Age affects white matter microstructure and episodic memory across the older adult lifespan
Introduction
White matter plays a crucial role in the transmission and coordination of neural impulses between gray matter regions (Salat, 2011). Significant and widespread white matter deterioration observed in normal aging results from demyelination, axonal shrinkage, decreased fiber density, and gliosis (Bartzokis, 2004; Bowley et al., 2010; Peters, 2019, 2002; Peters et al., 2010). Increases in the magnitude and extent of this white matter damage in advanced age (>90 years old) is thought to reflect pathologic processes associated with a higher prevalence of dementia and white matter disease in this age group (Corrada et al., 2010, 2008; Kawas et al., 2015; Wardlaw et al., 2015; Yang et al., 2013). However, few studies have assessed white matter aging across the older adult lifespan and whether these age effects are driven by individuals with or at risk for dementia through the tenth decade of life.
The microstructural composition of white matter can be assessed in vivo using diffusion tensor imaging (DTI), which measures the jitter (diffusion) of water molecules (Beaulieu, 2002; Jones, 2008; Mori and Zhang, 2006). In healthy white matter, microstructures such as axonal membranes and myelin restrict the diffusion of water, which causes the primary diffusion direction to occur along the length of these structures rather than perpendicular to them. DTI measures these diffusion properties to provide estimates of the degree of restricted diffusion (fractional anisotropy; FA) and the average rate of diffusion parallel (axial diffusivity; AD) or perpendicular (radial diffusivity; RD) to the primary diffusion direction (Beaulieu, 2002; Jones, 2008; Jones et al., 2013; Mori and Zhang, 2006).
DTI studies in healthy younger-old adults without dementia (i.e., ages 60–89) report a relatively consistent pattern of linear age-related decreases in FA and increases in both AD and RD, with the magnitude of these effects varying by fiber class. That is, the largest age-related differences are seen in the fornix (de Groot et al., 2016; Kochunov et al., 2007; Lövdé et al., 2013), a medial temporal region that connects the hippocampus to cortical regions, and the genu of the corpus callosum (Barrick et al., 2010; Lövdé et al., 2013), a callosal region that connects frontal cortex in the left and right hemispheres. Large age effects are also observed within association fibers that connect cortical gray matter regions within the same hemisphere (e.g., external capsule)(Cox et al., 2016; Lövdé et al., 2013). However, projection and thalamic fibers that connect cortical gray matter regions to the spinal cord (e.g., corona radiata) and thalamus (e.g., thalamic radiations), respectively, show minimal age effects (Cox et al., 2016; de Groot et al., 2016; Lövdé et al., 2013). These regional variations have also been observed in DTI aging studies across the lifespan (Bendlin et al., 2010; Cox et al., 2016; Giorgio et al., 2010; Hsu et al., 2010; Hugenschmidt et al., 2008; Isaac Tseng et al., 2020; Kennedy and Raz, 2009; Kochunov et al., 2012; Lebel et al., 2012; Malykhin et al., 2011; Michielse et al., 2010; Mooij et al., 2018; Stadlbauer et al., 2008a, 2008b; Westlye et al., 2010; Xie et al., 2016).
Of note, very few DTI studies of healthy older adults without dementia have included a sizeable number of individuals beyond 90 years of age (c.f., Beck et al., 2021; de Groot et al., 2016), where the high prevalence of dementia-related cognitive impairment and white matter disease may magnify the effect of aging on microstructure (Yang et al., 2013). We focused on nonagenarians in a previous study (n = 94; Bennett et al., 2017), finding the largest age-related microstructure differences (decreased FA, increased diffusivity) in medial temporal (fornix) and callosal (splenium) regions, comparable to what is seen in younger-old adults, except that it was the splenium and not genu of the corpus callosum that was affected within the oldest-old. Importantly, these age-microstructure relationships did not differ between cognitively normal oldest-old adults and those diagnosed with cognitive impairment no dementia (CIND). However, because this earlier study did not include a younger-old comparison group, it remains unknown whether age is linearly related to white matter microstructure across the full extent of the older adult lifespan or whether there are nonlinear age effects on microstructure that may reflect disproportionate increases in normal age or disease-related pathology in advanced age.
