|Titre :||The gateway hypothesis, common liability to addictions or the route of administration model a modelling process linking the three theories (2016)|
|Auteurs :||A. MAYET ; S. LEGLEYE ; F. BECK ; B. FALISSARD ; N. CHAU|
|Type de document :||Article : Périodique|
|Dans :||European Addiction Research (Vol.22, n°2, February 2016)|
|Article en page(s) :||107-117|
|Discipline :||EPI (Epidémiologie / Epidemiology)|
Thésaurus TOXIBASETHEORIE DE L'ESCALADE ; MODELE ; TABAC ; CANNABIS ; PRODUIT ILLICITE ; INITIATION ; VOIE D'ADMINISTRATION
BACKGROUND: The aim of this study was to describe the transitions between tobacco (T), cannabis (C) and other illicit drugs (OIDs) initiations, to simultaneously explore several substance use theories: gateway theory (GT), common liability model (CLM) and route of administration model (RAM).
METHODS: Data from 2 French nationwide surveys conducted in 2005 and 2010 were used (16,421 subjects aged 18-34). Using reported ages at initiations, we reconstituted a retrospective cohort describing all initiation sequences between T, C and OID. Transition probabilities between the substances were computed using a Markov multi-state model that also tested the effect of 2 latent variables (item response theory scores reflecting propensity for early onset and further substance use) on all transitions.
RESULTS: T initiation was associated with increased likelihood of subsequent C initiation, but the reverse relationship was also observed. While the most likely initiation sequence among subjects who initiated the 3 groups of substances was the 'gateway' sequence T -> C -> OID, this pattern was not associated with substance use propensity more than alternative sequences. Early use propensity was associated with the 'gateway' sequence but also with some alternative ones beginning with T, C or OID.
CONCLUSION: If the gateway sequence appears as the most likely pattern, in line with GT, the effects of early onset and substance use propensities were also observed for some alternative sequences, which is more in line with CLM. RAM could explain reciprocal interactions observed between T and C. This suggests shared influences of individual (personality traits) and environmental (substance availability, peer influence) characteristics.
|Domaine :||Drogues illicites / Illicit drugs ; Tabac / Tobacco|
|Refs biblio. :||55|
|Affiliation :||Centre d'épidémiologie et de santé publique des armées ; GSBdD Marseille Aubagne, Marseille, France|