by Emily Oster
The best evidence we have is from one (small) randomized trial of two hundred families.5 The results support some of the claims about baby-led weaning, but not all. Parents reported less food fussiness, and the infants in the baby-led weaning group were more likely to eat with their family. They were also likely to be breastfed longer, and the introduction of food was pushed later (i.e., to around six months rather than four).
On the other hand, this study did not find any differences in whether children were overweight or obese by the age of two, and they didn’t find any differences in the nutrients the children consumed or their total calorie intake. The researchers noted that this was hard to measure given the smearing around of food. The kids did eat slightly differently—the baby-led-weaning group was more likely to have meat and salt, for example—but these differences didn’t go in any systematic direction.
One of the main concerns with this approach is that it could lead to choking, if infants are unable to swallow big pieces of things. The study showed that it was no more common in the baby-led-weaning group than the traditional spoon-feeding group. Choking is, however, reasonably common in all babies, and people in the study were encouraged not to introduce foods that presented significant choking hazards. A four-month-old shouldn’t have large pieces of hard fruit, baby-led weaning or not.
This study followed two hundred people; clearly, learning detailed answers to these questions would require a lot more than that. If you do want to try baby-led weaning, there is nothing in the evidence to say it is a bad idea. If you do not, there is also nothing compelling to say you should go out and do it.
A final note on timing: There is some debate about the right time to introduce solid foods and, in particular, a question of whether introducing solids too early will lead to obesity later. What is the reason to wait for four months at all? Should you really be waiting for six months, or longer? The reasons to wait until four months are largely physiological—babies really cannot eat before this—but waiting longer than that doesn’t seem like it matters. There is some correlation between the timing of food introduction and childhood obesity, but it seems to be due to other factors, like parental weight and diet.6
DOES WHAT YOU FEED YOUR KIDS MATTER?
Deciding whether to start with purees is one thing, but there is a more important question here: What, exactly, should you be feeding your child? The bottom line is that more or less everyone on the planet eats, and they more or less all eat solid food, so regardless of how you introduce foods, you’re likely to end up with a child who eats something.
There is no guarantee, however, that your child will like a wide variety of foods, will eat healthily, and will be willing to try new things. Perhaps it isn’t difficult to produce a child who will eat chicken nuggets and hot dogs, but how do you end up with one who loves sautéed kale and kimchi with squid? Or at least one who will try them?
Let’s acknowledge: this issue may not be important to everyone. You may care that your child is willing to eat some vegetables, but you may not particularly care if they are picky or not. There is nothing wrong with a child who eats only broccoli and pasta, as long as that works for your family. Going further, you may not care if the child eats only pasta, figuring they’ll get into broccoli when they grow up. You will need to think more carefully, in this case, about how your child will get the necessary vitamins, but otherwise this is not obviously problematic.
How much you care about this is likely to depend on how your family eats. For a while I was making two dinners—one for Penelope and a later one for us—and it got to be too much. Ultimately, we altered both what we ate and what she ate so we could eat together. But many people are fine with the system of two dinners.
Let’s assume, however, that you do care about promoting a “healthy diet.” The good news is that there is plenty of research on this question. The bad news is that a lot of it is not very good.
Consider a paper from 2017 that got a lot of media attention.7 The authors followed 911 children from age nine months to six years and related their early diet to their later diet. They found that children who ate a varied diet—and in particular those who consumed a wide variety of fruits and vegetables—at nine months were also more likely to eat a varied diet with vegetables at age six.
The researchers concluded that tastes are formed early, and it is therefore important to expose children to a variety of foods early in life.
This is certainly one possible explanation for the results. But it is by no means the most likely one. A much more plausible explanation is that the parents who feed their children vegetables at age one are also likely to feed them vegetables at age six. This is just a very basic causality problem, and it is difficult to learn anything here.
However, we can get some clues about the true underlying relationships from smaller, more indirect studies.
Consider the following quite neat example. Researchers recruited a group of moms and randomized them into a “high-carrot” or “low-carrot” diet during pregnancy and lactation. The high-carrot moms were drinking a lot of carrot juice.
When their children were ready for rice cereal, the researchers offered them (the babies, not the moms) cereal made with water, or one flavored with carrots. The kids whose moms had eaten more carrots were more likely to prefer the carrot cereal (as evidenced by their consumption and their facial expressions, and presumably also whether they picked up the dish and threw it on the floor).8 This suggests that flavor exposure—in this case, thorough the placenta and through breast milk—affects whether children are receptive to new flavors.
Related to this, once children are starting to eat solid foods, there is randomized evidence that repeated exposure to a food—say, giving kids pears every day for a week—increases their liking of it. This works for fruits, but also for vegetables, even bitter ones.9 It reinforces the idea that children can get used to different flavors and that they like familiar ones.
