Symptoms of Depression, Eating Disorders, and Binge Eating in Adolescents With Obesity The Fast Track to Health Randomized Clinical Trial

Importance
Depression and eating disorders are heightened for adolescents with obesity. Clinical reviews alongside self-report questionnaires are important to ensure appropriate intervention.


Objective
To evaluate changes in self-report symptoms of depression, eating disorders, and binge eating in adolescents with obesity during the Fast Track to Health trial.


Design, Setting, and Participants
This was a randomized clinical trial conducted from 2018 to 2023. It was a multisite trial conducted at children's hospitals in Sydney, New South Wales, and Melbourne, Victoria, Australia, and included adolescents (13-17 years) with obesity (defined as adult equivalent body mass index ≥30; calculated as weight in kilograms divided by height in meters squared) and 1 or more related complications.


Interventions
Duration was 52 weeks including a very low energy diet for 4 weeks followed by intermittent energy restriction (IER) or continuous energy restriction (CER).


Main Outcomes and Measures
Self-report symptoms of depression (Center for Epidemiologic Studies Depression Scale-Revised 10-Item Version for Adolescents [CESDR-10]; scores 0-30), eating disorders (Eating Disorder Examination Questionnaire [EDE-Q]; scores 0-6), and binge eating (Binge Eating Scale [BES]; scores 0-46) were assessed. Adolescents were screened for depression and eating disorders (weeks 0, 4, 16, and 52) and monitored for the onset of new symptoms of disordered eating during dietetic consults.


Results
Of 141 adolescents (median [IQR] age, 14.8 [12.9-17.9] years; 71 male [50.4%]) enrolled, median baseline EDE-Q score was 2.28 (IQR, 1.43-3.14), median baseline CESDR-10 score was 9.00 (IQR, 4.00-14.50), and median baseline BES score was 11.00 (IQR, 5.00-17.00). There were no differences between groups for change in CESDR-10 (mean difference at week 52, 0.75; 95% CI, -1.86 to 3.37), EDE-Q (mean difference at week 52, 0.02; 95% CI, -0.41 to 0.45), or BES (mean difference at week 52, -2.91; 95% CI, -5.87 to 0.05). The within-group reductions at week 4 were maintained at week 52, for CESDR-10 and EDE-Q, indicating reduced symptoms of depression and eating disorders. Within-group reductions on the BES were maintained in the IER group only. Seventeen adolescents (12.1%) required support or referral for depression and/or disordered eating, including 7 (5%; 5 IER, 2 CER) adolescents who experienced the onset or reemergence of symptoms during the intervention.


Conclusions and Relevance
Results suggest that many treatment-seeking adolescents with obesity self-reported symptoms of depression and eating disorders. Although symptoms reduced for most, some required additional support. Obesity treatment is an opportune time to screen and monitor for depression and disordered eating.


Trial Registration
Australian New Zealand Clinical Trials Registry: ACTRN12617001630303.


Participants
Eligible adolescents were aged 13 to 17 years with obesity (defined as the adult equivalent BMI ≥30; calculated as weight in kilograms divided by height in meters squared) 28 and 1 or more cardiometabolic complications (eg, insulin resistance, hypertension). 26Agreement to participate was obtained from adolescents and written consent from parents/carers.Information on ethnicity was collected as per standard clinical practice for clinical trials.

Screening and Monitoring
Participants were screened for depression and eating disorders, using self-report questionnaires at baseline and week 4, 16, and 52 (Figure 1).Participants were monitored for disordered eating by trained study dietitians during each consultation according to a prespecified protocol (Figure 1). 26If support for depression or disordered eating was needed, this was provided as part of the study or referral to external services as per standard clinical care.Additional support visits for disordered eating were recorded.All adolescents were referred to their primary care practitioner for ongoing care at week 52.0-30), 29,30 Eating Disorder Examination Questionnaire (EDE-Q; scores 0-6), 31 and Binge-Eating Scale (BES; scores 0-46) 32 were used.CESDR-10 scores of 8 or greater indicate subthreshold symptoms of depression, with a possible, probable, and major depressive episode, defined by the presence of anhedonia, dysphoria, or irritability with additional symptoms (eTable 1 in Supplement 4). 30An EDE-Q cutpoint of 2.7 33 or the presence of disordered eating behaviors was used to identify adolescents at risk of eating disorders (Figure 1).On the BES, a score of 17 or less indicates no binge eating, scores 18 to 26 indicate mild/moderate binge eating, and scores of 27 or greater indicate severe binge eating. 32[35][36][37][38]

