Skip to main content
NCBI home page
The Journals of Gerontology Series B: Psychological Sciences and Social Sciences logoLink to The Journals of Gerontology Series B: Psychological Sciences and Social Sciences
. 2024 Apr 30;79(7):gbae075. doi: 10.1093/geronb/gbae075

The Effects of Computerized Cognitive Training in Older Adults’ Cognitive Performance and Biomarkers of Structural Brain Aging

Hyun Kyu Lee 1,✉,a, Chandramallika Basak 2,a, Sarah-Jane Grant 3, Nicholas R Ray 4, Paulina A Skolasinska 5, Chris Oehler 6, Shuo Qin 7, Andrew Sun 8, Evan T Smith 9, G Hulon Sherard 10, Adriana Rivera-Dompenciel 11, Mike Merzenich 12, Michelle W Voss 13
Editor: Vanessa Taler14
PMCID: PMC11165429  PMID: 38686621

Abstract

Objectives

Cognitive training (CT) has been investigated as a means of delaying age-related cognitive decline in older adults. However, its impact on biomarkers of age-related structural brain atrophy has rarely been investigated, leading to a gap in our understanding of the linkage between improvements in cognition and brain plasticity. This study aimed to explore the impact of CT on cognitive performance and brain structure in older adults.

Methods

One hundred twenty-four cognitively normal older adults recruited from 2 study sites were randomly assigned to either an adaptive CT (n = 60) or a casual game training (active control, AC, n = 64).

Results

After 10 weeks of training, CT participants showed greater improvements in the overall cognitive composite score (Cohen’s d = 0.66, p < .01) with nonsignificant benefits after 6 months from the completion of training (Cohen’s d = 0.36, p = .094). The CT group showed significant maintenance of the caudate volume as well as significant maintained fractional anisotropy in the left internal capsule and in left superior longitudinal fasciculus compared to the AC group. The AC group displayed an age-related decrease in these metrics of brain structure.

Discussion

Results from this multisite clinical trial demonstrate that the CT intervention improves cognitive performance and helps maintain caudate volume and integrity of white matter regions that are associated with cognitive control, adding to our understanding of the changes in brain structure contributing to changes in cognitive performance from adaptive CT.

Clinical Trial Registration

NCT03197454

Keywords: Age-related cognitive decline, Brain plasticity, Computerized cognitive training, Gray matter volume, White matter integrity

The aging population is growing rapidly, resulting in an increasing number of individuals requiring continuous patient care for age-related neurodegenerative illnesses (Beller, 2013). Given that the changes in brain function and structure that lead to dementia begin years before diagnosis, interventions designed to target brain plasticity in cognitively normal adults, such as cognitive training (CT), have been studied as an effective and accessible option for potentially delaying or preventing progression of age-related cognitive decline to dementia (Gates & Valenzuela, 2010; Lampit et al., 2014; Reijnders et al., 2013).

Although there are studies supporting the efficacy of CT in improving cognitive functioning among older adults (Ball et al., 2002; Basak et al., 2020; Edwards et al., 2005; Lee et al., 2020; Mahncke et al., 2006), these improvements are mostly in tasks related to the trained skills. The benefit to untrained skills or cognitive functioning in everyday life has been mixed, suggesting a limited generalization of CT to untrained skills or real-world performance (Melby-Lervåg et al., 2016; Simons et al., 2016).

Moreover, although there are existing studies on cognitive changes associated with brain aging (Fjell & Walhovd, 2010; Lövdén et al., 2013), further research is necessary to develop a comprehensive understanding of the structural brain changes resulting from CT (Boyke et al., 2008; Chen et al., 2021; Engvig et al., 2010; Nguyen et al., 2019; Strenziok et al., 2014). As cognitive plasticity is thought of as training-induced cognitive change in performance caused by training-induced structural changes in the brain, it is critical to demonstrate structural brain changes along with CT-induced performance improvement in large-sample controlled trials.

In terms of structural brain changes, aging is associated with a decrease in gray matter volume. There is a greater loss of gray matter volume in older adults in lateral prefrontal regions, followed by parietal lobe, indicating an anterior–posterior gradient of aging (DeCarli et al., 2005; Raz, 2005). Declines appear later for the posterior and subcortical brain regions, such as the caudate nucleus and temporal lobe including the hippocampus and entorhinal cortex, coupled with less significant atrophy in primary sensory cortices (Fjell et al., 2014; Raz et al., 2005). The caudate and hippocampus are particularly important structures in the aging brain, as they play critical roles in learning, episodic memory, and executive function (Dawe et al., 2020; den Heijer et al., 2010; Nedelska et al., 2012), and their degeneration is associated with an increased risk of developing Alzheimer’s disease (Bauer et al., 2015; Petersen et al., 2000). These reductions in gray matter volumes are accompanied by changes in neural activity and structural and functional connectivity, which may contribute to age-related declines in cognitive function (He et al., 2012; Paul et al., 2011).

Reduced brain connectivity with aging stems in large part from decreases in white matter integrity, demonstrated by reductions in white matter fractional anisotropy (FA; Abe et al., 2008). High FA reflects the directionality of axon bundles in white matter tracts and is a reliable marker of white matter structural integrity. The anterior-posterior gradient found in gray matter in the aging brain also exists in white matter, including major pathways for anterior-posterior connectivity. O’Sullivan et al. (2001) found that the difference between white matter FA in older and younger adults was most significant in anterior tracts, less significant in middle regions of the brain, and nonsignificant in posterior tracts. For long-range connections, the superior longitudinal fasciculus (SLF) has been identified as an important structure that experiences significant structural changes with age (Rizio & Diaz, 2016).

The Goals of the Current Study

In the current study, we investigated the short- and longer-term (6 months after training) effects of adaptive CT on cognitive performance, and the association of such effects with changes in gray matter volume and white matter integrity. We examined changes in gray matter volume within two subcortical regions that are associated with learning and memory and are susceptible to atrophy during the aging process: the caudate nucleus and the hippocampus. Also, we investigated changes in white matter FA in tracts derived from exploratory whole-brain diffusion tensor imaging (DTI) analysis.

We hypothesized that the online, adaptive CT would yield short- and longer-term benefits on cognitive performance. Additionally, we hypothesized that CT would be associated with positive changes in brain structure, and that these changes in brain metrics would be associated with cognitive gains at both the short- and longer-term follow-up time points, indicating a link between brain plasticity and sustained cognitive plasticity in older adults.

Materials and Methods

Design

This was a double-blinded, active-controlled, randomized trial conducted at two sites: the University of Iowa and the University of Texas at Dallas. The study was preregistered on ClinicalTrials.gov (https://clinicaltrials.gov/study/NCT03197454).

Participants

We recruited participants from the communities of Iowa City, Iowa, and Dallas, Texas, using flyers, newspapers, postcards, and online advertisements for a CT study aimed at age-related cognitive decline. Eligible participants were fluent English speakers, aged ≥65 years, with adequate sensorimotor capacity to perform training, no diagnosis of Alzheimer’s or related dementias, a Montreal Cognitive Assessment (MoCA) score > 22, no medical conditions predisposing them to imminent functional decline, and were eligible for magnetic resonance imaging (MRI; for detailed inclusion and exclusion criteria, refer to ClinicalTrials.gov). All participants signed an informed consent form approved by the Institutional Review Boards of the respective university.

Randomization

After screening and baseline testing, 136 participants were assigned to either CT or AC using the minimization method (Endo et al., 2006) with probability 0.8 to the group minimizing symmetric Kullback–Leibler discrepancy on four prognostic variables (age, education, gender, and MoCA). The largest allowable discrepancy in counts between the two groups was set to four participants.

