Evidence of an infectious aetiology for teenage and adult cancers – an analysis of seasonality of birth


Session type:

Richard Feltbower1, Marlous Van Laar1, Sally Kinsey2, Susan Picton2, Antony Moran3
1University of Leeds, Leeds, United Kingdom,2Leeds Teaching Hospitals NHS Trust, Leeds, United Kingdom,3North West Cancer Intelligence Service, Manchester, United Kingdom


There is increasing evidence that environmental factors, such as infections, occurring around the time of birth may affect the subsequent development of childhood cancer, especially leukaemia and certain Central Nervous System (CNS) tumours.  Few studies have examined whether an infectious aetiology might exist for tumours seen in teenagers and young adults (TYA).  Certain factors like infections may vary with season and we therefore investigated whether there was any evidence of seasonality of month of birth, based on data from 6,282 TYA diagnosed with leukaemia, lymphoma and CNS tumours aged 15-24 years.


Information on month and year of birth were extracted from the North West Cancer Intelligence Service (NWCIS), the lead UK registry for TYA cancers, from 15-24 year olds diagnosed between 1996-2005 in England. The following diagnostic categories were analysed: all cancers, all leukaemias and main subtypes (acute lymphoid leukaemia, acute myeloid leukaemia), all lymphomas and main subtypes (Hodgkin lymphoma, non-Hodgkin lymphoma (NHL)), all CNS tumours and main subtypes (astrocytoma, other glioma, ependymoma and medulloblastoma).  Seasonal variation was tested using (i) Walter and Elwood’s test and (ii) logistic regression analysis allowing for cyclical variation in month of birth.


No statistically significant evidence of seasonality was present for any diagnostic group except for all cancers combined (P=0.030). However, some seasonal trends were evident for lymphoma (mainly NHL) and CNS tumours (ependymoma, medulloblastoma) with a consistent predicted peak in incidence in mid-summer and trough in late-winter, whereas no clear seasonal patterns were exhibited for leukaemia or any specific subtypes.  All models fitted the data well.


These findings suggest that environmental factors occurring seasonally and around the time of birth are not associated with the totality of cancer in TYA.  However, possible seasonal links with NHL, ependymoma and medulloblastoma support an infectious aetiological hypothesis.