News article

Study on breastfeeding and intelligence is flawed

October 15, 2006

On October 4, the BMJ (formerly known as British Medical Journal), published a study by Der, Batty, and Deary concluding that breastfeeding has no effect on intelligence.

(Effect of breast feeding on intelligence in children: prospective study, sibling pairs analysis, and meta-analysis, 4 Oct 2006)

The conclusion, however, suffers from a serious methodologic flaw often present in other breastfeeding studies. It does not describe the amount of breastfeeding, judging “dose” only by duration, and even that the authors admit was “less reliable than whether a child was breastfed or not.”

In this study, a breastfed child could mean anything from child who breastfeeds only once a day, all the way to a child who breastfeeds 8-12 times a day, and gets no other food or drink besides breastmilk. In determining the effects of any drug or diet, the medical community does not generally tolerate publishing data without fully determining dose and duration. Why is breastfeeding often treated differently than any other factor? The degree of breastfeeding per day must be determined; duration alone is not a proxy for dose.

Furthermore, 3 months duration, the median in this study, is not very long, so it is premature to conclude whether breastfeeding has an effect or not. All major medical authorities recommend that babies get no other food or drink besides breastmilk for the first 6 months, with continued breastfeeding for the first 1-2 years of a child’s life. In this study, not only was median duration well below recommended levels, but the 95th percentile was only 14 months. Interestingly, a table in the study lists a significant association with intelligence for children breastfed more than 29 weeks, but that finding was not mentioned in the text of the paper.

To study whether breastfeeding has any effect on intelligence, one must not only control for the factors these authors controlled for, but must study the true “dose” of breastmilk, including at least a subset of infants who are exclusively breastfeeding, and the median duration must be at least 6 months.

This the second well-publicized study in recent years to suffer from a flaw in neglecting breastfeeding exclusivity, the other being the 2002 study by Sears et al. in the Lancet, on asthma. In that study, a presumably large but unspecified number of babies received formula in the first days of life, yet the authors concluded breastfeeding did not affect asthma risk, even though other literature clearly shows even a small exposure to formula in neonates can affect the body’s immune system.

We must demand consistent standards in research on breastfeeding, and the gold standard must be complete exclusivity.