Early changes in depression predict outcomes of inpatient adolescent anorexia nervosa

Weight gained during treatment predicts 6-month body mass index in a large sample of patients with anorexia nervosa using ensemble machine learning

OBJECTIVE

This study used machine learning methods to analyze data on treatment outcomes from individuals with anorexia nervosa admitted to a specialized eating disorders treatment program.

METHODS

Of 368 individuals with anorexia nervosa (209 adolescents and 159 adults), 160 individuals had data available for a 6-month follow-up analysis. Participants were treated in a 6-day-per-week partial-hospital program. Participants were assessed for eating disorder-specific and non-specific psychopathology. The analyses used established machine learning procedures combined in an ensemble model from support vector machine learning, random forest prediction, and the elastic net regularized regression with an exploration (training; 75%) and confirmation (test; 25%) split of the data.

RESULTS

The models predicting body mass index (BMI) at 6-month follow-up explained a 28.6% variance in the training set (n = 120). The model had good performance in predicting 6-month BMI in the test dataset (n = 40), with predicted BMI significantly correlating with actual BMI (r = .51, p = 0.01). The change in BMI from admission to discharge was the most important predictor, strongly correlating with reported BMI at 6-month follow-up (r = .55). Behavioral variables were much less predictive of BMI outcome. Results were similar for z-transformed BMI in the adolescent-only group. Length of stay was most predictive of weight gain in treatment (r = .56) but did not predict longer-term BMI.

CONCLUSIONS

This study, using an agnostic ensemble machine learning approach in the largest to-date sample of individuals with anorexia nervosa, suggests that achieving weight gain goals in treatment predicts longer-term weight-related outcomes. Other potential predictors, personality, mood, or eating disorder-specific symptoms were relatively much less predictive.

PUBLIC SIGNIFICANCE

The results from this study indicate that the amount of weight gained during treatment predicts BMI 6 months after discharge from a high level of care. This suggests that patients require sufficient time in a higher level of care treatment to meet their specific weight goals and be able to maintain normal weight.

An overview and investigation of relapse predictors in anorexia nervosa: A systematic review and meta-analysis

OBJECTIVE

An extensive number of predictors has been examined across the literature to improve knowledge of relapse in anorexia nervosa (AN). These studies provide various recovery and relapse definitions, follow-up durations and relapse rates. The current study summarizes these values and predictors of relapse in AN in a review and meta-analysis.

METHOD

The study was executed according to PRISMA guidelines. Different databases were searched and studies in which participants did not receive an official clinical diagnosis were excluded. A quality analysis was performed using the National Institute of Health's Study Quality Assessment Tool. Random-effects meta-analyses were conducted to summarize data.

RESULTS

Definitions of relapse and recovery were diverse. During an average follow-up period of 31 months an average relapse rate of 37% was found. Predictive variables from 28 studies were grouped in six categories: age and sex, symptoms and behaviors, AN subtype and duration, weight or weight change, comorbidity, and personality. The studies were characterized by non-significant and contradictory results. Meta-analyses were performed for the predictors age, AN duration, pre-treatment BMI, post-treatment BMI and depression. These yielded significant effects for post-treatment BMI and depression: higher pre-treatment depression (SMD = .40 CI [.21-.59] and lower post-treatment BMI (SMD = -.35 CI [-.63 to -.07]) increased relapse chances in AN.

DISCUSSION

Our results emphasized a lack of sufficiently powered studies, consistent results, and robust findings. Solely post-treatment BMI and pre-treatment depression predicted relapse. Future research should use uniform definitions, larger samples and better designs, to improve our understanding of relapse in AN.

PUBLIC SIGNIFICANCE

Knowledge about predictors is important to understand high relapse rates. Our study performed a review and meta-analysis of relapse predictors in AN. Related to the heterogeneity in studies examining predictors, an overview of relapse and recovery definitions, follow-up durations and relapse rates for AN was provided. Significant effects were found for post-treatment BMI and pre-treatment depression. More studies with uniform definitions are needed to improve clinical implications.

