Svanes C, Lie R T, Kvåle G, Svanes K, Søreide O
Department of Surgery, Haukeland University Hospital, University of Bergen, Norway.
Am J Epidemiol. 1995 May 1;141(9):836-44. doi: 10.1093/oxfordjournals.aje.a117519.
Previous reports have shown that peptic ulcer mortality follows birth cohorts. To the authors' knowledge, temporal variation in ulcer incidence has not been studied. Therefore, they present incidence data for a defined area of western Norway where 1,312 patients born between 1845 and 1975 were treated for ulcer perforation between 1935 and 1990. A rise and subsequent fall in incidence was observed in successive birth cohorts for both sexes, with the highest incidence observed for males born between 1900 and 1919 and females born between 1920 and 1929. Age-period-cohort analyses based on Poisson regression techniques were adapted to provide a statistical tool for testing specific cohort and period effects. Age-cohort models without period effects explained the variations in incidence for both sexes and all ulcer locations, suggesting cohort-dependent etiology. A cohort pattern in prevalence of smoking partly explained the cohort pattern in perforation risks for both sexes. No period effects were seen that could be attributed to the increase in the sale of non-steroidal anti-inflammatory drugs, to the introduction of antibiotics around 1950, or to World War II. Susceptibility to ulcer perforation seems to follow birth cohorts, and major etiologic factors should be sought in prenatal life, in childhood, or in life-style patterns that follow birth cohorts.
