5 December 2008. Widely seen as the molecular trigger for a cascade of neurological and behavioral changes leading to Alzheimer disease, Aβ lurks within the brains of many cognitively normal seniors, too. Does the presence of fibrillar amyloid in these folks foretell future dementia? Based on several studies presented at the Society for Neuroscience (SfN) annual meeting in Washington, DC, held 15-19 November, the answer hovers around a not-so-straightforward “probably.” Using a full arsenal of brain imaging technology including positron emission tomography (PET) and magnetic resonance imaging (MRI), the new investigations are pinpointing the neurological features and cognitive abilities associated with amyloid deposition, and may help determine whether these changes reflect normal aging or early signs of disease. Meanwhile, preliminary data from a study of “super agers” hints that staying mentally sharp in the golden years may depend less on Aβ and correlate more with the ability to stave off tau pathology.
In a slide talk, Elizabeth Mormino, a graduate student in the lab of William Jagust at the University of California, Berkeley, told the audience that live brain imaging using the PET radiotracer Pittsburgh Compound B (PIB) routinely detects amyloid in 10-40 percent of non-demented elderly. Comparing brain amyloid levels of normal controls from the Berkeley Aging Cohort with those found in an independent cohort of UC San Francisco AD patients, she said that “on the whole, amyloid levels are higher in the AD patients, but there is some overlap.” She and colleagues analyzed whether amyloid load correlated with reduced hippocampal volume and episodic memory in this group of 20 dementia-free Berkeley seniors, and in two other groups of non-demented elderly—17 normal controls and 39 PIB-positive mild cognitive impairment (MCI) patients—from the Alzheimer’s Disease Neuroimaging Initiative (ADNI). Mormino noted that only the PIB-positive MCI individuals were analyzed because PIB-negative MCI patients often convert to non-AD dementias, perhaps reflecting the presence of pathologies not related to amyloid.
Across all three cohorts, individuals with greater amyloid deposition had smaller hippocampi, Mormino reported. However, an association between higher PIB index and poorer episodic memory was less consistent—showing up in the Berkeley cohort and in the PIB-positive MCI ADNI group, but not in the ADNI normal control group. These observations led the researchers to suspect that “maybe these three variables are related but in a specific way,” Mormino said. “Maybe hippocampal volume mediates the relationship between PIB and episodic memory.” To test that idea, her team performed regression analysis looking at how PIB load contributes to episodic memory. When they corrected for hippocampal volume, PIB was no longer significantly associated with episodic memory, she reported. However, regardless of whether they accounted for PIB levels, hippocampal volume remained a significant predictor of episodic memory. Consistent with a model in which Aβ deposition, hippocampal atrophy, and episodic memory loss occur sequentially in nondemented elders, the findings suggest that the relationship between Aβ and episodic memory is indirect and possibly mediated by hippocampal damage.
During the same SfN slide session, Keith Johnson of Massachusetts General Hospital and Harvard Medical School in Boston presented PIB data hinting that amyloid in the brains of healthy older people may in fact spell impending doom. “What we attempted to do was to take a look at individuals who are normal, who have amyloid binding, and see after a follow-up period whether their neuropsychological function had changed,” Johnson said. In Washington, he expanded on findings described earlier this spring by Harvard Medical School colleague Dorene Rentz at the Human Amyloid Imaging conference in Chicago (see ARF related conference story). In the study, 31 non-demented older adults (15 with Clinical Dementia Rating [CDR] scores of 0, 16 with CDR 0.5) received PIB-PET scanning and cognitive tests at baseline, and follow-up cognitive assessment about a year later—test scores were adjusted for age, education, estimated IQ, and baseline cognitive performance. During the study’s short timeframe, greater amyloid deposition in the precuneus correlated with memory decline, most prominently in 30-minute delayed recall. “This is not a substantial, clinically apparent decline in memory function. It’s really memory performance in a very specific way that we’re detecting here,” he said, noting that tests of executive function, language function, and visuospatial function did not show this trend. Still, the results are intriguing given that memory decline of any sort showed up after just one year, and that the changes were related to amyloid buildup in the precuneus. “The precuneus is a very good proxy. It’s the region that leads the pack in terms of amyloid deposition,” said Johnson, adding that longitudinal follow-up is required to nail down whether brain amyloid in normal elderly predicts later progression to dementia.
