RESUMEN
Because the visibility of homeless persons congregating in urban areas has increased since the 1980's, the relationship between homelessness and mental illness has caused more and more concern. A multitude of epidemiological surveys have been organized in Great Britain, Australia, Canada, and mainly in United States, and have attempted to evaluate scientifically the psychiatric morbidity of this population. This literature review reveals disparity of epidemiological methods in assessing the type and extent of mental illness among homeless adults. The lack of consensual definition of homelessness, the choice of different settings in which the research is organized (street, health centres, shelters), and the use of diverse instruments of psychiatric evaluation (diagnosis by clinician, by scale or by structured diagnostic interview) lead to a great disparity of the results. Thus, 1/3 of the homeless adults had prior history of psychiatric hospitalisation. Rates of psychosis range to 70% and it is estimated that 4% to 74% of the homeless persons suffer from affective disorders. Substance abuse disorder remains a problem for a significant number of these individuals, with a high frequency of dual diagnosis. Such divergent data highlight the anglo-saxon debate between those who accuse desinstitutionnalisation as a reason of homelessness, and those who blame the socioeconomic background.