Study Design

All participants in the study underwent cognitive testing and MRI sessions, with the task order fixed within the cognitive session. To maintain blinding, the study was described in the informed consent form as a comparison of two distinct types of CT. Participants completed their assigned home-based training program on their own computers, logging into the training website using preassigned credentials. Depending on their group assignment, participants received either CT or AC games. The user experience was designed to be very similar between the two groups, with no differences in training program interface or design except for the content of the training exercises. Participants were asked to complete at least five training sessions within a 7-day period, with each session lasting 42 min. The training lasted for 10 weeks, with approximately 35 hr of total training across 50 sessions. Participants were allowed to train for more or less than five sessions per week as desired. An additional 4 weeks of training were provided if participants were unable to complete all training sessions within the assigned 10-week period. The general flow of the training program is presented in Supplementary Figure 1.

After completing their assigned training program, participants underwent a cognitive testing session and an MRI scan (posttraining). Access to the training programs was restricted after the 10-week training period. To assess whether the cognitive changes observed posttraining were maintained over time, a follow-up cognitive testing session was conducted 6 months after the completion of training.

MR Imaging Procedures

At the Dallas site, participants were scanned using a Philips Achieva 3 Tesla MRI magnet, and at the Iowa site, using a 3.0T General Electric 750 W MRI magnet. High-resolution anatomical images were acquired using a transverse MP-RAGE (Dallas)/SAG MP-RAGE (with PROMO; Iowa) T1-weighted sequence with the following parameters: TR = 8.46 (Dallas)/8.5 (Iowa) ms; TE = 3.7 (Dallas)/3.2 (Iowa) ms; flip angle = 8°; acquisition matrix = 256 × 230; voxel size = 1 mm3; 240 slices. The duration of this scan was 5 min 55 s (Dallas)/5 min 36 s (Iowa). Diffusion-weighted scans were performed using a single-shot spin-echo echo planar imaging sequence with the following parameters: flip angle = 90°; slice thickness = 2 mm; repetition time = 5,450 (Dallas)/9,000 (Iowa) ms; echo time = 91 (Dallas)/76.3 (Iowa) ms; matrix = 100 × 99 (Dallas)/256 × 256 (Iowa); field of view = 220  × 220 mm (Dallas)/256  × 256 mm (Iowa); voxel resolution = 1.96 × 1.96 × 2.20 (Dallas)/1 × 1 × 2 (Iowa) mm3; bandwidth = 3,116.2 (Dallas)/2,000 (Iowa) Hz/pixel; multiband factor = 2; acceleration = 2 phase. The diffusion-sensitizing gradients were applied along 60 noncollinear directions with a b value of 1,000 s/mm2, and 10 volumes were acquired without diffusion weighting (b = 0).

Preprocessing and Analytical Pipeline for Gray Matter Volumes

Gray matter data were available at both time points for 116 participants. Subcortical segmentation of gray matter volumes was performed using Freesurfer (v6.0) as a part of the fMRIPrep (v20.2.1) preprocessing pipeline (Esteban et al., 2019). The “aseg” atlas was used to estimate the probabilistic location of the subcortical structures (Fischl et al., 2002). The bilateral caudal and hippocampal volumes were extracted from the Freesurfer outputs for further analyses. Caudal and hippocampal volumes were corrected for the total estimated intracranial volume (eICV) for pretraining and posttraining data separately using the residuals method (Pintzka et al., 2015). The correction was performed separately for each study site due to significant baseline differences in raw eICV between the sites (both p < .05). After the correction, the left and right volumes were summed for the caudate and the hippocampus. The statistical models were applied to both lateralized and summed (across left and right) volumes.

Preprocessing and Analytical Pipeline for DTI Data

The diffusion-weighted images were preprocessed using FMRIB Software Library (FSL; Jenkinson et al., 2012; Smith et al., 2004). We first performed a brain extraction using the BET command (Smith, 2002) on each participant’s b0 volume, with a fractional intensity threshold of 0.3, to create a mask that is then registered and applied to the rest of the volumes using FLIRT (Jenkinson et al., 2002) and fslmaths. Each participant’s images were then subjected to an open-source quality control software called DTIprep (Liu et al., 2010), which performs several steps, including an eddy-current and motion artifact correction. The quality control process was in two phases: (1) a fully automatic phase for quality assessment and artifact correction/removal that implemented eddy-current and motion correction and (2) a visual assessment phase for both the diffusion weighted imaging (DWI) volumes and the reconstructed DTI data. Second, DTIFIT (Behrens et al., 2003) was used to fit a directional tensor model for each voxel. We used the Advanced Normalization Tools (ANTs) software to perform a nonlinear registration between each participant’s brain-extracted T1 image and the Montreal Neurological Institute (MNI) coordinates, and between each participant’s T1 and their brain-extracted b0 volume. These registrations are used at a later step to register any specific regions of interest (ROIs) created in MNI space to each participant’s b0.

Exploratory whole-brain DTI analysis used voxel-wise statistical analysis of FA using TBSS (Tract-Based Spatial Statistics), part of FSL (Smith et al., 2004). After the nonlinear registrations were completed, the mean FA image was created and thinned to create a mean FA skeleton, which represented the centers of all tracts common to the group. Each participant’s aligned FA data were then projected onto this skeleton and the resulting data were fed into General Linear Models to determine the changes from pre-to-post training (post > pre) for each group. The randomize function packaged with FSL was applied to implement nonparametric permutation-based group comparisons with a voxel and cluster-based thresholding at t > 2.33 and p < .05 (respectively) (5,000 permutations), which corrects for multiple comparisons across the TBSS skeleton by using the null distribution of the max (across the image) cluster mass.

After correcting for multiple comparisons, significant pre-to-post changes in FA were obtained only for the CT group, not for AC group. To determine specific ROIs from the resulting clusters in the CT group, clusters were binarized and multiplied with masks associated with the JHU ICBM-DTI-81 White Matter Labels Atlas (Mori et al., 2008). This resulted in three discrete white matter ROIs, namely left internal capsule (IC), left SLF, and left external capsule (EC).

Cognitive Assessments

Assessments administered before and after training were grouped into three categories: Processing Speed (Digit Symbol Substitution, Letter Comparison, Pattern Comparison), Memory (Visual Short-Term Memory, Face–Name Task, Selective Reminding Task), and Executive Function (Flanker Task, Task Switch Task, N-back Task). Our primary cognitive outcome measure was an overall cognitive composite score (i.e., mean of all normalized cognitive construct measure scores across pre-, post-, and 6-months follow-up). A detailed description of all assessments is provided in Supplementary Table 1.

Functional and Participant-Reported Outcome Measures

Functional status was assessed using Timed Instrumental Activities of Daily Living (TIADL). Psychological well-being was measured with General life satisfaction, Perceived stress, and Self-efficacy surveys, and the Center for Epidemiologic Studies Depression scale revised (CESD-20). A detailed description of all assessments is provided in Supplementary Table 1.

Cognitive Training Programs

Cognitive training

The CT program (BrainHQ, Posit Science) consisted of 17 exercises designed to improve speed and accuracy of information processing, manipulation of information held in memory, and the execution of higher-level cognitive tasks across multiple cognitive domains, including visual and auditory processing speed and accuracy, attention, memory, and executive function.

During each training session, participants were given seven exercises, and all exercises continually adapted task difficulty based on participants’ performance. This adaptive feature was crucial for inducing maximal performance improvement and brain plasticity by creating a discrepancy between an individual’s cognitive ability and cognitive demands of the training. The adaptivity was applied in three domains: within an exercise, between levels of each exercise, and between distinct exercises.

