Machine learning models for diagnosis and risk prediction in eating disorders, depression, and alcohol use disorder

Research output: Contribution to journalResearch articleContributedpeer-review

Contributors

  • Zuo Zhang - , King's College London (KCL), University of Birmingham (Author)
  • Lauren Robinson - , King's College London (KCL), South London and Maudsley NHS Foundation Trust, University of Oxford (Author)
  • Robert Whelan - , Trinity College Dublin (Author)
  • Lee Jollans - , Trinity College Dublin (Author)
  • Zijian Wang - , Donghua University (Author)
  • Frauke Nees - , Heidelberg University , University Hospital Schleswig-Holstein Campus Kiel (Author)
  • Congying Chu - , University of Chinese Academy of Sciences, CAS - Institute of Automation (Author)
  • Marina Bobou - , King's College London (KCL), University College London (Author)
  • Dongping Du - , King's College London (KCL), Virginia Tech College of Engineering (Author)
  • Ilinca Cristea - , King's College London (KCL) (Author)
  • Tobias Banaschewski - , Heidelberg University  (Author)
  • Gareth J. Barker - , King's College London (KCL) (Author)
  • Arun L.W. Bokde - , Trinity College Dublin (Author)
  • Antoine Grigis - , Université Paris-Saclay (Author)
  • Hugh Garavan - , University of Vermont (Author)
  • Andreas Heinz - , Berlin Institute of Health at Charité (Author)
  • Rüdiger Brühl - , Physikalisch-Technische Bundesanstalt (Author)
  • Jean Luc Martinot - , Université Paris-Saclay (Author)
  • Marie Laure Paillère Martinot - , Université Paris-Saclay, Sorbonne Université (Author)
  • Eric Artiges - , Université Paris-Saclay, EPS Barthélémy Durand (Author)
  • Dimitri Papadopoulos Orfanos - , Université Paris-Saclay (Author)
  • Luise Poustka - , University of Göttingen (Author)
  • Sarah Hohmann - , Heidelberg University  (Author)
  • Sabina Millenet - , Heidelberg University  (Author)
  • Juliane H. Fröhner - , Department of Psychiatry and Psychotherapy, Neuroimaging Center, Technische Universität Dresden (Author)
  • Michael N. Smolka - , Department of Psychiatry and Psychotherapy, Neuroimaging Center, Technische Universität Dresden (Author)
  • Nilakshi Vaidya - , Charité – Universitätsmedizin Berlin (Author)
  • Henrik Walter - , Berlin Institute of Health at Charité (Author)
  • Jeanne Winterer - , Charité – Universitätsmedizin Berlin, Free University of Berlin (Author)
  • M. John Broulidakis - , University of Southampton, Northeastern University (Author)
  • Betteke Maria van Noort - , MSB Medical School Berlin Hochschule für Gesundheit und Medizin (Author)
  • Argyris Stringaris - , University College London (Author)
  • Jani Penttilä - , Tampere University Hospital (Author)
  • Yvonne Grimmer - , Heidelberg University  (Author)
  • Corinna Insensee - , University of Göttingen (Author)
  • Andreas Becker - , University Medical Center Göttingen (Author)
  • Yuning Zhang - , University of Southampton (Author)
  • Sinead King - , King's College London (KCL), University of Galway, Royal College of Surgeons in Ireland (Author)
  • Julia Sinclair - , University of Southampton (Author)
  • Gunter Schumann - , Charité – Universitätsmedizin Berlin, Fudan University (Author)
  • Ulrike Schmidt - , King's College London (KCL), South London and Maudsley NHS Foundation Trust (Author)
  • Sylvane Desrivières - , King's College London (KCL) (Author)

Abstract

BACKGROUND: Early diagnosis and treatment of mental illnesses is hampered by the lack of reliable markers. This study used machine learning models to uncover diagnostic and risk prediction markers for eating disorders (EDs), major depressive disorder (MDD), and alcohol use disorder (AUD).

METHODS: Case-control samples (aged 18-25 years), including participants with Anorexia Nervosa (AN), Bulimia Nervosa (BN), MDD, AUD, and matched controls, were used for diagnostic classification. For risk prediction, we used a longitudinal population-based sample (IMAGEN study), assessing adolescents at ages 14, 16 and 19. Regularized logistic regression models incorporated broad data domains spanning psychopathology, personality, cognition, substance use, and environment.

RESULTS: The classification of EDs was highly accurate, even when excluding body mass index from the analysis. The area under the receiver operating characteristic curves (AUC-ROC [95 % CI]) reached 0.92 [0.86-0.97] for AN and 0.91 [0.85-0.96] for BN. The classification accuracies for MDD (0.91 [0.88-0.94]) and AUD (0.80 [0.74-0.85]) were also high. The models demonstrated high transdiagnostic potential, as those trained for EDs were also accurate in classifying AUD and MDD from healthy controls, and vice versa (AUC-ROCs, 0.75-0.93). Shared predictors, such as neuroticism, hopelessness, and symptoms of attention-deficit/hyperactivity disorder, were identified as reliable classifiers. In the longitudinal population sample, the models exhibited moderate performance in predicting the development of future ED symptoms (0.71 [0.67-0.75]), depressive symptoms (0.64 [0.60-0.68]), and harmful drinking (0.67 [0.64-0.70]).

CONCLUSIONS: Our findings demonstrate the potential of combining multi-domain data for precise diagnostic and risk prediction applications in psychiatry.

Details

Original languageEnglish
Number of pages11
JournalJournal of Affective Disorders
Publication statusE-pub ahead of print - 17 Dec 2024
Peer-reviewedYes

External IDs

Scopus 85212852225
ORCID /0000-0002-8493-6396/work/175758542
ORCID /0000-0001-5398-5569/work/175768379

Keywords

Keywords

  • Alcohol use disorder, Eating disorders, Major depressive disorder, Predictive modeling, Risk factors