Using Continuous Glucose Monitoring to Passively Classify Naturalistic Binge Eating and Vomiting Among Adults With Binge-Spectrum Eating Disorders: A Preliminary Investigation.

OBJECTIVE: Binge eating and self-induced vomiting are common, transdiagnostic eating disorder (ED) symptoms. Efforts to understand these behaviors in research and clinical settings have historically relied on self-report measures, which may be biased and have limited ecological validity. It may be possible to passively detect binge eating and vomiting using data collected by continuous glucose monitors (CGMs; minimally invasive sensors that measure blood glucose levels), as these behaviors yield characteristic glucose responses. METHOD: This study developed machine learning classification algorithms to classify binge eating and vomiting among 22 adults with binge-spectrum EDs using CGM data. Participants wore Dexcom G6 CGMs and reported eating episodes and disordered eating symptoms using ecological momentary assessment for 2 weeks. Group-level random forest models were generated to distinguish binge eating from typical eating episodes and to classify instances of vomiting. RESULTS: The binge eating model had accuracy of 0.88 (95% CI: 0.83, 0.92), sensitivity of 0.56, and specificity of 0.90. The vomiting model demonstrated accuracy of 0.79 (95% CI: 0.62, 0.91), sensitivity of 0.88, and specificity of 0.71. DISCUSSION: Results suggest that CGM may be a promising avenue for passively classifying binge eating and vomiting, with implications for innovative research and clinical applications.

Early change in gastric-specific anxiety sensitivity as a predictor of eating disorder treatment outcome.

Eating disorders (EDs) are often accompanied by gastrointestinal (GI) distress. Anxiety sensitivity is the tendency to interpret sensations of anxiety as threatening or dangerous, and includes both broad physical symptoms (e.g., elevated heartrate) and GI-specific symptoms. Physical and GI-specific anxiety sensitivity may be important risk and maintaining factors in EDs. This study tested the hypothesis that greater reductions in both types of anxiety sensitivity during the first month of treatment would predict lower ED symptoms and trait anxiety at discharge and 6-month follow-up. Patients (n = 424) in ED treatment reported physical and GI-specific anxiety sensitivity, ED symptoms, and trait anxiety at treatment admission, 1-month into treatment, discharge, and 6-month follow-up. Analyses were conducted with hierarchical linear regression with imputation, controlling for relevant covariates. Results indicated that early reduction in GI-specific but not general physical anxiety sensitivity predicted both lower ED symptoms and lower trait anxiety at discharge and 6-month follow-up. These findings demonstrate the importance of GI-specific anxiety sensitivity as a potential maintaining factor in EDs. Developing and refining treatments to target GI-specific anxiety sensitivity may have promise in improving the treatment not only of EDs, but also of commonly co-morbid anxiety disorders.

Early reduction in anxiety sensitivity predicts greater reduction in disordered eating and trait anxiety during treatment for bulimia nervosa.

OBJECTIVE: Anxiety sensitivity (AS), the trait-like fear of symptoms of anxiety, has been associated with eating disorder (ED) pathology broadly, bulimia nervosa (BN) symptoms specifically, and the anxiety disorders that are commonly comorbid with BN. AS, especially for physical symptoms specifically, maybe a risk and maintenance factor for BN and comorbid anxiety. METHOD: Adult participants with BN (n = 44) in a clinical trial comparing CBT to mindfulness and acceptance-based treatment (MABT) reported ED symptoms, trait anxiety, and AS through treatment and follow-up. We predicted that greater early reduction (i.e., within the first month of treatment) in AS would predict greater reductions in ED symptoms and trait anxiety at post-treatment and follow-up. RESULTS: Early reductions in AS for physical concerns predicted lower ED symptoms and trait anxiety at post-treatment but not follow-up. Exploratory analyses indicated that treatment groups did not differ in either early or total change in AS, controlling for baseline AS. DISCUSSION: Early reductions in AS may be an important treatment target for BN, and may additionally support reductions in anxiety. Future research should identify which components of CBT and MABT best target AS, to deliver these components early in treatment, when they can have maximum effect. PUBLIC SIGNIFICANCE: Anxiety sensitivity, the fear of symptoms of anxiety, is associated with eating disorders (ED). In this study, participants in treatment for bulimia nervosa reported ED symptoms, trait anxiety, and anxiety sensitivity through treatment and follow-up. Greater early reductions in anxiety sensitivity predicted lower ED symptoms and trait anxiety at post-treatment. Future research should identify which elements of treatment best target anxiety sensitivity, to deliver them early in treatment.

