Several years ago, a group of gut microbes went on a 14,000 kilometer-long trip. Having freshly emerged from an infant in southern Malawi, they were scooped up by eager scientists, frozen, stored in cold boxes, exchanged from one courier to another, flown across the Atlantic Ocean, driven to the Washington University in St. Louis, and finally transplanted into the bodies of germ-free mice that had been raised all their lives in a sterile bubble. Their epic voyage was part of a study by Jeffrey Gordon, showing how the microbes of our gut contribute to the problems of malnutrition—and how they might help to fix them.

Gordon is a leader in the study of the human microbiome—the trillions of microbes that share our bodies and influence our health. A few years ago, his team showed that malnourished children grow up with different gut microbiomes. These communities change during early life. Waves of species replace each other in predictable steps, much like new landscapes get colonized by lichens, then shrubs, then trees. Normally, it takes three years for the communities to reach an adult state. But this pattern of succession slows and stagnates in malnourished children, leaving them with a microbiological age that lags behind their biological one.

This hidden immaturity matters because our microbes help us to harvest energy from our diet. If they’re not doing that effectively, their hosts might suffer, especially when their diet offers very little energy to begin with. “Most people think about development from the perspective of our human cells and organs, but there’s another facet to it–our gut microbiota, an organ composed of microbes,” says Gordon. “It gives us a more transcendent view of human developmental biology.”

Now, following similar work in Bangladesh, Gordon’s graduate student Laura Blanton studied the changing microbiomes of 60 healthy Malawian infants, and devised an algorithm that calculated their microbiological age based on the species in their guts. She then applied the algorithm to another group of 259 babies, and found that their microbiological age scores at 12 months predicted how much weight they had put on by 18 months. That’s a sign that the microbes are influencing the babies’ growth, rather than merely reacting to them.

The team found another such sign by transplanting stool samples from Malawian infants into sterile baby mice. Even though all the rodents ate the same food, those that received microbes from an underweight infant put on less weight and developed weaker bones than those which received a healthy baby’s microbiome.

So, if certain gut microbes can stymie a baby’s growth, perhaps others can speed things up? To find out, Blanton simply implanted mice with microbes from either a healthy infant or an underweight one, and housed them together in the same cages. Mice willingly eat each other’s poop, and so regularly bombard their own microbiomes with those of cage-mates. And in these “Battles of the Microbiomes”, the healthy communities came out on top, invading and displacing the immature ones.

The team then identified five species of bacteria that were especially successful colonizers, and had a particularly strong influence on their host’s growth. When mice ate this quintet, even if they had been previously loaded with microbiomes from malnourished children, they gained normal amounts of weight. Thanks to the incoming bacteria, they were better at converting amino acids in their diet into flesh and muscle, rather than simply breaking down those nutrients for energy.

“This underscores the critical importance of early life gut microbiome composition and its capacity to transform dietary components into molecules that significantly influence childhood development,” says Susan Lynch from the University of California, San Francisco. And Fergus Shanahan from University College Cork agrees: “This is a beautiful piece of work,” he says. “It’s ground-breaking in showing how elements of the microbiota influence their host’s health and well-being.”

Of course, none of this explains why the microbiomes of malnourished infants stall in their development in the first place. There are many possible reasons including antibiotic exposures, gut diseases, and poor diets, which vary from person to person. Still, Gordon’s team has now found similar patterns in malnourished children from both Bangladesh and Malawi—two countries with very different cultural and dietary traditions. “That gives us hope that there are basic themes in normal gut microbiota development that operate across children,” he says. “That may allow us to devise interventions that affect large groups of individuals.”

An obvious solution would be to feed children with probiotics containing growth-promoting microbes, as Blanton did with her mice. But such strategies rarely work out as planned. The bacteria in probiotics are terrible at gaining a foothold in the gut, which is why these products fail to meet the hype that surrounds them. “We need to do a better job of installing these organisms,” says Gordon.

An alternative is to feed people with foods that will nourish the right microbes. “Ready-to-use therapeutic foods”—fortified peanut-based meals handed out by aid organizations—don’t do the trick. Gordon has found that they have no lasting effects on immature gut microbiomes, but he’s trying to find alternatives that do. “We’re trying to find the next generation of foods that are microbiota-directed, readily available, affordable, and culturally acceptable,” he says.

At least one such food already exists—breast milk. Around 10 percent of human milk comprises complex sugars called HMOs, which babies cannot digest. Instead, these sugars have evolved to feed a baby’s first microbes. They’re the means through which mothers ensure that their newborns are colonized by beneficial bacteria.

Mark Charbonneau, another of Gordon’s students, found that Malawian mothers with severely stunted infants produce different milk to those with normal babies. In particular, they made far lower levels of sialylated HMOs—those containing a building block called sialic acid.

To see if these missing substances can improve growth, Charbonneau ran experiments on two kinds of germ-free animals—mice and piglets. He loaded them with microbes from a stunted and malnourished Malawian infant, before feeding them with sialylated sugars taken from cow milk. In both cases, the animals put on more weight, developed thicker and healthier bones, and showed higher levels of substances that are involved in healthy brain development.

These benefits didn’t happen because the sialylated sugars changed the proportions of bacteria in the animals’ guts. Instead, they changed what those microbes were capable of. “They became more metabolically flexible,” explains Charbonneau. Healthy animals can switch to burning fat for energy when sugar isn’t available, but undernourished ones cannot. “With the milk sugars, that normal flexibility is restored.”

Given these benefits, it may be worth offering milk that’s fortified with these sugars to mothers who can’t naturally produce enough of them. Cow milk offers an obvious source. It contains a narrower range and lower abundance of microbe-feeding sugars than human breast milk, but it makes up for that with sheer volume. When the dairy industry makes cheese, the waste liquid that remains contains sialylated sugars. “There’s value in that waste stream,” says David Mills, a milk expert at the University of California, Davis who was involved in Charbonneau’s study.

And there’s an even bigger opportunity, he says. Cows have all the genes necessary for making the same range of breast milk sugars as humans do, at the same quantities. “If you can turn these on or breed them on in a non-GMO way, and then give that cow to different parts of the world ... Boom!” says Mills. “I’d pay extra for that milk.”

But as always, this strategy comes with caveats. Charbonneau found that two bacteria reacted especially vigorously to the influx of sialylated sugars. One, Bacteroides fragilis, devoured the sugars and released chemicals that fed the other, Escherichia coli. That’s notable because many strains of E. coli are harmless residents of the gut, but some can cause vicious diseases. “You really need to know which part of the microbiome you’re feeding or there’s the potential to cross-feed the wrong bugs in some poor child,” says Mills.

Together, these studies illustrate the big themes in microbiome research: how influential our microscopic companions are; how much potential there is for improving our health by manipulating them; and how carefully we must proceed in doing so.