“Virtually every mother can breastfeed” says UNICEF, and statements supporting the universal achievability of breastfeeding as a goal every mother should be capable of seem to be an underlying prerequisite for breastfeeding advocacy efforts around the world. At the same time however, surveys conducted in virtually any country you care to name – the UK, China, Canada, South Africa and more – confirm that a substantial proportion of mothers, wherever they are, stop breastfeeding because they think that they don’t have enough milk to feed their babies. So, how can we reconcile these two completely different perspectives? And what is the evidence base for our understanding of this problem? In this blog, I follow the trail of claims in the lay literature on breastfeeding insufficiency and try to pin down the reality from research studies.

In a recent media article, one UK researcher states “Physiologically speaking only around 2% of women should be unable to breastfeed”. No specific research study is quoted for this 2%, but the claim is linked to another article written by a La Leche League trained lactation consultant. The lactation consultant’s article does not mention the 2% figure anywhere, but does say that “it has been suggested that 5% of mothers are unable to produce enough breast milk to nourish their babies at the breast”, and then goes on to clarify that “…there was no research… the 5% had taken on a life of its own”. The consultant then quotes her own experience saying “in my practice… I have documented that only one in a thousand of new mothers simply do not produce any milk at all”, but does not offer peer-reviewed evidence for this claim.

In searching the lay literature further, I seemed to get a little warmer: a review for the National Childbirth Trust comments: “these medical contraindications [low or absent milk supply, and other reasons]… affect a negligible proportion of new mothers”, and a research study is cited for this statement. The study cited is:

The study by Huggins describes the characteristics of 34 women presenting with some degree of breast hypoplasia; a particular presentation of the breast tissue suspected as being a risk factor for lactation problems. However, the study is a case series. Crucially, it lacks any population denominator. That means we have no clue how many women this series of 34 was drawn from; they may have been drawn from a huge sample, or a tiny one; making the type of problems experienced by these 34 women potentially very common, or very rare – we simply have no way of knowing population prevalence from a case series. In addition, breast hypoplasia is only one risk factor for milk supply problems; other contributors, such as postpartum haemorrhage, retained placenta, prior breast surgery and others, are also thought to affect production. In addition, it is possible that some mothers have problems making enough milk without any of these prior risk factors in the picture.

The idea that “true” low milk supply is statistically rare seems to be such a common claim in the lay breastfeeding literature that I became obsessed by it. Desperate to get a real answer from a proper scientific study I embarked on the only way I knew to answer this question short of actually doing a study myself – I dived headlong into PubMed!

First, we should clarify what we are looking for here. As highlighted above, we are looking for “population prevalence” – how common is the problem as a proportion of mothers giving birth. Therefore, ideally we need to find a study starting out with a population-based birth cohort and then following forwards in time to see what happens to these mothers and babies – a prospective cohort study. We want some indicator of an objective measure of milk sufficiency; either actual measured production using test weighing or test expression, or serial measurements of baby’s weight. And, to make this relevant to the UK situation, we want to look at a sample from a developed-country setting, ideally with good support given to mothers so we know that they are not just producing small amounts of milk because they didn’t feed responsively or weren’t helped to attach the baby effectively. I didn’t consider studies in preterm or low birth weight babies, or in developing countries as they may not be relevant to the main question.

I conducted two main searches in PubMed

  • “insufficient milk breastfeeding” – 484 hits
  • “milk adequacy breastfeeding” – 126 hits

I extracted the PubMed output into .csv files, screened all the titles, and checked abstracts where the study looked relevant. I found seven studies that (based on the abstract) looked as if they might meet these criteria, and I checked full-text for all of these. From those seven, only one article provided data directly addressing how common low milk supply might be:

  • Neifert et al (1990). The influence of breast surgery, breast appearance, and pregnancy-induced breast changes on lactation sufficiency as measured by infant weight gain. Birth:17(1):31-8.

There were plenty of other interesting studies, but most either didn’t use an objective measure of milk production, or started out with a selected sample. For example, some were comparing characteristics of mothers who had breastfeeding problems against mothers who didn’t, and that type of study design doesn’t allow you to derive a population prevalence of milk supply problems. The closest I got other than Neifert et al was a study by Galipeau et al from Canada, which did seem to use a birth cohort, and where the authors report evaluating “actual insufficient milk supply” (AIMS). However, the paper doesn’t report the number of mothers in their sample with AIMS, so it’s not possible to calculate a population prevalence from their data (although it may be possible to obtain data on this by contacting the authors).

