Seasonal trends of diagnosis of childhood malignant diseases and viral prevalence in South Korea

Background: Several studies have reported a seasonal trend in the diagnosis of childhood cancer suggesting seasonal factors such as infection. The present study aimed to analyze the diagnosis pattern of childhood malignant diseases using public health data, and to compare this pattern with seasonal viral infection trends.

Method: Using the open data source of the Health Insurance Review and Assessment Service, we extracted data regarding all patients under 21 years of age and who had any cancer, aplastic anemia or myelodysplastic syndrome between September 2009 and December 2013. The positive detection rates of 11 viruses was collected from the surveillance data of Korea Centers for Disease Control and Prevention, and seasonality analysis were conducted with both data.

Results: In total, 9085 patients were diagnosed with malignant disease during the study period; there were about 175 new cases per month on average. Monthly stacked time series by year showed an apparent seasonal variation with the highest monthly average in January as 236, and the lowest in September as 120. In winter, significantly more patients were diagnosed with acute lymphoblastic leukemia, acute myeloid leukemia, neuroblastoma, and Hodgkin's lymphoma than in other seasons. There was a temporal correlation with the diagnostic trends of several diseases and the prevalence of recent human parainfluenza virus.

Conclusion: This study tentatively suggests that the diagnosis of childhood malignancy follows a seasonal trend in Korea, and has a possible correlation with viral prevalence in several diseases. Further long-term analysis of epidemiological data is needed to explore possible causality.