Socioeconomic factors and childhood cancer : survival and long-term consequences

Abstract: Childhood cancer is rare and survival rates have increased substantially over the past 60 years as a result of better diagnostics and treatments. Despite this, childhood cancer is the most common disease-related cause of death among children aged 1-14 years in Sweden. Moreover, there are indications that the progress in survival has not been gained by all groups equally and socioeconomic differences in childhood cancer survival have been observed across the world. However, if such differences are seen within Sweden, a country with universal healthcare that is free of charge for children, has not previously been investigated. The increased survival rates have also resulted in a growing population of adult childhood cancer survivors. Survivors may face various late health effects but less is known about the socioeconomic consequences of having had cancer as a child. To address these issues, the overall aims of this thesis were to determine if and how survival from childhood cancer varies by socioeconomic factors and to investigate long-term socioeconomic consequences, in particular educational attainment, in young adult survivors of childhood cancer. The research questions were addressed in four studies that utilized information in the Nordic population and health data registers. Study I-III investigated socioeconomic and familial factors in relation to survival from childhood cancer. Study I was a cohort study including children diagnosed with cancer in Sweden from 1991 to 2010. In this study we observed socioeconomic differences in overall survival – children of parents with a lower level of education had worse survival from childhood cancer, while no differences were observed for household income. The differences by parental education were seen already in the first year after diagnosis. These findings, together with the publication of a seminal study from the US, led us to investigate potential socioeconomic differences in early mortality from childhood cancer. In Study III, we included children diagnosed with cancer in Sweden and Denmark during 1991-2014 and assessed the association between parental socioeconomic factors and early mortality (defined as deaths occurring within 90 days after cancer diagnosis). We observed that children from disadvantaged backgrounds were at increased risk of early mortality, with parental education and maternal income showing the most pronounced associations. Associations with later mortality (defined as deaths occurring 1 to 5 years after cancer diagnosis), were attenuated or close to unity in this study. In Study II we investigated another aspect of social family circumstances, number of siblings and birth order, in relation to survival from acute lymphoid leukaemia (ALL) and acute myeloid leukaemia (AML) in children diagnosed 1991-2015 in Sweden. In this study we found no evidence supporting a previously suggested hypothesis of lower survival after leukaemia among children with siblings, but we rather observed the opposite. We also included clinical information for children diagnosed with ALL and saw that the superior survival among children with siblings was seen mainly within children with low-risk profiles. In Study IV, we compared attainment of upper secondary education in young adult survivors of childhood cancer, matched population comparisons, and the survivors’ siblings in Sweden, Denmark and Finland. We observed that delays in attaining this education level were more common among survivors of all cancer types compared to their peers. However, by the age of 25 years, many survivors had caught up with regard to upper secondary education. Particular risk groups were survivors diagnosed with central nervous system (CNS) tumours, and survivors diagnosed with ALL in the early time period. We also saw that survivors who had spent more time in hospital around the time of diagnosis or had hospital contacts, in particular for psychiatric diseases, in the age range 20-24 years were at increased risk of not having attained upper secondary education by age 25 years. In conclusion, we have observed social inequalities in survival from childhood cancer in the Nordic countries with universal healthcare access. The differences were seen already early in the disease course, and these findings need further attention. We have also shown that although survivors of childhood cancer may experience delays in their education, many catch up. However, there are vulnerable risk groups that need further support. The results in this thesis are based on information from nationwide, population and health data registers, which minimized the risk of bias from non-participation, loss to follow-up, and self-reports. Moreover, the included studies highlight the potential and importance of Nordic collaborations in research of rare diseases.