attention bias. For example, individuals with a short 5-hydroxytryptamine allele have been shown to be more vigilant toward threats (PergaminHight et al., 2012). Additionally the experience of childhood

Oxytocin attenuates attentional bias in AN care, personality, and interpersonal context, which impact the effect of oxytocin, were not considered (BakermansKranenburg and van Ijzendoorn, 2013). These limitations would have increased the variance and reduced the effect and so the preliminary findings suggest that more work should be done. The vigilance toward food and fat shapes in patients with AN is probably due to the activation of fear circuits. The findings from our study suggest that oxytocin moderates the activation of these fear circuits. Oxytocin has an effect on fear learning in humans. Extinction of the fear response improves when intranasal oxytocin is given at the same time as extinction learning, and fear conditioning to faces is nullified if oxytocin is given after conditioning (Petrovic et al., 2008; Acheson et al., 2013). A reduction in the fear circuitry may have underpinned the reduction in weight and shape concerns found in a proof-of-principle study in which patients with AN were given intranasal oxytocin over 6 weeks as a supplement to nutritional rehabilitation (Russell et al., 2012). We found no evidence of potential harm from this form of treatment in terms of any change in eating behavior. Therefore, further studies using oxytocin as a form of intervention for patients with AN are warranted. In conclusion, oxytocin attenuated vigilance to eating and fat shape stimuli in patients with AN. The effect was stronger in those with communication traits associated with autistic spectrum conditions. No effect on the consumption of juice was observed. Our findings suggest the potential for oxytocin nasal spray to reduce AN symptoms, and a future evaluation of oxytocin as a treatment for AN is merited.

Role of funding source The funders had no role on the study design, the collection, analysis, and interpretation of the data, nor in writing the manuscript or the decision on where to submit the manuscript for publication.

Conflict of interest All authors report no biomedical financial interests or potential conflicts of interest.

Acknowledgements We thank Professor Roz Shafran for offering the photos. This study was supported under the framework of the international cooperation program managed by the National Research Foundation of Korea (2011-0030914) and the Basic Science Research Program through the National Research Foundation of Korea, which is funded by the Ministry of Education (MOE) (NRF-2011-0024415) to Youl-Ri Kim. Janet Treasure is part funded by the National Institute for Health Research (NIHR) Mental Health Biomedical Research Centre at South London and Maudsley NHS Foundation Trust and King’s College London. The views expressed are those of the author(s) and not necessarily those of the NHS, the NIHR or the Department of Health and in addition the Swiss Anorexia Foundation contributed to this work in part the work of Valentina Cardi and Janet Treasure.

