Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2026 Feb 27.
Published in final edited form as: Neurobiol Aging. 2024 May 27;141:46–54. doi: 10.1016/j.neurobiolaging.2024.05.012

Cognitive reserve proxies are associated with age-related cognitive decline – Not age-related gait speed decline

Helena M Blumen a,b,*, Oshadi Jayakody a, Emmeline Ayers b, Nir Barzilai c, Christian Habeck d, Sofiya Milman b, Yaakov Stern d, Erica F Weiss b, Joe Verghese a,b
PMCID: PMC12944813  NIHMSID: NIHMS2141906  PMID: 38820770

Abstract

Cognition and gait share brain substrates in aging and dementia. Cognitive reserve (CR) allows individuals to cope with brain pathology and delay cognitive impairment and dementia. Yet, evidence for that CR is associated with age-related cognitive decline is mixed, and evidence for that CR is associated with age-related gait decline is limited. In 1,079 older (M Age = 75.4 years; 56.0% women) LonGenity study participants without dementia at baseline and up to 12 years of annual follow-up (M follow-up = 3.9 years, SD = 2.5 years), high CR inferred from cognitive (education years), physical (number of blocks walked per day; weekly physical activity days), and social (volunteering/working; living with someone) proxies were associated with slower rates of age-related decline in global cognition – not gait speed decline. Thus, cognitive, physical, and social CR proxies are associated with cognitive decline in older adults without dementia. The multifactorial etiology and earlier decline in gait than cognition may render it less modifiable by CR proxies later in life.

Keywords: Cognitive reserve, Brain maintenance, Cognitive decline, Gait decline, Hippocampal volume, Frontal cortical thickness

1. Introduction

Cognitive reserve (CR) is a construct that can be used to account for individual differences in the relationship between brain status and cognitive status (Stern, 2009, 2012; Stern et al., 2022; Stern et al., 2020). Cognitive reserve can be accumulated across the lifespan from participation in cognitive, physical, and social activities, but is likely influenced by genetic factors as well. Greater CR permits individuals to better cope with (or adapt to) age-related brain changes or pathologies and thereby delay the onset of cognitive impairment and dementia. Cognitive reserve is typically assessed with convenience proxies such as education years, occupational complexity, and self-reported participation in physical and social leisure activities. Cognitive reserve can also be measured directly using functional brain imaging measures (e.g., resting-state or task-based fMRI), yet such measures are constrained by their specific methodologies (Stern et al., 2020; Stern et al., 2005) – e.g., the implementation of CR in the brain may extend beyond resting-state functional networks, or a particular brain activation pattern during a specific cognitive task.

While cross-sectional examinations of CR have consistently observed positive associations between CR proxies and cognitive performance or status (Singh-Manoux et al., 2011; Wilson et al., 2009; Zahodne et al., 2015), the results from longitudinal examinations have been mixed (Barnes et al., 2004; Berggren et al., 2018; González et al., 2013; Jin et al., 2023; Karlamangla et al., 2009; Lane et al., 2017; Scarmeas et al., 2001; Singh-Manoux et al., 2011; Verghese et al., 2006; Verghese et al., 2003; Wilson et al., 2009; Wilson et al., 2019; Zahodne et al., 2011; Zahodne et al., 2015), and many have focused on the role of one CR proxy such as education years or occupational complexity (Berggren et al., 2018; Lane et al., 2017; Wilson et al., 2009; Wilson et al., 2019; Zahodne et al., 2011; Zahodne et al., 2015). A systematic review and meta-analysis concluded that the effects of education on cognitive decline are small and insignificant in older adults (Seblova et al., 2020). Yet, a recent large-scale study found that educational attainment reliably moderated the relationship between global, lobar, and regional brain structures and cognitive decline in middle-aged and older adults (Jin et al., 2023).

The first aim of this study was to examine the relationship between CR and cognitive decline using CR proxies that not only reflect participation in cognitive (education years) activities, but also physical (number of blocks walked per day; physical activity days per week) and social (working/volunteering, living with someone) activities. We examined this in a sample of 1,079 older adults without dementia (enrolled in the LonGenity study (Gubbi et al., 2017; Zhang et al., 2020)) who had completed up to 12 annual assessments. We also explored whether the relationship between structural brain measures (hippocampal volume and frontal cortical thickness) and cognition was moderated by CR in a subset of 112 older adults who had completed one (cross-sectional) MRI assessment. Consistent with the CR framework, we hypothesized that high CR would be associated with less cognitive decline. We further explored whether CR would moderate the relationship between hippocampal volume (and frontal cortical thickness) and cognition, such that a weaker association would be observed among those with high CR than those with low CR. A stronger association between structural brain measures and cognition among those with high than low CR would be more consistent with the concept of brain maintenance or reduced age-related brain changes or pathology as a function of lifetime experiences (Stern et al., 2022; Stern et al., 2020).

The second aim of this study was to examine the relationship between the CR proxies and gait speed decline. Gait and cognition are interrelated and share neural substrates in aging and dementia (see (Clouston et al., 2013; Cohen et al., 2016; Leisman et al., 2016; Scherder et al., 2007) for reviews) – including, but not limited to the hippocampus and the prefrontal cortex (Blumen et al., 2014; Blumen et al., 2019a; Blumen et al., 2019b; Ezzati et al., 2015; Holtzer et al., 2014; Kaup et al., 2011; O’Shea et al., 2016; Rosso et al., 2013; Rosso et al., 2017; Tripathi et al., 2019). Gait decline often precedes cognitive decline, mild cognitive impairment (MCI), and dementia (Beauchet et al., 2016; Buracchio et al., 2010; Jayakody et al., 2022; Quan et al., 2017). In fact, gait speed decline starts to accelerate about 14 years prior to a MCI diagnosis in men, and about 6 years before a MCI diagnosis in women (Buracchio et al., 2010). Our own research further suggests that age-related gait speed decline proceeds at a faster rate than age-related cognitive decline (Jayakody et al., 2022).

The relationship between CR proxies and gait speed decline is relatively unknown. One study reported that age-related gait speed decline proceeds at a slower rate in those with good cognitive function (executive function and episodic memory) than those with poor cognitive function – and that this relationship is stronger among those with high CR (as assessed with pre-morbid IQ) (Holtzer et al., 2012). Another study reported that pre-morbid IQ moderates the relationship between gait speed and incident mobility impairment, such that those with slower gait and lower IQ had an increased risk for mobility impairment (O’Brien and Holtzer, 2021). To our knowledge, the relationship between CR and gait speed decline has not been directly examined or contrasted with cognitive decline in aging. We hypothesized that CR proxies would be associated with gait speed decline. Due to the physical demands of walking, we further hypothesized that our physical CR proxies (number of blocks walked per day; physical activity days per week) would be more strongly associated with age-related gait speed decline than our social (working/volunteering, living with someone) and cognitive (education years) CR proxies. Because the hippocampus and frontal cortex has been linked to gait speed and gait speed decline (Blumen et al., 2019a; Callisaya et al., 2013; Callisaya et al., 2014; Manor et al., 2012; Rosano et al., 2007; Rosano et al., 2012; Rosso et al., 2017), we also explored whether CR proxies would moderate the relationship between cross-sectional measures of hippocampal volume and frontal cortical thickness and gait speed.