Statistical Analysis
Statistical analyses were performed using SPSS statistics, version 28.0 (IBM Corp) blinded to group allocation.All participant data were retained, consistent with intention to treat for longitudinal analysis.Linear mixed models, with an autoregressive first-order covariance structure and restricted maximum likelihood, were used to estimate the change in outcomes between baseline and week 4, 16, and 52.Assumptions of modeling were met.Analyses were repeated using an unstructured covariance model as part of sensitivity analyses.A pairwise post hoc comparison was conducted to compare means over time.Results are presented as the difference in estimated marginal means (EMM).Nine sibling pairs were recruited.Analyses were repeated removing 1 participant from each pair, with no differences in results; therefore, all data were retained in analyses.Independentsample t tests and Mann-Whitney U tests were used to test for differences between continuous variables at baseline, week 4, and week 16 between those who completed and those who did not complete the study.For categorical variables, descriptive statistics were used to describe the number of participants meeting screening criteria at each time point and those above and below prespecified cutpoints at baseline and week 52.These proportions are presented for those with complete data at each time point.The content of additional support visits provided to adolescents was independently coded by 2 authors, with agreement reached through discussion.

Depression Change Over Time
There was an initial reduction in CESDR-10 scores at week 4 (change in EMM IER, −4.23; 95% CI, −5.66 to −2.81; change in EMM CER, −4.99; 95% CI, −6.42 to −3.57) and maintained at week 52 (IER, −2.70; 95% CI, −4.95 to −0.45; CER, −3.87; 95% CI, −5.98 to −1.77) with no difference between groups (mean difference at week 52, 0.75; 95% CI, −1.86 to 3.37) (Table and Figure 2).There were no differences between those who completed and those who did not complete the study at any time point.3A).There were no differences between those who completed and those who did not complete the study at any time point.

EDE-Q Cutpoint
For those who completed the study and using an EDE-Q cutpoint of 2.7, a total of 51 adolescents (55%) remained below the cutpoint for the duration of the trial, and 8 (9%) remained above the cutpoint.There were 25 adolescents (27%) who had scores of 2.7 or greater at baseline but did not at week 52.total of 8 adolescents (9%; 3 IER, 5 CER) had an increase in scores from less than 2.7 at baseline to 2.7 or greater at week 52.Of the 8 adolescents who had increased scores, 4 had psychological comorbidity at baseline and were already engaged with, or referred to, external psychological support.The remaining 4 adolescents were not identified as having overt concerns with disordered eating, although strategies to manage overeating were discussed.

Outcome of Screening and Monitoring
Seventeen adolescents (12.1%) required support or referral for depression and/or disordered eating.

Screening Questionnaires
At baseline, 7 adolescents (5%) were identified as requiring support for disordered eating and 4 (2.8%) for depression.This was provided during the trial if they were not already engaged with external support.At week 52, 2 of the adolescents who were identified at baseline and 1 additional adolescent required referral for ongoing psychological support for disordered eating.

Dietetic Monitoring
Seven of 141 adolescents (5%, all female; 5 IER; 2 CER) were identified via dietetic monitoring as requiring additional support for disordered eating or body image concerns.They received a median of 3 (range, 1-9) additional sessions with the dietitian, clinical psychologist, and/or pediatrician, often with multidisciplinary input.Additional support was mostly provided during phase 3 of the study, when scheduled reviews reduced.Of those requiring additional support, 1 adolescent had mild body image concerns after a sports injury, with no disordered eating concerns.Two adolescents had support for meal skipping with mild body image concerns or secret eating, together with excess weighing, both in the context of requiring psychological support from baseline for anxiety and social

Adverse Events
Two adolescents had adverse events related to mental health and were withdrawn.One adolescent was withdrawn at week 14 after the reemergence of prior body image concerns and related poor self-esteem.The second adolescent was diagnosed with atypical anorexia nervosa associated with excess restriction and rapid weight loss (identified via monitoring protocol, as reported previously), despite transition to weight maintenance plan, and was withdrawn at week 22 as a severe adverse event.This occurred during the COVID-19 pandemic when review appointments were predominantly conducted virtually.Both adolescents received ongoing mental health care.

