Brain Aging 

Our understanding of the aging brain can be improved by identifying which brain regions are most vulnerable to aging, whether these age effects are constant across the older adult lifespan or exacerbated in advanced age, and whether they differ for cognitively normal versus impaired older adults. In addition to answering these questions, one line of my research seeks to identify MRI measures that are most sensitive to these age effects and the neural substrates that are captured by those MRI measures.

Gray Matter Microstructure. Our recent work uses diffusion imaging to assess microstructural properties of gray matter, not just white matter. We have demonstrated that multi-compartment diffusion imaging is more sensitive to age and memory performance compared to traditional single tensor diffusion approaches (DTI), with effects driven by both cellular and non-cellular sources of the diffusion signal. Ongoing work is examining the extent to which these diffusion metrics reflect age-related iron accumulation, gliosis, and tau pathology. Taken together, this line of work establishes gray matter microstructure as an important marker of brain aging and neurocognitive aging.


Locus Coeruleus. We also developed an ultra-high spatial resolution diffusion sequence that has allowed us to examine multi-compartment indices of gray and white matter microstructure at an unprecedented scale for in vivo imaging (<1mm). Here, we used this sequence to investigate age effects in the locus coeruleus (LC), a deep brainstem structure that is known to degrade with normal aging and Alzheimer’s disease. Significant age group differences were seen in LC integrity (increased fractional anisotropy [FA] and reduced mean diffusivity [MD] in older adults). Decreased LC integrity was also significantly related to worse memory performance in older adults. These findings support a role for LC in cognitive aging and may lead to early stage imaging biomarkers for Alzheimer’s diseases.


The Oldest-Old. One line of our work has focused on oldest-old adults over 90 years old. This is a particularly interesting population because they are underrepresented in the neurocognitive aging literature in spite of being the fastest growing segment of our population. The oldest-old are also unique in that the have a higher prevalence of dementia, well-documented hippocampal-specific pathology, as well as brain-wide white matter disease. In a recent literature review, we examined the extent to which current theories of brain and cognitive aging capture the patterns observed in advanced age.

Neurocognitive aging

Other lines of our research are focused on the contribution of brain aging to cognitive aging, with an emphasis on learning and memory processes as the loss of these abilities is a primary complaint of older adults. We are particularly interested in how medial temporal (hippocampal) and basal ganglia (caudate, putamen, globus pallidum) memory systems support these mnemonic processes and whether differences in the extent to which they degrade with aging contributes to declines in performance.

Mnemonic Discrimination. A fundamental component of memory is the ability to differentiate between new and previously experienced events, even when two events are highly similar. Previous studies have attributed this type of mnemonic discrimination to the hippocampus, but few have assessed the role of other brain regions, particularly those in cortex. Here, we used functional MRI to assess cortical activity while younger adults performed a memory task. We found activity consistent with mnemonic discrimination in both the hippocampus and visual cortex, but only hippocampal activity predicted mnemonic discrimination performance. These findings suggest that the hippocampus, but not cortex, is crucial for this ability to discriminate between highly similar events.


Associative Learning. The hippocampus and basal ganglia both support our ability to learn regularities in the environment. But they are differentially affected in aging and differentially engaged at various stages of associative learning, which may contribute to age-related declines in performance. Here, we examined integrity of these gray matter structures in relation to associative learning in younger and older adults. Independent of age, we found that better integrity of the caudate (decreased restricted diffusion) and hippocampus (decreased hindered diffusion) related to better learning earlier and later in the task, respectively. These findings converge with those of previous functional imaging studies, validating gray matter integrity as a viable marker of cognition in healthy adults.

Associative Memory. A critical feature of our memory for past events is that it is inherently associative. In addition to creating memories for individual items, we form associations between items (inter-item associative memory), between items and their elements (intra-item associative memory), and between items and their context (context or source memory). A key difference between item and associative memory is that the latter engages binding processes, which are thought to be mediated by medial temporal structures and server to link stimuli that are encountered in the same episode into a cohesive memory trace, although there is some evidence that the basal ganglia also supports associative memory. However, our current understanding of inter-item associative binding in episodic memory, and the extent to which it is impacted by aging, has been limited by a focus on associations between pairs of items. To address this gap, we developed a novel recognition task that parametrically manipulates associative load by having participants study word pairs, triplets, and quadruplets and then get tested on whether sets of each size were repeated, recombined, or novel. Our findings support the notion that associative memory is impaired at higher associative loads and in aging.

  • Franco, C., Alcaraz-Torres, A., Bennett, I.J. (2022). The QuadMax Task: A novel parametric manipulation of associative memory load in adults across the lifespan.