Breast cancer and colon cancer incidence in females in Scotland, 1960-84.

In view of the suggested similarity in risk factors for breast cancer and colon cancer in women, an analysis was undertaken to investigate age, period, and cohort influences on the risk of these diseases by use of incidence data for Scotland. Statistical modeling of data for the period 1960-84 revealed that both the age effects and the cohort effects were different for each site. While younger cohorts apparently had signs of reduced risk of both diseases, reduced risk was only apparent among the very youngest cohorts in the case of breast cancer, whereas there has been a gradual reduction in the risk of colon cancer among successive cohorts born since the beginning of the century. Clemmesen's hook was demonstrated in the age-incidence curve of breast cancer, which was clearly shown to have two different slopes: among premenopausal and among postmenopausal women. The declining risk of breast cancer in younger cohorts appeared contradictory to their apparently "high-risk" profile (in terms of present knowledge of etiologic factors) and suggested the possible existence of a factor that protects against breast cancer, at least at younger ages, but that remains, as yet, unidentified.