Discussion
The aim of these analyses was to evaluate changes in selfreported symptoms of depression, eating disorders, and binge eating during the Fast Track to Health trial.At baseline, over one-half of adolescents reported some symptoms of depression and/or eating disorders, with 21% reporting mild, moderate, or severe binge eating.Mean scores on the CESDR-10, EDE-Q, and BES scales were reduced at week 4, after the VLED phase, and maintained at week 52 for depression and eating disorder risk, with no difference between groups.The reduction in BES score at week 4 was only maintained in the IER group at 52 weeks.A proportion of adolescents (12.1%) required additional support for depression or disordered eating, identified through self-report questionnaires or dietetic monitoring, with support for appropriate early intervention.One adolescent developed an eating disorder, occurring during the COVID-19 lockdown restrictions, making it difficult to isolate the influence of the intervention.Overall, participation in intensive behavioral interventions using a prescriptive dietary approach did not worsen mental health or eating behaviors.Indeed, an improvement in self-report symptoms was seen alongside relative reductions in BMI, and improved blood pressure percentile, total cholesterol level, and triglyceride level in both groups. 27This is particularly important as the trial was predominantly undertaken during the COVID-19 pandemic, with intermittent lockdowns and other public health restrictions, which also impacted adolescent depression, disordered eating, and BMI. 6-100][41] Interestingly, the greatest reduction in scores occurred after the VLED phase, with no further reduction after transition to CER and IER.These outcomes have scarcely been measured dur-ing adolescent VLED interventions. 42Findings are consistent with adult VLED trials that show reduced binge eating, 43 but are in contrast to longitudinal studies in adolescents showing higher degrees of energy restriction, are associated with disordered eating. 23Given the short 4-week VLED phase, it is difficult to speculate why this initial improvement was seen.It is possible that the initial reduction in BMI, 27 together with frequent support, 11,26 improved mental well-being.However, eating disorders can take several years to emerge. 44Further research is needed on mental health effects of VLED interventions and maintenance beyond 1 year.
Although adolescence is a time of increased prevalence of depression and disordered eating, 3,5 and symptoms increased during COVID-19 pandemic, 6-9 few adolescents had increased symptoms during the trial.Those who experienced elevated symptoms often had other mental health concerns at enrollment.To our knowledge, few studies have reported the onset of eating disorders during adolescent obesity treatment.One trial 45 reported 2 cases of eating disorders from 125 adolescents receiving liraglutide, 3.0 mg, with lifestyle therapy, and another 46 reported 3 cases of binge eating disorder at 6-year follow-up of a behavioral program.More data are needed to understand the proportion of adolescents who may be at risk of eating disorders while receiving obesity treatment. 47Addressing participants' mental health within obesity programs, or enabling concurrent referral for psychological support, is important to support holistic health of adolescents during weight management. 1,48,49rescriptive dietary interventions, such as those used in this trial, are considered dieting behaviors that may precede eating disorder development. 12Dietary restraint scores increased during the intervention period, as might be expected due to the energy restriction targets, and returned to baseline at 52 weeks.Given that shape concern, weight concern, and depression all reduced during this time, the transient increase in dietary restraint is unlikely to represent ongoing problematic eating. 21,50Additionally, mean dietary restraint scores remained well below those of clinical samples of adolescents (4.00) 51 or adults (4.17) 52 with anorexia nervosa, even after the VLED phase of the intervention (2.18, IER group; 2.48, CER group) (Table ).At 52 weeks, mean dietary restraint scores had returned to baseline and were similar to community norms for adolescent girls. 53,54This suggests that dietary prescription in the context of a multidisciplinary care poses less risk than when undertaken without supervision in the community.We speculate that this is due to the multidisciplinary nature of the intervention and frequency of support provided by trained clinicians.However, further research is needed to directly compare different dietary interventions, no-treatment controls, and community samples to understand the association between different diets and eating disorder risk.
Use of self-report questionnaires to screen for depression and eating disorders, particularly at baseline, enabled adolescents requiring support to receive this early.Clinical monitoring for emergence of disordered eating during dietetic reviews enabled a subgroup of adolescents with potentially problematic behaviors to be identified and assessed.Of note, some symptoms identified during dietetic monitoring,

