Understanding heterogeneity, comorbidity, and variability in depression: Idiographic models and depression outcomes.

This study uses time-intensive, item-level assessment to examine individual depressive and co-occurring symptom dynamics. Participants experiencing moderate-severe depression (N = 31) completed ecological momentary assessment (EMA) four times per day for 20 days (total observations = 2480). We estimated idiographic networks using MDD, anxiety, and ED items. ED items were most frequently included in individual networks relative to depression and anxiety items. We built ridge and logistic regression ensembles to explore how idiographic network centrality metrics performed at predicting between-subject depression outcomes (PHQ-9 change score and clinical deterioration, respectively) at 6-months follow-up. For predicting PHQ-9 change score, R2 ranged between 0.13 and 0.28. Models predicting clinical deterioration ranged from no better than chance to 80 % accuracy. This pilot study shows how co-occurring anxiety and ED symptoms may contribute to the maintenance of depressive symptoms. Future work should assess the predictive utility of psychological networks to develop understanding of how idiographic models may inform clinical decisions.

Facing Eating Disorder Fears: An Open Trial Adapting Prolonged Exposure to the Treatment of Eating Disorders.

Eating disorders (EDs) are maintained by core fears, which lead to avoidance behaviors, such as food avoidance or compensatory behaviors. Previously tested exposure-based treatments for EDs have generally focused on proximal outcomes (e.g., food), rather than addressing core fears (e.g., fear of weight gain and its consequences). The current study tested the feasibility and initial clinical efficacy of 10 sessions of imaginal and in vivo exposure for core ED fears (termed "Facing Eating Disorder Fears"), mainly fear of weight gain and its associated consequences. Participants were 36 adults with anorexia nervosa (AN), bulimia nervosa, or other specified feeding and eating disorders determined by semistructured diagnostic interviews. ED symptoms, fears, and body mass index (BMI) were assessed at pretreatment, posttreatment, and 1-month follow-up. Treatment involved 10 sessions of imaginal and in vivo exposure to ED fears in combination with in vivo exposures to feared and avoided situations as homework. ED symptoms and fears decreased from pre- to posttreatment and at 1-month follow-up. BMI increased significantly from pre- to posttreatment, particularly for those with AN. Effect sizes ranged from small to very large. ED symptoms and fears decreased and BMI increased following exposure. Increases in BMI occurred without any direct intervention on eating, suggesting that weight gain can be achieved without a specific focus on food during ED treatment. Facing Eating Disorder Fears may be a feasible stand-alone intervention for EDs. Future research must test comparative efficacy through randomized controlled trials.

The Unique and Moderating Role of Social and Self-Evaluative Factors on Perinatal Eating Disorder and Depression Symptoms.

Pregnancy and postpartum represent a critical transition period for changes in eating disorder (ED) symptoms and depression. Past research has established a relationship between ED and depressive symptoms during pregnancy. However, changes in depression and ED symptom across stages of pregnancy and postpartum, and factors that influence this relationship, remain understudied. Social factors and self-evaluative factors may be important given rapidly changing social pressures and expectations during this transitional time. The current study (N = 454 pregnant women) examined (1) differences in ED and depressive symptoms across pregnancy and postpartum and (2) whether social factors (social appearance anxiety; social support) and self-evaluative factors (maladaptive perfectionism; self-compassion) moderate the relationship between depression and ED symptoms cross-sectionally and prospectively. Study aims, hypotheses, and data analysis were preregistered on the Open Science Foundation (osf.io). This study did not identify differences in ED or depression symptoms across women at different stages of pregnancy; however, depression symptoms significantly improved within individuals from pregnancy to postpartum. ED symptoms and all social and self-evaluative factors were uniquely associated with depression during pregnancy. ED symptoms, maladaptive perfectionism, social appearance anxiety, and self-compassion during pregnancy significantly predicted postpartum depression, when accounting for prenatal depression. During pregnancy, but not postpartum, when social support and self-compassion were low, and when maladaptive perfectionism was high, there was a stronger relationship between ED and depression symptoms. ED symptoms and social and self-evaluative factors could be targeted in routine medical care and stepped-care interventions to improve maternal mental healthcare and prevent postpartum depression.

Development and Validation of a Broad and Fear-Adaptable Measure of Fear Approach and Application to Common Eating Disorder Fears.

