The Atlantic magazine just published an article written by Hanna Rosin questioning the general perception that research strongly supports that breast-feeding is better than formula.
This is actually a pretty well researched journalistic piece about cognitive development and breastfeeding that raises a number political and scientific issues regarding breast-feeding.
In certain overachieving circles, breast-feeding is no longer a choice—it’s a no-exceptions requirement, the ultimate badge of responsible parenting. Yet the actual health benefits of breast-feeding are surprisingly thin, far thinner than most popular literature indicates. Is breast-feeding right for every family? Or is it this generation’s vacuum cleaner—an instrument of misery that mostly just keeps women down
Although the author seems to overstate the “weaknesses” of the literature on the benefits of breast-feeding, she touches on a very important issue: The difference between statistical vs. clinical significance. That is, even when a statistical difference is found, such difference may be clinically inconsequential. This is due to what “statistically difference” actually means (which varies depending on the statistical procedure used). When two groups are “statistically different” in one measure (let’s say for example ‘IQ scores’), this may only mean that the difference observed between the groups is very unlikely to be due to chance. It does not necessarily mean that the differences between the groups are “big,” –if by “big” we mean that the differences are actually noticeable or even predictive of any other factor (happiness, success, school performance, etc, etc). So many times we find statistically significant differences between groups that are actually not clinically meaningful.
Let’s take the breast feeding and IQ example. In 1999 Anderson and colleagues (Anderson et al. 1999. Breast-feeding and cognitive development: a meta-analysis. American Journal of Clinical Nutrition, Vol. 70, No. 4, 525-535) conducted a meta-analysis of 20 studies that examined the effects of breast-feeding on intellectual functioning. They found that among the studies that properly controlled for confounding variables (variables that could explain the group differences), kids who were breast-fed had higher cognitive scores than kids who were formula-fed, and this effect was stronger for low-birth-weight infants. However the effects were very mild, in that breast-feeding resulted in an average of 3 additional IQ-equivalent points. Given the nature (not so precise) of our IQ and cognitive scales, 3 points is well within the normal variation in IQ scores that we would expect from the same person from one day to another. Although, there are other statistical issues involved (the expected normal variation is actually mostly due to error, while error accounts for only a small part of the differences between the groups), the main issue is that while the 3 point difference was “statistically significant”, such 3 point difference may or may not be clinically meaningful. However, Anderson’s meta-analysis also showed that among low-birth-weight babies, the improvement was 5 points, and at this 5-point level we may begin to see some large clinically meaningful differences (in school performance, achievement, etc). This is also in line with a recent article on the effects of breast-feeding on very-low-birth-weight babies, which I will be reviewing soon.
Lastly, I am not in any way making an argument against breast-feeding. Cognitive performance is just one of a multitude of factors that have been found to benefit from best-feeding. The data are very consistent showing benefits in multiple domains. Hanna Rosin is correct in that the research is not conclusive, but such inconsistencies are inherent to most medical and psychological research. Yet, Hanna Rosin raised a very interesting issue regarding how we evaluate scientific findings.
James W Anderson, Bryan M Johnstone, & Daniel T Remley (1999). Breast-feeding and cognitive development: a meta-analysis American Journal of Clinical Nutrition,, 70 (4), 525-535 DOI: http://www.ajcn.org/cgi/content/abstract/70/4/525?ijkey=e9ed6c3046f324e9945ed829769c24c06cda94eb&keytype2;=tf_ipsecsha
- All Posts (279)
- Bullying (1)
- Child Psychology (250)
- Editorials (7)
- How To Guide (7)
- Parenting (53)
- All Posts (279)