[Sociodemographic and mental health risk factors: Global and local conditions according to the Survey results "mental health in the general population" in 18 international locations].

Introduction: Mental Health in General Population survey (MHGP) is a socio-anthropological and epidemiological multicentre research carried out by the WHO Collaborating Centre for Research and Training in Mental Health (Lille, France). It assessed the prevalence of major mental disorders in the general population with 15,747 people aged 18 years and above in 18 sites worldwide: 6 European sites, 4 sites in the Maghreb, 4 sites in the Indian Ocean, 2 sites in the Caribbean and two sites in the Pacific Ocean.

Objectives: To assess the risk of the presence of at least one mental disorder in terms of sociodemographic factors (gender, age, marital status, family income, education level, professional activity, religious practice and social isolation) and location (zone [Europe, North Africa, Indian Ocean, Caribbean Islands and the Indian Ocean Islands] and "investigation site").

Methodology: Statistical analysis was performed using data collected in 18 international sites of the MHGP survey. Logistic regression was used to model the relationship between sociodemographic and geographic factors and the presence of at least one disorder (mood disorder, anxiety disorder, psychotic disorder, abuse or dependence on alcohol or drugs), evaluated with the Mini International Neuropsychiatric Interview (MINI) diagnostic questionnaire.

Results: The prevalence of mental disorders rates vary among 18 sites, ranging from 15.5 % (Andorra) to 60 % (Algiers). The adjusted global epidemiological model (18 cluster sites) confirms a decreased level of risk of at least a psychiatric pathology due to a favorable sociodemographic "profile": marital status (married), family income (higher), age group (60 years and above), educational level (university), gender (male), practice of religion (among believers), employment (exercised). Analysis at geographical situation's level confirms existence of sub socio-geo-demographic models differentiated by ranking and levels of variables' modalities. Classification of variables and their modalities is clearly differentiated not only between 5 zones, but also within each of them depending on the sites that comprise it. This produces differentiated models for each of the 18 survey sites.

Conclusions: The impact of sociodemographic risk factors on mental health is confirmed regardless of World region. However, the implementation of action plans for the prevention of mental disorders requires a detailed understanding of people's needs in terms of the disorder's prevalence, nature and strength of risk factors, at regional and local levels. This observation provides incentives to develop this research axis in world francophone and Latin speaking areas. These epidemiological results can be refined thanks to the data collected in the MHGP surveys about each mental disorder and comorbid conditions, the recourse of populations to assistance or care, as well as results of the socio-anthropological axis.