In this large national cohort study, high fetal growth and family history of Hodgkin lymphoma were independent risk factors for Hodgkin lymphoma in childhood through young adulthood. The association between high fetal growth and Hodgkin lymphoma is consistent with a similar finding reported for boys but not girls <15 years of age in a smaller cohort study in Australia (26
). However, our findings are contrary to that study's null findings for birth weight as well as most others based mainly on case-control data. A recent meta-analysis of 5 case-control studies (2,660 cases and 69,274 controls, aged <18 years) and 2 cohort studies (278,751 children, aged <9 years) reported no overall association between high or low birth weight and Hodgkin lymphoma (18
). Unlike the current study, most of those studies were limited to children and/or adolescents and did not examine gestational age at birth. The current study overcomes many earlier limitations by means of its large population-based cohort design with more complete perinatal data and Hodgkin lymphoma ascertainment, longer follow-up, and greater statistical power.
In contrast to fetal growth, gestational age at birth was not associated with Hodgkin lymphoma in this cohort. Although fetal growth and birth weight produced similar results, our findings confirm more specifically that fetal growth is the relevant component of birth weight that accounts for its association with Hodgkin lymphoma. The mechanisms by which high fetal growth may affect the risk of Hodgkin lymphoma are unknown, but one hypothesis involves growth factor pathways such as insulin-like growth factor-I levels, which are correlated with fetal growth and have been shown to inhibit apoptosis and enhance tumor growth (20
). High levels of insulin-like growth factor-I and other growth factors increase cell division and growth rates that may augment fetal sensitivity to carcinogenic effects or prevent apoptosis in lymphoid cells that have already begun malignant transformation. Epigenetic assessments of polymorphisms in the insulin-like growth factor family may help to further elucidate these pathways and their potential carcinogenic effect on lymphoid cells in utero.
The strong association we found between family history and risk of Hodgkin lymphoma was based on small numbers of cases with affected relatives but is consistent with earlier findings and may reflect both genetic and shared environmental factors. The heritability of Hodgkin lymphoma in the Swedish population has been estimated to be 28% (2
). Many human leukocyte antigen alleles have been associated with Hodgkin lymphoma (42
), specifically areas within the human leukocyte antigen class I and class III regions, which may account for ethnic variation in susceptibility to Hodgkin lymphoma (6
We found no association between birth order and Hodgkin lymphoma, irrespective of age at diagnosis. This is consistent with several (32
) but not all (28
) previous studies, most of which were smaller case-control studies that varied widely in their adjustment for confounding. Low birth order has been hypothesized to be a risk factor for Hodgkin lymphoma by means of delayed infection with EBV and other infectious agents, due to fewer exposures as a result of the absence of older siblings. This in turn may prevent normal maturation of the immune system from T-helper cell type 2 to T-helper cell type 1 predominance in childhood (43
). Our findings do not support this hypothesis but, rather, are consistent with those of an earlier case-control study of 354 Hodgkin lymphoma patients and 1,718 healthy controls that reported no association between age at first occurrence of infectious disease and Hodgkin lymphoma (in contrast to non-Hodgkin lymphoma) (43
). In addition, a recent study of EBV serologies in 55 Hodgkin lymphoma patients or siblings with a history of infectious mononucleosis (a characteristic manifestation of delayed EBV infection) and 173 Hodgkin lymphoma patients or siblings without a history of mononucleosis reported that chronic or severe EBV infection was a risk factor for Hodgkin lymphoma, independent of mononucleosis history (44
). These findings suggest that underlying immune dysfunction is more etiologically relevant than age at EBV infection. Additional studies with more detailed information on infection history and childhood social environment are needed to clarify the complex relations between these factors and Hodgkin lymphoma.
An increasing incidence of the nodular sclerosis subtype has been reported in Denmark and Norway (9
) and elsewhere in Europe (11
) during 1978–1997. We found a similar birth cohort effect for this subtype, although a sensitivity analysis suggested that it might be explained by temporal changes in subtype reporting. There were no temporal trends in the risk of Hodgkin lymphoma overall in this cohort during the study period (1973–2009).
The most important strengths of this study were its national population-based cohort design and large sample size, enabling more robust and generalizable inferences. Linkage of national birth and cancer registries provided detailed information on perinatal factors and Hodgkin lymphoma incidence that was nearly 100% complete (21
). A cohort design prevented selection bias that may potentially occur in case-control studies, and the use of registry-based data prevented bias that may result from self-reporting. We were able to examine the specific contributions of fetal growth and gestational age at birth while accounting for other perinatal and familial factors. Family history of Hodgkin lymphoma was also based on registry data with virtually complete ascertainment, thus improving the reliability of those risk estimates.
Study limitations included the unavailability of information on infection history, immune-related disorders, smoking, and maternal weight or body mass index; hence, we were unable to examine the potentially important effects of these factors. Maternal body mass index, which increased in Sweden during this study period (45
), influences birth weight (46
) and warrants further investigation as a potential modifier of the association between fetal growth and Hodgkin lymphoma. Although statistical power was greater than in previous studies, the ability to detect associations with rarer histologic subtypes was still limited. Subtype data were also missing for some individuals, although there was no evidence that this occurred differentially with respect to perinatal factors or family history.
In summary, high fetal growth and family history of Hodgkin lymphoma were independently associated with Hodgkin lymphoma among individuals born in Sweden in 1973–2008. These findings suggest that perinatal factors including possible growth factor pathways may contribute to the risk of Hodgkin lymphoma in childhood through young adulthood. Further elucidation of these risk factors and their etiologic mechanisms may potentially facilitate the identification of high-risk individuals at young ages.