Strengths and Limitations
Our study has several strengths.This was the first, to our knowledge, in-depth study of depression and eatingdisorder symptoms in treatment-seeking, culturally diverse, Australian adolescents with obesity and concurrent complications.Structured screening and monitoring procedures were used with evidence-based, predefined cutpoints.This study also has limitations.Data presented are secondary analyses using self-report measures, therefore, results should be interpreted with caution.Two of the 4 subscales within the EDE-Q relate to body dissatisfaction, which is elevated in adolescents with obesity, and a third includes dietary restriction, which is a component of the intervention. 12,36Adolescents also reported misunderstanding questions on the EDE-Q.For example, participants reported occasions of overeating based on consuming a somewhat larger portion size at main meals compared with other family members or peers.Questions related to vomiting were sometimes misunderstood, eg, vomiting due to migraines or gastroesophageal reflux were reported but were not self-induced vomiting.Although the EDE-Q is validated for use in adolescents, a youth-specific version of this questionnaire 35 or more recently validated short eatingdisorder assessments for youth, 57,58 may improve screening in future trials.The trial was mainly undertaken during the COVID-19 pandemic.Therefore, an increase in self-reported depression and disordered eating during this time could be due to the intervention or the COVID-19 pandemic and differentiation between these is not possible.

Conclusions
In this randomized clinical trial, adolescents with obesity who were seeking treatment self-reported symptoms of depression, eating disorders, and/or binge eating.Symptoms reduced for most adolescents, concurrent with BMI reduction, highlighting a potential dual role of obesity treatment in improving mental as well as physical health.A subgroup of adolescents were identified as needing additional support for depression or disordered eating with some experiencing increased symptoms during the intervention.Approaches to screening and monitoring using a combination of questionnaires and clinician assessment appear important.Treatment practitioners should have mechanisms in place for identification and management of changes in psychopathology in adolescents seeking obesity treatment.Although obesity services may present an opportune time to support adolescent mental health, appropriate training, time, and resourcing is needing to implement this into practice.

Figure 1 .
Figure 1.Depression and Eating Disorder Screening and Monitoring Protocol

Figure 2 .Figure 3 .
Figure 2. Change in Symptoms of Depression Between Baseline and Week 52 12

Figure 4 .
Figure 4. Change in Shape Concern, Weight Concern, Eating Concerns, and Dietary Restraint Subscales of the Eating Disorder Examination Questionnaire (EDE-Q) Between Baseline and Week
Change Over Time At 4 weeks, there was a reduction in BES scores (IER, −4.38; 95% CI, −5.93 to −2.83; CER, −3.44; 95% CI, −5.00 to −1.88), maintained in IER at 52 weeks (−3.72; ing, with 8 (4 in each group) categorized as not binge eating at baseline moved to mild/moderate (n = 6) or severe (n = 2) binge eating at week 52.Two adolescents (CER group) increased from mild/moderate to severe binge eating.Of the 10 adolescents with an increase in binge eating, 5 had other mental health comorbidities identified at baseline or during the trial (eg, autism spectrum disorder, depression, anxiety, mild uncontrolled eating) and were referred for external psychological support or provided with additional support during the trial.On further assessment, 5 adolescents were not identified as having overt binge-eating symptoms; overeating behaviors were discussed with 3 adolescents.
, respectively.One adolescent had additional support to address body dissatisfaction and rigidity around diet and exercise.Three adolescents received additional support for restriction or rigidity around eating, with rapid weight loss; 2 of these adolescents had a medical review for amenorrhea, which resolved during the trial.