Dados do Trabalho

Título

Influence of sleep, socioeconomic and psychiatric factors in the risk for consumption of tobacco and alcohol

Introdução

Evidence shows that sleep, psychiatric and sociodemographic factors can separately affect substance use. Considering the contribution of each of these elements, it becomes necessary to investigate its adjoined impact on the risk on substance use.

Objetivo

This study aimed at building models that include the three factors simultaneously, in order to verify how they contribute to the risk of substance use.

Métodos

Volunteers from the EPISONO 2007, which was performed to investigate sleep patterns in the inhabitants of the city of São Paulo, Brazil, answered validated questionnaires regarding general health and underwent a polysomnographic exam. For the analyses here shown, scores in the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST) were evaluated considering polysomnographic findings and answers to questionnaires assessing sleep, psychiatric and socioeconomic factors. Using structured equation modelling (SEM) a theoretical model was built, uniting construct variables corresponding to sleep, mental health and socioeconomic factors and linking them to ASSIST scores. Additional logistic regression analyses were performed to observe which variables pertaining to sleep, psychiatric and socioeconomic factors could be considered associated factors to substance use risk.

Resultados

793 EPISONO participants (mean age=43.4 years; 452 women) were included. A successful model was built for tobacco consumption risk. It was found that subjective sleep quality, psychiatric symptoms and socioeconomic status influenced ASSIST scores for tobacco consumption. The logistic regression analyses also identified effects of anxiety (OR=1.07 [1.01 to 1.13]), on the alcohol consumption risk.

Conclusões

With our theoretical model, we successfully explained how sleep, mental health and social factors could form a network affecting tobacco consumption risk. However, models for other substances could not be built. This impossibility may be connected to the nature of our sample, which was a screening of the general population of large urban center, and therefore include a limited amount of individuals with a higher risk in substance use. Using logistic regression, effects of mental health factors were seen acting on alcohol consumption risk, suggesting more similar studies are performed using specific population samples.

Palavras -chave

Sleep; sociodemographic factors; psychiatric symptoms; substance use