Résumé
Objective To explore the risk factors on the symptoms of lumbar intervertebral disc herniation so as to develop a predictive model for the disease. Methods With a populationbased case-control study, 303 of 50 123 residents were diagnosed as having lumbar intervertebral disc herniation symptoms. 152 cases and 167 healthy controls, matched by gender and age, were randomly chosen as case and control groups. Questionnaires were used to collect information on the exposure to risk factors and logistic predictive model was then established. Results Through non-conditional logistic regression analysis, data showed that the positive family history of lumbar vertebra disorder, lumbar treatment or surgery, mental stress, acute low back injury, permanent work pose, and body mass index ≥23.0 kg/m2 were the risk factors among residents from the countryside. The area under the receiver operator characteristic curve of logistic predictive model was 0.809. When 0.4 was set as the classification cutoff, the total predictive correct rate, sensitivity, and specificity were 74.0%, 73.7%, and 74.3% respectively. Conclusion The occurrence of lumbar disk herniationcan in countryside population was affected by multi-variables including genetic and environmental, and could be predicted with the logistic regression model established by our group.The positive predictive results could be used to alarm the patients and doctors for prevention and treatment of the disease.