For example, during a 2–3-min exercise, the exposure duration of stimuli in the exercise would get shorter if participants were doing well and longer if they were doing poorly. The difficulty was continually adjusted so that participants maintained an accuracy level of approximately 75%–85%. Participants played the same exercise two to three times until they reached a certain performance criterion, and when they repeated the exercise, the initial difficulty was set based on their previous best performance to promote further improvement.

As levels progressed, participants were challenged with progressively more cognitively demanding tasks, such as increasing the number of distractors, target distractor similarity, or the number of items to be remembered. At the beginning of training, participants focused on improving lower-level processes that support higher-level cognition and executive function. As training progressed, participants built on their improvements with exercises targeting higher-level associative cognitive function. For a more detailed description of the training exercises, please refer to Supplementary Figure 2 and Supplementary Table 2.

Active control

We utilized 12 commercially available computer games designed to meet several criteria: (1) provide a face-valid approach to CT to ensure participant blinding to group affiliation; (2) match expectation-based influences on cognitive performance outcomes; (3) match the experimental program in overall program use intensity, staff interaction, reward, and overall engagement; and (4) not be continuously adaptive. Three out of the 12 games could increase in difficulty within a session by discrete levels. The user experience in the AC and CT groups was nearly identical, except for the games played. For a detailed explanation of the active control games and example exercises, please refer to Supplementary Table 3 and Supplementary Figure 3.

Interaction With Study Staff

The staff members who interacted with the participants during the training sessions (i.e., the training coach) were not blinded to the participants’ assigned training condition. This was necessary to handle any training-related issues effectively. However, the training coach was instructed to describe each program’s features as potentially beneficial to avoid biasing the participants toward one training program or the other. Posttraining assessments were administered by staff members who were separate from the training coaches and blinded to the participants’ training condition. The training coach regularly interacted with participants, reaching out once a week through their preferred communication method (email, text, or phone call). During these check-ins, the training coach addressed any training-related issues or questions and monitored the participants’ progress.

Statistical Analysis Plan

We tested our predictions with mixed linear regression, with an overall cognitive composite score as the dependent variable for cognitive measures, caudate and hippocampal volume as the dependent variable for gray matter measures, and FA of IC, SLF, and EC as the dependent variable for white matter integrity measures. Our analysis model for each outcome included training group as a fixed factor with site as a random effect, and age, education, gender, and baseline score as covariates. A group effect immediately after training ended estimated the shorter-term effect of CT on each outcome. The same model was used for all the other outcome measures except gray matter volume. As eICV correction was performed on the gray matter volume, age, education, and gender were not used as covariates. We made preplanned analyses of the overall cognitive composite score and exploratory analyses of the cognitive domain measures; and we made preplanned analyses of the hippocampal and caudate volume measures. We performed whole-brain analyses of white matter tracts and identified specific ROIs for further analysis as described above.

Definition of Transfer Effect

We operationalize near and far transfer based on the functional domain. Enhanced performance on untrained cognitive assessments (overall cognitive composite score) represents near transfer, as it assesses the cognitive abilities targeted by CT. The improvement in our functional outcome measure (TIADL) is considered far transfer, indicating the transfer of trained cognitive abilities to the behavioral functional context.

Results

Demographics

Figure 1 shows the consort flow. Of 136 randomized participants, seven participants voluntarily discontinued the study before baseline assessment and five participants voluntarily discontinued the study before completing the first training session. Therefore, 124 participants formed our predefined intent-to-treat (ITT, all randomized participants who completed at least one training session) population (CT = 60; AC = 64).

Figure 1.

Figure 1. Consort chart depicts the flow of participants in a randomized controlled trial, encompassing initial total participants, allocation to intervention and control groups, and post-training and follow-up assessment counts per group.

Open in a new tab

Consort chart. Applicants were first screened via email with a questionnaire that surveyed demographics (e.g., age, English language proficiency). If not excluded based on the survey, a phone interview was checked for medical and nonmedical conditions affecting neuropsychological and MRI testing.

At baseline, there were no group differences in demographic variables (age, gender, education) or MoCA (ps > .16, see Table 1). Also, there was no difference in number of participants, age, gender, education, and MoCA between study sites (ps > .25, see Supplementary Table 4 for details). However, there was a significant baseline difference in the overall cognitive composite score and CESD between the two groups.

Table 1.

Baseline of Demographic, Inclusion Criteria, Cognitive, Functional, and Participant-Reported Outcome Measures

Measures All ITT Cognitive training Active control p Value
N 124 60 64
Age, mean ± SD 71.75 ± 4.59 71.76 ± 4.64 71.75 ± 4.58 .98
Male, n (%) 51 (41.1%) 26 (43%) 25 (39%) .76
Years of education, mean ± SD 16.78 ± 2.40 16.73 ± 2.38 16.82 ± 2.44 .82
Baseline MoCA score, mean ± SD 27.29 ± 2.07 27.28 ± 2.09 27.29 ± 2.07 .97
Ethnic characteristics (non-Hispanic/Hispanic), n 117/7 57/3 60/4 .92
Racial characteristics (White/Black/Asian), n 118/3/3 58/2/0 60/1/3 .52
Overall cognitive composite, mean ± SD −0.14 ± 0.46 −0.23 ± 0.46 −0.06 ± 0.45 .04
TIADL average, mean ± SD 16.19 ± 8.44 16.67 ± 8.37 15.76 ± 8.62 .49
Self-efficacy, mean ± SD 36.75 ± 3.29 36.76 ± 3.45 36.73 ± 3.17 .79
Life satisfaction, mean ± SD 29.03 ± 4.57 28.83 ± 4.73 29.21 ± 4.44 .77
Perceived stress, mean ± SD 7.59 ± 5.17 7.23 ± 4.57 7.93 ± 5.68 .42
CESD-20, mean ± SD 5.14 ± 4.70 4.25 ± 3.28 5.98 ± 5.62 .04
Open in a new tab

Note: CESD-20 = Center for Epidemiologic Studies Depression scale; ITT = intent-to-treat; MoCA = Montreal Cognitive Assessment; TIADL = Timed Instrumental Activities of Daily Living.

Completion Rate

Regarding training adherence, 98.3% of ITT participants completed all 50 sessions of their training program (122 out of 124), 93.5% of participants completed immediate posttraining assessments (116 out of 124) and 4% of participants completed partial remote posttraining assessments due to the coronavirus disease 2019 (COVID-19) pandemic (5 out of 124). Remote assessments only included self-reported assessments (Perceived stress, Self-efficacy, Life satisfaction and CESD surveys) because neuropsychological assessments could not be administered online due to web-based issues regarding response latencies and stimulus presentation, display screen size, and reduced control over participant behavior. A total of 70.9% of participants completed the 6-month follow-up assessment (88 out of 124) and 24.2% of participants completed a partial remote 6-month follow-up assessment due to the COVID-19 pandemic (30 out of 124).

Cognitive Performance Changes Immediately After and 6 Months Following Training Completion

Full assessment results are reported in Table 2. Results showed a statistically significant advantage in the CT group over the AC group at posttraining on the overall cognitive composite score, tested by group differences (Z = 3.48 p < .01, d = 0.66). When conducting the same analysis for each cognitive domain as exploratory analysis, there was a significant advantage for CT over AC at posttraining in executive function after correcting for multiple comparisons with the Benjamini–Hochberg false discovery rate control procedure (Z = 2.55 p = .011, q = 0.03) and no advantage for CT over AC in processing speed (Z = 1.92, p = .054, q = 0.08).

Table 2.