Elevated interoceptive deficits in individuals with eating disorders and self-injurious thoughts and behaviors: A replication and extension.

INTRODUCTION: People with eating disorders (EDs) frequently report self-injurious thoughts and behaviors (SITBs). We aimed to replicate and extend findings demonstrating greater interoceptive deficits according to SITB severity. We predicted that interoceptive deficits would be greater among people with versus without lifetime SITBs and among those with multiple suicide attempts compared with single attempts or nonsuicidal self-injury. We also predicted that interoception would be increasingly impaired with greater current suicidality severity. METHOD: Adults (N = 118) seeking ED treatment reported current and lifetime SITBs upon admission to treatment. We used planned contrasts to compare interoceptive deficits (self-reported using the Eating Disorders Inventory) by SITB and current suicidality groups (assessed by self-report and clinical interview). RESULTS: Interoceptive deficits were greater among people with any SITB history compared to those without. People with multiple suicide attempts did not differ in interoceptive deficits from those with single attempts or only nonsuicidal self-injury. Interoceptive deficits were elevated among those with any current suicidality compared to those with no suicidal symptoms; interoceptive deficits did not differ according to severity of suicidality. CONCLUSION: Among those with EDs, interoceptive deficits are elevated among those with concurrent SITBs. Interventions aimed at improving interoception may reduce the risk for SITBs.

Effects of borderline personality disorder symptoms on dialectical behavior therapy outcomes for eating disorders.

Existing literature on the effects of borderline personality disorder (BPD) and eating disorder (ED) comorbidity in terms of clinical presentation and treatment outcome has been limited and inconclusive. The present study examined whether clients with EDs and varying levels of BPD symptoms presented with more severe ED symptoms at admission, and whether they responded to dialectical behavior therapy (DBT)-based treatment. Participants (N = 176) were adults in a DBT-based partial hospitalization program for EDs at an academic medical center. Participants completed self-report measures at admission, 1-month postadmission, discharge, and 6-month follow-up. Results suggested that patients with elevated BPD symptoms at admission had greater ED symptoms during treatment, evidenced by small to moderate effect sizes. However, patients with high BPD symptoms demonstrated steeper declines in binge eating, fasting, and parasuicidal behavior early during treatment compared to patients with low BPD symptoms. Individuals with high BPD symptoms at admission (i.e., probable BPD diagnosis) were as likely to meet remission criteria and relapse as individuals with low BPD symptoms, though this null finding may be influenced by small cell sizes. Our findings also suggest that DBT skills use does not predict changes in symptoms. In sum, our findings suggest that while clients with higher BPD symptoms may improve during DBT-based partial hospitalization, their ED symptoms may remain more severe. Future studies are needed to determine whether adjunctive treatments improve outcomes for clients with EDs and comorbid BPD symptoms in DBT programs and whether skills use quality is a better predictor of ED symptom changes. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Using clinical cutoff scores on the eating disorder examination-questionnaire to evaluate eating disorder symptoms during and after naturalistic intensive treatment.

Clinical cutoff scores for self-report measures provide a means of evaluating clinically significant pathology during and after treatment. A cutoff of 2.8 on the Eating Disorder Examination-Questionnaire (EDE-Q) has been recommended to screen for eating disorders (ED). We used this cutoff to assess ED symptoms in adolescents (n = 444) and adults (n = 592) through ED treatment and follow-up. Most patients scored above 2.8 at intake (adolescents 67%, M = 3.21; adults 78%, M = 4.20) and below 2.8 at discharge (adolescents 65%, M = 1.87; adults 66%, M = 2.67), with gains often maintained through follow-up (40% of adolescents and 35% of adults at 12-month follow-up). EDE-Q scores were higher in adults than adolescents and in patients with binge/purge disorders. Results suggest a cutoff of 2.8 on the EDE-Q effectively tracks ED symptom improvement through treatment and discharge. This supports the need for the development of culture-specific and empirically developed clinical cutoffs and their widespread use to evaluate program effectiveness.

Temporal dynamics of interoceptive attention and positive and negative affect in adults engaging in disordered eating and nonsuicidal self-injury.