To address this gap, the current study recruited 108 individuals across the older adult lifespan (65–98 years), including nonagenarians from The 90+ Study (Kawas and Corrada, 2006), who underwent diffusion imaging and completed an episodic memory task. Our first aim tested whether the effect of age on white matter microstructure was better explained by linear or nonlinear models. We hypothesized that more extensive white matter damage in advanced age would be seen as nonlinear effects of age on white matter microstructure, with the largest age-related differences in medial temporal and callosal fiber classes. Our second aim tested whether these relationships were affected by oldest-old adults diagnosed with CIND. We hypothesized that the age-microstructure relationships would not differ after excluding oldest-old adults diagnosed with CIND, consistent with our previous work in nonagenarians (Bennett et al., 2017), suggesting that these microstructure measures are capturing normal aging processes rather than pathology associated with early cognitive impairment. Our third aim was to assess the functional relevance of white matter aging, focusing on relationships between medial temporal microstructure and episodic memory given our interest in early cognitive impairment (Bastin and Salmon, 2014; Jahn, 2013).
Section snippets
Participants
We recruited a total of 110 older adults (65–98 years, 64 female). Seventy-nine younger-old adults (65–92 years, 46 female) from the Riverside community voluntarily responded to online and print advertisements. Thirty-one oldest-old adults (90–98 years, 18 female) were a subset selected from current participants in The 90+ Study, a longitudinal study of aging and dementia in the oldest-old (see Kawas and Corrada, 2006 for additional details), who had not previously received a diagnosis of
Linear effects of age on white matter microstructure
First, we conducted linear regressions to assess the effect of chronological age on white matter microstructure (Bonferroni corrected, p < 0.013), separately for each fiber class and diffusion metric. Results revealed that older age was linearly associated with decreased FA and increased AD and RD in each fiber class, R2 > 0.14, ps ≤ 0.001, except for AD in the medial temporal fiber class, p = 0.048 (Table 2 and Fig. 1; see Supplemental Table 1 for individual regions). Importantly, this pattern
Discussion
The current study examined how age affects white matter microstructure of 4 major fiber classes within an older adult lifespan sample that included a sizeable number of nonagenarians from The 90+ Study. Results revealed significant nonlinear age-related declines in microstructure (decreased FA, increased RD) of the medial temporal, association, callosal, and projection and/or thalamic fiber classes. Importantly, these effects of age on microstructure remained significant even after excluding
Acknowledgements
This work was supported by the following National Institute of Health grants: R00 AG047334 (Bennett), R21 AG054804 (Bennett), F31 AG071189 (Merenstein), R01AG053555 (Yassa), and R01 AG021055 (Kawas, Corrada). We thank Brianna Cabrera, Justino Flores, Brooke Jensen, Chelsea Evelyn, Dana Greenia, and Myra Larson for assistance with data collection. We also thank the 90+ Study participants and their families.
Declarations of competing interest
None.
CRediT authorship contribution statement
Jenna L. Merenstein: Data curation, Writing – original draft, Visualization, Investigation, Formal analysis. María M. Corrada: Funding acquisition, Writing – review & editing. Claudia H. Kawas: Funding acquisition, Writing – review & editing. Ilana J. Bennett: Conceptualization, Methodology, Funding acquisition, Supervision, Writing – review & editing.
References (95)
- et al.
Methodological considerations on tract-based spatial statistics (TBSS)
Neuroimage
(2014) - et al.
White matter structural decline in normal ageing: A prospective longitudinal study using tract-based spatial statistics
Neuroimage
(2010) Age-related myelin breakdown: a developmental model of cognitive decline and Alzheimer's disease
Neurobiol. Aging
(2004)- et al.
White matter microstructure across the adult lifespan: A mixed longitudinal and cross-sectional study using advanced diffusion models and brain-age prediction
Neuroimage
(2021) - et al.
Age-related white matter integrity differences in oldest-old without dementia
Neurobiol. Aging
(2017) - et al.
Disconnected aging: cerebral white matter integrity and age-related differences in cognition
Neuroscience
(2014) - et al.
Mnemonic discrimination relates to perforant path integrity: An ultra-high resolution diffusion tensor imaging study
Neurobiol. Learn. Mem
(2016) - et al.
Frequency of stages of Alzheimer-related lesions in different age categories
Neurobiol. Aging
(1997) - et al.
Age-related differences in white matter microstructure: Region-specific patterns of diffusivity
Neuroimage
(2010) - et al.
Associative memory and underlying brain correlates in older adults with mild cognitive impairment
Neuropsychologia
(2016)