Outcome Measure Analysis (ITT population) for Cognitive, Functional, and Participant-Reported Measures

Measures Baseline Within group change (at posttraining) Within group change (at follow-up) Between groups difference at posttraining (with covariates) Between groups difference at follow-up (with covariates)
CT AC CT AC CT AC
Baseline mean ± SD (range) Baseline mean ± SD (range) Change (95% CI) Change (95% CI) Change (95% CI) Change (95% CI) Change difference (95% CI) Z score Effect size, p value Change difference (95% CI) Z score Effect size, p value
Overall cognitive composite −0.23 ± 0.46 (−1.42–0.82) −0.06 ± 0.45 (−1.26–0.98) 0.36 (0.28, 0.43) 0.15 (0.08, 0.23) 0.31 (0.21, 0.41) 0.17 (0.07, 0.27) 0.19 (0.08, 0.29) 3.47 0.66, .001 0.12 (−0.02, 0.26) 1.68 0.36, .094
Processing speed −0.31 ± 0.67 (−1.74–1.02) −0.08 ± 0.70 (−1.75–1.34) 0.45 (0.32, 0.58) 0.25 (0.13, 0.38) 0.27 (0.11, 0.44) 0.29 (0.13, 0.45) 0.18 (−0.003, 0.36) 1.92 0.36, .054 −0.05 (−0.28, 0.18) −0.43 −0.09, .66
Memory −0.15 ± 0.47 (−1.47–0.83) −0.09 ± 0.53 (−1.50–0.93) 0.23 (0.12, 0.34) 0.13 (0.05, 0.22) 0.31 (0.19, 0.44) 0.21 (0.08, 0.34) 0.09 (−0.02, 0.20) 1.53 0.28, .13 0.12 (−0.03, 0.27) 1.59 0.33, .11
Executive function −0.23 ± 0.70 (−3.03–0.84) −0.01 ± 0.57 (−1.68–1.04) 0.37 (0.23, 0.52) 0.07 (−0.05, 0.19) 0.31 (0.13, 0.48) 0.006 (−0.16, 0.17) 0.22 (0.05, 0.39) 2.55 0.49, .011 0.22 (0.006, 0.44) 2.02 0.44, .044
TIADL average (seconds) 16.67 ± 8.37 (8.11–45.78) 15.76 ± 8.62 (7.0–40.69) −0.31 (−3.28, 2.65) 2.08 (−0.69, 4.86) −0.64 (−3.78, 2.50) 0.94 (−2.27, 4.16) −2.02 (−5.43, 1.39) −1.16 −0.22, .25 −0.86 (−4.45, 2.73) −0.47 −0.10, .64
Self-efficacy 36.76 ± 3.45 (28–40) 36.73 ± 3.17 (29–40) 0.16 (−0.31, 0.64) −0.51 (−1.04, 0.02) 0.02 (−0.56, 0.60) −0.83 (−1.52, 0.15) 0.69 (0.009, 1.38) 1.98 0.36, .047 0.87 (−0.01, 1.75) 1.93 0.35, .053
Life satisfaction 28.83 ± 4.73 (14–35) 29.21 ± 4.44 (13–35) 0.29 (−0.61, 1.19) 0.54 (−0.05, 1.13) 0.24 (−0.74, 1.22) 0.85 (0.21, 1.48) −0.28 (−1.30, 0.73) −0.55 −0.10, .58 −0.65 (−1.69, 0.38) −1.23 −0.22, .22
Perceived stress 7.23 ± 4.57 (0−21) 7.93 ± 5.68 (0−24) 0.42 (−0.73, 1.57) −0.14 (−1.12, 0.85) 0.84 (−0.40, 2.06) 0.61 (−0.49, 1.71) 0.24 (−1.07, 1.55) 0.36 0.08, .72 0.05 (−1.50, 1.61) 0.06 0.01, .95
CESD-20 4.25 ± 3.28 (0−14) 5.98 ± 5.62 (0–27) 0.39 (−0.61, 1.40) −0.07 (−1.20, 1.07) 0.13 (−0.76, 1.01) −0.27 (−1.81, 1.27) −0.04 (−1.50, 1.40) −0.06 −0.02, .95 0.017 (−1.77, 1.80) 0.02 0.00, .99
Open in a new tab

Notes: AC = active control; CESD-20 = Center for Epidemiologic Studies Depression scale; CT = cognitive training; ITT = intent-to-treat; TIADL = Timed Instrumental Activities of Daily Living.

For the mixed linear model, baseline score, age, gender, and education were covaried, and site was entered as a random effect. Change score, change difference score, and effect size are derived from mixed linear model. Higher score is better performance except TIADL, perceived stress and CESD-20.

When the same analysis was performed on the 6-month follow-up data, results showed no significant benefit for CT over the AC at 6 months from the completion of training compared to baseline on the cognitive composite score (Z = 1.68, p = .094, d = 0.36). When conducting the same analysis for each cognitive domain as exploratory analysis, there was no significant advantage for CT over AC at 6 months follow-up in executive function after correcting for multiple comparisons with the Benjamini–Hochberg false discovery rate control procedure (Z = 2.02, p = .044, q = 0.13).

When the same analysis was performed incorporating baseline CESD as one of the covariates, the results remained consistent. For detailed results, please refer to Supplementary Table 5.

Changes in Functional and Participant-Reported Outcome Measures

There were no significant effects observed on any of the secondary outcome measures, including TIADL, Life satisfaction, Perceived stress, and CESD. However, there was a significant advantage in Self-efficacy for the CT group over the AC group, immediately after training (Z = 1.98, p = .047), as well as a marginal benefit at 6 months after the training (Z = 1.93, p = .053).

Differential Changes to Gray Matter Volumes From Cognitive Training

Results showed that the CT group showed significant advantage over the AC group on bilateral caudate volume after 10 weeks of training (Z = 2.52, p = .013); this effect was mainly driven by the left caudate volume (Z = 2.97, p = .003). The difference in caudate volume after 10 weeks of training in the CT group over AC can be attributed to maintained volume in the CT group (pre: 6,786.68 mm³ to post: 6,711.87 mm³), whereas the AC group showed decreases in volume (pre: 6,733.42 mm³ to post: 6,545.96 mm³).

There was no advantage of the CT group over the AC group at posttraining for hippocampal volume (Z = 0.28, p = .78).

Differential Changes to White Matter Integrity From Cognitive Training

The results of the three white matter ROIs, derived from whole-brain TBSS analysis (left IC, left EC, left SLF), are presented in Table 3. A statistically significant group difference at posttraining in FA favoring CT, after covarying for pretraining FA, was observed for the left IC (Z = 2.15, p = .034, Figure 2A), which connects left caudate to left frontal cortex. A marginal difference, again favoring CT, was observed for the left SLF (Z = 1.81, p = .073, Figure 2B). There was no significant group difference at posttraining FA for left EC (p = .77).

Table 3.

