Brain clues to energy-efficient learning.

The smart brain may operate on the principle that "less is more," at least when it comes to learning new material. Preliminary data indicate that learning a complex task spurs marked drops in the brain's energy consumption, particularly among people who score highly on a standard intelligence test.

"It's possible that the brain learns over time what neural circuits not to use to perform a task, eventually relying only on certain important circuits," says neuro-psychologist Richard J. Haier of the University of California, Irvine. "Learning and intelligence may share some important information-processing components in the brain."

Haier's group reports the link between intelligence and learning-induced energy conservation in an article accepted for publication later this year in INTELLIGENCE.

The researchers employed positron emission tomography (PET) scans to chart brain metabolism. PET scans measure the intensity of brain activity directly, based on the quantity of a minute amount of a harmless, radioactively labeled substance -- such as glucose or oxygen -- injected into volunteers and absorbed by their brain cells. A previous PET-scan study conducted by the same research team revealed reduced brain metabolism among people who scored highly on an intelligence test that taps into abstract, nonverbal reasoning abilities (SN: 2/27/88, p.137).

In the new study, 24 men received injections of a radioactively tagged glucose compound. Eight of them then played a computer game for the first time. The game required players to use the computer keyboard to rotate and move floating shapes, each consisting of four blocks, in order to create solid rows of blocks across the computer screen. The remaining participants served as controls, passively watching, single-digit numbers flashed on the computer screen.

After 35 minutes, which allowed for absorption of the glucose by brain cells, the researchers took PET scans of each man. Those practicing the computer game displayed a substantially greater average rate of glucose consumption throughout their brains than did the controls.

But following four to eight weeks of daily practice with the game, the average rate of brain metabolism decreased significantly for computer players -- to a level slightly less than that seen in the controls.

Those players with the highest scores on a nonverbal intelligence test of abstract reasoning showed the largest decreases in overall brain metabolism.

However, a few areas of the brain consumed more glucose after practice, Haier notes. These regions, such as the right side of a small inner-brain structure called the hippocampus, may play key roles in memory and learning, he asserts.

Research with larger samples and different learning challenges must confirm the findings, Haier acknowledges. For now, his team theorizes that practice with a novel task, combined with innate intelligence, forges energy-saving neural pathways in the brain.