2. Methods

2.1. Participants

This study used data from the LonGenity study – an ongoing longitudinal cohort study of Ashkenazi Jewish older adults that started in 2008 and primarily examines the genetic and biological mechanisms of successful aging (Gubbi et al., 2017; Zhang et al., 2020). Participants are recruited from the community through public records (i.e., voter registration lists) and general advertising (i.e., Jewish newsletters, contacts at synagogues), excluding those with dementia (>8 on Blessed Mental Status Examination; >2 on AD8-item Informant Questionnaire), severe visual or hearing impairment, and siblings enrolled in the study. Participants are defined as either offspring of parents with exceptional longevity (OPEL; at least one parent lived to age 95 or older) or offspring of parents with usual survival (OPUS). The current study examined a subset of 1,079 LonGenity participants (selected from a sample of 1,096) who completed annual neuropsychological, gait, and CR-proxy assessments and had up to 12 years of annual follow-up testing (M follow-up = 3.9 years (SD = 2.5). The mean number of assessments was 3.6 (SD = 3.0). One participant with possible dementia at baseline was excluded prior to analysis. Sixteen participants with daily blocks walked daily >2 SD away from the mean were also excluded. MRI acquisition was added to the LonGenity study protocol in 2018. Cross-sectional MRI data from 112 LonGenity participants were explored in this study. Ethical clearance was obtained from the Committee on Clinical Investigations of the Albert Einstein College of Medicine. Written informed consent was obtained from all participants.

2.2. Exposures

The following CR proxies were obtained from self-report: education years, number of blocks walked per day (“How many city blocks (12 blocks=1 mile) or the equivalent do you walk each day?”), physical activity days per week (“How many times per week do you engage in any physical activity such as aerobic dancing, jogging, bicycling, etc., long enough to work up a sweat?), living with someone (“do you live alone?” yes/no; a structural measure of social connection (Holt-Lunstad, 2018; Holt-Lunstad et al., 2010)), paid or volunteer work (“Are you working now as either a volunteer or in a paid position?” yes/no). While it is helpful to broadly categorized these CR proxies as cognitive (education years), physical (number of blocks walked per day; physical activity days per week) and social (volunteering, living with someone) CR proxies, we recognize none is purely cognitive, physical, or social.

2.3. Outcomes

2.3.1. Cognitive assessments

A z-standardized global cognition composite was generated from z-standardized performance (z = (x-μ)/σ; where x is the raw score, μ is the sample mean, and σ is the standard deviation of the sample) on 10 individual neuropsychological tests (11 different scores): 1) Trail Making Test A and B (Reitan, 1958), 2) Digit Span and 3) Digit Symbol Substitution tests from the Wechsler Adult Intelligence Scale-III (WAIS-III) (Wechsler, 1955), 4) Boston Naming (15-item) test (Williams et al., 1989), 5) Category fluency and 6) Phonemic fluency tests (Ruff et al., 1996), 7) Figure copy and 8) Figure recall from the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS) (Duff et al., 2005), 9) Free Recall from the Free and Cued Selective Reminding; Test (FCSRT; (Grober et al., 1997; Zimmerman et al., 2015)) and 10) Logical Memory subtest from Wechsler Memory Scale-Revised (Elwood, 1991).

2.3.2. Gait assessment

Gait speed (cm/s) was assessed during usual pace walking on an 8.5-meter-long computerized GAITRite walkway (following a practice walk). To allow for initial acceleration and terminal deceleration, walking started 1.2 m before and ended 1.2 m after the walkway. Most participants wore comfortable footwear and walked unassisted in a quiet and well-lit hallway. A small number of participants likely walked with an assistive device, but this was not consistently recorded and therefore could not be considered in our statistical models.

2.3.3. Brain structure assessments

Overall (mean) hippocampal volumes in each hemisphere were quantified from T1-weighted 3D images (240x220x240 mm FOV; 1.0 mm isotropic resolution; TR/TE= 9.8/4.6 ms; SENSE factor= 2.5) acquired on a Philips 3 T Ingenia Elition, using FreeSurfer 6.0 (https://surfer.nmr.mgh.harvard.edu/fswiki/HippocampalSubfields; (Iglesias et al., 2015)). Frontal cortical thickness (mean of lateral orbitofrontal, medial orbitofrontal, pars opercularis, pars triangularis, pars orbitalis, rostral and caudal middle frontal, superior frontal, frontal pole, rostral and caudal anterior cingulate thicknesses) in each hemisphere were also quantified from T1-weighted images using FreeSurfer 6.0 cross-sectional pipeline (https://surfer.nmr.mgh.harvard.edu/fswiki/recon-all; (Dale et al., 1999; Fischl et al., 1999)).

2.4. Covariates

Demographic data (age, sex) and parental longevity (OPEL or OPUS) were obtained from self-report. A global health score was generated as the sum of nine physician-diagnosed medical conditions at baseline: hypertension, diabetes, chronic heart failure, arthritis, depression, stroke, chronic obstructive pulmonary disease, myocardial infarction, and angina. No participants had a mild cognitive impairment diagnosis at baseline. A sample-specific cognitive impairment covariate was generated from the consensus of two neuropsychologists when a participant scored ≥1.5 SD below the mean on any neuropsychological test (Tabert et al., 2006).

2.5. Statistical analyses

STATA (StataCorp LLC, College Station, TX, United States) version 17.0 was used in all analyses and the integrity of the analytic code was reviewed by two of the authors (H.M.B & O.J). Longitudinal associations between baseline proxies of CR and cognitive and gait decline were examined in separate linear mixed effects models (LMEMs) performed in all (1,079) eligible participants. Global cognition and gait speed were standardized into a z-score using the sample means and standard deviations across all follow ups. The mean of gait speed was 104.8 (SD = 20.9) and the mean of global cognition was 0.8 (SD = 6.4). Age was centered around the sample mean. “Age” was the time variable, and was therefore modeled as a fixed effect in the model for the means and as a random effect in the model for the variance (Hoffman, 2015). Modeling age as a fixed effect allowed us to determine if cognitive reserve moderates age-related cognitive and or gait decline (CRxAge) and modeling age as a random effect permits the slope to vary across participants. We only considered CR proxies at baseline, and therefore CR proxies were modeled as fixed effects. Interactions between CR proxies and centered age (e.g., education years x centered age) were used to determine if age-related changes in gait and cognition were moderated by CR proxies at baseline. These LMEMs also included a fixed quadratic term for age (as we have previously shown that age-related changes in gait and cognition accelerates over time (Jayakody et al., 2022)) and the following covariates: sex, parental longevity, baseline cognitive impairment, and global health. The inclusion of a fixed quadratic term for age permits us to determine whether there is an overall change (acceleration or deceleration) in the linear change with age but does not account for individual differences in the quadratic change with age. A simplified example of the linear mixed effect models used is: Global Cognition= b1iAge + b2iEducation + b3i(Education×Age) + b4i(Age×Age) + b5(Gender) + b6(Parental Longevity) + b7(Cognitive Impairment) + b8(Global Health).

Several sensitivity analyses were also performed to examine the potential impact 1) of excluding participants with incident dementia, and 2) of different follow-up times (less than 5 vs. 5 or more years of annual follow-up), and 3) of CR on age-related decline in performance on individual measures of executive functions (Trails B time) and episodic memory (free recall from the FCSRT). Examining the effect of different follow-up times is particularly important given that chronological age was our time variable, and those with the same age yet different follow-up times could systematically differ.

LMEMs were also used to explore if CR proxies moderated the relationship between cross-sectional measures of brain health (hippocampal volume; frontal cortical thickness) and cognition and gait speed in the MRI subsample. Only one MRI scan was available, and the MRI subsample had shorter follow-up time (M = 1.4 years) – therefore, the quadratic age term was removed, and the goal was to explore if the relationship between brain structures and gait and cognition (not decline) was moderated by CR proxies at baseline. Thus, these models included an interaction term between a structural brain measure and CR proxy (e.g., education years x hippocampal volume, education years x frontal thickness) and age, sex, parental longevity, cognitive impairment, and global health. Separate models were performed for each hemisphere. Our physical CR proxies (number of blocks walked per day; physical activity days per week) were not examined in the MRI subsample because only a limited number of assessments were administered during (or close to) the MRI visit. A simplified example of the linear mixed effect models used is Gait Speed= b1iLeft Hippocampus + b2iEducation + b3i(Left HippocampusxEducation) + b4i(Age) + b5(Gender) + b6(Parental Longevity) + b7(Cognitive Impairment) + b8(Global Health)

3. Results

The mean age of the sample (n=1,079) was 75.4±6.7 years (56.0 % female; see Table 1 for additional sample characteristics). A total of 200 participants died, and 240 participants dropped out of the study. The mean age of the MRI subsample (n=112) was 78.9±6.3 years (56.3% female; see Table 1). Table S1 (eTable 1a–b)provides a summary of sample size and mean age of participants at each follow up.

