8 May 2009. Figuring out what high levels of brain amyloid mean for otherwise healthy older people was a prime focus of the Human Amyloid Imaging (HAI) conference, held 24 April in Seattle. Along with new research suggesting correlations between high amyloid load and declining glucose metabolism (see Part 1 of this series), several other studies strengthen the case that amyloid-laden brains are on the wane. These data help relate live measurements of Alzheimer disease (AD) pathology to subtle changes in brain function that may be present years before standard tests pick up cognitive decline. Understanding what happens during the critical time window between early pathology and possible dementia onset should help open the doors to future preventive therapies for AD, researchers at the conference agreed.

Like several other speakers, John Becker, Massachusetts General Hospital, Boston, tackled this basic question: Does amyloid deposition correlate with diminishing brain metabolism in cognitively normal older people? Becker measured both features with positron emission tomography (PET)—the former with the amyloid tracer Pittsburgh Compound-B (PIB), the latter as fluorodeoxyglucose (FDG) metabolism—but his FDG-PET assessments focused on brain areas that make up large-scale functional networks. Among 53 participants with a Clinical Dementia Rating (CDR) of 0, 11 had high PIB uptake. As a group these folks had lower glucose metabolism in portions of the default-mode network, including the posterior cingulate-precuneus, inferior parietal, lateral temporal, and ventromedial areas. The default-mode network is a system of brain areas that fire up during rest and tone down when the person focuses on a mental task. These same brain regions are among those hit first in AD (see ARF related news story and Buckner et al., 2005), and loss of default-mode activity in the resting state seems to mark early AD (see ARF related news story and Greicius et al., 2004).

Trey Hedden, a Massachusetts General Hospital colleague and collaborator on Becker’s work, used a different approach to address a similar issue. His study did not use FDG-PET but rather functional magnetic resonance imaging (fMRI) to probe the relationship between increased Aβ burden (measured by PIB-PET) and changes in the default-mode network. In Seattle, Hedden reported that healthy seniors with high PIB uptake had reduced functional coherence in the default network. This trend appeared in fMRI scans taken during rest and mental tasks, albeit more strongly at rest. Furthermore, higher amyloid load correlated with lower default-mode connectivity, even though the high-PIB and low-PIB groups appeared equal on memory tests and clinical assessments.

In another study presented at HAI, subtle cognitive differences that went undetected on standard neuropsychological tests did in fact show up in a more challenging memory test. Dorene Rentz, of Brigham and Women’s Hospital in Boston, tested whether amyloid load correlates with cognitive performance, and whether cognitive reserve influences this relationship in normal elderly. Often measured by years of education, or in this study by the AMNART IQ test, “cognitive reserve” seems to help protect against dementia in the face of brain Aβ accumulation (ARF related news story).

In Rentz’s study, which included 66 normal elderly and 17 AD patients, those with higher amyloid load in the precuneus area of the brain tended to fare worse in neuropsychological testing. When considering only the normal seniors, though, precuneus Aβ burden did not correlate with test performance unless a more difficult measure (Memory Capacity Test) was used. The MCT requires subjects to learn two 16-word lists, and it was the second list learning that differentiated the high- and low-PIB normal groups, Rentz reported. In her study, the relationship between precuneus amyloid load and cognitive test performance was weakened by cognitive reserve. “People with high cognitive reserve were able to maintain their performance, whereas those with low reserve were doing worse,” Rentz said. However, “at a certain point, cognitive reserve is no longer protective” and probably “loses its protective value after a certain level of amyloid is endured,” she noted. Overall, her data underscore the potentially confounding effects of cognitive reserve in neuropsychological performance, and suggest that more challenging tests are needed to pick up subtle cognitive deficits that escape standard measures.—Esther Landhuis.

This is Part 2 of a four-part series. See Part 1, Part 3, and Part 4.

All attendees are invited to send additions and corrections to esther@alzforum.org.