Does Extended Bottle-Feeding Really Cause Obesity?

Obesity Campaign Poster

Obesity Campaign Poster (Photo credit: Pressbound)

At 21 months, Maya still really likes hitting the bottle. It’s a ritual — the first bottle of the morning — and a request as soon as I get home from work. She sits on my lap, we cuddle, and she relaxes a bit, her body getting softer and less tense. In the evenings, I don’t give her much milk because it will ruin her dinner. We both know it’s just the pose that matters, and the snuggles that are part of that nice, quiet pause.

So of course I was immediately concerned and even a bit perturbed when my pediatrician rather abruptly told me at our last visit to stop using bottles “cold turkey” because their use is linked to obesity. Her less-than-insightful suggestion was to just get rid of all our bottles at once, and thereby make it physically impossible for Maya to keep using one. At the time, I should have asked her if she wanted to come visit for that little period of self-inflicted hell, even if just to explain to my daughter that we are only depriving her of this small comfort in order to make sure she won’t eventually become overweight.

Regardless of her apparent cluelessness about the importance of easing children into changes in their lives, I had to take seriously the problem she raised about bottles. So I went and read what I could about the study linking bottle usage to obesity (the actual text of the study is $31, and IMHO, not such a good investment).

The study, from the Journal of Pediatrics in May of last year, made headlines at the time that carried its message, including articles titled like this one: “To Avoid Adult Obesity Stop Bottle-Feeding at 18 Months,” from Medical News Today, which intoned darkly:

If you want to reduce your baby’s chances of becoming an obese adult you should not continue bottle-feeding him/her beyond 12 to 18 months.

Who wants a fat kid, really? Or this one, from U.S. News, “Prolonged Bottle Feeding Boosts Kids’ Obesity Risk,” which begins:

Nearly one-quarter of 2-year-old bottle feeders were obese at age 5, researchers say.

Well, I suppose that’s clear enough. But what did the research really say? Here’s more detail from the abstract:

Data from the Early Childhood Longitudinal Study, Birth Cohort were analyzed for 6750 US children born in 2001. The outcome was obesity (body mass index ≥95th percentile) at 5.5 years, and the exposure was parental report of the child using a bottle at 24 months. The prevalence of obesity at 5.5 years was 17.6%, and 22.3% of children were using a bottle at 24 months. The prevalence of obesity at 5.5 years was 22.9% (95% CI, 19.4% to 26.4%) in children who at 24 months were using a bottle and was 16.1% (95% CI, 14.9% to 17.3%) in children who were not.

Prolonged bottle use was associated with an increased risk of obesity at 5.5 years (OR, 1.33; 95% CI, 1.05 to 1.68) after controlling for potential confounding variables (sociodemographic characteristics, maternal obesity, maternal smoking, breastfeeding, age of introduction of solid foods, screen-viewing time, and the child’s weight status at birth and at 9 months of age). [Emphasis added.]

I’m struck by several things right off the bat. First, although nearly 23 percent of bottle-feeders were obese at the age of 5 1/2, 16 percent of the rest of the population (i.e., not bottle users) also were, which is only a 7 point difference (though it’s true that the association appears to hold at this level of the analysis).

Second, the sample size is on the small side in terms of who’s left — i.e., 22 percent of the sample used a bottle, 23 percent of whom ended up overweight. That’s a total of 341 kids. If we subtract out the 16 percent that is the general rate of obesity in the remaining population, we’re down to 55 kids whose habits and body weight are driving the conclusions (because they make up that 7-percent spread). The authors say that is a statistically significant number, though, so let’s look at their assumptions more closely.

They used a data set with limited inputs, to be sure. The first glaring omission is that the study did not account for what was in the bottles. Apple juice, for example, does not fill the stomach the way that milk does, and it creates a taste for sugary drinks in children, making it easy to consume to excess. It also contains a significant number of calories (117 per cup).

Whole milk, on the other hand, may be higher in calories, but offers a host of essential fats, vitamins and calcium. It is harder (though certainly not impossible) to over-consume because it is both filling and satisfying. Water, obviously, has no calories.

Formula, much of which is loaded with sugars that stimulate appetite, unsurprisingly is also linked in previous studies to obesity. Researchers here indicate they controlled for breastfeeding as a variable. But the formula versus breastfeeding research is complicated by studies that show bottle-fed infants gain more weight even if the bottles contain breastmilk, meaning that merely controlling for breastfeeding may not be enough.

Given that children are frequently given juices (or even worse beverages like Kool-Aid) to drink, and the small number of families whose habits are driving the conclusions, this seems like an important caveat to the findings, and one that was notably missing from the official conclusion or from the reported coverage of the study.

Instead, the authors publicly suggest the opposite, as here, where one of them claims that the study accounted for “feeding practices during infancy.” Um, I don’t think so. The two variables “age of introduction of solid foods,” and “breastfeeding” are certainly tangentially related to overall infant feeding habits (and perhaps, health), but when a study is attempting to measure the impact of bottle-feeding, controlling for the contents of that bottle strike me, at least, as one of the more important variables to be included in the equation. After all, what a child is actually consuming has just got to be more important than whether it’s being delivered by bottle or cup.