Table 1.

Baseline Characteristics of the whole sample and the MRI sub-sample.

Whole sample
 MRI sub-sample
(n= 1,079) (n= 112)
Age (years), mean, SD 75.4  6.7 78.9   6.3
Female, n, % 604 56.0 63  56.3
OPEL n, % 574 53.2 67  59.8
Proxies for cognitive reserve
Education (years), mean, SD 17.5a  2.8 17.8   2.7
Blocks walked/day, mean, SD 14.5 13.8 -
Physical activity days/week, mean, SD 2.6  2.2 -
Living with someone, n, % 631 65.1 66  61.7
Engaged in paid/volunteer work, n, % 672 62.7b 85  75.9
Medical conditions
Hypertension, n, % 448 41.5 35  31.5
Diabetes, n, % 92  8.5 8   7.1
Chronic heart failure, n, % 10  0.9 0   0
Arthritis, n, % 428 39.7 59  54.1
Depression, n, % 209 19.7 1   9.0
Stroke, n, % 38  3.5 0   0
Chronic obstructive pulmonary disease, n, % 32  3 0   0
Myocardial infarction, n, % 59  5.5 0   0
Parkinson’s disease, n, % 13  1.2 1   0.9
Gait
Gait-speed (cm/s), mean, SD 110.1 19.9 102.7  20.0
Cognition
Global cognition, mean, SD 0.4c  5.8 2.0   6.8
Hippocampal volume
Left hippocampus (mm3), mean, SD - 2904.3 326.1
Right hippocampus (mm3), mean, SD - 2971.2 346.9
Frontal cortical thickness
Left frontal (mm), mean, SD - 2.5   0.12
Right frontal (mm), mean, SD - 2.5   0.12
Open in a new tab

OPEL, offspring of parents with exceptional longevity; BMI, body mass index; cm, centimeter; s, seconds; SD, standard deviation; mm, millimeter; mm3 cubic millimeter. Note: Gait and cognitive assessments were conducted annually since 2008. The mean number of assessments was 3.90 (SD= 2.50) and the mean follow-up time, mean was 3.93 (SD=3.25). MRI data collection started in 2018 and therefore has a different baseline and shorter follow-up time (M Follow-up = 1.4 years).

a

The range was 27 years. The 75th percentile was 20 years and the 25th percentile was 16 years. Only 9 participants had less than 12 years of education.

b

94% of those that engaged in paid or volunteer work, worked for pay and on average 19.04 h per week.

c

A z-standardized global cognition composite was generated from z-standardized performance on 10 different neuropsychological tests.

3.1. Cognitive reserve proxies and age-related decline in cognition

The longitudinal associations between CR proxies and global cognition and gait speed as a function of age are summarized in Table 2. All CR proxies were associated with the rate of decline in global cognition such that older adults with higher CR (more education years, greater number of city blocks walked per day, physical activity days per week, engaging in paid or volunteer work, and living with someone) was associated with a slower rate of the decline than those with lower CR (fewer education years, smaller number of blocks walked per day, fewer physical activity days per week; not engaging in paid or volunteer work, and living alone). These results are visualized in Fig. 1a–c.

Table 2.

Age-related decline in global cognition and gait speed as a function of CR proxies.

Education years Blocks walked per day Physical activity days per week Living with someone Paid or volunteer work
β (95% CI) p-value β (95% CI) p-value β (95% CI) p-value β (95% CI) p-value β (95% CI) p-value
Global Cognition
CR Proxy x Age (interaction) 0.002 (0.001, 0.004) 0.001 0.001 (0.001, 0.001) <0.001 0.004 (0.002, 0.006) <0.001 0.012 (0.004, 0.020)) 0.005 0.009 (0.001, 0.017) 0.024
Gait Speed
CR Proxy x Age (interaction) 0.001(−0.001, 0.003) 0.279 −0.000 (−0.001, 0.000) 0.614 −0.000 (−0.003, 0.002) 0.823 −0.006 (−0.017, 0.004), 0.248 0.003 (−0.007, 0.013) 0.510
Open in a new tab

Fig. 1.

Fig. 1.

Open in a new tab

a: Age-related changes in global cognition as a function of low (25th percentile = 16 years) and high (75th percentile = 20 years) education. The predicted longitudinal trajectory is based on data with a mean of 3.9 years of follow-up. b: Age-related changes in global cognition in those living with someone and those that live alone. The predicted longitudinal trajectory is based on data with a mean of 3.9 years of follow-up. c: Age-related changes in global cognition as a function of low (25th percentile = 0 days) and high (75th percentile = 4 days) levels of weekly physical activity participation. The predicted longitudinal trajectory is based on data with a mean of 3.9 years of follow-up.

3.2. Cognitive reserve proxies and age-related decline in gait speed

No CR proxies were associated with the rate of decline in gait speed (see Table 2).

3.3. Hippocampal volume, frontal thickness, CR, and cognition

The association between (right) hippocampal volume and global cognition was moderated by one CR proxy: living with someone (see Table 3) – such that the association between right hippocampal volume and global cognition was weaker among those who lived with a partner or spouse (high CR) than among those who lived alone (low CR). Finally, the association between frontal cortical thickness and global cognition in both hemispheres was significantly stronger among those who worked or volunteered (higher CR) than among those who did not work or volunteer (low CR; see Table 4).

Table 3.

Global cognition and gait speed as a function of CR proxies and hippocampal volumea.

Education years Living with someone Paid or volunteer work
β (95%CI) β (95%CI) β (95%CI)
p-value p-value p-value
Global Cognition
CR Proxy x Hippocampal Volume (mm3 Left hemisphere) −0.116 (−0.302, 0.070) −0.624 (−1.464, 0.216) 0.422 (−0.477, 1.321)
0.220 0.145 0.358
CR Proxy x Hippocampal Volume (mm3 Right hemisphere) −0.959 (−0.273, 0.082) −0.960 (−1.738, −0.174) −0.083 (−0.994, 0.828)
0.289 0.016 0.858
Gait Speed
CR Proxy x Hippocampal Volume (mm3, Left hemisphere) −0.064 (−0.266, 0.138) 0.505 (−0.497, 1.506) −0.000 (−1.019, 1.012)
0.532 0.323 0.994
CR Proxy x Hippocampal Volume (mm3, Right hemisphere) −0.0264 (−0.219, 0.166) 0.624 (−0.323, 1.570) 0.060 (−0.976, 1.096)
0.788 0.196 0.910
Open in a new tab
a

β and 95%CI are multiplied by 1000

Table 4.

Global cognition and gait speed as a function of CR proxies and frontal brain measures.

Education Living with someone Paid or volunteer work
β (95%CI) β (95%CI) β (95%CI)
P value P value P value
Global Cognition
CR Proxy x Frontal Cortical thickness (mm, Left hemisphere) 0.004 (−0.524, 0.532) 0.382 (−2.227, 2.991) 4.507 (1.800, 7.213)
0.988 0.774 0.001
CR Proxy x Frontal Cortical thickness (mm, Right hemisphere) 0.354 (−0.093, 0.801) −0.141 (−2.419, 2.702) 4.384 (1.761, 7.006)
0.120 0.914 0.001
Gait speed
CR Proxy x Frontal Cortical thickness (mm, Left hemisphere) 0.438 (−0.144, 1.019) 1.689 (−1.270, 4.648) 3.065 (−0.021, 6.150)
0.140 0.263 0.052
CR Proxy x Frontal Cortical thickness (mm, Right hemisphere) 0.562 (0.078, 1.046) 2.002 (−0.921, 4.925) 2.517 (−0.523, 5.558)
0.023 0.179 0.105
Open in a new tab

3.4. Hippocampal volume, frontal thickness, CR, and gait speed

The association between hippocampal volumes and gait speed were not moderated by any CR proxies. The association between right frontal cortical thicknesses and gait speed was moderated by education years – such that it was stronger among those with high CR (more education years) than among those with low CR (fewer education years).

