It seems that researchers looking at the “benefits” of breastfeeding (really the problems with formula) can’t catch a break these days.  In the past month, two prominent research studies have hit the news and yet detractors everywhere call “foul”.  For a long time, critics of the research showing favourable outcomes associated with breastfeeding (which really are normal outcomes with negative outcomes associated with formula feeding) argued that many other variables weren’t considered.  After all, more educated and higher SES women are more likely to breastfeed, at least in the USA where much of the research is conducted.  However, in recent years researchers have taken heed and include all these variables of interest – and more – and we’re ending up with the same results.

Take the recent studies, for example.  The first, out of Israel, found that breastfeeding significantly reduced the likelihood that a child will develop ADHD[1].  Specifically, when controlling for a host of factors that influence the development ADHD, infants who are not breastfed at 3 months of age are three times more likely to develop ADHD than children who were breastfed.  Notably, the control group for this study included both children with low risk for developing ADHD and children who were at-risk for developing ADHD (those who had siblings diagnosed), with the results being significant even when just comparing the at-risk control group with those who developed ADHD.

The problem?  Well, it’s not a longitudinal study.  Forget that no first study into a phenomenon ever is longitudinal because the time and resources (both in terms of personnel and money) are so huge that there has to be some impetus to start looking at things longitudinally.  Something the researchers actually plan to do now that they’ve found this rather interesting and significant result.

The second study, out of the US, found (once again) a positive association between breastfeeding and IQ[2].  This has been found many times over, but people have continuously fought it saying the results must be due to inherent factors in the families themselves.  For a while it was assumed it was actually feeding on demand[3] and not breastfeeding (though it’s true that feeding on a schedule, regardless of the method, leads to greater risk of lower intelligence compared to on-demand of the same feeding method).  But studies keep coming out finding this link between breastfeeding and intelligence.

This most recent study actually took care to examine the most cited factors as being potential mediators (mediator = a variable that explains the entire relationship between two other variables) and was longitudinal in nature, allowing for a causal interpretation.  In this study, the researchers found that up to a year (all they included), every month of formula feeding decreased verbal ability at age 3 by .21 points per month (for a total of 2.52 points on the Peabody Picture Vocabulary measure used).  At age 7, every month of formula use decreased verbal IQ by .35 points per month and nonverbal IQ decreased by .29 points per month.  For a child who either breastfeeds or uses formula for a year, the average difference is about four IQ points, all other factors being equal.

The problem?  Well, I’ll first say that at least the detractors who were interviewed for pieces covering the research[4] are no longer trying to argue that the results are due to other variables as the research pretty much covered that.  Now the issue is that the effect is too small to truly care about.  Or rather, too small to not just be random variation.  This argument I find highly amusing given that the point of running a significance test is to determine if the results are due to random variation or not based on the sample size you have, and the statistical significance is saying it is not due to random variation.  But is the effect too small?  This is a good question because often effects can be statistically significant but not practically so.

Let’s look at some examples and you be the judge on the significance.

While the effect will vary based on where children would be on the normal distribution normally (remember, IQ is normally distributed), a four-point decrease is the equivalent of a .27 standard deviation regardless of where they fall on the distribution.  This would take a child that was in the 60th percentile down to the 49th percentile.  A child who was in the 70th percentile down to the 60th percentile.  A child in the 80th percentile down to the 72nd percentile.  And a child in the 90th percentile down to the 84th percentile.  Only when we get to the extreme ends does the change differ only slightly.  For example, a child in 99th percentile will only drop to the 98th percentile (and likewise a child in the 2nd percentile would only drop to the 1st).  For the vast majority of children, however, the differences are, in my opinion, practically significant, though you can feel free to feel differently.  Important to note, though, is that the effect isn’t static.  That is, for some kids, there will be no effect and for others the effect will be even larger.  What we don’t know yet (and why more research is so critical) is what other factors influence this interaction and this is why people (and myself) speak about research findings at the societal level and the individual level.  As always, research speaks in generalities and does not predict what will happen, but provides some form of risk assessment.

All this on top of the increased risk for myriad diseases (including, but not limited to, cancer and diabetes) associated with formula use as outlined in dollars and lives in a 2010 Pediatrics article[5].  (A reminder for any that missed it: in the USA, not reaching the goal of 90% of moms breastfeeding exclusively for six months costs the health care system $13 billion and leads to the death of 911 children each year.)  There are also studies that have found a link between formula use and autism[6] and formula use and asthma[7].  Not to mention that breastfeeding has also been linked to lower rates of neglect, independent of risk factors[8].