Fear approach is a theorized mechanism of exposure treatment for anxiety-based disorders. However, there are no empirically established self-report instruments measuring the tendency to approach feared stimuli. Because clinical fears are heterogeneous, it is important to create a measure that is adaptable to person- or disorder-specific fears. The current study (N = 455) tests the development, factor structure, and psychometric properties of a self-report instrument of fear approach broadly and the adaptability of this measure to specific eating disorder fears (i.e., food, weight gain). Factor analyses identified a unidimensional, nine-item factor structure as the best fitting model. This measure had good convergent, divergent, and incremental validity and good internal consistency. The eating disorder adaptations retained good fit and strong psychometric properties. These results suggest that this measure is a valid, reliable, and adaptable measure of fear approach, which can be used in research and exposure therapy treatment for anxiety-based disorders.

An Intensive time series investigation of the relationships across eating disorder-specific fear responses and behavior urges in partially remitted anorexia nervosa.

Anorexia nervosa (AN) is a serious and persistent psychiatric illness. Many individuals with AN cycle between stages of remission (i.e., relapse), with research documenting that cognitive remission generally lags behind nutritional/weight restoration. Yet, little is known about which mechanisms promote movement from partial remission in AN (defined as nutritional, but not cognitive, recovery) to full remission. Fear-based processes, including avoidance and approach behaviors, likely contribute to the persistence of cognitive-behavioral AN symptoms after nutritional restoration. The current study used intensive longitudinal data to characterize these processes during partial remission (N = 41 participants with partially remitted AN; 4306 total observations). We aimed to a) characterize frequency of fear-based processes in real-time, b) investigate associations across fear-based processes and behavioral urges, and c) test if real-time associations among symptoms differed across commonly feared stimuli (e.g., food, social situations). On average, participants endorsed moderate fear and avoidance, with weight-gain fears rated higher than other feared stimuli. Momentary fear, avoidance, approach, and distress were all positively associated with AN behavior urges at one time-point and prospectively. Central symptoms and symptom connections differed across models with different feared stimuli. These findings provide empirical support for the theorized fear-avoidance-urge cycle in AN, which may contribute to the persistence of eating pathology during partial remission. Fear approach may be associated with temporary increases in urges, which should be considered during treatment. Future research should explore these associations in large, heterogeneous samples, and test the effectiveness of exposure-based interventions during partial remission.

Profiles of experienced and internalized weight-based stigma in college students across the weight spectrum: Associations with eating disorder, depression, and anxiety symptoms.

OBJECTIVE: Weight stigma comprises negative attitudes and weight-related stereotypes that result in rejection, discrimination, and prejudice against individuals in larger bodies. Both internalized and experienced weight stigma are associated with negative mental health outcomes, yet it remains unknown how types of stigmatizing experiences (e.g., systemic vs. intraindividual), internalized stigma, and weight status relate, or how profiles of weight stigma differentially impact mental health. METHOD: The current study (N = 1001 undergraduates) used latent profile analysis to identify weight stigma risk profiles and test whether profiles were cross-sectionally associated with eating disorder symptoms, depression, and social appearance anxiety. RESULTS: The best-fitting solution indicated a class high on all facets of weight stigma, a class low on all facets of weight stigma, and three groups with intermediate levels of weight, weight bias internalization, and experienced weight stigma. Gender, but not ethnicity, was associated with class membership. Classes with higher experienced and internalized stigma had higher eating disorder symptoms, depression, and social appearance anxiety. CONCLUSION: Findings support the utility of weight stigma profiles in identifying risk for negative mental health outcomes. These findings can inform initiatives to reduce weight stigma among college students, especially among high-risk groups.

A Pilot, Time-Series Investigation of Depression, Anxiety, and Eating Disorder Symptoms in Adults Experiencing Major Depressive Symptoms: The Need for Eating Disorder Assessment and Research in Depression.

Major Depressive Disorder (MDD) is a prevalent psychiatric disorder impacting 10-16% of Americans in their lifetime. Approximately 60% of individuals with MDD have comorbid anxiety disorders. Additionally, although scarce research has examined eating disorders (EDs) in depression, a bidirectional association exists between ED and MDD symptoms. The current pilot study (N = 31 individuals with moderate to severe depression) modeled networks of depressive, anxiety, and ED symptoms using intensive time-series data. This study also tested if temporal central symptoms predicted six-month clinical outcomes. The most central symptoms were guilt, self-dislike, lack of energy, and difficulty concentrating. Several anxiety and ED symptoms were also central, including physical anxiety, social anxiety, body dissatisfaction, and desire for thinness. The central symptom crying predicted six-month depression with a medium effect size. These findings suggest anxiety and ED symptoms may influence the day-to-day course of depression in some individuals with comorbid diagnoses, but predictors of symptoms across hours may differ from predictors across longer time scales (i.e., months). Time scale should be considered when conducting and interpreting research on MDD. Research, assessment, and treatment for MDD should continue to explore transdiagnostic approaches including anxiety and ED symptoms to optimize care for individuals with complex presentations.