141

References Acheson, D., Feifel, D., de Wilde, S., McKinney, R., Lohr, J., Risbrough, V., 2013. The effect of intranasal oxytocin treatment on conditioned fear extinction and recall in a healthy human sample. Psychopharmacology (Berl) 229, 199—208. Aspen, V., Darcy, A.M., Lock, J., 2013. A review of attention biases in women with eating disorders. Cogn. Emot. 27, 820—838. Bakermans-Kranenburg, M.J., van Ijzendoorn, M.H., 2013. Sniffing around oxytocin: review and meta-analyses of trials in healthy and clinical groups with implications for pharmacotherapy. Transl. Psychiatry 3, e258. Bar-Haim, Y., Lamy, D., Pergamin, L., Bakermans-Kranenburg, M.J., van Ijzendoorn, M.H., 2007. Threat-related attentional bias in anxious and nonanxious individuals: a meta-analytic study. Psychol. Bull. 133, 1—24. Baron-Cohen, S., Jaffa, T., Davies, S., Auyeung, B., Allison, C., Wheelwright, S., 2013. Do girls with anorexia nervosa have elevated autistic traits? Mol. Autism 4, 24. Baron-Cohen, S., Wheelwright, S., Skinner, R., Martin, J., Clubley, E., 2001. The Autism-Spectrum Quotient (AQ): evidence from Asperger syndrome/high-functioning autism, males and females, scientists and mathematicians. J. Autism Dev. Disord. 31, 5—17. Beck, A.T., Ward, C.H., Mendelson, M., Mock, J.E., Erbaugh, J., 1961. An inventory for measuring depression. Arch. Gen. Psychiatry 4, 561—571. Bradley, B.P., Hogg, K., White, J., Groom, C., de Bono, J., 1999. Attentional bias for emotional faces in generalized anxiety disorder. Br. J. Clin. Psychol. 38, 267—278. Brignell, C., Griffiths, T., Bradley, B.P., Mogg, K., 2009. Attentional and approach biases for pictorial food cues. Influence of external eating. Appetite 52, 299—306. Brooks, S., Prince, A., Stahl, D., Campbell, I.C., Treasure, J., 2011. A systematic review and meta-analysis of cognitive bias to food stimuli in people with disordered eating behaviour. Clin. Psychol. Rev. 31, 37—51. Cardi, V., Kan, C., Roncero, M., Harrison, A., Lounes, N., Tchanturia, K., Meyer, C., Treasure, J., 2012. Mealtime support in anorexia nervosa: a within-subject comparison study of a novel vodcast intervention. Psychother. Psychosom. 81, 54—55. Cardi, V., Lounes, N., Kan, C., Treasure, J., 2013. Meal support using mobile technology in anorexia nervosa. Contextual differences between inpatient and outpatient settings. Appetite 60, 33—39. Chiodera, P., Volpi, R., Capretti, L., Marchesi, C., d’Amato, L., De Ferri, A., Bianconi, L., Coiro, V., 1991. Effect of estrogen or insulin-induced hypoglycemia on plasma oxytocin levels in bulimia and anorexia nervosa. Metabolism 40, 1226—1230. Cohen, J., 1992. A power primer. Psychol. Bull. 112, 155—159. Cuthbert, B.N., Insel, T.R., 2013. Toward the future of psychiatric diagnosis: the seven pillars of RDoC. BMC Med. 11, 126. Demitrack, M.A., Lesem, M.D., Listwak, S.J., Brandt, H.A., Jimerson, D.C., Gold, P.W., 1990. CSF oxytocin in anorexia nervosa and bulimia nervosa: clinical and pathophysiologic considerations. Am. J. Psychiatry 147, 882—886. Dobson, K.S., Dozois, D.J.A., 2004. Attentional biases in eating disorders: a meta-analytic review of stroop performance. Clin. Psychol. Rev. 23, 1001—1022. Fairburn, C., Beglin, S., 1994. Assessment of eating disorders–—Interview or self-report questionnaire. Int. J. Eat. Disord. 16, 363—370. Fairburn, C.G., Cooper, Z., 1993. The Eating Disorder Examination, 12th ed. Guilford, New York. Faunce, G.J., 2002. Eating disorders and attentional bias: a review. Eat. Disord. 10, 125—139. First, M.B., Spitzer, R.L., Gibbon, M., Williams, J.B.W., 2002. Structured Clinical Interview for DSM-IV-TR Axis I Disorders-Patient Edition (SCID-I/P). Biometrics Research Department, New York State Psychiatric Institute, New York.

