A new functional MRI (fMRI) study shows how two regions of the brain prominently affected in Alzheimer disease cooperate to form new memories, and gives researchers a look at what goes wrong during aging-related memory loss. In the January 31 PNAS online, Reisa Sperling and colleagues at Harvard Medical School show a reciprocal activation of the hippocampus and deactivation of the medial parietal lobes when test subjects learn something new. In some aged people, this arrangement breaks down: deactivation of the medial parietal lobes fails, and the hippocampus becomes hyperactive. Previous studies have demonstrated deactivation and its loss in aging, MCI, or AD. The new work goes further to show that the extent to which the parietal region turns off—or fails to turn off—predicts whether the subjects remember what they learned 30 minutes later.

“This is the first demonstration that the degree to which you turn off the medial parietal area in older adults predicts whether they’re going to remember something later or not, and that that area is just as important as what happens in the hippocampus,” Sperling said. “Our study indicates that it is the coordination of the medial parietal area turning down its activity and the hippocampus turning up that predicts if you’re going to remember a face-name pair in 30 minutes,” Sperling said. “Older adults with even mild memory impairment have difficulty in turning off this medial parietal region.”

Besides giving insight into basic mechanisms of memory, and how they alter with age, the study could provide clues to what goes wrong in Alzheimer disease, where amyloid plaque accumulation in the parietal cortex and neurofibrillary tangle formation in the hippocampus mark the disease in its earliest stages. “This work makes a link between these two areas that are vulnerable to the pathology of AD, showing why they are important in memory and how they are vulnerable in older individuals who have even just a little bit of memory trouble,” Sperling said.

The finding that age-related memory impairments are primarily explained by loss of deactivation in medial parietal regions jibes with much recent research showing that what is turned off in the brain during memory tasks may be just as important as what is turned on. Both the ability to deactivate parts of the memory network and the connectivity between regions of the network is compromised in aging (see Lustig et al., 2003 and ARF related news story). Studies from Sperling’s lab, and others including Michael Greicius, Randy Buckner, and Cindy Lustig, have shown that people with mild cognitive impairment or AD show disruptions in the same hippocampal-parietal network (see ARF related news story). The parietal cortex is part of the default mode network, a distributed network involved in memory and heavily affected by AD pathological changes (see ARF related news story).

To study the activity in these regions during successful or failed memory making, lead author Saul Miller and colleagues compared patterns of activation/deactivation in 17 young and 17 old subjects while they were learning to pair faces with names. The investigators looked at the BOLD fMRI signals while subjects learned face-name pairs, and then tested their memory 30 minutes later. As expected, the older people performed worse, although in all cases their abilities fell within the range considered normal. Both groups showed higher activation of the hippocampus and frontal regions during successful trials than when they did not learn. In addition, the young subjects showed a decrease in signal in medial parietal regions during successful learning. Aged subjects who showed lower memory performance demonstrated a distinct failure of deactivation in the same regions.

At the same time, the lower-performing aged group showed a greater activation in the hippocampus during successful memory encoding. As Sperling explains it, “Even the low-performing older adults get things right sometimes, but to do it they have to drive their hippocampus really hard. That is important because in previous work on people with mild cognitive impairment and apoE4 carriers, we and others have seen this strange hippocampal hyperactivation. We see it all the time, and this new data provides evidence that it might be compensatory, that it is required for successful memory when there is a failure in the default mode system.”

The finding of hippocampal hyperactivation filled in an important piece of the puzzle, Sperling says, in her thinking about how the systems work together, and how they might break down. “When one part of the network is failing, meaning the posterior cingulate region that is so vulnerable to amyloid early on, that is when you start to see the hippocampus go into hyper drive.” Later on in AD progression, hippocampal activity is lost altogether, as is the ability to make new memories.

The fMRI measurement of activation/deactivation during face-name learning appears to be a highly sensitive gauge of memory trouble. This could give researchers a new tool to probe the effects of amyloid pathology on brain function. (For a review of this topic, see Dickerson and Sperling, 2007.) “We know from PET-PIB amyloid imaging and from autopsies that there are non-demented older adults walking around with a lot of amyloid. The next thing we are doing is trying to figure out if low-performing older adults that show the vulnerability are the ones who have amyloid in that region,” Sperling said. Studies planned in collaboration with Keith Johnson and Randy Buckner at Massachusetts General Hospital, incorporating both PET-PIB and fMRI, should provide the answer.

Sperling has some data from longitudinal studies suggesting that hippocampal hyperactivation during memory tasks predicts cognitive decline over the next few years. Nonetheless, she does not consider task-based fMRI to be a practical screening tool. The technique is too complex, for one thing. She does foresee its use in clinical trials, as a functional probe to monitor the effects of lowering brain amyloid. “I run a lot of clinical trials and we are giving therapies, some with potential toxicity, to move amyloid out of the brain. We really need to understand whether amyloid is the problem and how it causes memory impairment, and if we change amyloid levels, can we make the brain work better?” she said.

“One day we will want to ask the question of should we give a vaccine to every older person with amyloid even if they are clinically totally normal. Functional imaging can help us understand whether the presence of amyloid is the beginning of AD, and whether we need to be intervening even at that early point,” Sperling said.—Pat McCaffrey


  1. These findings seem consistent with how the neurons of various brain loci communicate. The parietal lobe has been found in recent studies utilizing PET-PIB scans to be a prominent figure in early effects of amyloid deposition and shows high correlation with hippocampus atrophy.

    fMRI studies further make a more significant correlation as well.

Make a Comment

To make a comment you must login or register.


News Citations

  1. New Technologies Help Define Old Age
  2. Network Diagnostics: "Default-Mode" Brain Areas Identify Early AD
  3. Tracing Alzheimer Disease Back to Source

Paper Citations

  1. . Functional deactivations: change with age and dementia of the Alzheimer type. Proc Natl Acad Sci U S A. 2003 Nov 25;100(24):14504-9. PubMed.
  2. . Functional abnormalities of the medial temporal lobe memory system in mild cognitive impairment and Alzheimer's disease: insights from functional MRI studies. Neuropsychologia. 2008;46(6):1624-35. PubMed.

Further Reading

Primary Papers

  1. . Age-related memory impairment associated with loss of parietal deactivation but preserved hippocampal activation. Proc Natl Acad Sci U S A. 2008 Feb 12;105(6):2181-6. PubMed.