[Critical review of the epidemiology of backache]

High frequency occurrence of back pain and the magnitude of its impact on society explain the large number of epidemiologic studies. Most investigators have considered back pain as a whole, without reference to different etiological types of back pain, probably for lack of an available classification of these types. Prevalence of back pain in general populations varies between 14 and 45% and annual incidence around 6%. Risk factors for low back pain are often social or cultural factors: smoking, driving, psychological stress. These factors seem to be far from the starting point of the disease process. Vagueness of case descriptions may explain in part the disappointing results of analytic surveys. Indistinctness of risk factors, especially workplace factors, is the principal reason for the poor results of intervention epidemiology: very few primary prevention programs and no educational programs ("low back school") have been shown to be really effective.