142 Friederich, H.C., Kumari, V., Uher, R., Riga, M., Schmidt, U., Campbell, I.C., Herzog, W., Treasure, J., 2006. Differential motivational responses to food and pleasurable cues in anorexia and bulimia nervosa: a startle reflex paradigm. Psychol. Med. 36, 1327—1335. Giel, K.E., Friederich, H.C., Teufel, M., Hautzinger, M., Enck, P., Zipfel, S., 2011. Attentional processing of food pictures in individuals with anorexia nervosa-an eye-tracking study. Biol. Psychiatry 69, 661—667. Gillberg, C.L., 1992. Autism and autistic-like conditions-subclasses among disorders of empathy. J. Child Psychol. Psychiatry 33, 813—842. Hambrook, D., Tchanturia, K., Schmidt, U., Russell, T., Treasure, J., 2008. Empathy, systemizing, and autistic traits in anorexia nervosa: a pilot study. Br. J. Clin. Psychol. 47, 335—339. Heo, S., Lee, M., Choi, Y., Sohn, C., Lee, H., Lee, Y., 2004. Reliability and factor analysis of the Korean version of eating disorder examination. J. Korean Soc. Study Obes. 13, 42—52. Hepworth, R., Mogg, K., Brignell, C., Bradley, B.P., 2010. Negative mood increases selective attention to food cues and subjective appetite. Appetite 54, 134—142. Johansson, L., Lundh, L.G., Andersson, G., 2005. Attentional bias for negative self-words in young women — The role of thin ideal priming and body shape dissatisfaction. Pers. Ind. Dif. 38, 723— 733. Kirsch, P., Esslinger, C., Chen, Q., Mier, D., Lis, S., Siddhanti, S., Gruppe, H., Mattay, V.S., Gallhofer, B., Meyer-Lindenberg, A., 2005. Oxytocin modulates neural circuitry for social cognition and fear in humans. J. Neurosci. 25, 11489—11493. Lawson, E.A., Donoho, D.A., Blum, J.I., Meenaghan, E.M., Misra, M., Herzog, D.B., Sluss, P.M., Miller, K.K., Klibanski, A., 2011. Decreased nocturnal oxytocin levels in anorexia nervosa are associated with low bone mineral density and fat mass. J. Clin. Psychiatry 72, 1546—1551. Lawson, E.A., Holsen, L.M., Santin, M., Meenaghan, E., Eddy, K.T., Becker, A.E., Herzog, D.B., Goldstein, J.M., Klibanski, A., 2012. Oxytocin secretion is associated with severity of disordered eating psychopathology and insular cortex hypoactivation in anorexia nervosa. J. Clin. Endocrinol. Metab. 97, E1898—E1908. MacDonald, E., Dadds, M., Brennan, J., Williams, K., Levy, F., Cauchi, A., 2011. A review of safety, side-effects and subjective reactions to intranasal oxytocin in human research. Psychoneuroendocrinology 36, 1114—1126. Macleod, C., Mathews, A., Tata, P., 1986. Attentional bias in emotional disorders. J. Abnorm. Psychol. 95, 15—20. Maguire, S., O’Dell, A., Touyz, L., Russell, J., 2013. Oxytocin and anorexia nervosa: a review of the emerging literature. Eur. Eat. Disord. Rev. 21, 475—478. Mansell, W., Clark, D.M., Ehlers, A., Chen, Y.P., 1999. Social anxiety and attention away from emotional faces. Cogn. Emot. 13, 673— 690. Mogg, K., Bradley, B.P., 1998. A cognitive-motivational analysis of anxiety. Behav. Res. Ther. 36, 809—848. Mogg, K., Bradley, B.P., Hallowell, N., 1994. Attentional bias to threat-roles of trait anxiety, stressful events, and awareness. Q. J. Exp. Psychol. A 47, 841—864. Nijs, I.M.T., Muris, P., Euser, A.S., Franken, I.H.A., 2010. Differences in attention to food and food intake between overweight/obese and normal-weight females under conditions of hunger and satiety. Appetite 54, 243—254. Oldershaw, A., Hambrook, D., Stahl, D., Tchanturia, K., Treasure, J., Schmidt, U., 2011. The socio-emotional processing stream in anorexia nervosa. Neurosci. Biobehav. Rev. 35, 970—988. Ott, V., Finlayson, G., Lehnert, H., Heitmann, B., Heinrichs, M., Born, J., Hallschmid, M., 2013. Oxytocin reduces reward-driven food intake in humans. Diabetes 62, 3418—3425.