Personalizing eating disorder treatment using idiographic models: An open series trial.

OBJECTIVE: Treatments for adults with eating disorders (EDs) only work in about 50% of individuals, and for some diagnoses (e.g., anorexia nervosa; atypical anorexia nervosa), there are no existing evidence-based treatments. Part of the reason that treatments may only work in a subset of individuals is because of the high heterogeneity present in the EDs, even within diagnoses. Manualized treatments delivered in a standard format may not always address the most relevant symptoms for a specific individual. METHOD: The current open series trial recruited participants with transdiagnostic ED diagnoses (N = 79) to investigate the feasibility, acceptability, and initial clinical efficacy of a 10-session network-informed personalized treatment for eating disorders. This treatment uses idiographic (i.e., one-person) network models of ecological momentary assessment symptom data to match participants to evidence-based modules of treatment. RESULTS: We found that network-informed personalized treatment was highly feasible with low dropout rates, was rated as highly acceptable, and had strong initial clinical efficacy. ED severity decreased from pre- to posttreatment and at 1-year follow-up with a large effect size. ED cognitions, behaviors, clinical impairment, worry, and depression also decreased from pre- to posttreatment. CONCLUSIONS: These data suggest that network-informed personalized treatment has high acceptability and feasibility and can decrease ED and related pathology, possibly serving as a feasible alternative to existing treatments. Future randomized controlled trials comparing network-informed personalized treatment for ED to existing gold standard treatments are needed. Additionally, more research is needed on this type of personalized treatment both in the EDs, as well as in additional forms of psychopathology, such as depression. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Eating Disorder Screening and Treatment in a Medically Underserved Southern State: Data Collected by a Legislative Eating Disorder Council and Implications for Statewide Assessment.

Healthcare providers play a key role in early identification of eating disorders (EDs), especially in underserved states where ED treatment resources are lacking. Currently, there is little known about ED screening and treatment practices in underserved states. The current study assessed current ED screening and treatment practices among healthcare providers in an underserved state using data collected by a government-formed state ED council. Healthcare providers (N = 242; n = 209 behavioral health providers; n = 33 medical providers) practicing in Kentucky completed a brief, anonymous survey on ED screening and treatment practices, comfort with screening for EDs, and interest in continued education. Over half of healthcare providers indicated screening for EDs, with the majority using a clinical interview. After identification of ED symptoms, providers reported a combination of treating in-house, referring out, or seeking consultation. In bivariate analyses, medical providers were significantly more likely than behavioral health providers to use a screening tool specifically designed for EDs. The majority of medical providers indicated that they received education about EDs and feel knowledgeable about ED screening tools, though most reported infrequent use of these screening tools in their practice. Nearly all behavioral health and medical providers expressed interest in continuing education on ED screening and treatment. These findings indicate a need for, and interest in, education on evidence-based ED screening and treatment resources in underserved states and demonstrate the utility of a state ED council to collect these data to inform future education and treatment strategies.

A Network Analysis Investigation of Disordered Eating Across Demographic and Developmental Subpopulations Using a National Epidemiological Sample of High School Students.

Disordered eating (DE) poses a large societal burden, yet limited research has examined DE from a developmental epidemiological perspective. It is important to consider how demographics influence DE symptoms to inform prevention and early intervention programs across diverse subpopulations. Therefore, we conducted network analyses using a large nationally representative epidemiological sample of high school students (Youth Risk Behavior Survey, United States; n = 59,582) to identify the most important symptoms and symptom relationships among six DE behaviors. We compared networks by sex, grade, and race to identify differences in symptom networks. Dieting for weight loss was highly central across networks. Networks significantly differed across sex, grade, and race. Our results suggest that dieting for weight loss may be an early intervention target for eating disorders, regardless of demographic and developmental factors. In addition, sex, race, and age should be accounted for when researching and developing prevention programs for DE and eating disorders. Public health officials, as well as mental health professionals, should present a more balanced message about dieting and weight loss to high school students to prevent the detrimental impact of DE on physical and mental health. Notably, this study is the first large, nationwide epidemiological sample using DE symptoms in network analysis.