Gray Matter Volumes and Fractional Anisotropy From Regions of Interest at Pre- and Posttraining for the CT and AC Groups, Along With the Mixed Linear Model Group Effect

Regions of interest Brain metric—CT groupa Brain metric—AC groupb Between group differences at post (with covariates)
Pre Post Pre Post Z score p Value
Gray matter, volume (mm³) ± SD
Caudate (L + R) 6,786.68 ± 771.6 6,711.87 ± 782.15 6,733.42 ± 651.15 6,545.96 ± 709.72 2.52 .013
 Left caudate 3,389.15 ± 417.59 3,366.14 ± 425.77 3,321.23 ± 340.68 3,223.70 ± 379.87 2.97 .003
 Right caudate 3,397.54 ± 414.36 3,345.73 ± 425.61 3,412.19 ± 380.95 3,322.26 ± 399.08 1.12 .26
Hippocampus (L + R) 7,097.06 ± 1,008.46 7,387.83 ± 669.53 7,365.68 ± 857.98 7,486.38 ± 613.76 0.28 .78
 Left hippocampus 3,464.60 ± 552.27 3,643.13 ± 321.68 3,608.40 ± 401.02 3,659.05 ± 291.19 0.82 .41
 Right hippocampus 3,632.46 ± 529.29 3,744.70 ± 390.67 3,757.28 ± 483.26 3,827.33 ± 362.20 −0.38 .70
White matter TBSS, Mean FA ± SD
Left internal capsule 0.597 ± 0.034 0.596 ± 0.033 0.596 ± 0.034 0.591 ± 0.032 2.15 .034
Left external capsule 0.365 ± 0.028 0.366 ± 0.031 0.366 ± 0.026 0.366 ± 0.031 −0.14 .89
Left SLF 0.445 ± 0.037 0.446 ± 0.04 0.436 ± 0.037 0.432 ± 0.038 1.81 .073
Open in a new tab

Notes: AC = Active control; CT = Cognitive training; FA = fractional anisotropy; SLF = Superior Longitudinal Fasciculus; TBSS = Tract-Based Spatial Statistics. For gray matter volume, baseline volume was covaried and site was entered as a random effect. For white matter FA, baseline FA, age, gender, and education were covaried, and site was entered as a random effect.

a n = 54 for gray matter volume; n = 53 for white matter FA.

b n = 62 for gray matter volume; n = 58 for white matter FA.

Figure 2.

Figure 2. Images illustrate Left Superior Longitudinal Fasciculus and Left Internal Capsule. Scatterplots show the correlation between changes in fractional anisotropy of Left Superior Longitudinal Fasciculus and changes in cognitive performance.

Open in a new tab

The relationship between changes in fractional anisotropy (FA) and changes in cognitive performance. (A) Left internal capsule, MNI (x, y, z) coordinates (−22, −9, 12). (B) Left superior longitudinal fasciculus, MNI (x, y, z) coordinates (−35, −19, 26). Scatter plots for both CT (filled circles, black line) and AC (unfilled circles, dashed line) groups showing relationships between changes in FA (post–pre). (C) Changes in overall cognitive composite score at immediately after the cognitive training (post–pre), and (D) changes in overall cognitive composite score at 6-months follow-up (follow-up–pre). The reported statistics are controlled for effects of age, gender, and education. AC = active control; CT = cognitive training.

The mean FAs from pre- to posttraining assessments suggest that CT maintained FA in left IC after training when compared to AC. This maintenance of white matter FA in left IC goes along with observed maintenance of gray matter volume of left caudate which was limited to CT.

Changes in Brain Structure as a Potential Mechanism for Training-Related Cognitive Gains

Maintenance in caudate volume and white matter integrity of left IC and possibly in left SLF, indexed by higher FA, could be one of the neural mechanisms for training-related cognitive plasticity. We therefore tested correlations between significant or marginally significant brain metrics and cognitive gains, after controlling for effects of age, gender, and education that can influence brain plasticity and/or cognitive plasticity. No significant correlation between changes in the caudate volume and cognitive improvement in CT group was observed either immediately after the training (r(49) = −0.198, p = .164) or at the 6-month follow-up from baseline (r(37) = −0.102, p = .518).

Of the two white matter ROIs, increased FA in left SLF in CT was positively and significantly correlated with cognitive gains at immediate posttraining from baseline (r(48) = 0.29, p = .041; Figure 2C) and at 6-month follow-up from baseline (r(36) = 0.409, p = .011; Figure 2D). The correlations were significant or marginally significant even after bootstrapping the data to 1,000 samples for 95% confidence interval (r(48) = 0.288; p = .08 at post, r(36) = 0.373; p = .021 at 6-month follow-up). These structure–cognition relationships were not significant for the AC group (ps > .1).

Discussion

The aim of this study was to assess the impacts of an adaptive CT in cognitively normal older adults. We hypothesized that 10 weeks of CT use would lead to greater cognitive improvement on untrained cognitive tasks compared to 10 weeks of casual game use. Furthermore, we hypothesized that the cognitive benefits would be associated with positive changes in both gray matter volume in the caudate and hippocampus, and in white matter integrity in associated fiber tracts.

Immediately after the completion of 10 weeks of training, the CT group demonstrated greater improvement on an overall cognitive composite score comprised of performance on untrained tasks compared to the AC group. When examined by each cognitive domain, we found benefits to executive function immediately after the completion of 10 weeks of training.

Although there was a numerical advantage of CT over AC group, the difference between the two groups did not reach a statistically significant level after 6 months from the completion of the training, suggesting that the benefits derived from CT might diminish over time. However, due to the COVID-19 pandemic, approximately 25% of participants were unable to complete the in-lab, computer-based 6-month follow-up assessment, and this reduced sample size likely contributed to the null effects observed at this 6-month follow-up assessment. It is possible that booster training sessions would be necessary to sustain these benefits in older adults over a longer period. To explore this possibility further, a study with longer-term assessments and the inclusion of booster training would be required.

In addition, our findings demonstrated that adaptive CT could promote positive brain plasticity by preserving the gray matter volume in the left caudate, a region in dorsal striatum, and the white matter integrity of left IC. The IC is in the anterior inferomedial part of the brain that has both ascending and descending axons connecting dorsal striatum and thalamus to cortex (especially frontal) and brain stem. The IC plays a major role in psychomotor and reward pathways that are closely linked to cognition (Kolb & Whishaw, 2009). As such, the IC holds relevance in the context of adaptive CT, which specifically involves psychomotor learning of complex stimulus–response relationships across various tasks and incorporates feedback from task scores as rewards.

The preservation of structural brain integrity immediately after training underscores the potential benefits of this type of training in promoting brain health in older adults. This result further reinforces the concept that the cognitive improvements observed in CT are not solely attributable to behavioral changes, such as strategy modifications or increased task familiarity. Instead, these gains are associated with alterations in brain structure that are pertinent to both brain aging and the transfer of cognitive skills through learning and skill acquisition.

There was no correlation between caudate volume change and performance change in CT group. However, it is not uncommon for changes in brain volume after an intervention to not be directly associated with cognitive performance improvements (Niemann et al., 2014).

In contrast to gray matter volume, increased white matter integrity in the left SLF in CT group was positively correlated with an improved overall cognitive composite score not only at immediate posttesting but also at the 6-month follow-up assessment. The left SLF association tract is of particular interest given its relationship with executive function and other fluid cognition, and aging. The SLF connects superior and medial parietal cortex with cingulate and premotor areas (Kamali et al., 2014), and its FA has been associated with both processing speed (Kerchner et al., 2012) and executive function (Bendlin et al., 2010; Hummer et al., 2015), which were two cognitive domains that were not only trained during CT but also contributed to the overall cognitive composite score.

Among the exploratory measures, Self-efficacy was the only measure demonstrating an advantage for CT over AC. Although CT was not explicitly designed to enhance Self-efficacy, there is a possibility that the improvement in cognitive performance might have indirectly contributed to the enhancement of Self-efficacy. However, when we examined the correlation between the improvement in the overall cognitive composite and Self-efficacy, no significant positive correlation was observed. Thus, rather than the outcome of training influencing efficacy, the adaptive nature of training progress and achievement, along with the tailored feedback mechanisms within the adaptive CT program may have played a role in enhancing Self-efficacy. However, additional research is necessary to clarify the connection between CT and the enhancement of Self-efficacy, including whether efficacy benefits from CT are strongest in the domain of cognitive function.

When interpreting the results on cognitive plasticity and its longer-term effects, there are some limitations to be noted. Regarding the breadth of training effects, the benefit was limited to near transfer (overall cognitive composite score) and no significant benefit was found for far transfer (TIADL), suggesting that the CT benefit might be limited to the cognitive abilities targeted by CT and might not be extended to the behavioral functional domain. However, the lack of far transfer could be attributed to our inclusion of healthy, cognitively normal older adults who generally exhibited close-to-ceiling performance on the TIADL. Therefore, more evidence is needed from other functional measures that have greater sensitivity in these cognitively normal populations.

