Background: In Norway and Sweden, the introduction of mammography screening programmes has been associated with about a 50% increase in breast cancer incidence for the screened age groups and almost stable incidence in higher age groups. This suggests that mammography screening results in a substantial degree of overdiagnosis.
Material and methods: In a prospective cohort study of the Norwegian mammography programme we recorded the incidence of breast cancer among women aged 50 to 69 years offered mammography screening as well as the incidence of interval cancer among those attending, and breast cancer among those not attending. We used these data to calculate the incidence of cancer detected by screening among those screened.
Results: The incidence of invasive cancer at second and third screening is 2.4 times higher than expected. We estimated that only 42% of the screening-detected invasive cancers would have developed into clinical disease in the absence of a screening programme. When ductal carcinoma in situ is added, only 34% of the screening-detected cancers would have developed into clinical disease.
Interpretation: Two thirds of breast cancers detected by screening are overdiagnosed.