This shouldn’t be too surprising. People eat differently in different cultures, and we know people continue to express preferences for the foods they ate as a child, even if they move to another location.10
Putting this together, on one hand, from a global public health perspective, I would be extremely hesitant to conclude that lack of exposure to vegetables at age one was the main problem with older children’s diets. The problem is more likely to be with the foods kids are offered at both ages. On the other hand, from the standpoint of an individual parent, if you want your child to eat a variety of foods, this suggests it is beneficial to expose them—repeatedly—to these flavors.
However, even if you eat all kinds of weird stuff while breastfeeding, and carefully expose your child to Brussels sprouts for weeks on end, they may still end up being somewhat picky about their food. Researchers classify this pickiness into two groups: food neophobia (fear of new foods) and picky/fussy eating, in which the child just doesn’t like a lot of different foods.
Before getting into these, and how you might fix them (hard), you should know that most kids become more picky around two and then slowly grow out of it in their elementary school years. This is sometimes a surprise to parents—your eighteen-month-old eats like a horse, then all of a sudden around two, they start being very selective and just generally not eating much. I have sat at many a dinner where one of my kids has taken one bite and said, “I’m done!”
This change can lead to unrealistic expectations from parents about how much their toddler and young child will eat. As a review article from 2012 notes, “The majority of children between one and five years of age who are brought in by their parents for refusing to eat are healthy and have an appetite that is appropriate for their age and growth rate.”11 The article goes on to note that the most useful treatment for this problem is parental counseling, not anything to do with the child. Thanks for the judgment, researchers.
This suggests that even if your child does
n’t eat that much some of the time, you probably shouldn’t be overly concerned, but it doesn’t answer the question of how you can treat or avoid general pickiness. This is a topic of some research interest. One study I like a lot followed sixty families of kids aged twelve to thirty-six months as they tried introducing a new food. The families videotaped their dinner interactions for a night so researchers could study what seemed to influence the new food adoption.12
This study reported what parents actually do rather than what they say they do. This is good, since none of us is especially good at reporting our actual behavior. The primary finding relates to how parents talk about the new food. Kids are more likely to try to eat it with what researchers call “autonomy-supportive prompts”—things like “Try your hot dog” or “Prunes are like big raisins, so you might like them.” In contrast, they are less likely to try things if parents use “coercive-controlling prompts”—things like “If you finish your pasta, you can have ice cream” or “If you won’t eat, I’m taking away your iPad!!”
Other studies show that parental pressure to try new foods or to eat in general is associated with more food refusal, not less.13 These studies also show that food refusals are more common in families where parents offer an alternative. That is, if your kid doesn’t eat broccoli and then you offer him chicken nuggets instead, he may learn that this is always the reward for not eating new foods. This problem is exacerbated by parents’ concern that their child isn’t eating enough (which, see above, is probably not true).
Putting this together leads to some general advice: offer your very young child a wide variety of foods, and keep offering them even if the child rejects them at first. As they get a little older, do not freak out if they don’t eat as much as you expect, and keep offering them new and varied foods. If they won’t eat the new foods, don’t replace the foods with something else that they do like or will eat. And don’t use threats or rewards to coerce them to eat.
This advice is easy to give but it can be hard to take. It is frustrating to sit at a meal that you know to be delicious with a four-year-old who screams that they hate it and will not eat anything. I don’t have a great solution for this, other than earplugs.
I also tried to train Finn to say “I don’t care for pot roast” rather than “I HATE POT ROAST,” since it at least sounds more polite, even if still combined with pushing the plate away and putting on an angry pouty face. (Parenting: It’s a long game.)
All this discussion is predicated on the assumption that your child doesn’t actually have a problem with weight gain or nutrition. If you are worried, this is what the pediatrician is for—they can check on weight gain, malnutrition, vitamins, and so on. For children who are malnourished, there is a whole other set of guidelines, most of which are more intense and involved, for increasing eating.
ALLERGENS
The story at the start of this chapter gives a sense of how the recommendations for peanuts have changed: introduce early, not later. What the story doesn’t convey is whether this translates more generally to allergenic foods, and exactly how you are supposed to introduce them.
On the first question, the answer is probably yes. The vast majority of allergies result from eight food types: milk, peanuts, eggs, soy, wheat, tree nuts, fish, and shellfish. The incidence of these allergies has grown over time, perhaps as a result of better hygiene (so less allergen exposure early on), and clearly due in part to a lack of early introduction.
Milk, eggs, and peanuts make up a large share even of these. We covered the peanut evidence earlier. Other research suggests a similar mechanism is at work for eggs and milk.14 The evidence on milk isn’t as convincing as the other two, but perhaps only because large studies have not yet been released.
All this points to the possible importance of introducing all these allergens early—probably as early as four months. (Milk can be introduced in the form of yogurt or cheese.)
Importantly, although the language here is about “introduction,” these studies include regular exposure as well. It is not enough to have your kid try peanut butter or eggs. You need to actually keep giving it to them regularly.