In USA Today’s piece on the article, another expert is quoted on the need to cease bottle use:

“Drinking your calories may not be as filling as eating them,” says Jennifer Shu, a pediatrician in Atlanta and the editor of, a consumer website of the pediatrics academy. “That’s where the obesity problem comes in. It’s so easy to drink the calories, but people often are still going to eat the same amount of food.”

This argument seemed reasonable to me at first glance, but actually doesn’t really hold up. Certainly, the regular visits I made to Jamba Juice during law school likely explain why my exercise regimen at the time yielded disappointingly paltry results. Yet I don’t observe that Maya eats the same amount of food if she is full from a bottle — in fact, I worry that milk will displace other calories because she won’t be hungry, and so we limit the amounts she can drink around meals.

And all this likely misses the point. Above, Shu appears to suggest that children will, in a sense, over-drink (or over-eat because they drank too much). But so long as what children are drinking is good for them, and they are drinking and eating solid foods in the right balance, it seems to me that we wouldn’t want them to drink less. In other words, if the issue is amount, what should it matter if the drink comes from a bottle or cup? Again, parental monitoring of what is consumed, and how much, should matter far more.

Two mice; the mouse on the left has more fat s...

Two mice; the mouse on the left has more fat stores than the mouse on the right. (Photo credit: Wikipedia)

Second, the study variables omit consideration of the kind of bottle being used, whether glass or plastic. Before you think I’ve gone off the deep-end on this one, consider that studies have shown that Bisphenol-A (BPA) likely plays a significant role in obesity, both by making our bodies produce insulin as though we are consuming twice the calories we actually are, and by helping to flip a genetic switch that predisposes us to be fat.

The study’s data-set spans from 2001 to 2006, a period in which most parents were unaware of the pernicious BPA-in-baby-bottles issue and most bottles still had BPA in them, and in which plastic bottles were the norm, as they still are today. It would need far more study, of course, but in my view it’s at least possible that this is yet another instance of a simplistic analysis of behavioral factors that leaves the possibility of harmful chemical influences utterly undiagnosed.

Third, the authors’ recommendations fail to account for countervailing values in child development that may lead some families and children to benefit from extended bottle use, at least as part of their repertoire. Here’s how one of them breezily put it in an article on the study:

Rachel Gooze [] notes that weaning children from the bottle by the time they are 1 year of age is unlikely to cause harm and may prevent obesity. The authors suggest that pediatricians and other health professionals work with parents to find acceptable solutions for stopping bottle use at the child’s first birthday.

Yet research unequivocally shows that strong bonding with caregivers and relaxation (i.e., low anxiety) is essential to healthy brain development, particularly in young children ages 0 to 3 years. While extended use of a bottle is certainly not an essential part of creating these bonds and a relaxing atmosphere, the act of feeding a child is intrinsically a nurturing moment, and so it may not be irrelevant either. The researchers should have at least considered the possible downsides here.

For our family, Maya never breastfed (which is another story entirely), and so our bonding over a bottle has replaced a rather fundamental missing piece. I’m not eager to let this go based on one study showing she could, maybe, have a slightly greater chance of being obese four years from now, especially given the care and intention I take with her overall diet and the monitoring we do generally of her health, including her weight.

Obesity Campaign Poster

Obesity Campaign Poster (Photo credit: Pressbound)

For example, back on what goes into the bottle (and the baby), Maya almost never has juice, or really concentrated sugar of any kind, including the supposedly “kid-friendly” (non)foods: fruit leather, sweetened yogurt or those mushy fruit slurries in suck-down containers. (I’ll write a post on the re-joined debate over sugar and it’s impact on the body soon.)

If continuing to use a bottle appeared to be causing cavities or hampering her speech development, that would be another issue entirely, and is a legitimate concern raised by dentists (those sugary beverages again) and speech pathologists. In Maya’s case, she now has (I would guess) about 300 words and more every day. She’s also never been very interested in a pacifier or thumb-sucking, either of which can also be a speech development blocker. Moreover, she eats a wide variety of fruits, proteins and vegetables, uses both sippy cups and regular cups, and is learning to use a straw, pursuant to the advice of speech experts.

The bottle is merely a respite from these other ways for her to drink, and I assume will drop away sometime when she’s moved beyond the need for that to be our daily form of checking in. If not, we’ll ease it out of use and replace it with another important bonding ritual we can invent.

In the end, I’m unconvinced by this study, and disappointed that both my pediatrician and the mainstream press appear to have taken its limited data and recommendations as gospel. Clinical advice from most doctors rarely seems to take account of the havoc that would be wreaked on families’ emotional lives by following their rigid approach. And the discourse around the obesity issue has reached such a fever pitch that, as parents, it seems we’re now in a position, essentially, to be bossed around by experts on “slim” evidence indeed.

I hope that parents think through the issue for themselves before feeling guilted into suddenly dropping the bottle, at least based only on this latest — and in my view rather dubious — pronouncement.


How does your family come down on this issue? Am I just making up excuses because I don’t want to face the music (or really, screaming)?

Did I miss something important about the study or its implications? Or do you agree with me that this is just another in a too-long line of simplistic anti-obesity messages that fail to grapple with the real issues?