3.5. Sensitivity analyses

Only 15 participants developed dementia during follow-up visits. When they were removed from our analyses a similar pattern of results was observed – but the association between education and global cognitive decline was attenuated (see eTable 2). When we performed separate (stratified) analyses of those that had less than 5 years and those that had 5 or more years of follow-up, however, we found that age-related cognitive decline was only moderated by CR proxies in those with 5 or more years of follow-up (see Table 5). Supplementary analyses of individual measures of executive functions (Trails B time) and episodic memory (free recall from the FCSRT) revealed that 1) age-related decline in executive function was moderated by both cognitive (education years) and physical (blocks walked per day, physical activity days per week) CR proxies, while 2) age-related memory decline was moderated by physical (blocks walked per day, physical activity days per week) CR proxies (see eTable 3). Finally, removing individuals without follow-up gait and or cognitive data did not make any substantial changes to our results (see eTable 4)

Table 5.

Age-related decline in global cognition and gait speed as a function of CR proxies in those with 5 or more years of follow up and those with.

Education years Blocks walked per day Physical activity days per week Living with someone Paid or volunteer work
β (95% CI) β (95% CI) β (95% CI) β (95% CI) β (95% CI)
p-value p-value p-value p-value p-value
Global Cognition
CR Proxy x Age in those ≥5 years of follow up 0.003 (0.001, 0.004) 0.001 (0.001, 0.001) 0.005 (0.002, 0.007) 0.012 (0.003, 0.021) 0.010 (0.001, 0.019)
0.001 <0.001 <0.001 0.010 0.022
CR Proxy x Age in those <5 years of follow up 0.001 (−0.002, 0.004) 0.001 (−0.000, 0.001) 0.001 (−0.005, 0.004) 0.001 (−0.020, 0.023)) 0.005 (−0.023, 0.013)
0.716 0.089 0.738 0.897 0.607
Gait Speed
CR Proxy x Age in those ≥5 years of follow up 0.001(−0.001, 0.003) −0.000 (−0.001, 0.000) −0.001 (−0.002, 0.004) −0.000 (−0.012, 0.011) 0.003 (−0.008, 0.014)
0.183 0.571 0.557 0.954 0.559
CR Proxy x Age in those <5 years of follow up. 0.001 (−0.003, 0.005) 0.000 (−0.001, 0.001) −0.004 (−0.009, 0.002) −0.021 (−0.046, 0.005) 0.006 (−0.017, 0.029)
0.552 0.574 0.203 0.114 0.615
Open in a new tab

4. Discussion

The key finding of this study was that higher CR when inferred from 1) cognitive (education years), 2) physical (blocks walked per day; physical activity days per week), and 3) social (volunteering, living with someone) CR proxies are associated with slower rates of age-related decline in global cognition – but not with the rate of decline in gait speed.

4.1. Cognitive reserve proxies are associated with slower rates of age-related cognitive decline

Consistent with the CR framework (Stern, 2009, 2012; Stern et al., 2022; Stern et al., 2020), higher CR was associated with a slower rate of age-related decline in global cognition, regardless of whether it was inferred from cognitive, physical, or social CR proxies. Thus, frequent participation in cognitive, physical, and social activities not only delay the onset of dementia (Kukull et al., 2002; Lindsay et al., 2002; Roe et al., 2011; Silva et al., 2020; Verghese et al., 2003), but is also associated with a slower rate of cognitive decline in older adults without dementia. These results are consistent with several previous observations (Alvares Pereira et al., 2022; Barnes et al., 2004; Jokinen et al., 2016; Wang et al., 2013; Zahodne et al., 2015), yet inconsistent with others – particularly those that used years (or level of) education as the sole CR proxy (Berggren et al., 2018; Karlamangla et al., 2009; Wilson et al., 2009; Zahodne et al., 2011). This discrepancy could be due to the fact that our sample was composed of an exceptionally educated group of older adults with a mean of 17.5 education years (i.e., at least some graduate-level education). The average in other studies ranged from 10.2 to 14.1 years (Berggren et al., 2018; Karlamangla et al., 2009; Wilson et al., 2009; Zahodne et al., 2011) i.e., some college education or less). Thus, our education variable reflects cognitive activity that took place well into adulthood to a greater extent than other studies – and therefore may be a better reflection of CR or lifelong (rather than childhood) participation in cognitive activities – albeit limiting the generalizability of our findings to average or less educated older adults. Stratified analyses in those with less than 5 years and those with 5 or more years of follow-up further suggest that a longer follow-up time may be needed to detect the effects of CR proxies on age-related cognitive decline. It is also possible that those with fewer years of follow-up were systematically different in some other way, e.g. in poorer health than those with more years of follow-up. When demographic and medical characteristics were contrasted in those with less than 5 and those with 5 or more years of follow-up, those with shorter follow up were more likely to be women and have a history of chronic heart failure, depression, and stroke, and less likely to live with someone and engaging in paid or volunteer work (see eTable 5). Finally, this discrepancy could be due to the fact that our sample were offspring of parents with exceptional longevity (e.g., more than 53.4 % had at least one parent who lived 95 years or more). Parental longevity is associated with a slower rate of cognitive decline (Dutta et al., 2014; Frederiksen et al., 2002) and better brain health (Murabito et al., 2014; Tian et al., 2020). Regardless, the results of this study suggest that cognitive, physical, and social CR proxies are associated with age-related cognitive decline in highly educated older adults without dementia who are enriched for longevity. Additional studies are needed to further clarify the moderating effects of different CR proxies on cognitive decline in relatively normal and successful aging (Jin et al., 2023)

Exploratory analyses of the smaller MRI sub-sample suggests that one social CR proxy (living with someone) influence the relationship between cross-sectional hippocampal volume and global cognition and one social CR proxy (paid or volunteer work) influenced the relationship between cross-sectional frontal cortical thickness and global cognition. The nature of these interactions, however, differed. The weaker association between hippocampal volume and cognition observed in those with high (than low) CR when inferred from living with someone is more consistent with the suggestion that those with high CR are better at adapting or coping with brain pathology to maintain cognitive function in aging (Stern et al., 2022; Stern et al., 2020) – although longitudinal brain measures are needed to confirm this conclusion. We also note that while older adults who live alone are more isolated from family members than those who live with someone, they have also been shown to engage in more social activities, and do not always show reduced cognition or accelerated decline in aging (Evans et al., 2019; Mazzuco et al., 2017). Thus, future research is needed to clarify the potential of living with someone as a good proxy for CR. Regardless, this finding is consistent with the previous observation that the association between cortical thickness and fluid reasoning is weaker among those that highly express a task-invariant CR brain pattern (derived from functional MRI) than among those who express this CR brain pattern less (Stern et al., 2018). The strengthened association between frontal cortical thickness (and percent hippocampal volume) and cognition observed in those with high (than low) CR when inferred from engaging in paid or volunteer work is more consistent with the concept of brain maintenance (reduced brain changes and pathologies as a function of lifetime experiences) than cognitive reserve. Yet again, longitudinal brain measures are needed to confirm this conclusion.

