Protective and sustaining mechanisms in anorexia nervosa: a latent profile analysis

Introduction

Anorexia nervosa (AN) remains one of the most complex and life-threatening mental disorders, characterized by severe food restriction, intense fear of weight gain, and a distorted body image [1]. Despite extensive research, the heterogeneity of symptoms and the course of recovery pose challenges for both clinical understanding and personalization of treatment. Recent advancement in latent profile analysis (LPA) allows for the identification of distinct psychological subgroups, enabling a more nuanced view of the factors that protect against or perpetuate eating disorders. This increases the personalization of treatment and its effectiveness by adjusting the measures taken to the patient’s symptoms and requirements [2]. The aim of the study was to identify latent profiles of people with AN based on denial of the illness, quality of life, psychological well-being, severity of symptoms, depression and anxiety. The study also aimed to examine how protective and maintaining factors differentiate profiles and contribute to mental resilience or symptom persistence.

Methods

In this cross-sectional study, 68 women diagnosed with anorexia nervosa (DSM-5, ICD-10) participated (18–37 years, M = 22.56, SD = 4.49). Their illness duration ranged from 1 to 14 years, with average duration 3 years and 8 months. LPA was conducted to identify subgroups based on clinical symptoms (EAT-26, HADS, ZOB), illness denial and avoidance(IDQ), and psychological well-being (PWB). Total scores from each instrument were used in the analysis. Analyses were conducted in R (tidyLPA package). Models specifying one to four latent profiles were estimated using Gaussian finite mixture modeling, as implemented in the mclust package in R. Model selection was guided by the Akaike Information Criterion (AIC), Bayesian Information Criterion (BIC), entropy, profile sizes, and substantive interpretability. Differences between profiles in QoL (SF-36) and mental well-being (MHC-SF) were examined using Kruskal-Wallis tests. When omnibus effects were significant, post hoc pairwise comparisons were conducted using Wilcoxon rank-sum tests with Holm correction for multiple comparisons. Statistical significance was set at p < .05.

Results

The LPA revealed three profiles labeled as sustaining (22.1%), intermediate (61.8%), and protective (16.1%). The sustaining profile showed the highest symptom severity, strong denial and avoidance, and the lowest psychological well-being and quality of life. The protective profile displayed the lowest symptom levels, minimal denial, and the highest well-being and life quality across all domains. Post-hoc comparisons confirmed significant differences in psychological, social, and environmental aspects of quality of life, with the protective profile consistently outperforming the sustaining one.

Conclusion: The findings reveal substantial heterogeneity in psychological adaptation among individuals with AN, highlighting the need for individualized interventions focused on mechanisms. Identifying profiles among mental disorders such as AN should contribute to improving the therapeutic process by individualizing the person-centered approach. Through this classification, a patient correctly assigned to a subgroup, can receive more personalized therapeutic intervention which can accelerate treatment and reduce the likelihood of relapses by adjusting the forms of psychological support. It is a promising path to personalized treatment, supporting recovery processes and extending the concept of well-being beyond mere symptom remission.

References

. American Psychiatric Association, 2013. Diagnostic and statistical manual of mental disorders, fifth ed. American Psychiatric Association Publishing, Washington. https://doi.org/10.1176/appi.books.9780890425596
. Aloi, M., Carcione, A., Lo Coco, G., Rania, M., Carbone, E. A., de Filippis, R., Segura-Garcia, C., Liuzza, M. T., 2024. Metacognition as a transdiagnostic factor across eating disorders: a latent profile analysis study. Front. Psychology 15, 1391715. https://doi.org/10.3389/fpsyg.2024.1391715