Efforts to probe the functional significance of fibrillar brain amyloid in normal older adults were also described in an SfN poster by Trey Hedden of Massachusetts General Hospital. Collaborating with Johnson and others, Hedden used a combination of neuroimaging and neuropsychological techniques to address how Aβ pathology in otherwise healthy seniors relates to various measures of cognition and neural function that change with age. The researchers compared three groups of people: 36 healthy university students (ages 18-27), and a cohort of 29 healthy older adults (ages 61-84) subdivided into PIB+ (n = 17) and PIB- (n = 12) groups. As determined by diffusion tensor MRI, the elder participants differed from their younger counterparts in certain measures of white matter integrity that typically change with age, but these alterations were seen regardless of PIB status.
On the other hand, amyloid load did seem to matter for functional assessments involving the default network, a set of brain areas that fire up when the mind is resting and tone down during focused mental tasks. Among older participants, who as a group fared worse than the younger adults, PIB+ individuals underperformed their PIB- counterparts on several attentional control tasks. In addition, MRI measurements of correlated brain activity (activity measured during a task but that has been temporally filtered to remove task effects), revealed age-related disruptions that were worsened by the presence of amyloid. Previous work has shown that AD patients and memory-impaired older adults have reduced default activity (see ARF related news story and ARF news story), and it may be that specific parts of this network are differentially affected by aging and AD. Interestingly, PIB-dependent differences did not show up in the attention-task MRI data. Hedden speculates that the correlated connectivity measurements could be “a more sensitive measure of aging or disease-related effects because they represent the spontaneous functional coherence of the brain.” He suggested that when faced with specific tasks, cognitively normal older adults might be able to recruit compensatory reserves to confront possible difficulties, which could mask differences that would appear in the unfocused, spontaneous state. The bottom line, suggested Harvard Medical School colleague and coauthor Reisa Sperling in an e-mail to ARF, is that “the presence of amyloid does disrupt normal function in the default mode network, similar to the disruption reported in early AD. Thus, PIB imaging may be particularly useful in better defining the process of normal versus pathologic aging.”
Other researchers are tackling this question from a different angle. Instead of studying what goes wrong in the brain to bring on dementia, they are investigating what goes right to stave off cognitive decline in select individuals. A small segment of the elderly population retains sharp memory even at age 80 and beyond, Changiz Geula of Northwestern University, Chicago, told this reporter at the SfN meeting. “What is special about these brains?” Hints emerged in a poster describing preliminary data from the university’s SuperAging study, headed by Geula. The findings thus far are based on postmortem analysis of five “super agers” 80 and above—three who performed like 50 year olds on standard neuropsychological tests, and two who showed stable cognitive stability for at least three years before death. Compared to control brain tissue from age-matched non-demented elderly, brains from the high-performing super agers had considerably lower numbers of tau tangles and pre-tangles in the entorhinal cortex, middle temporal gyrus, and cingulate cortex. On the other hand, super agers had greater numbers of amyloid plaques in these brain areas, relative to age-matched controls. Geula stressed that these data are very preliminary. He mentioned, for instance, that the tangle trend fades when data from the cognitively stable super agers are added to the analysis. Nevertheless, if reproduced with larger sample sizes, the new findings are intriguing because they suggest that super agers have particular characteristics that may help them compensate for the buildup of pathological amyloid. Identifying these factors is the long-term goal of the SuperAging project, Geula said. Education and other measures of cognitive reserve come to mind as possibilities (see ARF related news story), but Geula could not yet say whether these measures influenced the preliminary plaque and tangle findings. Forthcoming analyses should shed light on this issue, he said.—Esther Landhuis.