Oral contraceptive use
During the past year we have published results partly confirming the findings of the Collaborative group showing an increased risk for current use, however, we have also both in the Norwegian-Swedish cohort and the total NOWAC cohort found a lasting increased risk related to total duration. A more detailed analysis showed that the risk was more due to estrogens than to gestagens based on our very detailed exposure information. A study of the interaction between OCs and alcohol consumption found more of an additive than a multiplicative effect, comparable to what we have found previously for childbearing and breast cancer. On the other hand, we found a very clear negative association with total duration up till 15 years exposure and risk of ovarian cancer. In addition, we have published a more comprehensive analysis of breast, ovary, endometrial, colorectal and total cancer incidence which strongly showed the importance in a public health perspective to deal with more than one single outcome. We have started to study the risk related to other sites of cancer as colo-rectal cancer and malignant melanoma.

Hormonal replacement therapy (HRT)
The use of HRTs increased very rapidly throughout the nineties in Norway. The sex hormones used have been different from those used in the US. We have had no brands with equine estrogens on the marked and little use of the gestagen medroxy progesterone acetat (MPA) in contrast to the participants in Women’s Health Initiative.
We found that current use of HRT doubled the risk of breast cancer compared to non-use (Bakken et al. 2001). Since we have detailed information about the brands used, we found that women using brands with fixed combination of estrogen-gestagen had a higher breast cancer risk than those who used brands with sequential estrogen-gestagen combination. The attributable risk was estimated to be 27% partly as a consequence of high prevalence of use and the type of HRT. The strong increase in incidence of breast cancer in the 90’ies in Norway was more due to the use of HRT than the introduction of mammographic screening.

Diet and cancer
Our analysis of diet and cancer is mainly performed as part of the EPIC collaboration since the statistical power in NOWAC for the dietary information still is to weak (information collected from 1996 and onwards). Studies in EPIC have shown a decreased risk of colo-rectal cancer with increasing consumption of fibre (Bingham et al. 2003), increased risk with high intake of meat, specially processed meat, and decreased risk with high intake of fish (Norat et al. 2005), and a decreased risk for lung cancer with increasing fruit consumption (Miller et al. 2003).

We participate in the collaborative work with the European nutritional database (ENDB). In EPIC we were responsible for the analysis of fish and breast cancer. The statistical analysis was completed based on more than 7,000 cases of breast cancer. The analyses revealed no connection between fish consumption and breast cancer risk (Engeset et al. 2006). In NOWAC we have run and submitted a factor and cluster analysis of the dietary patterns in the cohort (Engeset et al. 2004). In the future we will look at the predictive value of pattern of consumption compared to single foods analysis. We have also as the first cohort estimated the overall effect) of consumption of farmed fatty fish (salmon and trout) on cancer risk and mortality in response to published analysis of persistent organic pollutants in wild and farmed fish (Lund et al. 2004).

Sun habits and pigmentation
As the first large prospective study we published on the relationship between sun habits, pigmentation and malignant melanoma (Veierod et al. 2003). The ongoing discussion on solarium habits and melanoma will have to wait for two more years of follow-up.

Social class
The results from our study suggest a clear positive gradient in risk for breast cancer by level of education, which can be fully explained by established breast cancer risk factors like parity, age at first birth, body mass index, height, age at menarche, menopausal status, use of OCs and alcohol consumption. We are continuing this work with analysis of the influence of social class on other major cancer sites. In addition, survival analysis of social class and risk of breast cancer are planned.

We have obtained experience with D-vitamin analysis in blood and estimation of dietary intake (Brustad et al. 2004). In Norway, dependent on the place of living D-vitamin production in the skin ceases during winter – “D-vitamin winter”. Low D-vitamin production is partly compensated by the high dietary intake of D-vitamin through fatty fish and cod liver oil. We have information available to estimate both sources of D-vitamin and to validate the information through blood samples. We have a unique position to study the relationship between cancer and D-vitamin through these natural differences in sun production and high dietary consumption. We plan to start the analyses of D-vitamin status and breast and colon cancer in the near future.

Thyroid cancer
Although thyroid cancer is rare Norway is a high incidence country. We know that radioactive fallout could be responsible for some of the early increase among children, but no other major risk factors have been found. The research group International collaboration of thyroid cancer, founded by NCI did a data collection of all previous case-control studies, but even this common effort failed to find risk factors that could explain the disease. We have access to all cases in EPIC (by 2003 about 370) and in NOWAC (about 100 incident cases).