Venue: Amphithéâtre du Centre hospitalier Charles Perrens
Defense in french
Stability of the psychometric characteristics of the DSM-5 diagnostic criteria for substance use disorders and the predictive value of craving for substance use according to treatment context. How does craving differ from other diagnostic criteria?
(Stabilité psychométrique des critères diagnostiques du DSM-5 pour les troubles de l’usage de substances et valeur prédictive du craving sur l’usage de substance selon le contexte de soin. Dans quelle mesure le craving diffère-t-il des autres critères diagnostiques ?)
Charlotte Kervran
Team Auriacombe (Addiction)
SanPsy
Thesis supervisor: Marc Auriacombe
Abstract
Introduction
Loss of control of use of a rewarding substance is the core of substance use disorders. The psychometric validity of the DSM-5 diagnostic criteria for substance use disorders was tested for several substances, but the addition of craving as a diagnostic criterion is still debated. A model of addiction has been proposed in which craving intensity fluctuation over time predicts later use and relapse supporting its addition as a diagnostic criterion of the loss of control dimension. However, these previous studies focused only on samples of substance users attempting to reduce use or abstain.
The main objective of this thesis project is to test the psychometric stability of the diagnostic criteria of the DSM-5 substance use disorders according to treatment context, paying particular attention to the role of craving as a diagnostic criterion and as a dynamic dimension. Substance users recruited in harm reduction programs (HR), a substance-use friendly environment and substance users seeking treatment (TX) to reduce or abstain from use were compared.
Methods
The stability of the 11 diagnostic criteria for alcohol, cocaine, opiates, cannabis and tobacco use disorders was tested. The Item Response Theory (IRT) and Differential Item Functioning (DIF) analyses made it possible to study the behavior of the diagnostic criteria between them and in relation to the underlying latent trait of the use disorder, and to test the functioning of the criteria between the two populations. The applicability of the model of addiction was tested among HR subjects using the EMA (Ecological Momentary Assessment), collecting craving intensity and substance use repeatedly in daily life, to test the prospective link using the HLM (Hierarchical Linear Model) method.
Results
Regardless of substances used and treatment context, the diagnostic criteria assessing loss of control worked correctly and more particularly the craving criterion, which was less the case for the diagnostic criteria related to consequences of use. As a diagnostic criterion, craving seems to be the most selective, because it fits better to the one-factor model, it is among the least difficult to be endorsed (more frequent), is the most discriminant and captures subjects with mild substance use disorders. The EMA study highlights the predictive role of craving on use, regardless of the context of use.
Conclusion
This work questions the relevance of the criteria of consequences of use, because of their psychometric characteristics, which are less effective than the loss of control criteria, and their low inter-sample stability. This work highlights the role of craving as a driving force for use and supports its interest as a diagnostic and prognostic measure of loss of control. These results could make it a specific and early marker of addiction that can be easily detected and used in clinical practice, and justify its being a target for therapeutic interventions.
Keywords: Addiction, craving, behavioral neuroscience, diagnosis, diagnostic criteria
Publication
Jury
- M. BENYAMINA Amine – Professeur – Univ. Paris Sud – Rapporteur
- M. CASAS Miquel – Professeur – Univ. Barcelone – Rapporteur
- M. AMAD Ali – Maître de Conf. – Univ. Lille – Examinateur
- M. AHMED Serge – Dir. de Recherche – Univ. Bordeaux – Examinateur
- M. ALLA François – Professeur – Univ. Bordeaux – Examinateur