Case report: Intramuscular ketamine or intranasal esketamine as a treatment in four patients with major depressive disorder and comorbid anorexia nervosa

Introduction

A comorbid diagnosis of a depressive disorder is a negative prognostic factor for individuals with AN, and novel treatments are needed to target depressive symptoms in this population. One emerging promising treatment for depressive disorders is ketamine, although there is less research investigating the use of ketamine for alleviating depression in people with AN.

Case report

This study reports on four patients with a lifetime diagnosis of AN and a comorbid diagnosis of major depressive disorder who received either intramuscular ketamine (n = 2) or intranasal esketamine (n = 2) treatment from a private psychiatric clinic. Depressive symptomatology (PHQ-9) was measured prior to (es)ketamine administration on every dosing session and adverse effects were recorded during and after dosing. All patients reported a subjective decrease in depression, although only those administered intranasal esketamine showed a reduction in PHQ-9 depression scores over time. Number of doses ranged from 3 to 23. All patients tolerated treatment well and no serious adverse effects emerged, however nausea/vomiting was experienced by one patient on one dosing session. Weight remained stable in all cases, although notably across all patients, weight at the beginning of treatment was within a "healthy" range.

Discussion

These findings suggest that (es)ketamine may reduce depressive symptoms in people with major depressive disorder and a comorbid diagnosis of AN. Future feasibility and pilot trials are warranted in order to elicit robust data on efficacy, acceptability, safety and tolerability.

Expressed emotion and treatment outcome in higher levels of care for eating disorders

OBJECTIVE

Expressed emotion (EE) among caregivers toward the affected offspring is a negative prognostic indicator for adolescent patients with eating disorders (EDs) in outpatient treatment. Less research has examined its impact on adolescents in higher levels of care (HLOC). The current study examined differences in caregiver EE according to the subtype of anorexia nervosa (AN) (restricting [AN-R] versus binge/purge [AN-BP]), and level of care (LOC). We also examined the main effects of baseline caregiver EE (emotional overinvolvement [EOI] or criticism), AN subtype, and their interaction on eating pathology and depression at discharge.

METHOD

Adolescent patients (N = 203) receiving treatment at HLOCs completed measures of ED pathology (Eating Disorder Examination-Questionnaire) and depression (Patient Health Questionnaire-9) at baseline and discharge, and one caregiver of each patient completed a measure of EE (Family Questionnaire) at baseline.

RESULTS

No differences in caregiver EE were found between patients with AN-R versus AN-BP, or relative to LOC. Caregiver EE did not predict outcome for ED symptoms or depression at discharge.

DISCUSSION

The impact of high caregiver EE may be less substantial at HLOCs than outpatient care given that caregivers are less involved in treatment at HLOCs. Future research is needed to determine if high caregiver EOI leads to poor treatment outcome for adolescents as it does for adults, or whether it is an appropriate expression of care for patients who are ill enough to require HLOC treatment.

PUBLIC SIGNIFICANCE STATEMENT

High caregiver EE was not found to predict treatment outcome for adolescents with eating disorders in higher levels of care (HLOCs), possibly due to the limited involvement of caregivers in HLOCs. However, patients step down to outpatient treatment, where high caregiver EE can have a significant negative impact on outcome. HLOCs should incorporate efforts to reduce high caregiver EE in anticipation of step-down to outpatient treatment.

Predictors of relapse in eating disorders: A meta-analysis

INTRODUCTION

Eating disorders (EDs) have high rates of relapse. However, it is still not clear which factors are the strongest predictors of ED relapse, and the extent to which predictors of relapse may vary due to study and individual differences.

OBJECTIVE

We conducted a meta-analysis to quantify and compare which factors predict relapse in EDs and evaluate various potential moderators of these relations (e.g., ED subtype, sample age, length of follow-up, timing of predictor assessment, relapse operationalization).

