- Solar and artificial UV exposure: We study UV exposure from natural sun and solariums in relation to incidence of cancer and mortality. This is collaboration with the Norwegian Women and Cancer study, NOWAC, University of Tromsø, and the Women’s Lifestyle and Health study, WLH, Karolinska Institutet.
- Causal inference in cancer epidemiology: The Norwegian cohort studies and health registries represent a unique treasure, which needs to be more extensively exploited with efficient modern analytical tools. We will in collaboration with the research group ‘Causal inference methods’ use techniques from modern causal inference to study the complex associations between UV exposure and melanoma.
- Bias in exposure-disease associations in epidemological studies. Exposure assessment is an important methodological problem in epidemiology. We estimate misclassification of exposure measured by reproducibility and recall bias, and study bias in risk estimates for exposure-disease associations.
- Optimal use of statistical methods in medical research. Clinical studies frequently involve non-trivial designs that give complex data structures. We study modern statistical methods and will show that by extracting more information from the data, we get more insight into the clinical problem.
- Diet, body mass index, weight change and physical activity in relation to risk of cancer. In the Norwegian County Study (Norwegian Institute of Public Health (NIPH)) , ~65 000 participants have been screened up to three times since 1974 (weight, height, blood samples, smoking habits, recreational and occupational physical activity, food frequency questionnaire). The main analytic tool is stratified Cox regression models with time-dependent covariates.
- Risk factors of cancer and cardiovascular disease. In the Norwegian County Study (see above), we have calculated dietary intake of trans fatty acids - from partially hydrogenated vegetable oils, partially hydrogenated fish oils and ruminant fat - and study the associations with mortality from cardiovascular disease and risk of cancer.
- Nausea and vomiting in pregnancy: Diet, birth outcomes and maternal biological markers in the Norwegian Mother and Child Cohort Study (MoBa). We study demographic factors and dietary changes of women who experience nausea alone or NVP in gestation as compared to those without symptoms. We also study what impact nausea or NVP have upon pregnancy and birth outcomes.