An article entitled “The Burden of Suboptimal Breastfeeding in the United States: A Pediatric Cost Analysis,” by Bartick and Reinhold, was published in Pediatrics 2010 April 5. According to this news report, it showed that 900 babies’ lives and billions of dollars could be saved every year in the U.S. if we could get 90% of mothers to breastfeed for at least 6 months. It says breastfeeding has been shown to reduce the risk of stomach viruses, ear infections, asthma, juvenile diabetes, Sudden Infant Death Syndrome and even childhood leukemia.
This new study did not provide any new evidence. It simply took risk ratios from a three year old government report, extrapolated, and estimated the costs.
The report it is based on, the 2007 breastfeeding report from the Agency for Healthcare Research and Quality, examined 43 primary studies on infant health outcomes, 43 primary studies of maternal health outcomes, and 29 systematic reviews and meta-analyses that covered some 400 other studies. They found that
a history of breastfeeding was associated with a reduction in the risk of acute otitis media, non-specific gastroenteritis, severe lower respiratory tract infections, atopic dermatitis, asthma (young children), obesity, type 1 and 2 diabetes, childhood leukemia, sudden infant death syndrome (SIDS), and necrotizing enterocolitis.
They found
no relationship between breastfeeding in term infants and cognitive performance. The relationship between breastfeeding and cardiovascular diseases was unclear. Similarly, it was also unclear concerning the relationship between breastfeeding and infant mortality in developed countries.
So how could they take a study that showed no clear relationship with mortality and re-interpret it to predict that 900 lives a year could be saved? They used statistical skullduggery. They went to other statistical sources to find the rates of breastfeeding and the overall death rates from diseases like asthma. Then they used their imagination to estimate how many of these deaths involved non-breastfed children. Then they combined those estimated death rates together with the odds ratios from the AHRQ study to do their calculations. That’s not kosher.
There are other factors to consider. One of the reported adverse effects, necrotizing enterocolitis, is largely a disease of newborns who are premature and have low birth weights. Some of the diseases are treatable and not usually serious, like otitis media. And the risk of otitis in bottle fed babies can be decreased by not letting the child hold the bottle or take it to bed. For some conditions like atopic dermatitis, the risk depends on the family history: in this study there was an increased risk of atopic dermatitis with breastfeeding when parents had no history of allergies.
It’s interesting to read all the caveats in the text of the AHRQ report, especially about the dangers of relying on systematic reviews and meta-analyses when the individual studies those reviews are based on may be flawed. 80% of the studies included in their analysis were surveyed only via these secondary sources. There are individual studies that contradict their findings for most of the conditions they studied. The report’s conclusion cautioned:
A history of breastfeeding is associated with a reduced risk of many diseases in infants and mothers from developed countries. Because almost all the data in this review were gathered from observational studies, one should not infer causality based on these findings. Also, there is a wide range of quality of the body of evidence across different health outcomes.
That’s not exactly a ringing endorsement of reliable data to base a cost assessment on.
One of the commenters on the news story said
There is no reason for a healthy well-fed mother not to breast feed her baby
I beg to differ. There are a lot of healthy well-fed mothers who have found what they think are valid reasons not to breastfeed. I chose not to breastfeed my babies because it was inconvenient, time-consuming, interfered with my sleep, and was incompatible with my job as a doctor working 24 hour shifts in the emergency room and as a flight surgeon on call. I suppose I could have pumped milk and planned ahead and found a way to do it, but it would have required heroic measures. I can imagine leaking breast milk all over my flight suit when I was on an emergency helicopter mission and simply couldn’t stop to pump. Moreover, I tried breastfeeding briefly with my first baby and frankly, I didn’t like it. All in all, I thought my babies were better off with a happy mother and a bottle.
This new study confirms what we already knew: that breastfeeding is better for a baby than bottle-feeding. The question is how much better, and this study really can’t answer that question. It consists of estimates based on estimates based on mixed data of varying quality. Considering the quality of the data and the pitfalls of epidemiological studies, it is likely that this new study overestimates the value of breastfeeding and the number of preventable deaths.
If we could accurately calculate the numbers needed to treat (NNT) with breastfeeding to save one baby’s life or prevent one ear infection, they would be very high numbers. Mothers should be given those numbers; but they should also understand that if they bottle-feed, the odds are good that their child will thrive.
Breastfeeding is clearly better for babies, and I strongly support it, but I think the facts leave us room to support those women who make an informed choice not to breastfeed. Some women can’t produce enough milk or have health problems that interfere with breastfeeding. Some women know the benefits of breastfeeding but choose not to do it. We may not agree with their choice, but we can respect their autonomy. Thank goodness we now have safe, nutritious infant formulas that give us a choice.
This article was originally published in the Science-Based Medicine Blog.
