23.19 THIS IS HOW WE DO IT: Can intense cognitive training induce brain growth?

“Use it or lose it.” A frequent assertion by weightlifters in reference to their muscles. But can the same be said about the memory-forming components of our brain? For many years, scientists believed that no new neurons are added to the human brain once development is completed in early childhood. While this is largely true, more recent experimental approaches suggest that some parts of the hippocampus are an exception: cells are added even into adulthood, possibly as a function of use.

There is a type of brain scan called functional MRI, or fMRI, that can visualize the size and activity of different brain structures. Using this scanning technology, researchers have investigated how individuals’ brains change over time. They began, as many scientists do, with an interesting question: Can environmental stimulation improve your brain’s abilities? This is an important question, but perhaps too general for tackling in a single, carefully controlled experiment

Using a common strategy, the researchers refined the broad question into something more testable: Does acquiring and mastering a large and complex body of knowledge measurably change your brain anatomy? Maybe the cleverest part of the research plan was their identification of a population well-suited to the study: London cab drivers. To become licensed as a cab driver in London, a trainee must master the layout of more than 25,000 streets within a six-mile radius of the central train station. This is known as acquiring “The Knowledge” and typically takes about four years, during which the trainee must pass a series of stringent exams. Among trainees, 50% to 60% fail to qualify for their license.

Armed with their powerful observational tool—fMRI for measuring brain structures—and an interesting pair of groups to compare—cabbies and non-cabbies—the researchers were able to make an exciting discovery. The posterior portion of the hippocampus was significantly larger among cabbies than among non-cabbies. (And its size was positively correlated with the number of years the cab driver had been on the job.)

These findings were consistent with the hypothesis that learning “The Knowledge” creates new neurons due to the greater demands placed on the posterior hippocampus as cabbies navigate and learn the streets of London. But the results are of interest to people far beyond the world of cab drivers. If enduring structural changes in the brain can be induced by behaviors, that bodes well for the potential long-term benefits of educational interventions, as well as for rehabilitation after illness or injury.

With such widespread interest in these findings, it’s not surprising that many people were cautious about immediately drawing broad conclusions. One issue in particular aroused skepticism. The study results are also consistent with an alternative (less interesting or exciting) hypothesis. Maybe cab driving doesn’t make your hippocampus bigger. Maybe, instead, people differ innately in posterior hippocampus size and those with a larger hippocampus are simply predisposed to be successful cab drivers. The original study did not allow researchers to distinguish between their original hypothesis and this alternative hypothesis.

Is there an experimental approach that would make this possible? Yes! It’s called a longitudinal study. What if you studied individuals as they trained to become cab drivers, measuring their posterior hippocampus size before and after they acquired The Knowledge?

If posterior hippocampus size is increased as a result of acquiring The Knowledge, the hypothesis that brain changes can be induced by behaviors is supported. If posterior hippocampus size is not increased by acquiring The Knowledge, the hypothesis supported is that some people have a larger hippocampus, and those people are predisposed to succeed as cab drivers.

We can answer this question with three words: faster, easier, cheaper. Based on the promising but ambiguous results from the initial experiment, however, the researchers decided to design and conduct a longitudinal study with 79 people training to become London cabbies (as well as 31 control participants, not undergoing the training.) For each participant, the researchers recorded several measurements, including assessments of memory. They also determined the volume of each individual’s hippocampal gray matter, using fMRI. They collected these data at two time points: T1, at the beginning of training, and T2, four years later, just after qualification for a license.

Of the original 79 trainees, 59 returned for the post-training T2 measurement: 39 who had qualified for their license and 20 who had not. Among the 20 original participants who did not return for the T2 measurement, 2 had qualified and the other 18 did not complete the training. All 31 of the control group returned for the T2 measurement.

Based on the trainees’ license examination performance, the participants could be divided into three groups: group Q, trainees who qualified for their license (having mastered The Knowledge); group F, trainees who failed to qualify for their license; and group C, controls who did not undergo the training.

No hippocampus size differences were detected among the three groups prior to the start of training. In other words, the groups started out on equal terms. At T2, however, some definitive and dramatic results from the longitudinal study became clear.

• 1. Those who qualified—group Q—had a greater posterior hippocampus volume at T2 than at T1 (and a greater volume than participants in groups F and C at T2).
• 2. Those in groups F and C showed no change in posterior hippocampus volume at T2 relative to volume at T1.

Additionally, those who qualified (Q) were found to have spent an average of twice as many hours per week in training as did those who did not qualify (F). The qualifiers also scored significantly higher on tests of spatial memory at T2 than did groups F and C (which did not differ from each other).

The impact on hippocampus size of acquiring The Knowledge is quite clear when presented graphically:

In contrast to the conclusions they could draw from the initial cross-sectional study, the researchers expressed much more confidence in their conclusions based on the longitudinal study. Acquisition of The Knowledge was associated with an increase in size of the posterior hippocampus and with better performance on memory assessments.

The researchers also thought a more general conclusion was warranted—that “specific, enduring, structural brain changes in adult humans can be induced by biologically relevant behaviors engaging higher cognitive functions such as spatial memory”—and speculated on the potential for memory-improvement training in educational and clinical settings.

Does your answer to this question support the idea that cross-sectional studies may have value even when it’s clear that a longitudinal study would be more powerful?

The researchers also reported that retired cabbies’ posterior hippocampus volume and performance on memory assessments were “normalized”—that is, indistinguishable from those of non-cabbies. Does this surprise you? Why or why not?