Socioeconomic, Reproductive and Lifestyle Factors and Risk of Breast Cancer in Women: Registry-based studies in Finland and other

Female breast cancer is the leading cancer and a global public health problem. Although extensive research work has been done on epidemiology of breast cancer, a gap has been identified in implementing a multifactorial approach with many etiological factors simultaneously included in analyses of large study populations. This doctoral dissertation aimed to evaluate breast cancer risk with a series of typical study settings in etiological research, starting with the search for etiological clues of risk factors with indirect indicators (Study I in this thesis), then concentrating on one direct risk factor at a time (Study II) or a specific subset of the population (Study III). Finally, Study IV combines many relevant elements of the research question available from the registries in Finland.

Study I of this thesis demonstrates the possibilities and difficulties in exploring and interpreting the etiology of breast cancer from the mortality pattern of breast cancer patients. We observed excess mortality for diseases of the circulatory system, as well as for cancers of the stomach, lung, and ovary. It is not plausible that this excess mortality would be to a large extent related to breast cancer or its treatment, and therefore the findings were interpreted as suggestions to joint etiology. Factors associated with causes of death among breast cancer patients, such as reproductive history, lifestyle factors including alcohol consumption, smoking, physical inactivity, and the use of hormonal therapy are also risk factors for breast cancer. Genetic mutations causing breast cancer may also increase risk of stomach cancer and cardiac dysfunction. Our findings indicated a possible joint etiology of breast cancer and cause of mortality among breast cancer patients.

Study II is a population-based cohort study based on data from the Nordic Occupational Cancer (NOCCA) study incorporating five Nordic countries, namely Denmark, Finland, Iceland, Norway, and Sweden. This study was applied to explore occupational variations in the incidence of female breast cancer in Nordic countries and included the follow up of 7.5 million females in the Nordic population. The standardized incidence ratio (SIR) of breast cancer was calculated for each of the 54 occupational categories expressed as the ratio of observed to the expected number of cases. The expected number of cases for each country was estimated by multiplying the incidence rates by the respective numbers of person years at risk in the NOCCA Study by country. Overall, the highest risk elevations were observed among military personnel, dentists and physicians, while the lowest risks were observed among gardeners, farmers, and woodworkers. Histologically, no substantial variation in the incidence of breast cancer was observed. Occupational groups such as welders, tobacco workers, and painters had elevated SIRs for breast cancer diagnosed at age
In Study III, a nested case control design was applied on a cohort of 47,479 Finnish women with at least five childbirths. This study assessed the role of the length of birth intervals on the risk of breast cancer and reflected on distinct differentiation and functional roles of lobular and ductal cells, and different responses to hormonal exposure. Shorter birth intervals had a protective effect on the lobular breast among postmenopausal women, while an opposite effect was observed for lobular breast cancer among premenopausal women.

Study IV explored the roles of numerous components of reproductive history, postmenopausal hormonal therapy (HT), socioeconomic position, and occupational history on the risk of breast cancer, stratified according to characteristics of breast cancer such as histology and age at diagnosis. The study included 19,253 cases of breast cancer between 1994 and 2013 and, 96,265 controls. High socio-economic status and occupational sedentariness were associated with elevated risk of breast cancer even after adjustment for the use of postmenopausal HT. Increasing age at first and last birth was associated with an increased risk of breast cancer. In general, the use of HT was associated with an increased risk of breast cancer. Combined estradiol-progestin HT was more strongly associated with breast cancer risk (OR 1.15, 1.37 and 2.26 for
Study IV also confirmed the dual effect theory of pregnancy: increasing parity was protective against breast cancer in the long term, preceded by a short-term increase in risk. Our data allowed us to estimate the heights and durations of breast cancer risk peaks after each childbirth. Among premenopausal women for example, the transient increase in risk of breast cancer was two-fold immediately after the first birth as compared to nulliparous women, and the effect turned protective 7–8 years after the birth of the child. After the fifth birth, there was only a 20% excess risk peak, and a protective effect began within 2–3 years.

This dissertation illustrates that the various components of the individual's reproductive history (number of children, age at first birth, age at last birth, birth intervals), socioeconomic status, occupational exposures, including occupational physical activity, and lifestyle factors such as postmenopausal HT selection, markedly affect the risk of breast cancer. Especially the possibility to model the dual effect theory of parity for all age groups as well as by age at diagnosis was a novel feature of this dissertation. We exemplified that in a situation where direct information on risk factors for breast cancer patients is not available, mortality patterns could be used as a surrogate to obtain information on possible etiology. In the Nordic countries, individual-level data on many risk factors can be found from various precise and complete registries, and hence the use of surrogate variables is not often needed. The use of the NOCCA job exposure matrix to estimate occupation-specific quantitative measure of physical activity is an example of combining aggregate level information of an occupational category with individual level information of durations of occupations held by the person.