Influence of infections in the aetiology of childhood leukaemia

Background: Leukaemia is the most important cancer among children in industrialized countries. Infections may play a role: the incidence of acute lymphocytic leukaemia (ALL) is higher in resource-rich countries, where infections early in life are less common compared to resource-poor settings; there is a sharp peak in incidence at 2-5 years of age; local clusters of cases have been described and there is seasonal variation in the diagnosis of ALL. Kinlen proposed that population mixing, i.e. large influxes of people into previously isolated areas, could explain clusters of childhood leukaemia. Childhood leukaemia may thus be a rare response to a common, yet unidentified, infection. Greaves proposed the delayed-immune hypothesis as an explanation for the peak incidence of ALL at 2-5 years: a lack of exposure to infections in early life could predispose the immune system to aberrant responses to subsequent ‘delayed’ infections.

Aims: To determine whether leukaemia (any leukaemia and ALL) diagnosed at age <20 years is associated with the following measures of population mixing at community level (Kinlen’s hypothesis): 1) volume and diversity of incoming migrants into communities, and 2) change in annual in-migration; or with proxy measures of exposure to infections (Greaves’ hypothesis): 3) birth order, 4) child density in the neighbourhood, and 5) extra-familial child-care.

Methods: The study uses data from the Swiss National Cohort (SNC), the Swiss Childhood Cancer Registry (SCCR), the childhood cancer cytogenetic database (CCD) and demographic data from the Swiss Federal Statistical Office. Primary outcomes are any leukaemia and ALL in particular. Secondary outcomes are the most prevalent immunophenotype, B-cell-precursor ALL, and cytogenetic subgroups like high hyperdiploidy and TEL-AML1 translocation. Main exposures are percentage of the population that moved into a community within the last 5 years, diversity of areas of origin (Shannon’s entropy), change in annual volume of incoming migrants relative to average in previous years, the rank of the child among all live births of the same mother, neighbourhood indices of child density developed using road network connectivity, and parental full-time employment as a proxy measure of extra-familial child-care.

Rationale and Significance: The possibility that childhood leukaemia might be associated with a specific infection or with later exposure to infections is highly relevant for prevention. Previous studies were limited by the lack of spatial or temporal precision and few studies were able to use both residential locations at birth and at diagnosis. In the present study only routine databases with national coverage will be used and precise geocoding of the address at both birth and at diagnosis will allow high temporal and spatial resolution.

Current status of project: Analyses is ongoing and results for a first publication are currently being written up.

Study team: Spycher BD, Kuehni CE, Lupatsch JE, Kreis C. Institute of Social and Preventive Medicine, University of Bern.

Funding: Cancer Research Switzerland (Grant No KFS-3049-08-2012)

Contact: Ben Spycher (


Lupatsch JE, Kuehni CE, Niggli F, Amman RA, Egger M, Spycher B. Population mixing and childhood leukaemia in Switzerland. Census-based cohort study.Eur J Epidemiol, DOI 10.1007/s10654-015-0042-5, 26 May 2015

© ISPM - University of Bern 2019