Link found between breastfeeding and autism but more research needed

the breast feeding of newborn

They say a picture is worth a thousand words and the image below is quite impressive. It’s the rate of autism over time compared to the rate of breastfeeding over time:


Over the past 4 decades, there has been a shocking rise in the prevalence of autism. Antivax activists have pointed out that there has been an increase in the number of vaccines that infants receive and conclude that vaccines cause autism. But as this graph shows there has also been a dramatic increase in breastfeeding rates. Indeed the two seem to rise in concert over time demonstrating a link between breastfeeding and autism.

[pullquote align=”right” cite=”” link=”” color=”” class=”” size=””]It is grossly irresponsible for any medical professional or organization to announce a link based on a temporal association.[/pullquote]

Obviously more research is needed. We should be urgently investigating whether breastfeeding causes autism. We should reconsider recommendations designed to encourage breastfeeding and ask if we are ignoring the harmful effects. In the meantime, we should direct lactation consultants, La Leche League and the Baby Friendly Hospital Initiative to inform new mothers of the very clear association between breastfeeding and autism. How can women make an informed decision about breastfeeding if they don’t know about the link?

If you read this far you’re probably asking yourself what has happened to my reasoning abilities. Just because two phenomena rise in concert doesn’t make them linked. Sure, it raises the possibility, but it is grossly irresponsible for any medical professional or organization to announce a link based merely on a temporal association. I must be joking, right?

Yes, I am joking but the folks at Baby Friendly UK are not. Here’s a tweet they published this morning:


Links also found between induction of labor and autism but more research needed #bfconf

The hashtag refers to the UNICEF/Baby Friendly Hospital Initiative conference now under way in Liverpool.

Correlation is not causation, though. Rising autism rates are correlated with rising induction rates AND rising breastfeeding rates. Unless the folks at the BFHI are willing to suggest that autism is linked with breastfeeding, they shouldn’t be suggesting that autism is linked with induction.

The above tweet was part of a series reporting on a lecture by lactation professional Karin Cadwell that demonized oxytocin and suggested that it interferes with breastfeeding.

Wait, what? A natural hormone produced in part to promote breastfeeding interferes with it? Well, sure, if you put it in an IV! Apparently the oxytocin knows where it came from no matter that it is chemically identical to the oxytocin produced by the mother’s body.

If the webpage is any indication, the lecture is an amalgam of the naturalistic fallacy (if it’s natural it must be good; if it’s technological it must be bad), the Panglossian paradigm (nothing can improve upon nature), and utter nonsense like this:

Then, one of our local hospitals hired a full time OB anesthesiologist and epidurals began. The nurses reported an immediate outbreak in flat nipples and babies who had trouble latching. One of the nurses told me that it was almost as though the baby didn’t know that the breast was there! Now, epidurals, labor induction and augmentation with synthetic oxytocin, operative deliveries and elective cesarean births are widespread with ubiquitous breastfeeding problems.

Prof. Cadwell appears to think this is what passes for “reasoning”:

I ask myself almost every day, how can something that is so fundamental to good health, in fact our survival as a species, be so difficult? How can it be that babies fail to latch when the characteristics of their mother’s Montgomery gland secretions are analogous to their unique amniotic fluid flavor? What has happened to otherwise healthy, full term babies in the process of being born that they would deny themselves the ultimate pleasure nursing?

It’s analogous to the “argument” that anti-vaxxers make about vaccines: Immunity is fundamental to good health and our survival as a species. Unvaccinated is the biological norm. Therefore, vaccines are unnecessary and possibly harmful.

It’s a ridiculous “argument” when made by anti-vaxxers and equally ridiculous when made by lactation professionals.

Unvaccinated may be the biological norm for infants and children but deaths from vaccine preventable diseases are also the biological norm. Natural immunity is indeed the product of hundreds of millions of years of evolution, but that didn’t make it perfect or even close to perfect.

Just because breastfeeding is the biological norm does NOT mean that all women produce enough breastmilk, that some women don’t have flat nipples, that some babies can’t latch. But, but, but if those things occurred in nature lots of babies would die! That’s right and until very recently, lots of babies DID die for precisely that reason. Human existence is perfectly compatible with massive rates of infant mortality.

So if correlation does not equal causation and natural is not necessarily best, how do we figure out if breastfeeding or induction causes autism.

To determine if Event A caused Disease B, we need to investigate whether it satisfies Hill’s Criteria. These are nine criteria, most of which much be satisfied before we can conclude that Event A is not merely correlated with Disease B, but Event A actually causes Disease B.

I’m not going to review all nine criteria here. I’ll highlight two that are most important in this setting.

Consistency: Have the findings that purport to show a relationship been replicated by other scientists, in other populations and at other times? If studies fail to consistently show the relationship, causation is very unlikely.

This is a critical point. One experiment or even a few studies is NOT enough to determine causation. A large number of studies that consistently show the same result is required.


Consideration of alternative explanations: In the case of breastfeeding (or vaccines) and autism, there is a very simple alternative explanation. Autism cannot be diagnosed before the age of 2 and breastfeeding (and most vaccines) are given before the age of 2.

Similarly, as autism is almost certainly genetic, there may be differences in the pregnancies or fetuses affected by autism that lead to a need for oxytocin induction or augmentation in labor. For example, there is some evidence that autism leads to larger head size. Larger head size leads to greater need for interventions in labor. Therefore any association between oxytocin administration and autism may be a consequence not a cause.

The correlation between induction and autism is probably weaker than the correlation between breastfeeding and autism. Unless and until the folks at Baby Friendly are willing to suggest that breastfeeding causes autism, they have no business suggesting induction does.