select search filters
roundups & rapid reactions
factsheets & briefing notes
before the headlines
Fiona fox's blog

expert reaction to study investigating breastfeeding and risk of childhood leukaemia

Potential health benefits related to breastfeeding have been studied for many years, and a paper published in JAMA Paediatrics has reported a lower incidence of childhood leukemia in those who were breastfed based on meta-analyses of eighteen previously published papers.


Prof. Eve Roman, Professor of Epidemiology, Epidemiology & Cancer Statistics Group, University of York, said:

“In addition to being based on self-reported data, this meta-analysis contains a number of errors. Most notably, the main result from the UK Childhood Cancer Study (references 23 and 44) has been counted twice, which will have distorted the overall result.

“The UK Childhood Cancer Study published the findings on breast-feeding in 2001. The paper by Crouch et al (ref 44) used a sub-set of the data to look at infectious illness patterns, so this was not new or additional data but was a follow-up paper looking at data already published, and should not have been counted in this study.

“In addition to this major error, there are a number of other significant limitations, both with respect to the appropriateness of the methodology (attempting an unadjusted analysis) and the interpretation of the results. I stick by the original conclusion of the UKCCS 2001 paper which states that it ‘It is unclear whether the apparent small reduction in the odds ratio for these various types of childhood cancer is a generalized effect of breastfeeding or whether it reflects some systematic bias in the majority of studies that have investigated the question’.”


Prof Jean Golding, Emeritus Professor of Paediatric and Perinatal Epidemiology, University of Bristol, said:

“This study involves combining results from 18 studies of children who had leukaemia and their controls. The disadvantage with such case-control studies is that (a) the information on whether, and how long, the mother had breast fed the child is often obtained long after the child was born, and may not be wholly accurate; (b) as the authors note, there are often biases in the comparison group (the controls), such that the more educated mothers are more likely to take part – such more educated mothers are more likely to have breast fed, and consequently women who breastfed are probably overrepresented in the control group (children who didn’t get leukaemia) which might distort the results in terms of suggesting a bigger proportion of children who do not get leukaemia are breastfed.

“Consequently, and particularly in the light of the fact that the apparent benefits shown are small, I am not convinced that breast feeding has a protective effect on childhood leukaemia. This is in contrast to the overwhelming evidence of benefits of breast feeding in regard to protection from early infections and of the benefits on the child’s later intellectual ability.”


Prof. Mel Greaves, Director of the Centre for Evolution and Cancer, The Institute of Cancer Research, London, said:

“This new study confirms and extends previous studies indicating that breast feeding of infants for more than six months reduces the risk of subsequent childhood acute lymphoblastic leukaemia (ALL) by around 20 per cent. But it’s important to remember that the risk of childhood leukaemia is very low – around one in 2,000.

“This study also fits in with increasing evidence that infants who are exposed to infections – through attending day care or contact with older siblings – are less vulnerable to this type of leukaemia. The most plausible interpretation is that absence of early immune system exposure can result in later very abnormal responses to common infection that may, in susceptible individuals, trigger cancer. This is called the ‘delayed infection’ hypothesis for ALL.

“Extended breast feeding and natural infectious exposures in infancy were part and parcel of human life until relatively recently. The increasing incidence of ALL in modern, affluent and hygienic societies may be an unintended consequence of these social changes.

“In the absence of a vaccine that efficiently mimics natural infection in infancy, the available data suggests that risk of childhood leukaemia, which remember is very low, can be diminished even further by both prolonged breast feeding and increased social contacts in the first year of life. We need further research to understand if there are other ways to reduce the risk of childhood cancers.”


Prof. Valerie Beral, Professor of Epidemiology and Director of Cancer Epidemiology Unit, University of Oxford, said:

“The suggestion that children who had been breastfed might be at a reduced risk of childhood leukaemia has been around for decades. However previous research, and that described in this new report, are not sufficiently rigorous to confirm or refute such a claim.

“Some time after children were diagnosed with leukaemia their parents were asked to try to remember what they had done at around the time when the children were born. What the parents remembered was then compared with what parents of healthy children said. It is possible that parents’ memory of what they had done years ago was affected by the knowledge that their child had leukaemia. The researchers had no way of assessing such an important potential problem, which could well have distorted their results.”


Prof. Chris Bunce, Research Director, Leukaemia & Lymphoma Research, said:

“We do not want mothers who are unable to or choose not to breastfeed to worry as a result of this research. Childhood leukaemia is thankfully very rare and evidence for the protective benefits of breastfeeding is not as clear-cut as suggested by this latest study.

“Although this research represents a comprehensive analysis of many different studies, it can still only tell us about associations rather than proven causes, which may be complicated by other background factors. Only a tiny percentage of children – breastfed or not – will ever get leukaemia. Other research suggests that there are many different factors that are likely to play a role in its development.

“There is a theory that development of a strong immune system early in life could help provide a defence against childhood leukaemia for those already at risk. The most comprehensive ever UK study of childhood leukaemia cases, conducted in the 1990s, showed a slight association between breastfeeding and reduced rates of leukaemia – but not on the scale suggested by this latest research, and it was still only an association.”


Dr Colin Michie, Chair of the Nutrition Committee, Royal College of Paediatrics and Child Health, said:

“This work brings to the fore the possibility that some cancers can be prevented in the first year of life. The paper showed a significant effect which the authors describe as stronger in those infants breastfed for more than 6 months. In this group they calculated a 19% lower risk of childhood leukaemia. There are some challenges to the methodology, which are statistical, because it relies on the studies examined being free of bias, something that is not possible to measure accurately. Further the authors could not distinguish between exclusive breast feeding or mixed feeding as this information was not always available. However their observations are similar to those made by previous researchers.

“In addition, although the authors outline several ideas, we still do not clearly understand the mechanism of how some cancers can be prevented. Therefore more research is needed. Clearly for those who cannot breastfeed their infants as well as those working with leukaemia patients, this new publication must also drive research into the mechanisms by which breast milk works.

“The benefits of breastfeeding to populations of children are well established. Breastfed babies are less likely to contract chest and ear infections, suffer from sickness and diarrhoea or become obese. Earlier this year, a study highlighted that breastfeeding was associated better school performance, a higher IQ and higher incomes in later life. So this latest development is not only a very important discovery, but further adds weight to the benefits of breastfeeding.

“Across the UK, the number of women breastfeeding their baby at six weeks (55%), and at six months (34%) is increasing. Although breastfeeding rates are improving, we know there is wide variation, with younger mothers and those living in areas of higher deprivation least likely to breastfeed. This new information means we now have ever more reason to encourage as many women as possible to breastfeed. But efforts must be targeted where they are needed most so the benefits can potentially be felt on a much larger scale.”


‘Breastfeeding and childhood leukemia incidence, a meta-analysis and systematic review’ by Efrat L. Amitay et al. published in JAMA Paediatrics at on Monday 1 June 2015. 


Declared interests

Prof. Eve Roman: “No interests to declare.”

Prof. Jean Golding: “No interests.”

Prof. Mel Greaves: “I have no declarations of interest to declare.”

Prof. Valerie Beral: “I have no conflicts to declare.”

Prof. Chris Bunce: “I have no interests to declare.”

Dr Colin Michie: As well as sitting on the College’s Nutrition Committee, Colin attends meetings of the International Breast Feeding Association and the Academy Nutrition group. He also sits on two NICE committees in relation to child nutrition and is on the advisory board for LaLeche too, but this position has never involved in him doing any actual work as yet.

in this section

filter RoundUps by year

search by tag