Longitudinal group and individual networks of eating disorder symptoms in individuals diagnosed with an eating disorder.

Eating disorders (EDs) are serious psychiatric illnesses with high mortality and societal cost. Despite their severity, there are few evidence-based treatments, and only 50% of individuals respond to existing treatments. This low response rate may be due to the fact that EDs are highly heterogeneous disorders. Precision treatments are needed that can intervene on individual maintenance factors. The first step in such treatment development is identification of central treatment targets, both at the group (i.e., on average) and individual level. The current study (N = 102 individuals with an ED) utilized intensive longitudinal data to model several types of group-level and individual network models. Overall, we identified several group-level central symptoms, with the most common central symptoms of fear of weight gain, desire for thinness, feeling like one is overeating, thinking about dieting, and feeling guilty. We also found that these symptoms, specifically fear of weight gain, a desire to be thinner, thinking about dieting, feeling like one is overeating, and feeling guilty, predicted ED severity at a 1- and 6-month follow-up. We modeled 97 individual networks and found that central symptoms were highly heterogeneous, regardless of ED diagnosis. This work adds to the growing literature using intensive longitudinal data to model ED pathology and implicates fear of weight gain, thinking about dieting, and feelings of guilt as symptoms needing further treatment development work. Additionally, this work contributes essential knowledge on how group and individual network modeling can be used to conceptualize the maintenance of EDs on average and at the individual level. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

Eating disorder and social anxiety symptoms in Iranian preadolescents: a network analysis.

PURPOSE: Network studies of eating disorder (ED) symptoms have identified central and bridge symptoms in Western samples, yet few network models of ED symptoms have been tested in non-Western samples, especially among preadolescents. The current study tested a network model of ED symptoms in Iranian preadolescents (ages 9 to 13), as well as a model of co-occurring social anxiety disorder (SAD) and ED symptoms. METHOD: Preadolescent boys (n = 405) and girls (n = 325) completed the Children Eating Attitudes Test-20 and Social Anxiety Scale for Children. We estimated two network models (ED and ED/SAD networks) and identified central and bridge symptoms, as well as tested if these models differed by sex. RESULTS: We found that discomfort eating sweets were the most central symptoms in ED networks. Concern over being judged was central in networks including both ED and SAD symptoms. Additionally, concern over being judged was the strongest bridge symptoms. Networks did not differ by sex. CONCLUSION: Future research is needed to test if interventions focused on bridge symptoms (i.e., concern over being judged) as primary intervention points target comorbid ED-SAD pathology in preadolescents at risk for ED and SAD. LEVEL OF EVIDENCE: Level III; Evidence obtained from well-designed observational study, including case-control design for relevant aspects of the study.

Using individual networks to identify treatment targets for eating disorder treatment: a proof-of-concept study and initial data.

BACKGROUND: Eating disorders (EDs) are severe mental illnesses, with high morbidity, mortality, and societal burden. EDs are extremely heterogenous, and only 50% of patients currently respond to first-line treatments. Personalized and effective treatments for EDs are drastically needed. METHODS: The current study (N = 34 participants with an ED diagnosis collected throughout the United States) aimed to investigate best methods informing how to select personalized treatment targets utilizing idiographic network analysis, which could then be used for evidence based personalized treatment development. We present initial data collected via experience sampling (i.e., ecological momentary assessment) over the course of 15 days, 5 times a day (75 total measurement points) that were used to select treatment targets for a personalized treatment for EDs. RESULTS: Overall, we found that treatment targets were highly variable, with less than 50% of individuals endorsing central symptoms related to weight and shape, consistent with current treatment response rates for treatments designed to target those symptoms. We also found that different aspects of selection methods (e.g., number of items, type of centrality measure) impacted treatment target selection. CONCLUSIONS: We discuss implications of these data, how to use idiographic network analysis to personalize treatment, and identify areas that need future research. TRIAL REGISTRATION: Clinicaltrials.gov, NCT04183894. Registered 3 December 2019-Retrospectively registered, https://clinicaltrials.gov/ct2/show/NCT04183894 . NCT04183894 (ClinicalTrials.gov identifier).

