Functional imaging studies of episodic memory in Alzheimer's disease: a quantitative meta-analysis
Introduction
Alzheimer's disease (AD) is a progressive neurodegenerative disorder affecting approximately 4.5 million people in the US alone, a number expected to triple by 2050 (Hebert et al., 2003). Pathology begins with the formation of neurofibrillary tangles in the medial temporal lobe (MTL) including entorhinal cortex and hippocampus. Association regions are eventually involved with the development of neuritic plaques and tangles, and neuronal death (Braak et al., 1999). Clinically, the earliest cognitive deficits in AD patients are noted in episodic memory, though the progressive nature of the disease means that all manner of cognition, motor function and personality are eventually affected (Braak et al., 1999).
With an increasing understanding of novel targets for disease-modifying therapy (Golde, 2006), the early detection of AD is of growing importance not only from a practical standpoint (e.g., life planning), but also to allow for potential slowing of the disease course. Additionally, a means for tracking endogenous response to therapy is lacking. To this end, a number of methodologies have been applied to the search for a clinically useful biomarker (Ewers et al., 2007, Pike et al., 2007, Ringman et al., 2007). While structural imaging of the MTL and other sites has been extensively studied (Ramani et al., 2006), there is evidence that functional imaging techniques may identify abnormal activation signatures before structural changes are seen (Mondadori et al., 2006, De Santi et al., 2001).
Positron emission tomography (PET) and blood-oxygenation-level-dependent functional MRI (BOLD fMRI) methods have the ability to measure local task-related changes in metabolism or oxygenation as surrogates for neural activity, and have enjoyed widespread use for investigating cognitive function in the intact brain. In recent years, these techniques have been applied to the hunt for dysfunction in neuropsychiatric disorders, including AD. Such work has been carried out across the spectrum of AD, from cognitively intact adults with increased risk (e.g., Bookheimer et al., 2000), to patients with mild cognitive impairment (MCI) at high risk of conversion to AD (e.g., Johnson et al., 2006a, Mandzia et al., 2007), to patients with clinically diagnosed probable AD. Across these studies, much attention has been focused on the MTL, given its established role in memory and its early pathological changes in the disease process. A number of studies have reported increased MTL activity during memory tasks in at-risk persons compared to controls (Bookheimer et al., 2000, Mondadori et al., 2006, Bassett et al., 2006, Fleisher et al., 2005). However, other studies have found the opposite pattern in this population (Trivedi et al., 2006, Lind et al., 2006). Such findings may be complicated by the interaction of ApoE genotype and family history risk factors in predicting activation abnormalities, which appear to be more complex than a simple additive relationship (Johnson et al., 2006b). Studies of MCI patients have also been inconsistent, with some authors showing increased MTL activity (Dickerson et al., 2004, Dickerson et al., 2005) and others reduced activity compared to age-matched controls (Machulda et al., 2003, Mandzia et al., 2007, Johnson et al., 2006a, Johnson et al., 2006b). These inconsistencies may result from the heterogeneous nature of MCI, such that a given sample of patients may show variable patterns of MTL dysfunction, depending on whether they will subsequently convert to AD. Indeed, it has been suggested that at least two subtypes of MCI may be classified based solely on their contrasting patterns of MTL activity (Small et al., 1999).
There is better consensus regarding MTL dysfunction during memory tasks in patients with AD. A number of studies have found reduced activation in comparison to controls (Golby et al., 2005, Pariente et al., 2005, Rémy et al., 2005, Rombouts et al., 2000, Machulda et al., 2003, Dickerson et al., 2005, Sperling et al., 2003). Such findings make intuitive sense given the evidence of pathological degeneration and memory dysfunction in AD. However, there is less of a consensus regarding the pattern of activity in neocortical areas. In some studies, MTL dysfunction is matched to concomitant increases in activity in prefrontal, parietal or other sites. This has led to the suggestion that such changes may represent compensatory increases as a result of pathology in core circuits (Becker et al., 1996, Grady et al., 2003, Pariente et al., 2005). Indeed, Grady et al. (2003) were first to demonstrate a direct relationship between such activation and task performance, while Pariente et al (2005) found that activity in these regions even correlated with MMSE scores, suggesting a relationship to general cognitive status. However, such findings are not universal (Rombouts et al., 2000, Kato et al., 2001, Hamalainen et al., 2007), and there is evidence that apparent compensatory increases may be largely accounted for by the greater difficulty patients have completing such tasks (Gould et al., 2005, Gould et al., 2006). Gould et al. (2005) found that when subjective difficulty was equated in a paired-associate learning task, AD patients showed a similar activation pattern to controls. These conflicting findings are reflected in the interpretations made by recent qualitative reviews (Krishnan et al., 2005, Prvulovic et al., 2005, Wierenga and Bondi, 2007).