Y.-R. Kim et al. Parr, L., Modi, M., Siebert, E., Young, L., 2013. Intranasal oxytocin selectively attenuates rhesus monkeys’ attention to negative facial expressions. Psychoneuroendocrinology 38, 1748—1756. Pergamin-Hight, L., Bakermans-Kranenburg, M.J., van Ijzendoorn, M.H., Bar-Haim, Y., 2012. Variations in the promoter region of the serotonin transporter gene and biased attention for emotional information: a meta-analysis. Biol. Psychiatry 71, 373—379. Petrovic, P., Kalisch, R., Singer, T., Dolan, R.J., 2008. Oxytocin attenuates affective evaluations of conditioned faces and amygdala activity. J. Neurosci. 28, 6607—6615. Placanica, J.L., Faunce, G.J., Job, R.F.S., 2002. The effect of fasting on attentional biases for food and body shape/weight words in high and low eating disorder inventory scorers. Int. J. Eat. Disord. 32, 79—90. Posner, M.I., Snyder, C.R.R., Davidson, B.J., 1980. Attention and the detection of signals. J. Exp. Psychol. Gen. 109, 160—174. Rieger, E., Schotte, D.E., Touyz, S.W., Beumont, P.J.V., Griffiths, R., Russell, J., 1998. Attentional biases in eating disorders: a visual probe detection procedure. Int. J. Eat. Disord. 23, 199—205. Russell, J., Guastella, A., Maguire, S., O’Dell, A., Adam, B., 2012. Intranasal Oxytocin Reduces Cognitive Rigidity in Anorexia Nervosa? Eating Disorders Research Society, Porto, 55 pp. Sabatier, N., Leng, G., Menzies, J., 2013. Oxytocin, feeding, and satiety. Front. Endocrinol. 4, 35. Shafran, R., Lee, M., Cooper, Z., Palmer, R.L., Fairburn, C.G., 2007. Attentional bias in eating disorders. Int. J. Eat. Disord. 40, 369— 380. Shafran, R., Lee, M., Cooper, Z., Palmer, R.L., Fairburn, C.G., 2008. Effect of psychological treatment on attentional bias in eating disorders. Int. J. Eat. Disord. 41, 348—354. Smeets, E., Roefs, A., Jansen, A., 2009. Experimentally induced chocolate craving leads to an attentional bias in increased distraction but not in speeded detection. Appetite 53, 370—375. Spielberger, C., Gorsuch, R., Lushene, R., Vagg, P., Facobs, G., 1983. Manual for the State-Trait Anxiety Inventory, STAI (Form Y). Consulting Psychologists Press, Palo Alto. Steinglass, J.E., Sysko, R., Glasofer, D., Albano, A.M., Simpson, H.B., Walsh, B.T., 2011. Rationale for the application of exposure and response prevention to the treatment of anorexia nervosa. Int. J. Eat. Disord. 44, 134—141. Strober, M., 2004. Pathologic fear conditioning and anorexia nervosa: on the search for novel paradigms. Int. J. Eat. Disord. 35, 504— 508. Succu, S., Sanna, F., Cocco, C., Melis, T., Boi, A., Ferri, G.L., Argiolas, A., Melis, M.R., 2008. Oxytocin induces penile erection when injected into the ventral tegmental area of male rats: role of nitric oxide and cyclic GMP. Eur. J. Neurosci. 28, 813—821. Sunday, S.R., Halmi, K.A., Einhorn, A., 1995. The Yale-Brown-Cornell eating disorder scale: a new scale to assess eating disorder symptomatology. Int. J. Eat. Disord. 18, 237—245. Treasure, J., Corfield, F., Cardi, V., 2012. A three-phase model of the social emotional functioning in eating disorders. Eur. Eat. Disord. Rev. 20, 431—438. Veenstra, E.M., de Jong, P.J., 2012. Attentional bias in restrictive eating disorders. Stronger attentional avoidance of high-fat food compared to healthy controls? Appetite 58, 133—140. Vitousek, K.B., Hollon, S.D., 1990. The investigation of schematic content and processing in eating disorders. Cog. Ther. Res. 14, 191—214. Werthmann, J., Roefs, A., Nederkoorn, C., Mogg, K., Bradley, B.P., Jansen, A., 2011. Can(not) take my eyes off it: attention bias for food in overweight participants. Health Psychol. 30, 561—569. Zucker, N.L., Losh, M., Bulik, C.M., Labar, K.S., Piven, J., Pelphrey, K.A., 2007. Anorexia nervosa and autism spectrum disorders: guided investigation of social cognitive endophenotypes. Psychol. Bull. 133, 976—1006.