Parental age and risk of acute lymphocytic leukaemia and embryonal tumours in the Piedmont Region, Italy

Abstract

Relative risks (RRs) for maternal and paternal ages were estimated using log-binomial regression adjusting for sex and year of birth. Paternal and maternal age effects were mutually adjusted. Out of a population of more than 630000 children followedup from birth to age five, we observed 229 cases of ALL and 284 cases of embryonal tumours. Older maternal age was associated with ALL risk, irrespective of adjustment for paternal age (Table 1). Conversely, an apparent association with paternal age disappeared when maternal age was included in the model. Both maternal and paternal ages were independently associated with the risk of embryonal tumours, although the RR estimates were attenuated when both variables were introduced in the model. Subgroup analyses were carried out for cancer types with at least 30 observed cases. The strongest association of paternal age was found with embryonal tumours of the kidney (maternal age-adjusted RR for each 5-year increase in paternal age: 1.29, 95% confidence intervals: 0.98–1.68). In our study, ALLs were analysed separately since they are the most frequent type of tumour among children aged less than five. Other types of tumours were grouped and analysed together, under the assumption that parental age might have a stronger effect on tumours with an embryonal origin. Maternal age was associated with both leukaemia and embryonal tumour risk, whereas paternal age, which could relate to accumulation of germ cell mutations, 4 was associated with embryonal tumours only. Conflict of interest. None declared.