Eating disorders are severe psychiatric illnesses that carry high mortality, morbidity, and societal and personal burden. Treatments for eating disorders only work in 50% of patients, signifying a great need to improve treatments. One reason that treatments may not work, is because eating disorders vary substantially from one individual to the next, which existing treatments do not fully consider. The current study (N = 34 participants with an eating disorder diagnosis) uses a new modeling technique to identify which symptoms should be targeted in treatment in a personalized manner. As expected, we found that, using this modeling technique, symptoms that should be targeted in treatment vary considerably. We discuss how to use this modeling technique to identify individual treatment targets and ways in which the field can use this strategy to improve existing and create new treatments.

Network Analysis of PTSD Symptoms in a Sample of Polyvictimized Youth.

Polyvictimization (i.e., the experience of multiple types of victimization) increases the risk for post-traumatic stress disorder (PTSD) relative to single-type victimization. Network analysis has been used to conceptualize PTSD among children and adolescents exposed to specific types of victimization (e.g., maltreatment, natural disasters), but not among those theorized to be at highest risk of poor outcomes (i.e., polyvictimized youth). The current study used network analysis to enhance our understanding of PTSD symptoms in a diverse sample of adolescents exposed to polyvictimization. Data were drawn from the Longitudinal Studies on Child Abuse and Neglect, a multi-site study of children and adolescents living in the United States (ages 0-18). Using indicators of seven different types of age-16 victimization, k-means cluster analysis identified a subgroup of polyvictimized youth. A PTSD symptom network was estimated for polyvictimized youth, with depersonalization, self-blame, and sadness emerging as central symptoms. Regression analyses indicated that depersonalization prospectively predicted age-18 anxiety symptoms, p = .033. Central symptoms did not predict age-18 PTSD or depression symptoms. To date, this is the first network study on a polyvictimized sample. While preliminary, our findings suggest that dissociation and self-blame may be central to polyvictimized youth and that dissociation may represent an important prevention target for anxiety.

The unique and moderated relationships across repetitive negative thinking and disordered eating behaviors in adolescent girls.

OBJECTIVE: Repetitive negative thinking (RNT) is associated with eating disorder (ED) symptoms, but few studies have evaluated how RNT may relate to specific ED behaviors (e.g., fasting, binge eating). Furthermore, little research has examined RNT in adolescent girls, a critical population for ED development. Concern over mistakes perfectionism (i.e., preoccupation with and overvaluation of errors) and social appearance anxiety are transdiagnostic risk factors for EDs that may be amplified by RNT, contributing to the development of ED behaviors. METHOD: The current study (N = 332 female high school students) explored RNT in an adolescent sample and tested whether RNT was uniquely associated with fasting/binge eating and moderated the relationships between social appearance anxiety/perfectionism and fasting/binge eating. RESULTS: RNT was not associated with fasting and did not significantly moderate the relationship between social appearance anxiety/perfectionism and fasting. RNT was positively associated with binge eating but did not moderate the relationship between social appearance anxiety/perfectionism and binge eating. CONCLUSIONS: RNT may contribute differentially to specific ED behaviors such as fasting and binge eating during ED development. If replicated in prospective data, these findings suggest targeting RNT in adolescence may decrease binge eating.

Network investigation of eating disorder symptoms and positive and negative affect in a clinical eating disorder sample.

OBJECTIVE: Growing literature suggests that emotions influence the maintenance of eating disorder (ED) symptoms. However, most research has studied the relationship between ED symptoms and affect broadly (i.e., negative affect [NA], positive affect [PA]), rather than examining models comprised of multiple specific affective states (e.g., upset, proud). METHOD: The current study (N = 196 individuals with EDs) used network analysis to examine the most interconnected (i.e., central) NA and PA states in EDs and test the complex associations between specific NA, PA, and ED symptoms. We estimated two networks: one with affective states only and another with affective states and ED symptoms. RESULTS: Feeling distressed, afraid, attentive, and determined were the most central symptoms in the affect-only network. ED symptoms related to overvaluation of weight and shape, including affect-based ED symptoms (i.e., guilt about eating), were central in the network of affect and ED symptoms. Guilt about eating and shame were central bridge symptoms across affect and ED symptom clusters, meaning that they were each strongly connected across clusters, and may represent important pathways among affect and ED symptoms. DISCUSSION: Limitations include the cross-sectional and between-person nature of these analyses, from which we cannot derive causal or within-persons processes. Clinical interventions that target central and bridge symptoms (e.g., fear, shame) may disrupt the reinforcing cycle of NA in EDs that may contribute to ED behaviors. Future research should examine relationships among affective states and ED symptoms in longitudinal and intraindividual network models to develop more effective treatments for EDs.