4.2. Cognitive reserve proxies are not associated with age-related gait speed decline

Although gait and cognition are interrelated and share neural substrates in aging and dementia (Clouston et al., 2013; Cohen et al., 2016; Leisman et al., 2016; Scherder et al., 2007) no CR proxies examined here were associated with age-related gait speed decline. Yet, exploratory analyses of the MRI sub-sample indicated that the association between cross-sectional frontal cortical thickness and gait speed was stronger among those with more education years (high CR) than those with fewer education years. Longitudinal brain measures are needed to determine the presence of CR, more directly. Thus, the lifestyle factors that render older adults resistant to cognitive decline are likely different from those that render them resistant to gait speed decline. It is also possible that the multifactorial etiology (e.g., musculoskeletal, cardiopulmonary and metabolic (Ferrucci et al., 2000; Rosso et al., 2013)) and the relatively earlier decline and impairment in gait speed than cognition (Beauchet et al., 2016; Buracchio et al., 2010; Jayakody et al., 2022; Quan et al., 2017) may render gait speed less modifiable by lifestyle factors later in life. Yet, additional studies are needed to better understand such differences and inferences. Education years, for example, is typically treated as a covariate rather than a variable of interest in longitudinal studies of gait speed (Atkinson et al., 2007; Jayakody et al., 2022). In the few studies that have examined the influence of education on gait speed decline directly, the evidence have been mixed (Elbaz et al., 2013; Kyrönlahti et al., 2021) – and significant associations have been attributed to variables such as body mass index (BMI) and physical workload (Kyrönlahti et al., 2021).

The relationships between additional physical and social CR proxies and age-related gait speed decline also need to be further explored. Here, we found that the number of blocks walked per day and weekly physical activity levels were not associated age-related gait speed decline, but others utilizing more comprehensive self-reported measures of physical activities (e.g., exercise and non-exercise related walking, formal exercise classes, gardening, housework) have reported significant associations (White et al., 2012). Note, however, that although longitudinal associations between physical CR proxies and gait speed decline were not observed in this study, cross-sectional associations between physical CR proxies and gait speed at baseline were significant (b=.02, p<.0001 (blocks walked per day) b=.12, p<.0001 (physical activity days per week)). We also found that living with someone and working/volunteering was not associated with age-related gait speed decline, but recently reported that social support – particularly tangible social support (e.g., someone to help you if you were confined to bed) – was associated with gait speed decline in aging (Pollak et al., 2023). Others have reported significant associations between loneliness and motor function decline in general (Buchman et al., 2009; Buchman et al., 2010) – and a pilot clinical trial suggested that a volunteering program implemented in older adults improved gait speed (Chen et al., 2020). Thus, additional studies are needed to capture and or identify additional lifestyle factors that protect against age-related gait speed decline, and the influence of cognitive reserve and brain maintenance.

4.3. Strength and limitations

Directly comparing the effects of cognitive, physical, and social CR proxies on age-related cognitive and gait decline is novel. Key strengths of this study include a large sample size of older adults with extensive follow-up and cross-sectional brain health measures to explore if any association between brain status (hippocampal volume and frontal cortical thickness) and cognition and gait were moderated by CR proxies. The MRI sample in this study was smaller and cross-sectional, however, and with only limited follow-up for gait and cognition. Therefore, these results should be interpreted with caution. Longitudinal MRI measures in a larger sample are needed to confirm the conclusions related to CR and brain maintenance reported here. Our study sample was also composed of an exceptionally educated group of older Caucasian adults that are enriched for longevity (53.4% had at least one parent that lived 95 years or more), which may have contributed to our results, and limits the generalization of our findings. Future studies that utilize more diverse samples of older adults with different genetic and educational backgrounds are needed. Finally, more direct measures of the neural implementation of CR (e.g., resting-state or task-based fMRI) are needed.

4.4. Conclusion

This study suggests that cognitive, physical, as well as social CR proxies are associated with a slower rate of cognitive decline – but not gait speed decline – in aging. The multisystem causes and relatively earlier decline in gait than cognition may render it less modifiable by CR proxies in aging.

Supplementary Material

Supplementary

Acknowledgements

This work was supported by grants from National Institute of Health (NIH): National Institute on Aging (NIA): R01AG062659–01A1 (PI: Helena M. Blumen), R01AG057548–01A1 (PI: Joe Verghese), R01AG061155–01 (PI: Sofiya Milman), and R01AG057909 (PI: Nir Barzilai). We would also like to thank Ying Jin for her statistical advice.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.neurobiolaging.2024.05.012.

Footnotes

Submission Verification

The work described in this manuscript has not been previously published elsewhere and has not been submitted simultaneously for consideration/publication elsewhere. Its publication has been approved by all authors and the institution where the work was carried out. If accepted, it will not be published elsewhere in the same form (including electronically) without written consent from the copyright holder.

Disclosures

The authors report no disclosures relevant to the manuscript. The authors have no financial relationships or activities to disclose that could appear to have influenced the submitted work.

CRediT authorship contribution statement

Joe Verghese: Writing – review & editing. Erica F. Weiss: Writing – review & editing, Validation, Supervision. Yaakov Stern: Writing – review & editing. Sofiya Milman: Writing – review & editing, Validation, Supervision, Resources, Methodology, Funding acquisition, Data curation. Christian Habeck: Writing – review & editing. Nir Barzilai: Writing – review & editing, Resources, Methodology, Funding acquisition. Emmeline Ayers: Writing – review & editing, Data curation. Oshadi Jayakody: Writing – review & editing, Visualization, Validation, Formal analysis, Data curation. Helena Blumen: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Funding acquisition, Formal analysis, Data curation, Conceptualization.