Beyond the issue of controlling for task-performance effects, there are other challenges in imaging the AD patient. Fundamental questions about the validity of functional imaging results are based on evidence that the BOLD response may be altered in these patients (Rombouts et al., 2005, D'Esposito et al., 2003, Buckner et al., 2001), and that cortical atrophy may impact activation (Johnson et al., 2000). AD patients also show abnormal default-mode and resting activity, which may exaggerate task-related findings due to lower baseline activity in select regions (Greicius et al., 2004, Lustig et al., 2003). Any of these factors could lead to apparent signal change in brain regions which could be erroneously attributed to differences in task-related activity.
Despite the host of technical, pathological and psychological variables that make imaging with AD patients challenging, the current literature suggests that there is some agreement regarding memory-related activity across studies and laboratories. Further, it has recently been demonstrated that activation in some areas correlates well with commonly used measures of disease severity, strengthening the argument for functional imaging as a method for tracking decline or treatment response (Diamond et al., 2007).
Many of the inconsistencies noted between studies may reflect patient factors, the specific tasks employed, and statistical analysis protocols. Because no single paradigm can be completely sensitive to and specific for a given operation, it is difficult to draw conclusions about higher order functions from a single sample of patients with a single task. Meta-analyses in neuroimaging offer the opportunity to combine disparate studies which target the same higher level constructs, and to determine those regions of the brain which are consistently associated with these functions (Fox et al., 1998). A recently developed technique, Activation likelihood estimation (ALE; Turkletaub et al., 2002, Laird et al., 2005), describes the stereotactic coordinates of the brain most consistently active across studies. This method has been previously used to examine consistent activation associated with higher cognitive function in other populations (Achim and LePage, 2005, Glahn et al., 2005, Dickstein et al., 2006). Through its application here, we hope to summarize the state of consensus in functional imaging of AD. In doing so, we not only attempt to confirm generally accepted findings, but also aim to quantitatively elucidate patterns that may not be readily apparent from reading individual studies. In the present paper, we completed an ALE meta-analysis of episodic memory studies in AD patients and compared them to findings in age-matched controls, in an effort to tease out consistent findings in this important research area.
Section snippets
Literature collection and criteria
MEDLINE and PubMed searches were completed to collect functional imaging studies of episodic memory in AD patients. Reference lists of relevant papers were reviewed for inclusion of studies not retrieved in the original searches. Papers were included in the meta-analysis if they met the following criteria:
(1) Reported findings of episodic encoding and/or retrieval paradigms compared to some baseline task in patients with AD using PET or fMRI techniques. All papers were required to follow
Healthy controls
During encoding, healthy controls demonstrated elevated activation likelihood in a range of prefrontal, parietal and limbic sites. Within the frontal lobe, they showed elevated values in the bilateral precentral gyrus (BA4), medial surface (BA6), left rostrolateral prefrontal cortex (RL-PFC; BA10) and ventral lateral prefrontal cortex (VL-PFC; BA47, BA45), in addition to bilateral dorsolateral prefrontal cortex (DL-PFC; BA9). Increased activation likelihood in the parietal lobe was seen in the
Discussion
To our knowledge, this is the first meta-analysis of functional neuroimaging in Alzheimer's Disease. We chose to focus on papers examining episodic memory, currently the most extensively studied domain in task-related functional imaging of AD. Our results demonstrate that consistent patterns are found between patients and controls across studies and laboratories.
When healthy age-matched controls and patients were compared at encoding and retrieval, consistent differences between groups were
Acknowledgments
The authors wish to thank Dr. Cheryl Grady for thoughtful comments on an earlier version of the manuscript. GCS is supported by an MD/PhD scholarship from the Canadian Institutes of Health Research and the Mclaughlin Center for Molecular Medicine.
References (81)
- et al.
Task difficulty in a simultaneous face matching task modulates activity in face fusiform area
Cogn. Brain Res.
(2005) - et al.
The cognitive control network: integrated cortical regions with dissociable functions
NeuroImage
(2007) - et al.
Hippocampal formation glucose metabolism and volume losses in MCI and AD
Neurobiol. Aging
(2001) - et al.
Cerebellar contributions to episodic memory encoding as revealed by fMRI
NeuroImage
(2007) - et al.
Beyond the single study: Function/location metaanalysis in cognitive neuroimaging
Curr. Opin. Neurobiol.
(1998) What does the cerebellum really do?
Curr. Biol.
(2007)- et al.
Task-related activity in prefrontal cortex and its relation to recognition memory performance in young and old adults
Neuropsychologia
(2005) - et al.
Increased fMRI responses during encoding in mild cognitive impairment
Neurobiol. Aging
(2007) - et al.
