Background: The Norwegian "Women and cancer" study has shown that when combination hormone replacement products were used for more than five years the relative risk of breast cancer was almost tripled. The population attributable risk was 27 %. In a nested case-control study we have updated the risk estimates for hormone replacement therapy in post-menopausal women since 2002.
Material and method: We selected an age-matched control for each of the 589 women (aged 46-63) who were found in the period 2004-8 to have invasive breast cancer or cancer in situ. These women responded in 2003-6 to a questionnaire on menstruation and the use of hormone replacement therapy. The data from this and earlier questionnaires were compared and analysed using logistical regression.
Results: 226 (26 %) of the women were using hormone replacement at the time of the survey. The average time of use was ten years. Current users had a higher risk of breast cancer than never-users: adjusted OR 2.1 (95 % CI 1.5-3.0). The use of combination therapy for more than five years resulted in the highest risk: OR 3.0 (95 % CI 1.9-4.7). Earlier use of oestrogen products did not result in a higher risk. Neither oestradiol nor tibolon caused a statistically significant increase in risk. 232 cases of breast cancer (17 %) in women aged 45-64 could be attributed to the use of hormones.
Interpretation: Long-term use of combination products increases the risk of breast cancer. Relative risk estimates have undergone little change since the previous study, but population attributable risk has fallen along with the decline in new users.