Onto Neifert et al: this study recruited a birth cohort of 319 mothers who were all highly motivated to exclusively breastfeed for a month or longer. All were advised on best practices to achieve breastfeeding (rooming in, breastfeeding responsively and avoidance of supplementation) with most nursing within an hour or two of birth, nursing on demand at least every 2-3 hours, and nursing a minimum of 8 times in the first 24 hours. Lactation sufficiency was determined to have been achieved when the baby had a weight gain of at least 28.5g/day between two consecutive visits before day 21 postpartum. The authors found that 15% of babies didn’t meet this marker of adequate weight gain. In a more stringent analysis, defining adequacy as just 20g weight gain per day, 10.7% of babies were deemed to be not getting sufficient milk.

Image credit: Selbe Lynn, flickr. CC BY-NC-ND 2.0

So there we have it; this truly seems to be the only rigorous data I can get on how often breastfeeding just does not functionally work, despite motivation and good support. It’s miles away from 0.1%, way above 2%, and a lot higher than 5%. Crucially, these data suggest that difficulties in achieving exclusive breastfeeding (even when well supported) cannot be described as “statistically rare” – accepted terminology for describing an outcome that affects over 10% of an at-risk population would be “very common”, and between 1 and 10% would be “common”.

It’s clear though that this is just one study, from many years ago, and we can fully expect that prevalence of such a multifactorial problem will vary between settings. Therefore it would be hugely valuable to have estimates of functional milk insufficiency from a variety of settings and countries. Evidently, we can’t assume something this complex is just a “single statistic”.

What’s clear is that although there are a lot of individual risk factors for not being able to make enough milk, the list of currently understood risk factors (prior breast surgery, hypoplasia, blood loss etc) don’t collectively “add up” to the totality of the problem. A fascinating review brings together recent discoveries relating to the influence of many hormonal, genetic and other factors on lactation. One of the researchers, Shannon Kelleher, describes multiple genetic variants (many of which are surprisingly common), that have recently been implicated in humans, affecting either different hormonal or genetic pathways involved in milk production or which have influences on the adequacy of milk nutritional composition.

Evidently, science is only beginning to scratch the surface of the problem. I completely agree with obstetrician-gynecologist Alison Stuebe when she says:

“mothers are suffering, whether they lack glandular tissue and or they lack self-efficacy and support. We need mothers for whom lactation doesn’t work to know that they are not alone. And we need to demand research to develop the tools that will identify the underlying problems and allow us to implement the appropriate treatment”.

And Peter Hartmann, another lactation physiologist, comments:

the utter void of knowledge about the breast is downright astonishing”.

Where does this leave mothers who hear the message (e.g. here from the RCPCH): “With the right support and guidance, the vast majority of women should be able to breastfeed”? Policy guidance which omits to acknowledge both the fact that insufficiency is common, and our lack of detailed knowledge about it (ie, exactly why it happens and how it can be fixed), sets these mothers up for feelings of failure from the start. Good support, therefore, means being clear that we know non-ignorable proportions of mothers in developed country settings are likely to experience physiological problems with lactation. We can combine the information from studies such as those of Neifert and Kelleher with decades of research from breastfeeding advocates such as Kathryn Dewey, who have documented time and again that substantial proportions of mothers in developed country cohorts experience biological problems such as delayed onset of lactation, even when well supported and in the context of best current practice (eg: here, here and here).

We need to be honest with mothers, and avoid giving bold impressions of certainty where none exists. We can admit that some research shows that in developed country settings, many mothers may indeed have difficulties producing enough milk. We need to be clear that best practices and good management can optimise the chances of success, but do not guarantee it. We can reassure mothers wanting to breastfeed that it’s not about making the great the enemy of the good; she can continue to breastfeed alongside giving what supplementation is needed to provide for the baby’s needs. And in the absence of high quality evidence to support any specific interventions for improving milk supply, we must avoid giving mothers laundry lists of non-scientific or unproven interventions (cookies, herbal therapies…), which will only engender distrust in breastfeeding support services, or make her feel she needs to jump through ever escalating hoops in search of her desperate goal.

Finally in the words of one of my favourite bloggers on infant research, the neonatologist Keith Barringon, we can send the message that:

“A mother who feels good about herself and her attempts to give breast milk to her baby, as much as she can for 3 or 4 months if she can, will be much more likely to be successful than if she is made to feel inadequate by policies or by individuals that demand perfectly exclusive breastfeeding for 6 months.

Let her give… a few bottles of formula, or pumped breast milk… are not only harmless, they are better than harmless, they may help the mother and her baby to get over the hump and carry on breastfeeding, rather than giving up all together”.

I doubt this will appeal to some advocates out there, but let’s not throw the baby out with the bathwater, and try to be both mother- and baby-friendly in our advocacy and support.