METHODS

A total of 35 papers (effects = 315) were included. We used a multilevel random-effects model to estimate summary study-level effect sizes, and multilevel mixed-effects models to examine moderator effects.

RESULTS

Higher level of care, having psychiatric comorbidity, and higher severity of ED psychopathology were associated with higher odds of relapse. Higher leptin, higher meal energy density/variety, higher motivation for change, higher body mass index/weight/body fat, better response to treatment, anorexia nervosa-restricting (vs. anorexia nervosa-binge purge) subtype diagnosis, and older age of ED onset were associated with lower odds of relapse. Several moderators were identified.

DISCUSSION

A variety of variables can predict ED relapse. Furthermore, predictors of ED relapse vary among ED subtypes, sample ages, lengths of follow-up, timing of predictor assessments, and relapse operationalization. Future research should identify the mechanisms by which these variables may contribute to relapse.

Treatment response in children and adolescents with anorexia nervosa: a naturalistic, case-control study

PURPOSE

Although a few recent articles describe adults with treatment-resistant anorexia nervosa (TR-AN), no study addresses the specific features of subjects not responding to treatment in the developmental age. This study reports on the clinical and psychopathological variables that distinguish children and adolescents who did not respond to treatment (here "TR-AN") from good-outcome controls, in a multidisciplinary hospital treatment setting.

METHODS

Naturalistic, case-control study conducted on individuals showing lack of response to treatment and good-outcome controls. TR-AN was defined as two or more incomplete admissions and no complete admissions, consistently with studies in adults. Good-outcome was defined as complete first admission, availability for follow-up visit after 6 months, and maintaining at follow-up a %BMI > 70% in the absence of binging or purging in the preceding 3 months. Psychopathological (Eating Disorders Inventory-3 EDI-3; Beck Depression Inventory-II), clinical, and treatment variables at admission were compared. Significant differences in the univariate analyses were included in an exploratory binary logistic regression.

RESULTS

Seventy-six patients (30 TR-AN, 46 good-outcome AN controls) were enrolled (mean age 14.9 ± 1.9 years, F = 94.7%). TR-AN individuals had a higher age at admission and higher EDI-3 Eating Disorder Risk (EDRC) scores, were treated less frequently with a nasogastric tube (NGT), and achieved a lower BMI improvement at discharge than good-outcome controls. A predictive model for TR-AN status was found (X² = 19.116; Nagelkerke-R² = 0.478, p < 0.001), and age at admission (OR = 0.460, p = 0.019), EDI-3 EDRC (OR = 0.938, p = 0.043), and NGT (OR = 8.003, p = 0.019) were associated with a TR-AN status.

CONCLUSIONS

This is the first report on the psychopathological and clinical characteristics of children and adolescents not responding to treatment. These patients showed higher age and eating disorder scores, and were less frequently fed with NGT than controls. Despite the multiple incomplete admissions of our subjects, the short included follow-up limits the possibility for direct comparisons with adult samples of treatment-resistant patients. Thus, the specific features of children and adolescents with TR-AN should be assessed in longitudinal studies.

LEVEL OF EVIDENCE

III, Observational, case-control study.

Comorbidity Between Anorexia Nervosa and Depressive Disorder: A Narrative Review

OBJECTIVE

Anorexia nervosa (AN) is a disorder with a significantly high prevalence. Affective disorders, including depressive disorder (DD), often coexist with this eating disorder (ED). In this review we will focus on its prevalence, associated vulnerability factors, clinical manifestations, possible etiological factors and its prognosis.

METHODS

A bibliographic search was carried out in the PubMed database selecting those articles that approached the subject of comorbidity between AN and DD. The search was limited to articles published from January 1990 to December 2021.

