Blood vessels in the brain as they fill after neurons fire. This process causes the "lighting up" that's seen in fMRI scans. (Hillman Lab / Zuckerman Institute)

At some point, every pop-neuroscience story mentions a study where [insert some part of the brain] “lights up” whenever [insert some action or thought process] happens.

The concept has given us empirical evidence that we should listen to more music, have more sex, eat more chocolate, and most other things we’d like to be told to do. These functional magnetic resonance imaging (fMRI) studies make great stories. If you give a TED talk, I believe they’re legally required.

Still no place is darker than inside our heads. The images don’t actually measure brain activity; they detect blood flow. That flow makes an image light up. The approach is predicated on the idea that blood flow couples with brain activity. That leap is made, aptly, in our own brains, where we collectively assume that the areas getting a lot of blood flow are working harder.

Recent studies have called that foundational assumption to question, including a unique approach today in the Journal of Neuroscience, where a team at Columbia University's Zuckerman Institute released findings that upend the traditional fMRI model for young brains. Adult responses do not occur in newborn brains, the researchers found: Brain cells fire, but blood flow does not increase. Coupling happens later.

How, then, do infant brains function so effectively without the blood-supply pattern that is so critical to children and adults? And is this a key to the capabilities of the infant brain that could be mined for our own benefit? Possibly for future TED talks?

*   *   *

Some interesting clues exist. PET scans measure patterns of metabolism in the human brain, and it’s long been known that infant brains don’t reach adult patterns of glucose utilization until one year of age. This led some people to suggest that the newborn brain is largely anaerobic, that it can fuel itself without using oxygen, like some sort of zombie baby.

That seems not to be entirely incorrect.

“To some extent, our results suggest that the brain didn't need to increase blood flow because it didn't need the oxygen,” said Elizabeth Hillman a biomedical engineer at Columbia, lead author of today’s study. “And to some extent, our results do suggest that it doesn't need that much oxygen.”

Hillman began her career in physics and got into engineering before bringing her joint expertise to neuroscience. From this multidisciplinary vantage, her team has been attempting to understand the early stages of human brain development.

Hillman’s team used mice in their current experiments, as they would have been impossible in humans—the researchers had to cut open the heads of live mice to verify brain activity. Using transgenic mice with jellyfish DNA, the team is able to watch brain cells fluoresce when they are activated. (Brains literally lighting up.)

The left column shows neuronal activity in a mouse brain during hind-paw
stimulation. The right column shows blood vessel dilation.
(Hillman Lab / Zuckerman Institute)

Watching the neurons and the blood flow in real time, what the researchers found was that the young mice just did not show a correlation between the two.

This could explain why fMRI is unpredictable in human infants and, more importantly, challenge the idea that blood flow coupling is the foundation of brain development.

Hillman’s hypothetical explanation comes from her own children, remembering that as  toddlers, they would  run themselves into the ground until they were so hungry they couldn't function,“and then you have to stuff goldfish crackers in their mouth. I see a lot of analogies in the brain. The newborn brain is sort of caught by surprise—where neural activity happens and it doesn't have a blood-flow response there ready and waiting to stop the brain from running out of oxygen. I have this sort of weird little anthropomorphic interpretation, where the brain is sort of figuring it out as it develops. It’s saying, ‘Okay, this time you tricked me, but next time I'm going to make sure I have more blood ready.’”

It’s true, in adults, that people who have sleep apnea (during the night, usually because of obese physiology compressing the airways, the brain keeps experiencing hypoxia) are statistically more likely to do better than their peers after they have a stroke. This is, apparently, because their brains are conditioned to tolerate periods of impaired oxygen supply.

As the researchers write in the journal, the lack of early coupling of blood flow and neural activity “is demonstrated to result in a unique metabolic environment in the developing brain in which neural events drive local oxygen depletion.” So, through certain patterns of use, temporary low-oxygen environments could play an integral part in brain development—signaling where blood vessels will need to grow in the future.

“This gets me into why you should’t park a two-year-old in front of a television,” Hillman moralized semi-seriously, acknowledging that this is still supposition.

Other research has shown that enzymes involved in anaerobic metabolism are more common in neonatal brains, and that axons of mouse-brain cells lose their ability to function in the absence of oxygen by about eight months of age. It also may be necessary as a way of keeping us alive.

“It's shocking what a baby goes through during labor and delivery,” Hillman recalls from experience. “What it has to tolerate—the amount of head squeezing and decelerations of the heart—these things that plague almost every single ‘normal’ delivery. Yet, by and large, it comes out, and the baby is healthy.”

Even once it does make it out, suddenly the infant’s circulation has to transition from the umbilical cord to its own lungs and heart. “If that brain isn't ready to deal with periods of tens of seconds of inadequate oxygen supply,” she explains, “then none of us would be here.”

For those less interested in babies, there seem to be implications here for older people with conditions that result from oxygen deprivation—strokes and cerebral vascular disease—as well.

“I do feel that if we can understand this period, and it was something that we could switch on in the adult brain,” Hillman said, pausing, “it’s very exciting for exactly that reason.”

We want to hear what you think about this article. Submit a letter to the editor or write to letters@theatlantic.com.