And yet with all this people still adamantly insist that formula is “just as good”.  I will say that in our society as it currently is there are cases where formula is absolutely the best choice for the family regardless of the evidence above.  I don’t think it means that we say that formula is “just as good” though.  That’s misleading, and frankly, wrong, based on the evidence we have.  What we can say is that formula will overall be a better choice for some and that there are incredibly valid reasons not to breastfeed.

But I would go further and say what we really ought to be saying is that we need to shift a lot in our society.  I’ve argued this before and I will say it again and again until we start seeing change.  We should never say formula is “just as good” to alleviate possible guilt a mother may feel because she couldn’t or chose not to breastfeed.  Sorry, but that’s not a good response.  What we ought to be doing is making substantial changes so that over 40% of women don’t fail to meet the goals they set for themselves[9].  Cut out the booby traps that prohibit so many women from doing what they want to do.

And for those who cannot or actively choose not to breastfeed?  Let’s end the stigma around donor milk and make sure families have access to affordable donor milk for their child so that their baby can receive the many benefits of breastmilk.  Formula should not be the go-to second option.  It just shouldn’t.  Formula should be last.  Available for those who need and want it, but last on the list of what parents look to when deciding how to feed their babies.  This could provide the win-win situation that we need for families.  Instead of pitting babies’ needs against mothers’, can we not work to find a solution that works for everyone with the least amount of compromise for anyone?

One final thought:  I actually appreciate the critical eye that is levied on all research by critics.  Including breastfeeding research.  I appreciate it because we should know what exactly is influencing the results we get, nothing should just be taken at face value.  However, when it gets to the stage where meta-analysis after meta-analysis and review after review finds the same results, we have to start questioning why people are so adamant to criticize.  When one line of criticism fails (e.g., there is no effect of breastfeeding on IQ, it’s all in other variables) and detractors then turn to the next (e.g., it’s just not a big enough effect), then we have to question what agenda is there.  Often I think the agenda is actually not inherently a bad one – it’s there to try and alleviate any guilt moms feel about using formula.  The problem is that this agenda means that we don’t strive to find solutions that offer the best to everyone.  And I do have a real problem with that.

[Image Credit: Me and my daughter]


[1] Mimouni-Bloch A, Kachevanskaya A, Mimouni FB, Shuper A, Raveh E, Linder N. Breastfeeding may protect from developing Attention-Deficit/Hyperactivity disorder. Breastfeeding Medicine 2013; DOI: 10.1089/bfm.2012.0145.

[2] Belfort MB, Rifas-Shiman SL, Kleinman KP, Guthrie LB, Bellinger DC, Taveras EM, et al.  Infant feeding and childhood cognition at 3 and 7 years:  effects of breastfeeding duration and exclusivity.  JAMA 2013; DOI: 10.1001/jamapediatrics.2013.455.

[3] Iacovou M, Sevilla A.  Infant feeding: the effects of scheduled vs. on-demand feeding on mothers’ wellbeing and children’s cognitive development.  European Journal of Public Health 2012. DOI: 10.1093/eurpub/cks012.

[4]http://online.wsj.com/article/SB10001424127887324809004578635783141433600.html?mod=WSJ_article_MoreIn_Health%26Wellness (accessed July 29, 2013)

[5] Bartick M, Reinhold A.  The burden of suboptimal breastfeeding in the United States: a pediatric cost analysis.  Pediatrics 2010; 125: e1048-e1056.

[6] Schultz ST, Klonoff-Cohen HS, Wingard DL, Akshoomoff NA, Macera CA, Ji M, Bacher C.  Breastfeeding, infant formula supplementation, and Autistic Disorder: the results of a parent survey.  International Breastfeeding Journal 2006; 1: 16.

[7] Sonnenschein-van der Voort AMM, Jaddoe VVW, van der Valk RJP, Willemsen SP, Hofman A, Moll HA, et al. Duration and exclusiveness of breastfeeding and childhood asthma-related symptoms. European Respiratory Journal, 2011; DOI: 10.1183/09031936.00178110.

[8] Strathern L, Mamun AA, Najman JM, O’Callaghan MJ.  Does breastfeeding protect against substantiated child abuse and neglect? A 15-year cohort study.  Pediatrics 2009; 123: 483-493.

[9] Perrine CG, Scanlon KS, Li R, Odom E, Grummer-Strawn LM.  Baby-friendly hospital practices and meeting exclusive breastfeeding intention.  Pediatrics 2012; 130: 54-60.