Wednesday, April 08, 2015

Breast Feeding Might Not Reduce Childhood Obesity: Gasp

Breast milk might not be the new Slim Fast for the stroller set. A new paper suggests breastfeeding might not greatly reduce risk of childhood obesity. Yes, a study that questions one of the oft-hyped benefits of breast milk. Note it on your calendar. Note too that the media ignored it.

So let me fill you in...


Researchers from The Cincinnati Children's Medical Center reviewed studies relevant to a link between breastfeeding and childhood obesity. They report some research shows a link but caution the evidence is riddled with some serious limitations (e.g., confounding factors, publication bias, analysis bias). They argue breastfeeding likely impacts obesity in a more indirect and moderate manner than previous research has suggested. This makes a lot of sense. It’s not like breast milk marches into the infant, finds the obesity switch and flips it to slim.

The routes from boob to BMI, they argue, are more circuitous and involve factors like maternal weight, early taste preferences and digestion/gut flora. For instance, breast milk appears to provide more good gut bacteria which might play a role in later obesity. Breastfed babies also tend to have fewer ear infections and respiratory illnesses so this might lead to less exposure to antibiotics that have been linked to obesity. Or take diet preferences. Breastfed babies have been shown to like a wider range of healthy food. Research has not yet provided a link between breastfeeding, food preference and later childhood diet but it's plausible. 

So what do the authors conclude?

The best observational evidence to date, compiled across over 80 separate studies conducted over at least 20 years, suggests that breastfeeding, especially for longer durations or more exclusively, is associated with a 10–20 % reduction in obesity prevalence in childhood. However, the complex nature of the relationships between breastfeeding and obesity, including the fact that human milk is variable between women and breastfeeding may have differing effects in specific population subgroups, suggests that the concept of promotion of breastfeeding as a front-line strategy for the primordial prevention of obesity is not supported by the literature. [my bold]

In other words breastfeeding in and of itself will not greatly prevent childhood obesity. So don’t look for the Skinny Bitch Breast Milk Latte quite yet.

Speaking of products, the lead author receives funding from Mead Johnson, yes Big Formula. Does this instantly discredit this study? No. The authors report they believe a true link exists between breast milk and weight and advocate for more research elucidating the link. To some degree it’s in the formula industry’s best interest to identify (and then replicate) the seemingly magical ingredients in breast milk. The authors have provided a balanced review and addressed the limitations of existing research, a rarity in the breastfeeding literature. 

Some of these flaws include:

Establishing this relationship between breastfeeding and obesity protection has been notably difficult for many reasons, including imprecise recall of breastfeeding timing and intensity, incomplete accounting for covariates that may affect both the decision to breastfeed and obesity propensity, and long intervals between the exposure (human milk or breastfeeding) and outcome (obesity development).

Again, these concerns apply to a large swatch of BF research.

The closing, however is rather curious.

One assumption of this review is that a relationship between breastfeeding and obesity development in the offspring is real, and even plausibly causal. However, what if breastfeeding and human milk are actually not the drivers of differential obesity development in offspring? This is also possible and would shape a different set of questions relating to the relationship between breastfeeding and obesity risk. In this instance, breastfeeding behavior needs to be considered as embedded in a fabric of other positive attributes. For example in the USA, the decision to breastfeed may be a marker for mothers otherwise pursuing a healthy lifestyle, with knowledge about healthy choices, and the time, income and social support to translate that knowledge into the non-trivial choice to initiate and continue breastfeeding, in order to accrue the many known benefits for their infant’s and their own health. Thus, encouraging breastfeeding among those not otherwise inclined, without improvement in maternal diet, physical activity and other healthy lifestyle factors, may have limited impact on childhood obesity, as the often-cited PROBIT study suggests.
So one of the authors is funded through Mead Johnson and might have some motivation to downplay the link between breastfeeding and obesity. But also maybe some motivation to find the link, if it’s really there!

Some will bash this paper based solely on the formula connection. I don’t think it’s fair. It presents research in a balanced light and takes a more skeptical approach to the literature.

And let me pose a question. 

What about the heavily cited study from the Harvard researcher claiming breastfeeding would save 900 babies from dying every year in the U.S., mind you, the same woman who also actively advocates for banning infant formula? 

How is she less biased than the present authors? At least the latter disclosed their conflict on the paper, as per new requirements and practices that explicitly refer to funding. So readers know about the funding when they get to the end of the paper where the disclosure appears. The Harvard suboptimal breastfeeding paper, by the way, has been cited by the Surgeon General, received gobs of media attention and contains no such mention of activism or possible conflict of interest (and is not required to do so). The present authors also don't make any extreme claims - like the outlandish one that more breastfeeding would have saved children in the United States from dying, for instance, from asthma. 

Nor do the present authors hide behind any scientific goodbledygook or mysterious calculations (not fully explained in the paper) that might deter readers from asking too many questions or readily understanding or even identifying their outcome measures and thus ultimately the meaning of their results. Ahem. 

Thoughts?


You can read the present study in a PDF for free or find the summary at Science Daily.

No comments: