Lisa Lyngsie Hjalgrim1,2, Klaus Rostgaard1, Gerda Engholm3, Eero Pukkala4,5, Tom Børge Johannesen6, Elínborg Ólafsdóttir7, Henrik Hjalgrim1. 1. a Department of Epidemiology Research , Statens Serum Institut , Copenhagen S , Denmark. 2. b Department of Paediatrics and Adolescent Medicine , Rigshospitalet , Copenhagen Ø, Denmark. 3. c Department of Documentation and Quality , Danish Cancer Society , Copenhagen Ø, Denmark. 4. d Finnish Cancer Registry, Institute for Statistical and Epidemiological Cancer Research , Helsinki , Finland. 5. e School of Health Sciences, University of Tampere , Tampere , Finland. 6. f Cancer Registry of Norway , Majorstuen , Norway. 7. g Icelandic Cancer Registry, Icelandic Cancer Society , Reykjavik , Iceland.
Abstract
BACKGROUND: In epidemiological studies, Hodgkin lymphomas (HL) in children younger than 15 years and HL in adolescents and younger adults age 15-35 years has traditionally been studied separately, under the assumption that HL at age 0-14 constitute a homogeneous entity. However, the continued validity of this research practice in affluent settings may be questioned. Specifically, the boundary at age 15 years may not be epidemiologically justified, and therefore also questionable clinically. We therefore updated and further characterised recent HL incidence patterns among Nordic children. MATERIAL AND METHODS: We obtained HL incidence data in children aged 0-14 years for the period 1978-2010 from the five nationwide Nordic cancer registries. The data were analysed by log-linear and/or a mixture of Poisson regression models. RESULTS: The analyses showed statistically significantly decreasing and increasing HL incidence rates in children younger and older than eight years, respectively during the study period. Statistical modelling suggested that cases in children age 0-6 years constituted a disease entity of its own, whereas cases in older children were more likely to belong to the younger adult HL entity. CONCLUSION: Diverging incidence trends and statistical modelling suggest that HL in children age 0-14 years cannot be assumed to constitute a homogeneous disease entity in affluent settings.
BACKGROUND: In epidemiological studies, Hodgkin lymphomas (HL) in children younger than 15 years and HL in adolescents and younger adults age 15-35 years has traditionally been studied separately, under the assumption that HL at age 0-14 constitute a homogeneous entity. However, the continued validity of this research practice in affluent settings may be questioned. Specifically, the boundary at age 15 years may not be epidemiologically justified, and therefore also questionable clinically. We therefore updated and further characterised recent HL incidence patterns among Nordic children. MATERIAL AND METHODS: We obtained HL incidence data in children aged 0-14 years for the period 1978-2010 from the five nationwide Nordic cancer registries. The data were analysed by log-linear and/or a mixture of Poisson regression models. RESULTS: The analyses showed statistically significantly decreasing and increasing HL incidence rates in children younger and older than eight years, respectively during the study period. Statistical modelling suggested that cases in children age 0-6 years constituted a disease entity of its own, whereas cases in older children were more likely to belong to the younger adult HL entity. CONCLUSION: Diverging incidence trends and statistical modelling suggest that HL in children age 0-14 years cannot be assumed to constitute a homogeneous disease entity in affluent settings.