As the primary cognitive outcome encompasses measures from various cognitive domains, the observed benefit in the overall cognitive composite score should not be equated with an improvement in general cognition. When each cognitive domain was separately examined, the observed overall benefit primarily stems from improvements in executive function. However, it remains uncertain whether adaptive CT confers a distinctly stronger advantage on executive function compared to other cognitive constructs or if the measures employed for those cognitive constructs lacked the sensitivity to detect such benefits.

Also, the benefits from training were not observed after 6 months from the training. It is common from multiple studies that the benefits of CT wear off over time (Rebok et al., 2014; Zelinski et al., 2011). However, as the longer-term benefit was only measured 6 months from the completion of training without further follow-up, the time course of transfer of CT to cognitive function is not yet clear. Longer-term assessments should also be examined to determine whether training modifies negative effects of aging on cognitive and functional declines as well as on structural brain atrophy.

Finally, even with our effort to recruit diverse populations from two geographically different sites, our study population was still primarily non-Hispanic, and White race dominant, which limits generalizing study findings to other racial/ethnic groups.

Our principal finding was that CT improved cognitive performance on untrained cognitive tasks immediately after the training and slowed age-related atrophy of gray matter volume in left caudate and white matter integrity in the left IC. Furthermore, better maintenance of left SLF correlated with both immediate- and 6-month follow-up cognitive benefits. Results from this study set the stage for future studies with longer-term follow-up on cognitive and brain biomarkers of aging to investigate the predictors and modifiers of persisted benefit from CT to maximize long-term cognitive health in older adults and prevent future accelerated cognitive decline or impairment.

Supplementary Material

gbae075_suppl_Supplementary_Materials
gbae075_suppl_supplementary_materials.docx (3.6MB, docx)

Acknowledgment

The authors would like to thank Lauren Bullard, Anna Thompson, Margaret A. O’Connell, Bhargavi Bhaskar, Maisha Razzaque, and Taylor Quance for data collection and recruitment.

Contributor Information

Hyun Kyu Lee, Department of Research and Development, Posit Science Corporation, San Francisco, California, USA.

Chandramallika Basak, Department of Psychology, University of Texas at Dallas, Dallas, Texas, USA.

Sarah-Jane Grant, Department of Research and Development, Posit Science Corporation, San Francisco, California, USA.

Nicholas R Ray, Department of Psychology, University of Texas at Dallas, Dallas, Texas, USA.

Paulina A Skolasinska, Department of Psychology, University of Texas at Dallas, Dallas, Texas, USA.

Chris Oehler, Department of Psychological and Brain Sciences, University of Iowa, Iowa City, Iowa, USA.

Shuo Qin, Department of Psychology, University of Texas at Dallas, Dallas, Texas, USA.

Andrew Sun, Department of Psychology, University of Texas at Dallas, Dallas, Texas, USA.

Evan T Smith, Department of Psychology, University of Texas at Dallas, Dallas, Texas, USA.

G Hulon Sherard, Department of Psychology, University of Texas at Dallas, Dallas, Texas, USA.

Adriana Rivera-Dompenciel, Department of Psychological and Brain Sciences, University of Iowa, Iowa City, Iowa, USA.

Mike Merzenich, Department of Research and Development, Posit Science Corporation, San Francisco, California, USA.

Michelle W Voss, Department of Psychological and Brain Sciences, University of Iowa, Iowa City, Iowa, USA.

Vanessa Taler, (Psychological Sciences Section).

Funding

This work was supported by the National Institute on Aging (AG047722).

Conflict of Interest

H. K. Lee and S.-J. Grant are salaried employees of Posit Science.

Data Availability

Study protocol and collected data are available in the study repository (https://github.com/reon7902/supplement). The study was preregistered on ClinicalTrials.gov (NCT03197454).

Author Contributions

The authors confirm their contribution to the paper as follows: study conception and design: H. K. Lee, C. Basak, M. Merzenich, and M. W. Voss; data collection: S.-J. Grant, C. Oehler, N. R. Ray, P. A. Skolasinska, A. R. Dompenciel, S. Qin, E. T. Smith, and A. Sun; analysis and interpretation of results: H. K. Lee, C. Basak, M. W. Voss, M. Merzenich, N. R. Ray, P. A. Skolasinska, S. Qin, and G. H. Sherard; draft manuscript preparation: H. K. Lee, C. Basak, N. R. Ray, C. Oehler, M. W. Voss, M. Merzenich, and P. A. Skolasinska. All authors reviewed the results and approved the final version of the manuscript.

