|Home | About | Journals | Submit | Contact Us | Français|
New trial shows no reduction in risk
The possibility that breast feeding might protect against allergy and asthma has generated interest for 70 years. In this week's BMJ, a cluster randomised trial by Kramer and colleagues assesses whether exclusive and prolonged breast feeding reduces the risk of asthma and allergy at 6 years of age.1 It found no significant difference in allergy and asthma symptoms reported by parents or the results of allergy skin prick tests.
Hospitals in Belarus were randomised to promotion of breast feeding or usual care, and mothers intending to breast feed were eligible. The intervention increased the total duration of breast feeding and exclusive breast feeding in the intervention group. Six years later, parents answered seven questions about wheezing, hay fever, itchy rash, and whether their child had ever had asthma or eczema. The children also had skin prick tests to determine hypersensitivity to five airborne allergens. Overall, 10% of parents reported that their child ever wheezed, 5% that they ever had symptoms of hay fever, and 1% that they ever had asthma, with no significant difference between intervention and control groups. Positive skin prick tests were more common, with 27% of children having more than one positive test, but again there was no significant difference between the two groups.
The trial overcomes many of the challenges inherent in studying the influence of breast feeding on health outcomes. Assigning mothers to breastfeeding promotion or usual care eliminates the confounding inherent in observational studies. The cluster design allows better estimation of effects within each intervention group. Furthermore, the design includes prospective collection of high quality data on feeding when the children were 3, 6, 9, and 12 months, with standardised definitions for exclusive and any breast feeding.
The limitations of this study include a highly selected sample, comparison of two relatively similar breastfeeding groups, and the validity of the outcome measures. It is appropriate to select mothers intending to breast feed when testing the efficacy of a programme to promote breast feeding as this improves the duration of total and exclusive breast feeding. However, it limits external validity, because women who choose to breast feed may differ from those who do not in characteristics related to allergy and asthma outcomes, such as geography and socioeconomic status.
Although large differences were seen between the duration of breast feeding in the two groups, all women started breast feeding, and even in the control group 36% were still breast feeding at 6 months. Only 6.4% of the control group were exclusively breast feeding at 3 months compared with 44.3% of the intervention group, but many more may have been exclusively breast feeding at an earlier time point, such as 6-8 weeks. Hypothetically, exclusive breast feeding in the early weeks might be protective. It is possible that the groups were not divergent enough to answer the question of whether breast feeding protects against allergy and asthma.
The outcome measures also need to be considered. The reported prevalence of asthma was five times lower than the expected rate in the United Kingdom or the United States.2 3 Possible explanations include a lower prevalence of childhood asthma in this sample from Belarus compared with the UK and US; under-reporting or underdiagnosis of asthma in this sample; or lower prevalence of asthma in both the intervention group and the control group related to a common factor, such as the high initial breastfeeding rate. The second outcome, positive skin prick tests, is also problematic. Skin prick tests are better negative predictors than positive predictors and in clinical practice are recommended only as confirmatory tests for people with symptoms.4 A test with a positive predictive value of 11.9% for hay fever may not have adequate specificity to determine if breast feeding is associated with allergy.5
The finding that promoting breast feeding did not reduce hay fever, eczema, or asthma reported by parents or result in fewer positive skin prick tests despite large increases in the duration of exclusive breast feeding calls into question previous findings of associations between breast feeding and decreased risk of allergy and asthma. Although this study must be interpreted cautiously—taking into account its limitations—previous work on this question is conflicting.6 7
For the moment, promotion of breast feeding should include evidence that it reduces the incidence of a wide range of infectious diseases, including diarrhoeal diseases and lower respiratory tract infections.8 9 Evidence that it reduces the incidence of other conditions including diabetes, obesity, and some cancers is emerging.10 11 12 13 Furthermore, breast feeding has health benefits for the mother. Therefore, there is already ample evidence to promote breast feeding as a public health measure. None the less, the claim that breast feeding reduces the risk of allergy and asthma is not supported by evidence.
Competing interests: None declared.
Provenance and peer review: Commissioned; not externally peer reviewed.