When it comes to weaning, humans are weird.
Our closest relatives, chimpanzees and gorillas, breast-feed their offspring for several years. Some baby orangutans nurse until they are 7 years old.
But modern humans wean much earlier. In preindustrial societies, babies stop nursing after about two years. Which raises the question: How did we get that way? When did we make the evolutionary shift from apelike parenting to the short breast-feeding period of humans?
Scientists combed the fossil record for clues, but they came up empty until one researcher decided to play tooth fairy.
Manish Arora studies tooth chemistry at the Icahn School of Medicine at Mount Sinai in New York. He knows that teeth hold many secrets.
"You can almost visualize tooth development in terms of growth rings that you would see in a tree," says Arora.
Like tree rings, the layers of enamel and dentin that accumulate day after day mark the passage of time. And the compounds in the enamel can tell scientists a lot about the tooth owner's early growth and development.
But Arora was looking for something very specific: a marker in the tooth that would reveal the timing of weaning. So, he meticulously recorded the breast-feeding habits of women and their babies.
Years later, when the children started to place their lost baby teeth under their pillows, Arora was there to collect them. Luckily, this tooth fairy stand-in had a research grant.
"They do get a small reimbursement for every tooth they donate," Arora says.
When Arora and his colleagues started to analyze the chemical makeup of the teeth, they noticed an interesting pattern in the distribution of the element barium.
"During the period of breast-feeding, the barium levels in teeth were higher," Arora says. "At weaning, the levels of barium in teeth started to drop."
Barium is calcium's cousin (it's in the same column on the periodic table), and it goes where calcium goes. Over the years, it accumulates in our bones. When a mother begins nursing, some of that barium migrates into her breast milk and eventually into her baby's teeth. Arora had found his marker.
He could take a tooth and tell when that tooth's owner stopped nursing. His colleague Tanya Smith, who studies human evolution at Harvard, knew just the tooth to test first.
"It's a first molar tooth from a Neanderthal from a site in Belgium called Scladina," Smith says.
The tooth is 100,000 years old and perfectly preserved. Analyzing the tooth's barium distribution, the researchers determined that this Neanderthal started weaning after about 7 months, and then transitioned to a mixed diet. At 15 months, the barium signal dropped abruptly, as if mother and child had been separated.
The results were published in Nature.
Smith says applying the same technique to other fossilized teeth will help paint a clearer picture of the evolution of human weaning. The shorter nursing times would have given mothers the freedom to reproduce more frequently and gather food for the group. Now we're a step closer to finding out when and how this shift happened.
Arora adds that there's another important application: studying human nursing today. Most breast-feeding studies are questionnaire-based and can be foiled by the faulty memories of their subjects. But with Arora's method, scientists have access to a dental record that always tells the truth.