Breastfeeding - Does Science Mislead Parents & Professionals?

"Breastfeeding tied to stronger lungs, less asthma"

Was a headline that caught my eye this week.

"New studies show that breastfeeding strengthens babies' lungs and could reduce risk of developing asthma later in life."

It continued.

Naturally after a scan I reposted on my Facebook wall - but at the same time I flipped the language the correct way up:

"New studies show not breastfeeding tied to weaker lungs & more asthma"

One reply in particular caught my attention. A mum saying that she couldn't find anything that actually said not breastfeeding was tied to weaker lungs and more asthma, what the article said is that breastfeeding is tied to stronger lungs and less asthma, just because a study finds something it does not mean that the opposite is true.

Huh?

The study in question found:

" FEF50 was higher by 130 and 164 ml in children breastfed for 4-6 and >6 months respectively "

FEF50 is the Forced Expiratory Flow at 50% - ie the flow rate at the halfway point of the total volume exhaled. The larger this number is the better.

This means if you split the babies into 3 groups, and give them easy hypothetical values eg - those not breastfed (Group A) have a FEF50 value of 1, those breastfed for 4+ months  (Group B) score 2, and then those breastfed for more 6+ months (Group C) score 3  - we can report this in a number of ways.

We could say group A have a reduced FEF compared to group B & C, OR we could say group C have an increased FEF compared to group A & B.  Both statements are factually true and correct - it all depends which you choose to compare against.

What doesn't make sense is to say we agree group C has a higher score than group A, but then argue there is no evidence group A has a lower score than C.

It's like saying Bob is taller than Jim, but there's no evidence Jim is shorter than Bob!

We know human milk is the norm for human infants - therefore this has to be the basis for comparison surely? How do alternative substances compare to the real thing?

I often use the analogy of a manufactured blood substitute, because breastmilk is like blood in that it is a live bodily fluid.  If research showed people who had been given manufactured blood during a transfusion were at increased risk of say cancer, would we not want the blood thoroughly testing to ensure it wasn't the case? Would we be saying "real blood reduces risks of cancer"?

The other issue beyond giving a false belief that non breastfed children have a "normal risk" compared to those breastfed a magic elexir that "reduces" risks of disease, is that it can be misleading:

"A team led by Karen Silvers found that each month of exclusive breastfeeding (without any formula mixed in) was tied to a nine-percent drop in asthma risk."

One mum on Facebook asked, does this mean babies are born with a high asthma risk and each month of breastfeeding reduces that further?  I think it's a fair question given the language researchers chose to use.

Of course what they actually meant was each month of exclusive breastfeeding was tied to a nine-percent drop in asthma risk COMPARED TO THOSE THAT ARE NOT BREASTFED.

It's misleading.

Diane Wiessinger was I believe the first person to discuss watching our language back in 1996 (a must read if you haven't) - a piece which highlights perfectly how inflating the milk of own on species to "best", perpetuates the normalisation of infant formula.  Something I've previously blogged about here and here - and in no place is it more important than science.

A 2008 study examined exactly this:

"Do Journal Titles And Abstracts Accurtely Convey Findings On Differential Health Oucomes For Formula Fed Infants"

They wondered whether health professionals have access to unbiased information on the health implications of formula feeding.

They asked:

"Does the information conveyed to the casual reader of relevant scientific journals accurately convey their findings that formula-fed infants have worse health and development outcomes than the norm, that is, breastfed infants?

Or alternatively, is breastfeeding dammed with faint praise, and/or formula feeding treated like Harry Potter’s nemesis Voldemort – as ‘He Who Must Not Be Named’."

Method: The 2005 AAP Policy Statement on Breastfeeding and the Use of Human Milk cites 78 scientific studies as evidence that breastfeeding is protective against a range of infectious and chronic diseases. We systematically analysed the information content of titles and abstracts for these key studies on the health benefits of breastfeeding.

Their findings were interesting:

• Only 4-6 percent refer to infant formula in the title, for example, ‘Differences in morbidity between breastfed and formula-fed infants’, ‘Cow’s milk exposure and type I diabetes mellitus’
• 63-67 per cent had only a neutral statement in the title, or referred to the protection conferred by breastfeeding, reinforcing the cultural norm of breastfeeding as ideal rather than usual or ordinary
• 29-36 per cent misleadingly associated breastfeeding with illness or disease, through statements implying guilt through association, such as ‘Breastfeeding and risk of post neonatal death in the United States’, “Breastfeeding and the sudden infant death syndrome’.
• 16-22 per cent compare breastfeeding to artificial feeding with conclusions couched in terms of the ‘advantages’ of breastfeeding
• 72-74 per cent make no mention of artificial infant formula, and would not challenge a reader’s erroneous belief or assumption that artificial feeding carries no increased health risks for infants

They conclude:

The AAP Policy Statement on breastfeeding and human milk stated that ‘exclusive breastfeeding is the reference or normative model against which all alternative feeding methods must be measured with regard to growth, health , development and all other short-and long term outcomes’.

This has implications for research design which are rarely considered. Most studies examined above hypothesized formula feeding as the norm and tested for differential outcomes from breastfeeding defined as the ‘exposure/intervention’. 

This approach can bias research through how the research hypothesis is specified, and through poor specification of infant feeding categories, with a tendency to underestimation of risk associated with non human milk feeding.

In recent years commentators have also highlighted the bias and negative effects on breastfeeding practices of normalizing artificial feeding, referring for example, to ‘the benefits of breastfeeding’ rather than, for example, ‘the risks of formula feeding’.

Nevertheless, surveys reveal considerable cultural ambivalence and ignorance about the health consequences of artificial infant feeding. For example, some 30 per cent of mothers surveyed by the United States’ ‘Babies Were Born To Be Breastfed!’ Campaign agreed with a statement that ‘infant formula is as good as breastmilk’, and only a minority of the survey population agreed that ‘a breastfed baby is less likely to get ear infections or respiratory illness’.

 Likewise, a clear majority of public opinion in the United States supports the view that ‘breastfeeding is healthier for babies’, yet substantially more than half of the surveyed population disagree that ‘feeding a baby formula instead of breastmilk increases the chances the baby will get sick’. 

If exclusive breastfeeding was the norm against which other methods are measured, breastfeeding would not be ‘protective’ and breastfed infants would not enjoy ‘lower risks of ill health’; they would instead be referred to as ‘normal’, while formula fed infants are in fact ‘exposed’ to increased risk of poor health and development.

Whilst we often blame the media for twisting research, and indeed at times they clearly do! Perhaps we need to go further still back down the chain and start by ensuring the science is sound?