RESULTS

Of the 1891 abstracts reviewed, 33 studies met inclusion criteria. The prevalence of this comorbidity was extremely variable between studies, which exposed their heterogeneity. As to symptomatology this comorbidity presents itself with more severity, greater expression of psychological traits and greater cognitive impairment. Certain personality traits are postulated as vulnerability factors. Genetic factors such as neurochemicals seem to be involved in its pathogenesis.

CONCLUSION

The comorbidity between DD and ED have important influence in its symptomatic expression, severity and prognosis. Some of the analyzed studies provide consistent data, but there are others that are contradictory. It would be necessary to increase the number of studies and use a unified methodology.

Variability in the antioxidant MSRA gene affects the psychopathology of patients with anorexia nervosa

The objective is to determine whether variability in the MSRA gene, related to obesity and several psychiatric conditions, may be relevant for psychopathological symptoms common in Anorexia Nervosa (AN) and/or for the susceptibility to the disorder. A total of 629 women (233 AN patients and 396 controls) were genotyped for 14 tag-SNPs. Psychometric evaluation was performed with the EDI-2 and SCL-90R questionnaires. Genetic associations were carried out by logistic regression controlling for age and adjusting for multiple comparisons (FDR method). Two tag-SNPs, rs11249969 and rs81442 (with a pairwise r2 value of 0.41), were associated with the global EDI-2 score, which measures EDI-related psychopathology (adjusted FDR-q = 0.02 and 0.04, respectively). Moreover, rs81442 significantly modulated all the scales of the SCL-90R test that evaluates general psychopathology (FDR-q values ranged from 4.1E-04 to 0.011). A sliding-window analysis using adjacent 3-SNP haplotypes revealed a proximal region of the MSRA gene spanning 187.8 Kbp whose variability deeply affected psychopathological symptoms of the AN patients. Depression was the symptom that showed the strongest association with any of the constructed haplotypes (FDR-q = 3.60E-06). No variants were found to be linked to AN risk or anthropometric parameters in patients or controls. Variability in the MSRA gene locus modulates psychopathology often presented by AN patients.

The effect of group cognitive behavior therapy on Chinese patients with anorexia nervosa: an open label trial

BACKGROUND

The high cost of treatment for anorexia nervosa (AN) and lack of trained specialists have resulted in limited accessibility of effective treatment to patients with AN, which is particularly problematic in China. To increase the accessibility of evidence-based treatment and reduce the cost of treatment, this study aimed to explore the feasibility and efficacy of group cognitive behavior therapy (G-CBT) adapted from enhanced cognitive behavior therapy for eating disorders (CBT-E) in Chinese AN patients.

METHOD

A total of 78 patients with AN were assigned to G-CBT or individual outpatient treatment (IOT) and received three months of treatment for AN in each condition. Measures of eating pathology, depression and anxiety were administrated to both intervention groups at three time points: baseline, one month of treatment, and end of treatment; results were compared between groups and over time.

RESULTS

There were 70 participants included in the final analysis. Both G-CBT and IOT groups showed significant improvement in eating pathology and associated psychopathology (ps < .001) over the course of treatment, but no significant difference in symptom improvement was found between the two groups (ps > .05). G-CBT resulted in additional significant improvement in ED psychopathology over the last two months of treatment, and its overall therapeutic effect was influenced by baseline weight and early symptom improvement.

CONCLUSION

Preliminary findings from this open label trial suggest that G-CBT adapted from CBT-E is feasible in an outpatient setting and as effective as IOT in facilitating weight regain and reducing psychopathology in Chinese AN patients with little evidence for the superiority of either intervention.

