Franceschi Sand Boyle P. Risk factors for gallstone disease requiring surgery. International Journal of Epidemiology 1991,20: 209-215. Risk factors for cholelithiasis were investigated in a case-control study conducted in northern Italy on 195 incident cases of gallstone disease requiring surgery (119 females, 76 males) and 1122 controls in hospital for acute, nondigestive tract, non-neoplastic conditions. There was no consistent association with socio-demographic indicators, such as marital status, education and social class nor with smoking and coffee consumption. The relative risk of cholelithiasis decreased with increasing alcohol consumption: compared with non-drinkers, the odds ratio (OR) was0.8 for one to three drinks per day and 0.5 for over three. A direct association was observed with measures of body weight: relative to leaner individuals, the multivariate ORs were 1.2, 2.1 and 2.4 for subsequent levels of body mass. These trends in risk were statistically significant, consistent in the two sexes, and not apparently modified by adjustment for major identified potential confounding factors. History of hepatitis and liver cirrhosis were reported more frequently by cases, but it is difficult to assess the role of recall bias on these risk factors. No association was found with diabetes, thyroid disease and several digestive tract disorders. For females, no consistent pattern of risk was observed up to four births, but women with five or more births had an OR of 2.9 (95% confidence interval (Cl) -1.1-7.3). The risk decreased with increasing age at first and last birth, both trends being statistically significant. Overweight and alcohol consumption were the most important risk and protective factors respectively for cholelithiasis requiring surgical intervention in this population.Cholelithiasis is one of the most frequent causes of morbidity and surgery in industrialized countries. The reported frequency of the disease and the composition of the stones varies in different countries. In Okinawa, Japan, the overall prevalence was 3.2%,' while the prevalence plus ten-year cumulative incidence in the Framingham study was 8.2% } Prevalence of over 10% were found in other countries or in selected ethnic groups, such as Pima Indians. 3 " 9 Some of these differences may be accounted for by different methods and accuracy of case ascertainment or by varying definition (i.e. clinical or surgical series as opposed to ultrasound or autopsy-based studies), but it is unlikely that this can explain all of the variation which is observed between different populations. 5 " 16 Knowledge of the causes of gallstone disease is poor.