Selection of Cases and Controls for the Nested Case-control Study within the JACC Study: the First-wave

In this paper, we describe the methodology of the case and control selection for the first-wave nested casecontrolstudy within the JACC Study. Among the subjects participating in the cohort, serum samples of 42,249subjects (including 39,242 subjects aged between 40 and 79 at the baseline) were suitable for biochemical analysis.We here selected those who had died by 1997 or who were diagnosed with cancer with sera until 1994 as cases.For each case, 3 to 4 controls with sera were randomly selected, with matching for gender, age (as near aspossible) and residential area. As a result, 3,144 cases and 10,661 controls (2,867 cases and 10,351 controls were40 to 79 years old at the baseline) were selected to measure serum IGF-I, IGF-II, IGFBP-3, TGF-β1 and sFasvalues and total SOD activity. Cases were older and more likely to be men than the JACC Study subjects.Moreover, they were much older than controls because of the age-dependence of susceptibility to death, especiallyamong men. There were more smokers among cases compared with controls, though drinkers at the baselinewere fewer. Among deceased cases, cancer was the leading cause of death, followed by cardiovascular diseases.Lung cancer was most frequent among deceased cancer cases and the next most common site was the stomach.The leading cause of cancer incidence was stomach cancer followed by lung cancer. Simple comparison of meansand distribution of IGF-I, IGF-II, IGFBP-3, TGF-β1, sFas and total SOD activity between cases and controlsrevealed total SOD activity and sFas levels of cases to be higher than controls, while for the other componentsthe opposite was found.