References

  1. Abe, O., Yamasue, H., Aoki, S., Suga, M., Yamada, H., Kasai, K., Masutani, Y., Kato, N., Kato, N., & Ohtomo, K. (2008). Aging in the CNS: Comparison of gray/white matter volume and diffusion tensor data. Neurobiology of Aging, 29(1), 102–116. 10.1016/j.neurobiolaging.2006.09.003 [DOI] [PubMed] [Google Scholar]
  2. Ball, K., Berch, D. B., Helmers, K. F., Jobe, J. B., Leveck, M. D., Marsiske, M., Morris, J. N., Rebok, G. W., Smith, D. M., Tennstedt, S. L., Unverzagt, F. W., & Willis, S. L.; Advanced Cognitive Training for Independent and Vital Elderly Study Group (2002). Effects of cognitive training interventions with older adults: A randomized controlled trial. Journal of the American Medical Association, 288(18), 2271–2281. 10.1001/jama.288.18.2271 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Basak, C., Qin, S., & O’Connell, M. A. (2020). Differential effects of cognitive training modules in healthy aging and mild cognitive impairment: A comprehensive meta-analysis of randomized controlled trials. Psychology and Aging, 35(2), 220–249. 10.1037/pag0000442 [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bauer, E., Toepper, M., Gebhardt, H., Gallhofer, B., & Sammer, G. (2015). The significance of caudate volume for age-related associative memory decline. Brain Research, 1622, 137–148. 10.1016/j.brainres.2015.06.026 [DOI] [PubMed] [Google Scholar]
  5. Behrens, T. E. J., Woolrich, M. W., Jenkinson, M., Johansen-Berg, H., Nunes, R. G., Clare, S., Matthews, P. M., Brady, J. M., & Smith, S. M. (2003). Characterization and propagation of uncertainty in diffusion-weighted MR imaging. Magnetic Resonance in Medicine, 50(5), 1077–1088. 10.1002/mrm.10609 [DOI] [PubMed] [Google Scholar]
  6. Beller, G. A. (2013). The United States has lower life expectancy than other developed nations, despite having highest health care costs. Journal of Nuclear Cardiology, 20(2), 177–178. 10.1007/s12350-013-9692-4 [DOI] [PubMed] [Google Scholar]
  7. Bendlin, B. B., Fitzgerald, M. E., Ries, M. L., Xu, G., Kastman, E. K., Thiel, B. W., Rowley, H. A., Lazar, M., Alexander, A. L., & Johnson, S. C. (2010). White matter in aging and cognition: A cross-sectional study of microstructure in adults aged eighteen to eighty-three. Developmental Neuropsychology, 35(3), 257–277. 10.1080/87565641003696775 [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Boyke, J., Driemeyer, J., Gaser, C., Buchel, C., & May, A. (2008). Training-induced brain structure changes in the elderly. Journal of Neuroscience, 28(28), 7031–7035. 10.1523/JNEUROSCI.0742-08.2008 [DOI] [PMC free article] [PubMed] [Google Scholar]
  9. Chen, Q., Baran, T. M., Turnbull, A., Zhang, Z., Rebok, G. W., & Lin, F. V. (2021). Increased segregation of structural brain networks underpins enhanced broad cognitive abilities of cognitive training. Human Brain Mapping, 42(10), 3202–3215. 10.1002/hbm.25428 [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Dawe, R. J., Yu, L., Arfanakis, K., Schneider, J. A., Bennett, D. A., & Boyle, P. A. (2020). Late-life cognitive decline is associated with hippocampal volume, above and beyond its associations with traditional neuropathologic indices. Alzheimer’s & Dementia, 16(1), 209–218. 10.1002/alz.12009 [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. DeCarli, C., Massaro, J., Harvey, D., Hald, J., Tullberg, M., Au, R., Beiser, A., D’Agostino, R., & Wolf, P. A. (2005). Measures of brain morphology and infarction in the framingham heart study: Establishing what is normal. Neurobiology of Aging, 26(4), 491–510. 10.1016/j.neurobiolaging.2004.05.004 [DOI] [PubMed] [Google Scholar]
  12. den Heijer, T., van der Lijn, F., Koudstaal, P. J., Hofman, A., van der Lugt, A., Krestin, G. P., Niessen, W. J., & Breteler, M. M. B. (2010). A 10-year follow-up of hippocampal volume on magnetic resonance imaging in early dementia and cognitive decline. Brain, 133(4), 1163–1172. 10.1093/brain/awq048 [DOI] [PubMed] [Google Scholar]
  13. Edwards, J. D., Wadley, V. G., Vance, D. E., Wood, K., Roenker, D. L., & Ball, K. K. (2005). The impact of speed of processing training on cognitive and everyday performance. Aging & Mental Health, 9(3), 262–271. 10.1080/13607860412331336788 [DOI] [PubMed] [Google Scholar]
  14. Endo, A., Nagatani, F., Hamada, C., & Yoshimura, I. (2006). Minimization method for balancing continuous prognostic variables between treatment and control groups using Kullback–Leibler divergence. Contemporary Clinical Trials, 27(5), 420–431. 10.1016/j.cct.2006.05.002 [DOI] [PubMed] [Google Scholar]
  15. Engvig, A., Fjell, A. M., Westlye, L. T., Moberget, T., Sundseth, O., Larsen, V. A., & Walhovd, K. B. (2010). Effects of memory training on cortical thickness in the elderly. Neuroimage, 52(4), 1667–1676. 10.1016/j.neuroimage.2010.05.041 [DOI] [PubMed] [Google Scholar]
  16. Esteban, O., Markiewicz, C. J., Blair, R. W., Moodie, C. A., Isik, A. I., Erramuzpe, A., Kent, J. D., Goncalves, M., DuPre, E., Snyder, M., Oya, H., Ghosh, S. S., Wright, J., Durnez, J., Poldrack, R. A., & Gorgolewski, K. J. (2019). fMRIPrep: A robust preprocessing pipeline for functional MRI. Nature Methods, 16(1), 111–116. 10.1038/s41592-018-0235-4 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Fischl, B., Salat, D. H., Busa, E., Albert, M., Dieterich, M., Haselgrove, C., van der Kouwe, A., Killiany, R., Kennedy, D., Klaveness, S., Montillo, A., Makris, N., Rosen, B., & Dale, A. M. (2002). Whole brain segmentation: Automated labeling of neuroanatomical structures in the human brain. Neuron, 33(3), 341–355. 10.1016/s0896-6273(02)00569-x [DOI] [PubMed] [Google Scholar]
  18. Fjell, A. M., McEvoy, L., Holland, D., Dale, A. M., & Walhovd, K. B.; Alzheimer's Disease Neuroimaging Initiative (2014). What is normal in normal aging? Effects of aging, amyloid and Alzheimer’s disease on the cerebral cortex and the hippocampus. Progress in Neurobiology, 117, 20–40. 10.1016/j.pneurobio.2014.02.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  19. Fjell, A. M., & Walhovd, K. B. (2010). Structural brain changes in aging: Courses, causes and cognitive consequences. Reviews in the Neurosciences, 21(3), 187–221. 10.1515/revneuro.2010.21.3.187 [DOI] [PubMed] [Google Scholar]
  20. Gates, N., & Valenzuela, M. (2010). Cognitive exercise and its role in cognitive function in older adults. Current Psychiatry Reports, 12(1), 20–27. 10.1007/s11920-009-0085-y [DOI] [PubMed] [Google Scholar]
  21. He, J., Wong, V. S. S., Fletcher, E., Maillard, P., Lee, D. Y., Iosif, A. -M., Singh, B., Martinez, O., Roach, A. E., Lockhart, S. N., Beckett, L., Mungas, D., Farias, S. T., Carmichael, O., & DeCarli, C. (2012). The contributions of MRI-based easures of gray matter, white matter hyperintensity, and white matter integrity to late-life cognition. American Journal of Neuroradiology, 33(9), 1797–1803. 10.3174/ajnr.a3048 [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hummer, T. A., Wang, Y., Kronenberger, W. G., Dunn, D. W., & Mathews, V. P. (2015). The relationship of brain structure to age and executive functioning in adolescent disruptive behavior disorder. Psychiatry Research, 231(3), 210–217. 10.1016/j.pscychresns.2014.11.009 [DOI] [PubMed] [Google Scholar]
  23. Jenkinson, M., Bannister, P., Brady, M., & Smith, S. (2002). Improved optimization for the robust and accurate linear registration and motion correction of brain images. Neuroimage, 17(2), 825–841. 10.1016/s1053-8119(02)91132-8 [DOI] [PubMed] [Google Scholar]
  24. Jenkinson, M., Beckmann, C. F., Behrens, T. E. J., Woolrich, M. W., & Smith, S. M. (2012). FSL. Neuroimage, 62(2), 782–790. 10.1016/j.neuroimage.2011.09.015 [DOI] [PubMed] [Google Scholar]
  25. Kamali, A., Flanders, A. E., Brody, J., Hunter, J. V., & Hasan, K. M. (2014). Tracing superior longitudinal fasciculus connectivity in the human brain using high resolution diffusion tensor tractography. Brain Structure and Function, 219(1), 269–281. 10.1007/s00429-012-0498-y [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Kerchner, G. A., Racine, C. A., Hale, S., Wilheim, R., Laluz, V., Miller, B. L., & Kramer, J. H. (2012). Cognitive processing speed in older adults: Relationship with white matter integrity. PLoS One, 7(11), e50425. 10.1371/journal.pone.0050425 [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kolb, B., & Whishaw, I. Q. (2009). Fundamentals of human neuropsychology (6th ed). Worth Publishers. [Google Scholar]