OBJECTIVE: Interoception (the sense of the body's state) is associated with eating disorders and nonsuicidal self-injury, potentially through relationships with affect. We examined the relationship between interoceptive attention and both positive and negative affect. METHODS: Participants (N = 128) who reported recent self-harm (i.e., disordered eating and/or non-suicidal self-injury) completed ecological momentary assessment for 16 days. Participants completed multiple daily assessments of affect and interoceptive attention. We then tested the temporal dynamic relationship between interoceptive attention and affect. RESULTS: There was a relationship between positive affect and interoceptive attention such that people with higher average positive affect, and times when people were above their typical positive affect, were associated with higher interoceptive attention. There was a negative relationship between negative affect and interoceptive attention, such that people with higher average negative affect, and times when people were above their own typical negative affect, were associated with lower interoceptive attention. CONCLUSIONS: Better mood may associate with greater willingness to attend to body sensations. Our findings support active inference models of interoception and highlight the importance of refining our understanding of the dynamic nature of interoception and its relationship with affect.

Confirmatory factor analysis of Acute Suicidal Affective Disturbance in a sample of treatment-seeking eating disorder patients.

INTRODUCTION: We examined the factor structure of Acute Suicidal Affective Disturbance (ASAD) in individuals with eating disorders. ASAD is a proposed diagnosis with four symptoms: rapidly increasing suicidal intent, social or self-alienation, hopelessness regarding the previous symptoms, and overarousal. METHODS: Patients with eating disorders (N = 378) completed self-report symptom questionnaires at admission to a partial hospitalization program for eating disorders. Using proxy measures, we conducted a confirmatory factor analysis (CFA) of the one-factor model of ASAD. RESULTS: Initial model fit was poor. We iteratively revised the model to include theoretically-justified correlated residuals (i.e., those between items intended to measure the same ASAD symptom). After these modifications, model fit remained mediocre. We next conducted a multiple-group CFA to compare ASAD between individuals with the restricting subtype of anorexia nervosa (the "ANR" model, n = 145) and individuals with binge/purge disorders (i.e., bulimia nervosa and the binge-purge subtype of anorexia nervosa; the "BP" model, n = 234). The final model had mediocre fit with partial invariance between subgroups; the ASAD factor mean was higher in the BP model. We additionally tested a bifactor model, with similar findings. LIMITATIONS: Our use of proxy measures of ASAD items may have resulted in imprecise measurement of ASAD symptoms, pointing to the importance of future research using validated measures of ASAD. CONCLUSIONS: We found moderate support for ASAD in an eating disorder sample, with invariance between eating disorder phenotypes. Our findings suggest that the ASAD diagnosis may not fully generalize to eating disorder samples.

Feeling the urge: Using ecological momentary assessment to test the relationship of interoception with disordered eating and nonsuicidal self-injury.

We used ecological momentary assessment to test the temporal relationship between interoception (the perception of internal body states) and self-injurious behavior (SIB), specifically eating disorders (ED) and nonsuicidal self-injury (NSSI). Adults (N = 128) reported SIB and interoceptive attention several times a day for 14 days. Higher average interoception is associated with lower risk for SIB, whereas higher momentary interoception is associated with higher risk for SIB. Furthermore, interoceptive attention appears to fluctuate rather than change, and greater instability of interoceptive attention is associated with SIB. Our results demonstrate that an individual's absolute level of interoceptive attention is less predictive of SIB than fluctuation in interoceptive attention. These findings point to the importance of future research examining the possible etiological role of interoceptive variability in SIB, as well as the development of interventions to target interoceptive lability as a modifiable risk factor for ED and NSSI. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Associations of suicide risk with emotional reactivity, dysregulation, and eating disorder treatment outcomes.

