As people age, their muscles and joints often stiffen up, as do their brains. Dynamic modulation of brain activity—the activation and deactivation of networks in response to different cognitive challenges—diminishes with age and with Alzheimer’s disease, leading to poorer memory and other cognitive deficits. Now, a new study of healthy adults suggests that this happens faster in people who carry the apolipoprotein E4 (APOE4) allele, the strongest genetic risk factor for late-onset AD. The work, from Chris Foster, Karen Rodrigue, and Kristin Kennedy at the University of Texas at Dallas, appeared online June 26 in the Journal of Neuroscience.

The APOE4 allele increases the risk of age-related memory loss, accumulation of Aβ and amyloid, and thinning of the cerebral cortex and hippocampus. APOE4 also affects brain function. Normally, when the brain is at rest, activity hums along in a distributed set of brain regions called the default mode network. When faced with a job to do, such as remembering a bit of information, the brain dials down activity in default areas, and reallocates brainpower to regions needed for the task at hand. Aging, AD, and APOE4 all disrupt the shift, resulting in less deactivation where needed and worse learning and memory. 

APOE Effects.

Carriers of the APOE allele modulate neural activity differently in the bilateral precuneus (blue, shown in L and R medial view of on lower line; top shows lateral view). [Courtesy of Foster et al., The Journal of Neuroscience 2017.]

Rodrigue and Kennedy have been studying how this loss of deactivation proceeds as people age. They put subjects in a scanner and ask them to judge the distance between a dot and a line, and determine if it is more or less than a previously viewed reference distance. Because the effects on deactivation might be subtle, they set up the task with three levels of difficulty: In the easy task, the dot is very close to or very far from the line, while in the hardest version, the dot is near to the reference distance. The challenge for the brain is to ramp up deactivation as the task jumps from easy to hard, or down when it goes the opposite way. Recently, the researchers reported that people's ability to vary the extent of deactivation with increasing task difficulty diminished with age. They studied a large sample of 181 healthy adults spanning young adulthood to old age. The decline in modulation, which primarily occurred in regions in the default mode network, correlated with worse performance on the task (Rieck et al., 2017). 

In the new study, first author Foster drew on the same participants to analyze brain function by APOE status: 31 APOE4 carriers between 20 and 86 years old were matched with 31 non-carriers for age, sex, and years of education. While in the MRI scanner, the subjects performed multiple runs of the easy, medium, and hard versions of the dot and line test. APOE status affected neither the accuracy nor time to answer. In keeping with this, both APOE4 carriers and non-carriers increasingly deactivated a large cluster of voxels in the prefrontal cortex, the lateral temporal gyrus, and the precuneus as the test got harder.

Would parsing the data by age bring out any APOE4 effect? Again the researchers used voxel-wise analysis to identify areas where deactivation was modulated, this time comparing APOE4 carriers and non-carriers by age. They discovered a cluster of voxels in the precuneus that deactivated poorly in older APOE4 carriers as they tried to respond to more challenging tasks (see image above). In this region, the non-carriers of all ages maintained steady levels of modulation.

The diminished “scaling of deactivation to difficulty” predicted worse performance on the dot and line task, indicating that the changes detected by fMRI affected cognition, at least for spatial tasks. To see if memory was similarly affected, the researchers tested subjects, outside the scanner, on their ability to pair names with faces. This memory task critically relies on deactivation of the precuneus (Feb 2008 news), and performance similarly correlated with ability to modulate deactivation.

“This is a nice study showing very early physiological precuneus alterations in APOE4 carriers that precedes clinically evident cognitive impairment,” said Richard Caselli, Mayo Clinic, Phoenix, in an email to Alzforum. “It is fascinating, logical, and consistent with what we know about the precuneus region.”

The precuneus is a hot spot for early amyloid deposition, which is accelerated by APOE4. As such, Caselli noted that in the absence of Aβ biomarker data, it’s impossible to know if the change stems directly from APOE4 plus aging, or from early APOE4-induced AD.

Eric Reiman, Banner Research, Phoenix, had a similar take. “The data support and extend findings from other task-dependent functional MRI studies,” he wrote in an email. He wondered how the reduced precuneus deactivation would compare to measures of amyloid burden or other AD biomarkers in predicting subsequent cognitive decline.

The authors have amyloid PET scans, but only for subjects over 55. “We could not correlate age-related activation changes to amyloid load since we do not have the latter information on any of the young adults,” they wrote to Alzforum. But they agreed that amyloid could play a role. “It will be important for future research to disentangle this possibility,” Foster, Rodrigue, and Kennedy wrote. They also pointed out the need for longitudinal follow-up to determine whether the group differences in brain function presage future cognitive decline.

When does the APOE4 effect kick in? Because of their approach, the authors can’t say precisely at what age the two groups diverge. “Visually it appears that modulation remains similar into middle age and significant differences may not occur until older age, however, we can only speculate,” they wrote.

Sterling Johnson of the University Wisconsin, Madison, expressed concern that the three-way interaction the researchers found between age, task difficulty, and ApoE4 status might be difficult to replicate in a different lab, or even in the same lab. “A large proportion of fMRI results, particularly results about group differences, do not replicate well,” he wrote to Alzforum. “Going forward, it will be important to reproduce the findings,” he said.— Pat McCaffrey


  1. This interesting study found an association between the difficulty of a visuospatial judgment task and deactivation of the precuneus and other regions in the default mode network, a preferential relationship between age and less task-dependent DMN deactivation in APOE4 carriers, and an association between age-related task-dependent DMN deactivation and performance in the APOE4 carrier group.

    The findings support and extend prior findings from other task-dependent functional MRI studies, and they raise a number of equally interesting questions. For instance, to what extent are the associations with task difficulty generalizable to those tasks unrelated to visuospatial perception, memory, or judgment? To what extent are the findings temporally associated with preclinical evidence of Aβ pathology, which typically begins in the same regions? And how does the reduced precuneus deactivation compare to task performance, PET or CSF measures of amyloid burden, or other AD biomarkers in predicting a person’s subsequent cognitive course?

  2. I found this study fascinating, logical, and consistent with what we know about early cortical regions involved in preclinical Alzheimer’s disease, i.e., the precuneus region. This physiological “decline” (quotes reflect the cross-sectional nature of the study so decline is inferred from the age interaction) may well represent early cortical physiological disruption due to early AD-related pathology, and while this is fascinating, it also reflects the major weakness of this study, namely, the lack of biomarker data. Is this an APOE ε4-age effect or is this actually an APOE ε4 induced preclinical AD effect? Is this a cause of the early pathology or an effect of it? The authors address this nicely in their discussion. Also, while this study again shows how sensitive image-related measures of cortical physiology can be to early AD changes, we lack longitudinal neuropsychological data, so whether this precedes longitudinally assessed cognitive decline or not is unknown. Bottom line, this is a nice study showing very early precuneus physiological alteration in APOE4 carriers that precedes clinically evident cognitive impairment.

Make a Comment

To make a comment you must login or register.


News Citations

  1. Deactivation Flaws Predict Memory Troubles

Paper Citations

  1. . Age-related reduction of BOLD modulation to cognitive difficulty predicts poorer task accuracy and poorer fluid reasoning ability. Neuroimage. 2017 Feb 15;147:262-271. Epub 2016 Dec 13 PubMed.

Further Reading

No Available Further Reading

Primary Papers

  1. . Differential Aging Trajectories of Modulation of Activation to Cognitive Challenge in APOE ε4 Groups: Reduced Modulation Predicts Poorer Cognitive Performance. J Neurosci. 2017 Jul 19;37(29):6894-6901. Epub 2017 Jun 26 PubMed.