The relationship between fMRI activation and cerebral atrophy: comparison of normal aging and Alzheimer disease
Neuroimage
(2000) - et al.
Activation of brain regions vulnerable to Alzheimer's disease: the effect of mild cognitive impairment
Neurobiol. Aging
(2006)
Cortical activation during cholinesterase-inhibitor treatment in Alzheimer disease: preliminary findings from a pharmaco-fMRI study
Am. J. Geriatr. Psychiatry
Current status of functional MR imaging, perfusion weighted imaging, and diffusion tensor imaging in Alzheimer's disease diagnosis and research
Neuroimaging Clin. N. Am.
Functional activation imaging in aging and dementia
Psychiatry Res.: Neuroimaging
Verbal episodic memory impairment in Alzheimer's disease: a combined structural and functional MRI study
NeuroImage
Delayed rather than decreased BOLD response as a marker for early Alzheimer's disease
NeuroImage
Regional brain activity after prolonged cholinergic enhancement in early Alzheimer's disease. Magn. Reson. Imaging
Imaging
Meta-analysis of Go/No-go tasks demonstrating that fMRI activation associated with response inhibition is task-dependent
Neuropsychologia
Effects of aging on recognition of intentionally and incidentally stored words: an fMRI study
Neuropsychologia
Episodic memory-related activation in schizophrenia: meta-analysis
Br. J. Psychiatry
Brain regions associated with episodic retrieval in normal aging and Alzheimer's disease
Neurology
Familial risk for Alzheimer's disease alters fMRI activation patterns
Brain
Compensatory reallocation of brain resources supporting verbal episodic memory in Alzheimer's disease
Neurology
Patterns of brain activation in people at risk for Alzheimer's disease
N. Engl. J. Med.
Neuropathology of Alzheimer's disease: what is new since A. Alzheimer?
Eur. Arch. Psychiatry Clin. Neurosci.
Functional brain imaging of young, nondemented, and demented older adults
J. Cogn. Neurosci.
Hemispheric asymmetry reduction in older adults: the HAROLD model
Psychol. Aging
Imaging cognition II: an empirical review of 275 PET and fMRI studies
J. Cogn. Neurosci.
Age-related differences in neural activity during memory encoding and retrieval: a positron emission tomography study
J. Neurosci.
Alterations in memory networks in mild cognitive impairment and Alzheimer's disease: an independent component analysis
J. Neurosci.
The frontopolar cortex and human cognition: evidence for a rostrocaudal hierarchical organization within the human prefrontal cortex
Psychobiology
Evaluating self-generated information: anterior prefrontal contributions to human cognition
Behav. Neurosci.
Effects of healthy aging on hippocampal and rhinal memory functions: an event-related fMRI study
Cereb. Cortex
Alterations in the bold fMRI signal with ageing and disease: a challenge for neuroimaging
Nat. Rev., Neurosci.
Relationship of fMRI activation to clinical trial memory measures in Alzheimer disease
Neurology
Medial temporal lobe function and structure in mild cognitive impairment
Ann. Neurol.
Increased hippocampal activation in mild cognitive impairment compared to normal aging and AD
Neurology
The neural correlates of attention deficit hyperactivity disorder: an ALE meta-analysis
J. Child Psychol. Psychiatry
‘In the course of time’: a PET study of the cerebral substrates of autobiographical amnesia in Alzheimer's disease
Brain
Multicenter assessment of CSF-phosphoyrlated tau for the prediction of conversion of MCI
Neurology
Identification of Alzheimer disease risk by functional magnetic resonance imaging
Arch. Neurol.
Cited by (131)
End-to-end automatic pathology localization for Alzheimer's disease diagnosis using structural MRI
2023, Computers in Biology and MedicineEpisodic Memory–Related Imaging Features as Valuable Biomarkers for the Diagnosis of Alzheimer's Disease: A Multicenter Study Based on Machine Learning
2023, Biological Psychiatry: Cognitive Neuroscience and NeuroimagingGABAergic dysfunction, neural network hyperactivity and memory impairments in human aging and Alzheimer's disease
2021, Seminars in Cell and Developmental BiologyDeep Brain Stimulation for Alzheimer's Disease: Tackling Circuit Dysfunction
2021, NeuromodulationCitation Excerpt :Task-based deactivation of the so-termed “task-negative” DMN is thought to represent a switch of focus towards internal mentation and contribute to cognitive deficits in AD (15,60,64,65,68,69). Functional activity in these studies correlates with cognitive and neuropsychiatric symptoms and future cognitive deterioration (64,71,72). On the other hand, discrepancies exist between spatial patterns of functional activity, Aβ and tau pathology, and atrophy in multimodal imaging studies, obfuscating the link between pathology and function in these networks (58,61–63,74–76).