A network analysis of eating disorder symptoms and co-occurring alcohol misuse among heterosexual and sexual minority college women.

Eating disorders and alcohol misuse are common problems among college women. Individually, both have high prevalence rates and are associated with a significant economic burden. Yet eating disorders and alcohol misuse also frequently present simultaneously, which may increase symptom severity and related impairment. These associations are especially important to test in sexual minority populations, as symptoms may present differently, and the prevalence and personal cost of these disorders may be even higher for this group. The present study (N = 1072 undergraduate college women) used network analysis to identify pathways, central symptoms, and bridge symptoms across alcohol misuse and eating disorder symptoms. A network comparison test was used to determine if the network structure differed between heterosexual women (n = 923) and sexual minority women (n = 149). For the overall network, cognitive restraint, excessive exercise, and frequency of binge drinking, were the most central symptoms. Bridge symptoms included drinking in the morning, purging, alcohol-related guilt, and muscle building. Heterosexual and sexual minority women did not differ significantly in network structure or global strength. Regardless of sexual orientation, prevention efforts for eating disorders and alcohol misuse among college women should target central and bridge symptoms.

Latent profile analysis of impulsivity and perfectionism dimensions and associations with psychiatric symptoms.

BACKGROUND: Impulsivity and perfectionism are transdiagnostic personality factors that have been studied extensively and shown to relate to externalizing and internalizing pathology respectively. Typically, these personality factors are antithesized, with impulsivity characterized by lack of control and perfectionism characterized by rigid overcontrol. METHODS: The current study (N = 1,353 undergraduate students) used latent profile analysis to identify subgroups based on impulsivity and perfectionism dimensions and tested the relations of these subgroups with the symptomatology of ten prevalent types of psychopathology (depression, worry, social anxiety, attention-deficit hyperactivity disorder inattentive subtype, attention-deficit hyperactivity disorder impulsive-hyperactive subtype, alcohol use, obsessive-compulsive disorder, posttraumatic stress disorder, restrictive eating pathology, and binge eating pathology). RESULTS: The latent profile analysis identified four meaningful subgroups: high perfectionism, high impulsivity, combined impulsivity and perfectionism, and low impulsivity and perfectionism. The combined group was the most prevalent, comprising almost half of the sample. Further, the perfectionism group had the highest scores for obsessive-compulsive disorder, worry, and restrictive eating pathology, the impulsivity group had the highest scores for alcohol use disorder, and the combined group had the highest or second-highest scores across all types of psychopathology. LIMITATIONS: Limitations include the undergraduate sample, self-report, cross-sectional study design, and high bivariate residuals. CONCLUSIONS: These findings suggest impulsivity and perfectionism can co-occur. Further, the co-occurrence of these personality traits may heighten risk for psychopathology and help explain comorbidity across internalizing and externalizing disorders. Future research should continue to investigate the presentation, prevalence, and treatment for individuals high in both perfectionism and impulsivity.

A network examination of risky behaviours in a state-level and national epidemiological sample of high school students.

AIM: Engagement in risky behaviours, including substance use, disordered eating, suicidal behaviour, and peer victimization/violence, during adolescence is becoming increasingly prevalent. These risky behaviours are highly comorbid and associated with long-term consequences for health, relationships, and socioeconomic status, representing an important public health concern. Past research has primarily investigated risky behaviours in adolescence using latent variable models, which are based on assumptions that may limit insight into the complex reality of these behaviours. METHODS: The current study uses network analysis to examine adolescent substance use, disordered eating, suicide risk, and peer victimization/violence in a national (N = 29 008) and state-level (Kentucky; N = 3455) epidemiological dataset. We calculated central and bridge symptoms and compared network structure based on demographic factors (race, sex, grade) and sample (state vs. nation). RESULTS: The most central symptoms were suicidal ideation and attempts, stimulant drug use, and prescription drug misuse. The most central bridge symptoms were depression, methamphetamine use, peer violence, and suicide attempts. There were no differences in network structure between samples or across demographic factors in the Kentucky sample. There were differences in network structure across sex and race in the national dataset. CONCLUSIONS: These findings suggest stimulant use, suicidal ideation, depression, and peer violence may contribute to the high rates and co-occurrence of risky behaviours in adolescence. Based on network theory, these symptoms may represent important targets for intervention. Due to network differences, special considerations may be necessary to adapt such interventions to meet the needs of students from different backgrounds.