Which leads to the question: How?
This is a setting in which going slowly is a good idea. Try a little bit at first—only one allergenic food in a given day—and see how they react. If nothing, give them a little bit more. And so on until you get up to a normal amount.
And then keep these foods in the rotation.
This is a lot, especially since most babies don’t really eat much food anyway. To consistently expose them to peanuts and yogurt and eggs on top of everything else (what about the peas?) requires some logistical work. If you are daunted, and especially if you’re very concerned about these issues, there are some (new) products that contain powdered forms of these foods and are meant to be mixed with breast milk, formula, or cereals.
OTHER FORBIDDEN FOODS
Beyond allergens, there are a few other foods on the “forbidden foods” list: cow’s milk, honey, choking hazards, and sugar-sweetened beverages. Do these belong there?
The last one is obviously not just about infancy. Soda is strongly discouraged for infants and children (and adults). Your six-month-old does not need a Coke. Juice is more controversial (and, indeed, I recall a childhood dominated by orange juice), but generally, young children should have formula, breast milk, or (once they start eating solid foods) water. Whole fruits or fruit purees are preferable to fruit juice.
Choking hazards—nuts, whole grapes, hard candies—are also to be avoided, for obvious reasons. Babies and toddlers do choke, and these foods are more likely to lead to choking. Grapes are okay in pieces, nuts are okay in nut-butter form, and hard candies are not recommended for other reasons.
Cow’s milk is probably the most complicated recommendation, partly because it interacts with the allergen issues above. It is important to introduce some milk-based foods—yogurt, cheese—to avoid allergies. But milk itself is forbidden.
The concern is that cow’s milk is not a complete infant nutrition system, and if your infant drinks a lot of milk, it will restrict formula or breast milk intake. In particular, infants who have cow’s milk as their primary milk source are more likely to be iron deficient.15 The evidence says only that you shouldn’t replace formula or breast milk with cow’s milk. As an addition to, say, oatmeal or cereal, it isn’t a problem.
Finally, honey. The concern with honey is that it could lead to infant botulism. Infant botulism is a serious disease—basically, a toxin interferes with neurological functions, including affecting the infant’s ability to breathe. It is most common under the age of six months and it is treatable, with a very high success rate. Still, the treatment is not easy: the baby typically needs to be hooked up to a breathing machine for a few days until they are able to breathe on their own again.
The toxin that causes this, Clostridium botulinum, is found in soil and elsewhere, including in honey. This, combined with the fact that there were multiple case reports from the 1970s and ’80s in which infants who developed botulism had consumed honey, led to the recommendation against honey through the first year of life (sometimes even two or three).
The question of how important honey is as a source of botulism is an open one, though. Although the ban on honey has been widely publicized over the past decades, there has been basically no change in the rate of infant botulism.16 This suggests that other sources of botulism are more important in practice. So maybe this is overkill, but the downsides of avoiding honey are also limited.
VITAMIN SUPPLEMENTATION
People spend a lot of time telling you how perfect breast milk is, how it’s the most amazing food on the planet and contains everything your baby needs! Then, in pretty much the next breath, they hand you a bottle of vitamin D drops and tell you that, actually, breast milk doesn’t have enough vitamin D and you’d better remember to give your kid these drops
every day, or they might get rickets.
I would describe remembering these drops as a “challenge” for our family. Many a yelled conversation across the house concerned whether someone had given the drop or not that day. The days blur. Was it yesterday, or three weeks ago?
Perhaps we should consider ourselves lucky that Penelope and Finn did not get rickets.
Then again, perhaps this risk is overblown.
The general wisdom of vitamin supplementation (for anyone—adults, children, babies) is complicated. It is true that if you are deficient in particular vitamins, it can cause serious problems. Vitamin D deficiency causes rickets. Vitamin C deficiency famously causes scurvy, as was first recognized in sailors who went months without eating any fresh vegetables or fruit. However, if you eat a typical varied diet—even one that’s pretty unhealthy by many standards—you are very unlikely to be seriously deficient in any of these vitamins.
Your toddler or young child does not generally need a multivitamin (no Flintstones gummies for them). If they eat only a very limited diet, it is possible a multivitamin would be necessary, but this would be unusual. Even a child who seems like a very picky eater will be getting enough vitamins to sustain them. A baby who is breastfed will get most vitamins this way as well.
The two possible exceptions to this are vitamin D and iron.
Vitamin D is not present in many foods, and is not present in high concentrations in breast milk. People do get vitamin D through sun exposure, but since many of us live in houses in cold places and not on the savanna, sun exposure isn’t always consistent.
As a result, a lot of infants and children are considered deficient in vitamin D. This could be as much as a quarter or more of white children, and higher among children of color (darker skin lowers vitamin D absorption from the sun).17 Deficiency here is defined as having a blood concentration of vitamin D below some cutoff level.