References

  1. Alvares Pereira G, Silva Nunes MV, Alzola P, Contador I, 2022. Cognitive reserve and brain maintenance in aging and dementia: An integrative review. Appl. Neuropsychol.: Adult 29 (6), 1615–1625. 10.1080/23279095.2021.1872079. [DOI] [PubMed] [Google Scholar]
  2. Atkinson HH, Rosano C, Simonsick EM, Williamson JD, Davis C, Ambrosius WT, Rapp SR, Cesari M, Newman AB, Harris TB, Rubin SM, Yaffe K, Satterfield S, Kritchevsky SB, 2007. Cognitive function, gait speed decline, and comorbidities: the health, aging, and body composition study. J. Gerontol. Ser. A: Biol. Sci. Med. Sci 62 (8), 844–850. 10.1093/gerona/62.8.844. [DOI] [PubMed] [Google Scholar]
  3. Barnes LL, De Leon CM, Wilson RS, Bienias JL, Evans DA, 2004. Social resources and cognitive decline in a population of older African Americans and whites. Neurology 63 (12), 2322–2326. 10.1212/01.WNL.0000147473.04043. [DOI] [PubMed] [Google Scholar]
  4. Beauchet O, Annweiler C, Callisaya ML, De Cock A-M, Helbostad JL, Kressig RW, Srikanth V, Steinmetz J-P, Blumen HM, Verghese J, 2016. Poor gait performance and prediction of dementia: results from a meta-analysis. J. Am. Med. Dir. Assoc 17 (6), 482–490. 10.1016/j.jamda.2015.12.092. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Berggren R, Nilsson J, Lövdén M, 2018. Education does not affect cognitive decline in aging: A Bayesian assessment of the association between education and change in cognitive performance. Front. Psychol 9, 1138. 10.3389/fpsyg.2018.01138. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Blumen HM, Allali G, Beauchet O, Lipton RB, Verghese J, 2019a. A gray matter volume covariance network associated with the motoric cognitive risk syndrome: a multi-cohort MRI study. J. Gerontol. Ser. A: Biol. Med. Sci 74 (6), 884–889. 10.1093/gerona/gly158. [DOI] [PubMed] [Google Scholar]
  7. Blumen HM, Brown LL, Habeck C, Allali G, Ayers E, Beauchet O, Callisaya M, Lipton RB, Mathuranath PS, Phan TG, Pradeep Kumar VG, Srikanth V, Verghese J, 2019b. Gray matter volume covariance patterns associated with gait speed in older adults: a multi-cohort MRI study. Brain Imaging Behav. 13 (2), 446–460. 10.1007/s11682-018-9871-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Blumen HM, Holtzer R, Brown LL, Gazes Y, Verghese J, 2014. Behavioral and neural correlates of imagined walking and walking while talking in the elderly. Hum. Brain Mapp 35 (8), 4090–4104. 10.1002/hbm.22461. [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Buchman AS, Boyle PA, Wilson RS, Fleischman DA, Leurgans S, Bennett DA, 2009. Association between late-life social activity and motor decline in older adults. Arch. Intern. Med 169 (12), 1139–1146. 10.1001/archinternmed.2009.135. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Buchman AS, Boyle PA, Wilson RS, James BD, Leurgans SE, Arnold SE, Bennett DA, 2010. Loneliness and the rate of motor decline in old age: the rush memory and aging project, a community-based cohort study. BMC Geriatr. 10, 77. 10.1186/1471-2318-10-77. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Buracchio T, Dodge HH, Howieson D, Wasserman D, Kaye J, 2010. The trajectory of gait speed preceding mild cognitive impairment. Arch. Neurol 67 (8), 980–986. 10.1001/archneurol.2010.159. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Callisaya ML, Beare R, Phan TG, Blizzard L, Thrift AG, Chen J, Srikanth VK, 2013. Brain structural change and gait decline: A longitudinal population-based study. J. Am. Geriatr. Soc 61 (7), 1074–1079. 10.1111/jgs.12331. [DOI] [PubMed] [Google Scholar]
  13. Callisaya ML, Beare R, Phan TG, Chen J, Srikanth VK, 2014. Global and regional associations of smaller cerebral gray and white Mmatter volumes with gait in older people. PLoS ONE 9 (1), e84909. 10.1371/journal.pone.0084909. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Chen J-C, Chang Q-X, Liang C-C, Hsieh J-G, Liu PP-S, Yen C-F, Loh C-H, 2020. Potential benefits of environmental volunteering programs of the health of older adults: a pilot study. Arch. Gerontol. Geriatr 90, 104113 10.1016/j.archger.2020.104113. [DOI] [PubMed] [Google Scholar]
  15. Clouston SA, Brewster P, Kuh D, Richards M, Cooper R, Hardy R, Rubin MS, Hofer SM, 2013. The dynamic relationship between physical function and cognition in longitudinal aging cohorts. Epidemiol. Rev 35 (1), 33–50. 10.1093/epirev/mxs004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Cohen JA, Verghese J, Zwerling JL, 2016. Cognition and gait in older people. Maturitas 93, 73–77. 10.1016/j.maturitas.2016.05.005. [DOI] [PubMed] [Google Scholar]
  17. Dale AM, Fischl B, Sereno MI, 1999. Cortical surface-based analysis: I. Segmentation and surface reconstruction. Neuroimage 9 (2), 179–194. 10.1006/nimg.1998.0395. [DOI] [PubMed] [Google Scholar]
  18. Duff K, Beglinger LJ, Schoenberg MR, Patton DE, Mold J, Scott JG, Adams RL, 2005. Test-retest stability and practice effects of the RBANS in a community dwelling elderly sample. J. Clin. Exp. Neuropsychol 27 (5), 565–575. 10.1080/13803390490918363. [DOI] [PubMed] [Google Scholar]
  19. Dutta A, Henley W, Robine JM, Llewellyn D, Langa KM, Wallace RB, Melzer D, 2014. Aging children of long-lived parents experience slower cognitive decline. Alzheimer’S. Dement 10 (5 Suppl), S315–322. 10.1016/j.jalz.2013.07.002. [DOI] [PubMed] [Google Scholar]
  20. Elbaz A, Vicente-Vytopilova P, Tavernier B, Sabia S, Dumurgier J, Mazoyer B, Singh-Manoux A, Tzourio C, 2013. Motor function in the elderly: evidence for the reserve hypothesis. Neurology 81 (5), 417–426. 10.1212/WNL.0b013e31829d8761. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Elwood RW, 1991. The Wechsler Memory Scale—Revised: psychometric characteristics and clinical application. Neuropsychol. Rev 2 (2), 179–201. 10.1007/BF01109053. [DOI] [PubMed] [Google Scholar]
  22. Evans IEM, Llewellyn DJ, Matthews FE, Woods RT, Brayne C, Clare L, 2019. Living alone and cognitive function in later life. Arch. Gerontol. Geriatr 81, 222–233. 10.1016/j.archger.2018.12.014. [DOI] [PubMed] [Google Scholar]
  23. Ezzati A, Katz MJ, Lipton ML, Lipton RB, Verghese J, 2015. The association of brain structure with gait velocity in older adults: a quantitative volumetric analysis of brain MRI. Neuroradiology 57 (8), 851–861. 10.1007/s00234-015-1536-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Ferrucci L, Bandinelli S, Benvenuti E, Di Iorio A, Macchi C, Harris TB, Guralnik JM, 2000. Subsystems contributing to the decline in ability to walk: Bridging the gap between epidemiology and geriatric practice in the InCHIANTI study. J. Am. Geriatr. Soc 48 (12), 1618–1625. 10.1111/j.1532-5415.2000.tb03873.x. [DOI] [PubMed] [Google Scholar]
  25. Fischl B, Sereno MI, Dale AM, 1999. Cortical surface-based analysis: II: inflation, flattening, and a surface-based coordinate system. Neuroimage 9 (2), 195–207. 10.1006/nimg.1998.0396. [DOI] [PubMed] [Google Scholar]
  26. Frederiksen H, McGue M, Jeune B, Gaist D, Nybo H, Skytthe A, Vaupel JW, Christensen K, 2002. Do children of long-lived parents age more successfully? Epidemiology 13 (3), 334–339. 10.1097/00001648-200205000-00015. [DOI] [PubMed] [Google Scholar]
  27. González HM, Tarraf W, Bowen ME, Johnson-Jennings MD, Fisher GG, 2013. What do parents have to do with my cognitive reserve? Life course perspectives on twelve-year cognitive decline. Neuroepidemiology 41 (2), 10. 10.1159/000350723. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Grober E, Merling A, Heimlich T, Lipton RB, 1997. Free and cued selective reminding and selective reminding in the elderly. J. Clin. Exp. Neuropsychol 19 (5), 643–654. 10.1080/01688639708403750. [DOI] [PubMed] [Google Scholar]
