Reconstruction of retrospective cohorts on drug use using cross-sectional data.
Introduction.—In France, the repeated cross-sectional surveys ESCAPAD (Survey on Health and Behaviors during the Defense Readiness Call-up) and Baromètre santé measure the prevalence of health behaviors among adolescents and adults. However, these surveys are, in theory, not suited to the study of temporal phenomena such as trajectories of psychoactive substance use. This study reports on an exploration of transitions between the use of tobacco, alcohol, cannabis, and other illicit drugs (ADI) through the reconstruction of retrospective cohorts. Methods.—Data from ESCAPAD and the Health Barometers (2005 and 2010) were used. The reported ages of initiation for the various substances allowed for the definition of stages of use (levels of use for a given substance or sequences of use across different substances). The analysis of transitions between uses was conducted using survival analyses with multistate models that accounted for certain covariates. Results.- The collected data allowed for a precise description of psychoactive substance use processes and the exploration of several hypotheses, such as escalation theory and the common vulnerability model for addictions. The results, which were generally robust and reproducible across surveys, met certain criteria for causality: strength and stability of the observed associations, dose-response effects, consistency with the literature, and temporal relationships. Conclusion.—The reconstruction of retrospective cohorts using large samples from national prevalence surveys appears to be a reliable and cost-effective method for exploring the trajectories of health events. With regard to psychoactive substance use, recall bias appeared to be relatively limited. This retrospective approach cannot replace prospective cohort studies but may prove very useful, due to its good cost-effectiveness ratio, for exploring certain hypotheses or conducting international studies comparing several countries. Source: poster cited in the "Revue d'épidémiologie et de santé publique," Supplement 4, Issue 62.
Francophone Conference on Clinical Epidemiology (EPI-CLIN), Bordeaux, 14-16 mai 2014
Author(s): Mayet A, Legleye S, Beck F, Falissard B, Chau N
Publishing year: 2014
In relation to
Our latest news
news
Sexual Health Week 2026: Screening and Prevention Remain Essential
news
10 years of commitment and action in the service of public health
news