  28. Lampit, A., Hallock, H., & Valenzuela, M. (2014). Computerized cognitive training in cognitively healthy older adults: A systematic review and meta-analysis of effect modifiers. PLoS Medicine, 11(11), e1001756. 10.1371/journal.pmed.1001756 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Lee, H. K., Kent, J. D., Wendel, C., Wolinsky, F. D., Foster, E. D., Merzenich, M. M., & Voss, M. W. (2020). Home-based, adaptive cognitive training for cognitively normal older adults: Initial efficacy trial. Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 75(6), 1144–1154. 10.1093/geronb/gbz073 [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Liu, Z, Wang, Y, Gerig, G, Gouttard, S, Tao, R, Fletcher, T, & Styner, M. (2010). Quality control of diffusion weighted images. Proceedings of SPIE–the International Society for Optical Engineering, 7628, 76280. 10.1117/12.844748 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Lövdén, M., Wenger, E., Mårtensson, J., Lindenberger, U., & Bäckman, L. (2013). Structural brain plasticity in adult learning and development. Neuroscience & Biobehavioral Reviews, 37(9), 2296–2310. 10.1016/j.neubiorev.2013.02.014 [DOI] [PubMed] [Google Scholar]
  32. Mahncke, H. W., Bronstone, A., & Merzenich, M. M. (2006). Brain plasticity and functional losses in the aged: Scientific bases for a novel intervention. Progress in Brain Research, 157, 81–109. 10.1016/S0079-6123(06)57006-2 [DOI] [PubMed] [Google Scholar]
  33. Melby-Lervåg, M., Redick, T. S., & Hulme, C. (2016). Working memory training does not improve performance on measures of intelligence or other measures of “far transfer”: Evidence from a meta-analytic review. Perspectives on Psychological Science, 11(4), 512–534. 10.1177/1745691616635612 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Mori, S., Oishi, K., Jiang, H., Jiang, L., Li, X., Akhter, K., Hua, K., Faria, A. V., Mahmood, A., Woods, R., Toga, A. W., Pike, G. B., Neto, P. R., Evans, A., Zhang, J., Huang, H., Miller, M. I., van Zijl, P., & Mazziotta, J. (2008). Stereotaxic white matter atlas based on diffusion tensor imaging in an ICBM template. Neuroimage, 40(2), 570–582. 10.1016/j.neuroimage.2007.12.035 [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Nedelska, Z., Andel, R., Laczó, J., Vlcek, K., Horinek, D., Lisy, J., Sheardova, K., Bureš, J., & Hort, J. (2012). Spatial navigation impairment is proportional to right hippocampal volume. Proceedings of the National Academy of Sciences of the United States of America, 109(7), 2590–2594. 10.1073/pnas.1121588109 [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Nguyen, L., Murphy, K., & Andrews, G. (2019). Cognitive and neural plasticity in old age: A systematic review of evidence from executive functions cognitive training. Ageing Research Reviews, 53, 100912. 10.1016/j.arr.2019.100912 [DOI] [PubMed] [Google Scholar]
  37. Niemann, C., Godde, B., Staudinger, U. M., & Voelcker-Rehage, C. (2014). Exercise-induced changes in basal ganglia volume and cognition in older adults. Neuroscience, 281, 147–163. 10.1016/j.neuroscience.2014.09.033 [DOI] [PubMed] [Google Scholar]
  38. O’Sullivan, M., Jones, D. K., Summers, P. E., Morris, R. G., Williams, S. C. R., & Markus, H. S. (2001). Evidence for cortical “disconnection” as a mechanism of age-related cognitive decline. Neurology, 57(4), 632–638. 10.1212/wnl.57.4.632 [DOI] [PubMed] [Google Scholar]
  39. Paul, R., Lane, E. M., Tate, D. F., Heaps, J., Romo, D. M., Akbudak, E., Niehoff, J., & Conturo, T. E. (2011). Neuroimaging signatures and cognitive correlates of the Montreal cognitive assessment screen in a nonclinical elderly sample. Archives of Clinical Neuropsychology, 26(5), 454–460. 10.1093/arclin/acr017 [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Petersen, R. C., Jack, C. R., Xu, Y. -C., Waring, S. C., O’Brien, P. C., Smith, G. E., Ivnik, R. J., Tangalos, E. G., Boeve, B. F., & Kokmen, E. (2000). Memory and MRI-based hippocampal volumes in aging and AD. Neurology, 54(3), 581–587. 10.1212/wnl.54.3.581 [DOI] [PubMed] [Google Scholar]
  41. Pintzka, C. W. S., Hansen, T. I., Evensmoen, H. R., & Håberg, A. K. (2015). Marked effects of intracranial volume correction methods on sex differences in neuroanatomical structures: A HUNT MRI study. Frontiers in Neuroscience, 9, 9. 10.3389/fnins.2015.00238 [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Raz, N., Lindenberger, U., Rodrigue, K. M., Kennedy, K. M., Head, D., Williamson, A., Dahle, C., Gerstorf, D., & Acker, J. D. (2005). Regional brain changes in aging healthy adults: General trends, individual differences and modifiers. Cerebral Cortex, 15(11), 1676–1689. 10.1093/cercor/bhi044 [DOI] [PubMed] [Google Scholar]
  43. Rebok, G. W., Ball, K., Guey, L. T., Jones, R. N., Kim, H. Y., King, J. W., Marsiske, M., Morris, J. N., Tennstedt, S. L., Unverzagt, F. W., & Willis, S. L. (2014). Ten-year effects of the advanced cognitive training for independent and vital elderly cognitive training trial on cognition and everyday functioning in older adults. Journal of the American Geriatrics Society, 62(1), 16–24. 10.1111/jgs.12607 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Reijnders, J., van Heugten, C., & van Boxtel, M. (2013). Cognitive interventions in healthy older adults and people with mild cognitive impairment: A systematic review. Ageing Research Reviews, 12(1), 263–275. 10.1016/j.arr.2012.07.003 [DOI] [PubMed] [Google Scholar]
  45. Rizio, A. A., & Diaz, M. T. (2016). Language, aging, and cognition: Frontal aslant tract and superior longitudinal fasciculus contribute toward working memory performance in older adults. Neuroreport, 27(9), 689–693. 10.1097/WNR.0000000000000597 [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. Simons, D. J., Boot, W. R., Charness, N., Gathercole, S. E., Chabris, C. F., Hambrick, D. Z., & Stine-Morrow, E. A. (2016). Do “brain-training” programs work. Psychological Science in the Public Interest, 17(3), 103–186. 10.1177/1529100616661983 [DOI] [PubMed] [Google Scholar]
  47. Smith, S. M. (2002). Fast robust automated brain extraction. Human Brain Mapping, 17(3), 143–155. 10.1002/hbm.10062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Smith, S. M., Jenkinson, M., Woolrich, M. W., Beckmann, C. F., Behrens, T. E. J., Johansen-Berg, H., Bannister, P. R., De Luca, M., Drobnjak, I., Flitney, D. E., Niazy, R. K., Saunders, J., Vickers, J., Zhang, Y., De Stefano, N., Brady, J. M., & Matthews, P. M. (2004). Advances in functional and structural MR image analysis and implementation as FSL. Neuroimage, 23, S208–S219. 10.1016/j.neuroimage.2004.07.051 [DOI] [PubMed] [Google Scholar]
  49. Strenziok, M., Parasuraman, R., Clarke, E., Cisler, D. S., Thompson, J. C., & Greenwood, P. M. (2014). Neurocognitive enhancement in older adults: Comparison of three cognitive training tasks to test a hypothesis of training transfer in brain connectivity. Neuroimage, 85 Pt 3, 1027–1039. 10.1016/j.neuroimage.2013.07.069 [DOI] [PubMed] [Google Scholar]
  50. Zelinski, E. M., Spina, L. M., Yaffe, K., Ruff, R., Kennison, R. F., Mahncke, H. W., & Smith, G. E. (2011). Improvement in memory with plasticity-based adaptive cognitive training: Results of the 3-month follow-up. Journal of the American Geriatrics Society, 59(2), 258–265. 10.1111/j.1532-5415.2010.03277.x [DOI] [PubMed] [Google Scholar]

Associated Data

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

Supplementary Materials

gbae075_suppl_Supplementary_Materials
gbae075_suppl_supplementary_materials.docx (3.6MB, docx)

Data Availability Statement

Study protocol and collected data are available in the study repository (https://github.com/reon7902/supplement). The study was preregistered on ClinicalTrials.gov (NCT03197454).

Articles from The Journals of Gerontology Series B: Psychological Sciences and Social Sciences are provided here courtesy of Oxford University Press

RESOURCES