Older brains seem to hold on to more of a visual memory than they can use

Visual short-term memory is what lets you hold a picture in mind for a few seconds: where the cars were before you glanced in the mirror, or which way a ball was rolling. It gets steadily worse from early adulthood onwards. An old question is why. When we forget, is the memory itself gone, or is it still in the brain but hard to get at?

Brain scans can now help answer this. With fMRI, which tracks blood flow as a stand-in for brain activity, and computer programs that learn to recognise patterns, researchers can tell from activity alone what a person is holding in mind. This study used that approach to compare younger and older adults.

What they did

The team scanned 115 healthy adults aged 25 to 87 from Cam-CAN, a large study of healthy aging based in Cambridge, UK. Inside the scanner, people watched three patches of moving dots, one after another. One to three of them moved in a straight line and had to be remembered; the others just rotated and could be ignored. After an 8-second pause, a coloured circle showed which patch to report, and people turned a pointer to show the direction it had moved.

The researchers then trained a computer program to read the remembered direction from activity patterns in two sets of brain areas: the visual areas at the back of the head, and a network in the front and upper sides of the brain (the frontoparietal areas) that becomes active whenever a task gets demanding. They looked separately at the moment of seeing, the 8-second wait and the moment of recall, and asked how these brain "copies" of the memory related to age and to how accurate people were.

What they found

• Memory got worse with age. As expected, errors grew with age and with the number of patches to remember. Older people also pulled their answers more towards diagonal directions.
• The visual copy faded. The remembered direction could be read most clearly from visual areas. This copy was weaker in older people, especially at recall. A weaker visual copy statistically explained 26% of the effect of age on accuracy at recall, compared with 3% while seeing and 7% during the wait.
• But not always in step. At older ages, accuracy was less tightly tied to the visual copy. Some older people had a stronger copy in visual areas than their poor answers would suggest, as if the memory were partly there but not being used.
• The front of the brain helped out. The frontoparietal areas held a weaker copy of the direction, but during the wait this copy was stronger in older people, and a stronger copy went with fewer errors. Statistically, this partly offset the decline, which the researchers read as compensation.

What to keep in mind

• One scan per person. The authors point out that comparing people of different ages at one moment mixes the effects of aging with differences between generations. Following the same people over time would be needed to confirm the picture.
• Statistics, not proof of mechanism. "Mediation" and "compensation" here are statistical patterns. They fit a story about what the brain is doing, but don't show cause and effect.
• A task within easy limits. People remembered at most three directions. The authors say the findings may not carry over to tasks that overload memory, where people use other strategies.
• The design had limits. The task was not built to separate the phases of a trial cleanly, and the team could not read out the colours or the ignored patches from brain activity.
• Modest size, healthy volunteers. 115 people is reasonable for a brain-scan study but small for comparing ages. The paper is a preprint that has not yet been peer-reviewed.

Why it matters

The study suggests that part of what older people lose in short-term visual memory may not be the memory itself, but the ability to read it out at the moment of recall, and that the brain partly compensates by keeping the information in other areas. That is a more hopeful picture than simple decay. If it holds up, the authors suggest, training or other approaches might aim to strengthen the helping areas during the wait, or to improve how older people tap memories that are still there. Those ideas have not been tested yet.