The new CDC report on breastfeeding puts the cart before the horse


The new CDC report on breastfeeding, Improvements in Maternity Care Policies and Practices That Support Breastfeeding — United States, 2007–2013, has been getting a lot of press.

According to the report

…[P]ractices supportive of breastfeeding are improving nationally; however, more work is needed to ensure all women receive optimal breastfeeding support during the birth hospitalization.

Implications for Public Health Practice: Because of the documented benefits of breastfeeding to both mothers and children, and because experiences in the first hours and days after birth help determine later breastfeeding outcomes, improved hospital policies and practices could increase rates of breastfeeding nationwide, contributing to improved child health.

There just one serious problem: The documented benefts of breastfeeding in the US are trivial and there’s no evidence at all that experiences in the first hours and days after birth determine later breastfeeding outcomes.

[pullquote align=”right” color=”#555555″]US breastfeeding rates have no impact on child health.[/pullquote]

The authors of the report have made a very serious error that undergirds everything they have written. They’ve confused correlation for causation.

Simply put, breastfeeding in the US is associated with higher socio-economic status (higher income, greater education, better access to healthcare). Therefore, the fact that breastfed children have better health outcomes is more likely to be the result of higher socio-economic status (and there are reams of papers demonstrating this fact) than with breastfeeding (on which the evidence is weak, conflicting and plagued with confounders.)

Suppose I did a study comparing two groups of children to determine if breastfeeding increases children’s height. Imagine further that I found the children from Group A, which contains a high proportion of exclusively breastfed infants, turn out to be several inches taller at age 5 than the children from Group B, who never received breastmilk.


Did breastfeeding make the children in group A taller? We can’t say unless we have more information.


Now we can see that the mothers in group A are actually taller than the mothers in group B. It is likely genetics that made the children in group A taller, not breastmilk.

In the case of breastfeeding, the mothers in the US who breastfeed (group A) are more likely to be privileged than the mothers who don’t (group B). It is that privilege that makes their children healthier, not breastfeeding. Breastfeeding does not cause better health outcomes, both breastfeeding and better health outcomes are the result of privilege.

The same phenomenon applies to experiences in the first hours and days. Babies whose mothers are strongly committed to exclusive breastfeeding (group A) are going to have different experiences than babies whose mothers are not as committed (group B). The babies in group A are more likely to be breastfed within the first hour, for example, and are far less likely to receive supplementation with formula. Those experiences don’t cause an increase in breastfeeding rates, they reflect mothers’ commitment to breastfeeding which is the cause of differential rates of breastfeeding months later.

Why do the CDC researchers put the cart (the conclusions) before the horse? It’s probably because of white hat bias, the tendency to reach socially approved conclusions. In 2015, “everyone knows” that breastfeeding is good for babies so white hat bias leads breastfeeding researchers to ignore the privilege (greater income, greater education, better access to healthcare) that leads women to breastfeed and ascribe the benefits to breastfeeding itself.

It is this fundamental error that is responsible for a curious outcome of breastfeeding promotion. Despite millions of dollars spent promoting breastfeeding, and a dramatic rise in breastfeeding initiation, there has been no improvement in indicators of child wellbeing. There’s been no drop in infant mortality, no increase in life expectancy and no change in IQ. As far as I am aware, there is not a single return on our massive investment in breastfeeding promotion.

And that’s just what you would expect if researchers confused correlation for causation. Increasing breastfeeding rates won’t change indicators of child health because breastfeeding doesn’t lead to healthier babies; privilege does. We are literally wasting millions of dollars on promoting a practice that has a trivial impact on health, and feeling virtuous for doing so.

Don’t get me wrong. Breastfeeding is a good thing; I breastfed four children because I believed breastfeeding to be a good thing. But scientific evidence is far more important than feeling virtuous and the scientific evidence is pretty definitive: the benefits of breastfeeding in the US are trivial, and there’s no evidence at all that so called “baby-friendly” hospital practices have any impact on breastfeeding rates.

Let’s stop wasting money, not to mention stop pressuring women into making a choice with trivial benefits and stop judging hospitals by that choice. Breastfeeding rates have no impact on child health.

If anyone, including the CDC researchers, believe otherwise they must provide scientific evidence that breastfeeding impacts child health, not the wishful thinking of white hat bias that leads them to put the cart far before the horse.