INTRODUCTION: Emotional processes play a role in both suicide risk and eating disorders (EDs), which are often comorbid. However, limited research has explored how emotional processes relate to suicide risk in EDs and the prognostic value of suicide risk for ED treatment. Thus, the current study examined associations between emotion dysregulation and reactivity with suicide risk in patients with EDs, and determined if suicide risk predicts ED treatment outcomes. METHODS: Participants (n = 201) were adults in an ED partial hospitalization program who completed measures at admission, 1-month post-admission, and discharge. RESULTS: When controlling for depressive symptoms, limited access to adaptive emotion regulation strategies, difficulties engaging in goal-oriented behaviors, and engaging in impulsive behavior when experiencing negative emotions (i.e., emotion dysregulation) were associated with suicide attempt frequency. Depressive symptoms were associated with suicide risk severity, while emotion dysregulation and reactivity were not. Importantly, patients with elevated suicide risk at admission improved comparably to other risk categories across treatment, despite presenting with greater ED symptoms at admission. CONCLUSION: Emotion dysregulation and depression are salient factors when examining suicide risk in patients with EDs. Suicide risk and attempt history may not negatively impact ED treatment outcomes when using emotion-focused treatment.

Examining psychometric properties of the Interpersonal Needs Questionnaire among college students in India.

BACKGROUND: With the second-highest population in the world, suicide-related deaths in India are high, and adults under 30 are particularly at an increased risk. However, empirical examinations of factors contributing to suicide in India and assessments of reliability and validity of self-report measures assessing these constructs are rare. AIMS: The present study examined the psychometric properties of the Interpersonal Needs Questionnaire (INQ). MATERIALS & METHODS: Undergraduate students in India (N = 432) completed the INQ and questionnaires assessing suicidal ideation, depression, fearlessness about death, and pain tolerance. RESULTS: Confirmatory factor analyses of the 15-item INQ indicated that after removing three items assessing perceived burdensomeness, the two-factor structure of INQ demonstrated acceptable fit with good internal consistency for each of the subscales (α = .84-.90). In line with the interpersonal-psychological theory of suicidal behavior (IPTS), thwarted belongingness and perceived burdensomeness interacted to predict suicidal ideation. Additionally, these constructs were positively associated with suicidal ideation and depression, and weakly correlated with fearlessness about death and pain tolerance. DISCUSSION: Results support the relevance of the IPTS for understanding suicidal ideation among college students in India. CONCLUSION: The results suggest that modified INQ demonstrates strong internal consistency, as well as good construct, criterion, and discriminant validity among Indian college students.

No pain, no gain? Associations of athletic participation with capability for suicide among college students.

INTRODUCTION: Suicide is a leading cause of death among college students. In order to advance understanding of suicide risk factors in this population, the present study provides a theoretically grounded test of associations between experiences in sports and both facets of capability for suicide (i.e., fearlessness about death and pain tolerance). Secondarily, the present study aimed to explore possible distinctions between subjective and objective pain tolerance. METHODS: Undergraduate students (N = 153) completed the study at a single timepoint. A pressure algometer was used to measure objective pain tolerance, and self-report questionnaires assessed subjective pain tolerance, fearlessness about death (FAD), sports participation, sports-related injuries, level of contact in sport, and weight manipulation practices for sport. RESULTS: Athletes had greater capability for suicide (i.e., higher objective and subjective pain tolerance and greater FAD) relative to non-athletes. Sports-related injuries were positively associated with both FAD and subjective pain tolerance, and level of contact in sport was positively associated with FAD. Objective and subjective pain tolerance were moderately correlated. CONCLUSIONS: Sports-related injuries and level of physical contact in athletics are associated with the facets of capability for suicide among college student athletes. Discrepancies in results between subjective and objective pain tolerance suggest important distinctions between these constructs.

Preliminary development of an implicit association test to measure body dissatisfaction and predict disordered eating behaviors.

We conducted a preliminary test of a novel implicit association test for body dissatisfaction (BD-IAT). We predicted that BD-IAT would correlate with self-reported BD and predict later eating disorder (ED) attitudes. Female undergraduates (N = 145) self-reported BD and ED attitudes. In the BD-IAT, participants sorted words related to "satisfaction" and "dissatisfaction" to opposite sides of a computer screen. Participants categorized pictures of themselves to the side of the screen with the word "Me," which appeared with the "satisfied" or "dissatisfied" category in separate blocks. The BD-IAT measured the strength of the association between a participant's own body and dissatisfaction. The BD-IAT correlated positively with the EDI-BD (r = 0.21, p < 0.05), supporting its convergent validity. The BD-IAT predicted ED attitudes at one-week follow-up, controlling for explicit BD and baseline ED (b = 0.58, SE = 0.23, t = 2.57, p = .01); IAT scores uniquely accounted for 4.7 % of the variance in ED at one-week follow-up. These results provide preliminary evidence for the BD-IAT as an implicit measure of BD and suggest that it may predict future ED attitudes above and beyond explicit BD. Directions for future work aimed at continued development of this task are discussed.