  29. Gubbi S, Schwartz E, Crandall J, Verghese J, Holtzer R, Atzmon G, Braunstein R, Barzilai N, Milman S, 2017. Effect of exceptional parental longevity and lifestyle factors on prevalence of cardiovascular disease in offspring. Am. J. Cardiol 120 (12), 2170–2175. 10.1016/j.amjcard.2017.08.040. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Hoffman L, 2015. Longitudinal Analysis: Modeling Within-person Fluctuation And Change. Routledge. [Google Scholar]
  31. Holt-Lunstad J, 2018. Why social relationships are important for physical health: A systems approach to understanding and modifying risk and protection. Annu. Rev. Psychol 69, 437–458. 10.1146/annurev-psych-122216-011902. [DOI] [PubMed] [Google Scholar]
  32. Holt-Lunstad J, Smith TB, Layton JB, 2010. Social Relationships and Mortality Risk: A Meta-analytic Review. PLOS Med. 7 (7), e1000316 10.1371/journal.pmed.1000316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Holtzer R, Epstein N, Mahoney JR, Izzetoglu M, Blumen HM, 2014. Neuroimaging of mobility in aging: a targeted review. J. Gerontol. Ser. A: Biomed. Sci. Med. Sci 69 (11), 1375–1388. 10.1093/gerona/glu052. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Holtzer R, Wang C, Lipton R, Verghese J, 2012. The protective effects of executive functions and episodic memory on gait speed decline in aging defined in the context of cognitive reserve. J. Am. Geriatr. Soc 60 (11), 2093–2098. 10.1111/j.1532-5415.2012.04193.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Iglesias JE, Augustinack JC, Nguyen K, Player CM, Player A, Wright M, Roy N, Frosch MP, McKee AC, Wald LL, 2015. A computational atlas of the hippocampal formation using ex vivo, ultra-high resolution MRI: application to adaptive segmentation of in vivo MRI. Neuroimage 115, 117–137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Jayakody O, Breslin M, Ayers E, Verghese J, Barzilai N, Milman S, Weiss E, Blumen HM, 2022. Relative trajectories of gait and cognitive decline in aging. J. Gerontol. Ser. A: Biomed. Sci. Med. Sci 77 (6), 1230–1238. 10.1093/gerona/glab346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Jin Y, Lin L, Xiong M, Sun S, Wu S.-c, 2023. Moderating effects of cognitive reserve on the relationship between brain structure and cognitive abilities in middle-aged and older adults. Neurobiol. Aging 128, 49–64. 10.1016/j.neurobiolaging.2023.04.003. [DOI] [PubMed] [Google Scholar]
  38. Jokinen H, Melkas S, Madureira S, Verdelho A, Ferro JM, Fazekas F, Schmidt R, Scheltens P, Barkhof F, Wardlaw JM, Inzitari D, Pantoni L, Erkinjuntti T, 2016. Cognitive reserve moderates long-term cognitive and functional outcome in cerebral small vessel disease. J. Neurol., Neurosurg., Psychiatry 87 (12), 1296–1302. 10.1136/jnnp-2016-313914. [DOI] [PubMed] [Google Scholar]
  39. Karlamangla AS, Miller-Martinez D, Aneshensel CS, Seeman TE, Wight RG, Chodosh J, 2009. Trajectories of cognitive function in late life in the United States: Demographic and socioeconomic predictors. Am. J. Epidemiol 170 (3), 331–342. 10.1093/aje/kwp154. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Kaup AR, Mirzakhanian H, Jeste DV, Eyler LT, 2011. A review of the brain structure correlates of successful cognitive aging. J. Neuropsychiatry Clin. Neurosci 23 (1), 6–15. 10.1176/jnp.23.1.jnp6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Kukull WA, Higdon R, Bowen JD, McCormick WC, Teri L, Schellenberg GD, van Belle G, Jolley L, Larson EB, 2002. Dementia and Alzheimer disease incidence: A prospective cohort study. Arch. Neurol 59 (11), 1737–1746. 10.1001/archneur.59.11.1737. [DOI] [PubMed] [Google Scholar]
  42. Kyrönlahti SM, Stenholm S, Raitanen J, Neupane S, Koskinen S, Tiainen K, 2021. Educational differences in decline in maximum gait speed in older adults over an 11-Year follow-up. J. Gerontol. Ser. A: Biomed. Sci. Med. Sci 76 (4), 703–709. 10.1093/gerona/glaa196. [DOI] [PubMed] [Google Scholar]
  43. Lane AP, Windsor TD, Andel R, Luszcz MA, 2017. Is occupational complexity associated with cognitive performance or decline? Results from the Australian longitudinal study of ageing. Gerontology 63 (6), 550–559. 10.1159/000475559. [DOI] [PubMed] [Google Scholar]
  44. Leisman G, Moustafa AA, Shafir T, 2016. Thinking, walking, talking: Integratory motor and cognitive brain function. Front. Public Health 4, 94. 10.3389/fpubh.2016.00094. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Lindsay J, Laurin D, Verreault R, Hébert R, Helliwell B, Hill GB, McDowell I, 2002. Risk factors for Alzheimer’s disease: A prospective analysis from the Canadian study of health and aging. Am. J. Epidemiol 156 (5), 445–453. 10.1093/aje/kwf074. [DOI] [PubMed] [Google Scholar]
  46. Manor B, Newton E, Abduljalil A, Novak V.J.Dc, 2012. The relationship between brain volume and walking outcomes in older adults with and without diabetic peripheral neuropathy. Diabetes Care 35 (9), 1907–1912. 10.2337/dc11-2463. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Mazzuco S, Meggiolaro S, Ongaro F, Toffolutti V, 2017. Living arrangement and cognitive decline among older people in Europe. Ageing Soc. 37 (6), 1111–1133. 10.1017/S0144686X16000374. [DOI] [Google Scholar]
  48. Murabito JM, Beiser AS, Decarli C, Seshadri S, Wolf PA, Au R, 2014. Parental longevity is associated with cognition and brain ageing in middle-aged offspring. Age Ageing 43 (3), 358–363. 10.1093/ageing/aft175. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. O’Brien C, Holtzer R, 2021. Cognitive reserve moderates associations between walking performance under single- and dual-Task conditions and incident mobility impairment in older adults. J. Gerontol. Ser. A: Biol. Sci. Med. Sci 76 (10), e314–e320. 10.1093/gerona/glab178. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. O’Shea A, Cohen RA, Porges EC, Nissim NR, Woods AJ, 2016. Cognitive aging and the hippocampus in older adults. Front. Aging Neurosci 8, 298. 10.3389/fnagi.2016.00298. [DOI] [PMC free article] [PubMed] [Google Scholar]
  51. Quan M, Xun P, Chen C, Wen J, Wang Y, Wang R, Chen P, He K, 2017. Walking pace and the risk of cognitive decline and dementia in elderly populations: a meta-analysis of prospective cohort studies. J. Gerontol. Ser. A: Biol. Sci. Med. Sci 72 (2), 266–270. 10.1093/gerona/glw121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Reitan RM, 1958. Validity of the trail making test as an indicator of organic brain damage. Percept. Mot. Skills 8 (3), 271–276. [Google Scholar]
  53. Roe CM, Fagan AM, Grant EA, Marcus DS, Benzinger TLS, Mintun MA, Holtzman DM, Morris JC, 2011. Cerebrospinal fluid biomarkers, education, brain volume, and future cognition. Arch. Neurol 68 (9), 1145–1151. 10.1001/archneurol.2011.192. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. Rosano C, Aizenstein HJ, Studenski S, Newman AB, 2007. A regions-of-interest volumetric analysis of mobility limitations in community-dwelling older adults. J. Gerontol. Ser. A: Biol. Sci. Med. Sci 62 (9), 1048–1055. 10.1093/gerona/62.9.1048. [DOI] [PubMed] [Google Scholar]
  55. Rosano C, Studenski SA, Aizenstein HJ, Boudreau RM, Longstreth WT Jr., Newman AB, 2012. Slower gait, slower information processing and smaller prefrontal area in older adults. Age Ageing 41 (1), 58–64. 10.1093/ageing/afr113. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. Rosso AL, Studenski SA, Chen WG, Aizenstein HJ, Alexander NB, Bennett DA, Black SE, Camicioli R, Carlson MC, Ferrucci L, Guralnik JM, Hausdorff JM, Kaye J, Launer LJ, Lipsitz LA, Verghese J, Rosano C, 2013. Aging, the central nervous system, and mobility. J. Gerontol. Ser. A: Biol. Sci. Med. Sci 10.1093/gerona/glt089. [DOI] [PMC free article] [PubMed] [Google Scholar]