TRIAL REGISTRATION

The current study was registered at clinical trials.gov on September 23, 2018 (registration number NCT03684239). People with anorexia nervosa (AN) are known to be unmotivated for treatment and prone to relapse. Recovery from AN often needs intensive, long-term treatment from a specialized multidisciplinary team, which is not accessible for most people in China. Given the increasing incidence of AN and lack of eating disorder (ED) specialists in China, it is important to develop short-term cost-effective treatments for AN. In this study, we explored the feasibility and efficacy of group cognitive behavior therapy (G-CBT) adapted from enhanced cognitive behavior therapy (CBT-E) for people with AN from China. We found that G-CBT was as effective as individual outpatient treatment (IOT) typically provided to AN patients at the research site in facilitating weight regain, improving eating behaviors, and reducing ED and other symptoms. We also found that patients receiving G-CBT made more improvements in cognitive symptoms of the ED, which might help maintain treatment gains and prevent relapse in the long run. This potential long-term advantage of G-CBT needs to be verified in long-term follow-up.

Short-Term Outcome of Inpatient Treatment for Adolescents with Anorexia Nervosa Using DSM-5 Remission Criteria

This study evaluated the short-term outcome of a multimodal inpatient treatment concept for adolescents with anorexia nervosa (AN). In this prospective observational study, a cohort of 126 female adolescents with AN (age range: 11–17, mean age: 14.83) was longitudinally followed from admission to discharge (average duration of stay: 77 days). We used gold-standard clinical interviews and self-report data, as well as DSM-5 remission criteria, to evaluate the treatment outcome. From admission to discharge, body-mass-index (BMI) significantly improved by 2.6 kg/m². Data from clinical interviews and self-reports yielded similar improvements in restraint eating and eating concerns (large effects). Lower effects were observed for variables assessing weight/shape concerns and drive for thinness. At discharge, 23.2% of patients showed full remission of AN, 31.3% partial remission, and 45.5% no remission according to DSM-5 criteria. Differences in remission groups were found regarding AN severity, age at admission, and use of antidepressant medication. Living with both parents, longer duration of inpatient treatment and the use of antipsychotic medication were significantly associated with higher BMI change. The findings provide evidence for the short-term effectiveness of our inpatient treatment concept. We recommend using DSM-5 based remission criteria to evaluate the treatment outcome to improve the comparability of studies.

Post-hospitalization Daycare Treatment for Adolescents With Eating Disorders

Background: There are several possible facilities for the treatment of eating disorders (EDs). Specifically, there is the issue of the use of specialized daycare and ambulatory services over inpatient settings and the place of daycare programs following inpatient treatment. Aim: We sought to examine the contribution of post-hospitalization daycare program to the treatment of adolescents hospitalized with an ED. Methods: We assessed 61 female adolescents hospitalized with an ED. All but three were diagnosed with clinical or subthreshold anorexia nervosa (AN). Three were diagnosed with bulimia nervosa. Thirty-seven patients continued with a post-hospitalization daycare program for at least 5 months, whereas 24 did not enter or were enrolled in the program for <5 months. Patients completed on admission to, and discharge from, inpatient treatment self-rating questionnaires assessing ED-related symptoms, body-related attitudes and behaviors, and depression and anxiety. Social functioning was assessed 1 year from discharge using open-ended questions. One-year ED outcome was evaluated according to the patients' body mass index (BMI) and according to composite remission criteria, assessed with a standardized semistructured interview. To be remitted from an ED, patients were required to maintain a stable weight, to have regular menstrual cycles, and not to engage in binging, purging, and restricting behaviors for at least eight consecutive weeks before their assessment. Results: BMI was within normal range at follow-up, whether completing or not completing daycare treatment, and around 75% of the patients had menstrual cycles. By contrast, when using comprehensive composite remission criteria, less than a quarter of former inpatients not entering/not completing daycare program achieved remission vs. almost a half of the completers. In addition, a greater percentage of completers continued with psychotherapy following discharge. Fifty percent of both groups showed good post-discharge social functioning. No between-group differences were found in the BMI and the scores of the self-rating questionnaires at admission to, and discharge from, inpatient treatment. Conclusion: Adolescent females with EDs can maintain a normal-range BMI from discharge to 1-year follow-up, even if not completing daycare treatment. By contrast, completion of a post-hospitalization daycare program may improve the 1-year follow-up ED-related outcome of former ED inpatients.