Establishing the psychometric properties and construct validity of the Painful and Provocative Events Scale-Revised.

BACKGROUND: Engagement in painful and provocative events is central to hypotheses for how capability for suicide develops. However, the existing measure of painful and provocative events is not psychometrically sound. We developed a measure with improved psychometric properties: the Painful and Provocative Events Scale (PPES)-Revised. METHOD: In study 1, 447 adults (53.5% women, mean age = 35.4 years) answered 77 items describing painful and/or provocative experiences. Exploratory factor analysis (EFA) was performed. In study 2, 403 adults (55.1% women, mean age = 38.1 years) answered the retained items and confirmatory factor analysis (CFA) was performed. The scale's factorial invariance across gender was examined. Estimates of convergent and discriminant validity were obtained. RESULTS: EFA yielded a two-factor structure, which was confirmed with CFA. The factor structure was invariant across men and women. Estimates of convergent and discriminant validity were promising. LIMITATIONS: Limitations include non-weighted items, additional need for CFA among high-risk groups, lack of assessment of person-specific painful and provocative events, and need for prospective research to establish the scale's predictive validity. CONCLUSIONS: By assessing painful and provocative events more uniformly and reliably than the existing measure, the PPES-Revised has the potential to advance the understanding of capability for suicide.

Perfectly imperfect: The use of cognitive bias modification to reduce perfectionism.

BACKGROUND AND OBJECTIVES: Perfectionism is a transdiagnostic risk and maintenance factor for psychopathology. The current study developed and evaluated a cognitive bias modification, interpretation retraining (CBM-I) intervention targeting maladaptive perfectionistic beliefs. METHODS: Participants were undergraduate students randomized to complete the perfectionism CBM-I (n = 33) or control condition task (n = 36) at two time points. Additionally, participants completed measures of perfectionistic interpretations and trait perfectionism, as well as an impossible anagram task designed to elicit perfectionistic concerns. RESULTS: Results indicated that after the intervention, participants who completed the perfectionism CBM-I endorsed fewer perfectionistic interpretations than participants in the control condition. Furthermore, although the study groups self-reported comparably low confidence in their anagram task performance, participants who completed the perfectionism CBM-I reported wanting to re-do significantly fewer anagrams than participants in the control condition, suggesting greater acceptance of imperfect performance following the intervention. Moreover, supporting a key hypothesized mechanism of effect in CBM-I, reductions in perfectionistic interpretations mediated the effect of condition on the desire to re-do anagram task items. LIMITATIONS: The study results should be viewed in light of limitations, including the short time-span of the study, and the use of a relatively small, non-clinical, and demographically homogenous convenience sample. CONCLUSIONS: Further research and development of the perfectionism CBM-I intervention are needed, but the present findings add to a nascent evidence base that suggests CBM-I holds promise as an accessible and transdiagnostic intervention for perfectionism.

Are Eating Disorders and Related Symptoms Risk Factors for Suicidal Thoughts and Behaviors? A Meta-analysis.

This meta-analysis addressed whether eating disorders (EDs) are risk factors (i.e., longitudinal predictors) for suicidal thoughts and behaviors. We identified 2,611 longitudinal studies published through August 1, 2017. Inclusion required studies include at least one longitudinal analysis predicting suicide ideation, attempt, or death using an ED diagnosis and/or symptom. Fourteen studies (42 prediction cases) met criteria. Results indicated that clinically diagnosed EDs and disordered eating symptoms were significant but weak predictors of suicide attempts but not death. Effects remained weak when moderators were considered. By reviewing the methodological limitations of previous research, these results highlight avenues for future research.

Trait rumination and response to negative evaluative lab-induced stress: neuroendocrine, affective, and cognitive outcomes.