  57. Rosso AL, Verghese J, Metti AL, Boudreau RM, Aizenstein HJ, Kritchevsky S, Harris T, Yaffe K, Satterfield S, Studenski S, 2017. Slowing gait and risk for cognitive impairment: The hippocampus as a shared neural substrate. Neurology 89 (4), 336–342. 10.1212/WNL.0000000000004153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Ruff R, Light R, Parker S, Levin H, 1996. Benton controlled oral word association test: Reliability and updated norms. Arch. Clin. Neuropsychol 11 (4), 329–338. [PubMed] [Google Scholar]
  59. Scarmeas N, Levy G, Tang MX, Manly J, Stern Y, 2001. Influence of leisure activity on the incidence of Alzheimer’s disease. Neurology 57 (12), 2236–2242. 10.1212/wnl.57.12.2236. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Scherder E, Eggermont L, Swaab D, van Heuvelen M, Kamsma Y, de Greef M, van Wijck R, Mulder T, 2007. Gait in ageing and associated dementias; its relationship with cognition. Neurosci. Biobehav. Rev 31 (4), 485–497. 10.1016/j.neubiorev.2006.11.007. [DOI] [PubMed] [Google Scholar]
  61. Seblova D, Berggren R, Lövdén M, 2020. Education and age-related decline in cognitive performance: Systematic review and meta-analysis of longitudinal cohort studies. Ageing Res. Rev 58, 101005 10.1016/j.arr.2019.101005. [DOI] [PubMed] [Google Scholar]
  62. Silva NCBS, Gill DP, Nagamatsu LS, Owen AM, Petrella RJJEG, 2020. Systolic blood pressure dipping may be associated with mobility impairment and brain volume in community-dwelling older adults: An exploratory study. Exp. Gerontol 141, 111100 10.1016/j.exger.2020.111100. [DOI] [PubMed] [Google Scholar]
  63. Singh-Manoux A, Marmot MG, Glymour M, Sabia S, Kivimäki M, Dugravot A, 2011. Does cognitive reserve shape cognitive decline? Ann. Neurol 70 (2), 296–304. 10.1002/ana.22391. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Stern Y, 2009. Cognitive reserve. Neuropsychologia 47 (10), 2015–2028. 10.1016/j.neuropsychologia.2009.03.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Stern Y, 2012. Cognitive reserve in ageing and Alzheimer’s disease. Lancet Neurol. 11 (11), 1006–1012. 10.1016/s1474-4422(12)70191-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  66. Stern Y, Albert M, Barnes C, Cabeza R, Pascual-Leone A, Rapp P, 2022. A Framework for concepts of reserve and resilience in aging. Neurobiol. Aging 124, 100–103. 10.1016/j.neurobiolaging.2022.10.015. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Stern Y, Arenaza-Urquijo EM, Bartrés-Faz D, Belleville S, Cantilon M, Chetelat G, Ewers M, Franzmeier N, Kempermann G, Kremen WS, Okonkwo O, Scarmeas N, Soldan A, Udeh-Momoh C, Valenzuela M, Vemuri P, Vuoksimaa E, 2020. Whitepaper: Defining and investigating cognitive reserve, brain reserve, and brain maintenance. Alzheimer’S. Dement 16 (9), 1305–1311. 10.1016/j.jalz.2018.07.219. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Stern Y, Gazes Y, Razlighi Q, Steffener J, Habeck C, 2018. A task-invariant cognitive reserve network. Neuroimage 178, 36–45. 10.1016/j.neuroimage.2018.05.033. [DOI] [PMC free article] [PubMed] [Google Scholar]
  69. Stern Y, Habeck C, Moeller J, Scarmeas N, Anderson KE, Hilton HJ, Flynn J, Sackeim H, Van Heertum R, 2005. Brain networks associated with cognitive reserve in healthy young and old adults. Cereb. Cortex 15 (4), 394–402. 10.1093/cercor/bhh142. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Tabert MH, Manly JJ, Liu X, Pelton GH, Rosenblum S, Jacobs M, Zamora D, Goodkind M, Bell K, Stern Y, 2006. Neuropsychological prediction of conversion to Alzheimer disease in patients with mild cognitive impairment. Arch. Gen. Psychiatry 63 (8), 916–924. [DOI] [PubMed] [Google Scholar]
  71. Tian Q, Pilling LC, Atkins JL, Melzer D, Ferrucci L, 2020. The relationship of parental longevity with the aging brain-results from UK Biobank. Geroscience 42 (5), 1377–1385. 10.1007/s11357-020-00227-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  72. Tripathi S, Verghese J, Blumen HM, 2019. Gray matter volume covariance networks associated with dual-task cost during walking-while-talking. Hum. Brain Mapp 40 (7), 2229–2240. 10.1002/hbm.24520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  73. Verghese J, LeValley A, Derby C, Kuslansky G, Katz M, Hall C, Buschke H, Lipton RB, 2006. Leisure activities and the risk of amnestic mild cognitive impairment in the elderly. Neurology 66 (6), 821–827. 10.1212/01.wnl.0000202520.68987.48. [DOI] [PMC free article] [PubMed] [Google Scholar]
  74. Verghese J, Lipton RB, Katz MJ, Hall CB, Derby CA, Kuslansky G, Ambrose AF, Sliwinski M, Buschke H, 2003. Leisure Activities and the risk of dementia in the elderly. N. Engl. J. Med 348 (25), 2508–2516. 10.1056/NEJMoa022252. [DOI] [PubMed] [Google Scholar]
  75. Wang HX, Jin Y, Hendrie HC, Liang C, Yang L, Cheng Y, Unverzagt FW, Ma F, Hall KS, Murrell JR, Li P, Bian J, Pei JJ, Gao S, 2013. Late life leisure activities and risk of cognitive decline. J. Gerontol. Ser. A: Biol. Sci. Med. Sci 68 (2), 205–213. 10.1093/gerona/gls153. [DOI] [PMC free article] [PubMed] [Google Scholar]
  76. Wechsler D, 1955. Manual for the Wechsler adult intelligence scale. [Google Scholar]
  77. White DK, Neogi T, Nevitt MC, Peloquin CE, Zhu Y, Boudreau RM, Cauley JA, Ferrucci L, Harris TB, Satterfield SM, Simonsick EM, Strotmeyer ES, Zhang Y, 2012. Trajectories of Gait Speed Predict Mortality in Well-Functioning Older Adults: The Health, Aging and Body Composition Study. J. Gerontol. Ser. A: Biol. Sci. Med. Sci 68 (4), 456–464. 10.1093/gerona/gls197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  78. Williams BW, Mack W, Henderson VW, 1989. Boston naming test in Alzheimer’s disease. Neuropsychologia 27 (8), 1073–1079. 10.1016/0028-3932(89)90186-3. [DOI] [PubMed] [Google Scholar]
  79. Wilson RS, Hebert LE, Scherr PA, Barnes LL, Mendes de Leon CF, Evans DA, 2009. Educational attainment and cognitive decline in old age. Neurology 72 (5), 460–465. 10.1212/01.wnl.0000341782.71418.6c. [DOI] [PMC free article] [PubMed] [Google Scholar]
  80. Wilson RS, Yu L, Lamar M, Schneider JA, Boyle PA, Bennett DA, 2019. Education and cognitive reserve in old age. Neurology 92 (10), e1041–e1050. 10.1212/wnl.0000000000007036. [DOI] [PMC free article] [PubMed] [Google Scholar]
  81. Zahodne LB, Glymour MM, Sparks C, Bontempo D, Dixon RA, MacDonald SWS, Manly JJ, 2011. Education does not slow cognitive decline with aging: 12-Year evidence from the Victoria longitudinal study. J. Int. Neuropsychol. Soc 17 (6), 1039–1046. 10.1017/S1355617711001044. [DOI] [PMC free article] [PubMed] [Google Scholar]
  82. Zahodne LB, Stern Y, Manly JJ, 2015. Differing effects of education on cognitive decline in diverse elders with low versus high educational attainment. Neuropsychology 29 (4), 649–657. 10.1037/neu0000141. [DOI] [PMC free article] [PubMed] [Google Scholar]
  83. Zhang WB, Aleksic S, Gao T, Weiss EF, Demetriou E, Verghese J, Holtzer R, Barzilai N, Milman S, 2020. Insulin-like growth factor-1 and IGF binding proteins predict all-cause mortality and morbidity in older adults. Cells 9 (6), 1368. 10.3390/cells9061368. [DOI] [PMC free article] [PubMed] [Google Scholar]
  84. Zimmerman ME, Katz MJ, Wang C, Burns LC, Berman RM, Derby CA, L’Italien G, Budd D, Lipton RB, 2015. Comparison of “word” vs.“picture” version of the free and cued selective reminding test (FCSRT) in older adults. Alzheimer’s & Dementia: Diagnosis. Assess. Dis. Monit 1 (1), 94–100. 10.1016/j.dadm.2014.11.006. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplementary
NIHMS2141906-supplement-Supplementary.docx (40.8KB, docx)

RESOURCES