Theoretical models of depression posit that, under stress, elevated trait rumination predicts more pronounced or prolonged negative affective and neuroendocrine responses, and that trait rumination hampers removing irrelevant negative information from working memory. We examined several gaps regarding these models in the context of lab-induced stress. Non-depressed undergraduates completed a rumination questionnaire and either a negative-evaluative Trier Social Stress Test (n = 55) or a non-evaluative control condition (n = 69), followed by a modified Sternberg affective working memory task assessing the extent to which irrelevant negative information can be emptied from working memory. We measured shame, negative and positive affect, and salivary cortisol four times. Multilevel growth curve models showed rumination and stress interactively predicted cortisol reactivity; however, opposite predictions, greater rumination was associated with blunted cortisol reactivity to stress. Elevated trait rumination interacted with stress to predict augmented shame reactivity. Rumination and stress did not significantly interact to predict working memory performance, but under control conditions, rumination predicted greater difficulty updating working memory. Results support a vulnerability-stress model of trait rumination with heightened shame reactivity and cortisol dysregulation rather than hyper-reactivity in non-depressed emerging adults, but we cannot provide evidence that working memory processes are critical immediately following acute stress.

Examining Patterns of Change in the Acquired Capability for Suicide Among Eating Disorder Patients.

OBJECTIVE: Several models of suicidal behavior posit that, to transition from suicidal ideation to attempting suicide, individuals must have an acquired capability for suicide (ACS), comprised of fearlessness about death (FAD) and pain tolerance. ACS is hypothesized to increase monotonically through exposure to painful and provocative experiences. However, recent research suggests that ACS can decrease, bringing into question the hypothesis of monotonic increase. This study examined the nature of change in ACS over time within a sample of patients with eating disorder (ED). We predicted that there would be two classes of change in ACS: one high and increasing and one moderate and stable. METHOD: One hundred female patients with ED reported on ACS at admission and weekly during treatment. RESULTS: Growth mixture modeling to test models of FAD and pain tolerance identified that, for both factors, a one-class intercept-only model was the best-fitting model, suggesting that patients entered treatment with midlevel ACS and experienced no significant linear change over the course of treatment. CONCLUSIONS: Acquired capability for suicide demonstrated stability in this study; results highlight the need for additional research examining ACS across different timescales and in varied populations.

Which Comes First? An Examination of Associations and Shared Risk Factors for Eating Disorders and Suicidality.

PURPOSE OF REVIEW: This narrative review evaluates recent literature on the associations between eating disorders and suicidality and discusses potential shared mechanisms that may account for these relationships. Additionally, the review highlights shortcomings with the literature to date and suggests avenues for future research. RECENT FINDINGS: Individuals with anorexia nervosa, bulimia nervosa, and binge eating disorder experience elevated rates of suicidality compared to the general population. Suicide risk is higher when eating disorders occur with other psychological conditions. Additionally, genetic factors, emotion dysregulation, trauma, stressful life events, and lack of body regard may have roles in the development of both eating disorders and suicidality. Much of the risk for suicidality in eating disorders appears to be driven by comorbid psychopathology and genetic factors. However, the lack of longitudinal research makes it difficult to draw conclusions about the directionality or temporality of these relations; thus, novel methods are needed.

Implicit attitudes toward eating stimuli differentiate eating disorder and non-eating disorder groups and predict eating disorder behaviors.

OBJECTIVE: The current study tested whether people with and without eating disorders (EDs) varied in their implicit attitudes toward ED-relevant stimuli. Additionally, the study tested whether implicit evaluations of ED-relevant stimuli predicted ED symptoms and behaviors over a 4-week interval. METHOD: Participants were people without EDs (N = 85) and people seeking treatment for EDs (N = 92). All participants completed self-report questionnaires and a version of the affect misattribution procedure (AMP) at baseline. The AMP indexed implicit evaluations of average body stimuli, eating stimuli, and ED-symptom stimuli. Participants with EDs completed weekly follow-up measures of ED symptoms and behaviors for 4 weeks. RESULTS: Contrary to predictions, the anorexia nervosa (AN) group did not differ from the no ED group on implicit attitudes toward ED-symptom stimuli, and the bulimia nervosa (BN) group had less positive implicit attitudes toward ED-symptom stimuli relative to the no ED group. In line with predictions, people with AN and BN had more negative implicit attitudes toward average body and eating stimuli relative to the no ED group. In addition, among the ED group more negative implicit attitudes toward eating stimuli predicted ED symptoms and behaviors 4 weeks later, over and above baseline ED symptoms and behaviors. DISCUSSION: Taken together, implicit evaluations of eating stimuli differentiated people with AN and BN from people without EDs and longitudinally predicted ED symptoms and behaviors. Interventions that increase implicit liking of eating-related stimuli may reduce ED behaviors.