Integrated cognitive and physical fitness training enhances attention abilities in older adults

The study addresses whether an integrated, closed-loop cognitive-physical training program can enhance attention abilities and physical fitness in older adults more effectively than an active, expectancy-matched placebo control. Prior work shows mixed evidence for cognitive training benefits and strong evidence for physical fitness benefits in older adults, with growing interest in combining approaches to achieve broader gains efficiently. However, many combined-intervention studies lack rigorous controls (e.g., expectancy matching), mechanistic measures, and long-term follow-up. Building on prior closed-loop, adaptive video game interventions that improved cognitive control and related neural markers, the authors developed an exergame (Body-Brain Trainer, BBT) that simultaneously adapts cognitive difficulty and physical intensity via motion capture and real-time heart rate to personalize training. The purpose is to test BBT’s efficacy on sustained attention (behavioral and EEG markers), assess physical fitness outcomes, compare trained older adults to untrained young adults, and evaluate 1-year maintenance of effects.

- Cognitive training in older adults has shown benefits in some studies and reviews, including improvements in non-trained abilities, but findings are inconsistent and often limited by inadequate controls and lack of follow-up. - Physical exercise reliably benefits general health in older adults and has shown potential benefits for executive functions/attention, suggesting that combined cognitive-physical interventions might yield enhanced cognitive outcomes. - Combined interventions can be time- and resource-efficient and may be more enjoyable than single-domain or sequential training. Yet, prior combined studies rarely used expectancy-matched controls and often required longer training periods (12–30 weeks). - Closed-loop adaptive interventions (including video games and meditation-based approaches) have improved sustained attention and related neural markers (e.g., midline frontal theta) across age groups. Prior work by the authors has shown closed-loop games can normalize older adults’ cognitive control performance to young adult levels and produce neural changes that sometimes persist post-intervention. - Few combined-intervention studies incorporate neuroimaging/EEG to probe mechanisms; existing work suggests changes in metabolism and resting-state connectivity accompany cognitive gains. - Gaps identified: need for rigorous expectancy-matched controls, integrated adaptive cognitive and physical challenges within a single platform, objective neural markers, and long-term follow-up to assess durability.

Design: Double-blinded, randomized, placebo-controlled trial with expectancy-matched active control. Participants and assessors were blinded; analyses were conducted blinded to group. Participants: 49 healthy older adults (mean age 68.5 ± 6.36; 26 females) randomized to BBT (n=24) or Mind-Body Trainer (MBT; active, expectancy-matched placebo; n=25). A subset returned at 1-year follow-up (BBT n=21; MBT n=20). Inclusion criteria included normal/corrected vision, no stroke/significant brain injury/psychiatric history, no psychotropic/hormonal/cardiovascular medications, low videogame use (<2 h/month), and medical clearance for physical activity. Baseline cognitive screening included MoCA, GDS, PHQ-9, and a neuropsychological battery to ensure performance within normal limits for age. Interventions: - BBT (experimental): An exergame using Microsoft Kinect motion capture and heart rate monitoring to simultaneously adapt cognitive difficulty and physical intensity in real time. Modules targeted attention (visual search with distractors), working memory (spatial WM and multiple object tracking), and task switching (cognitive flexibility via morphing stimuli). Adaptive algorithms maintained ~80% accuracy; physical difficulty adjusted to keep heart rate within a personalized target zone. Schedule: in-lab, 3 sessions/week, ~1 hour per visit (36 min on-task) for 8 weeks. - MBT (control): Expectancy-matched home-based training using three iOS apps (Duolingo, Tai Chi Step by Step, 100 Logic Games), ~30 minutes/day, 5 days/week for 6 weeks (total ~15 hours). Participants were seated and trained independently with guidance materials and remote support. Apps were selected via an expectancy-matching process showing comparable expected improvements to BBT on key cognitive outcomes. Outcome measures: - Primary cognitive outcome: Customized Continuous Performance Task (CPT) measuring sustained attention; behavioral metric: ex-Gaussian tau (response time tail). - Neural measure: Midline frontal theta power (4–7 Hz) via 64-channel EEG during CPT; analyses focused on frontal electrodes (e.g., Fz, F3, F4, AFz). - Secondary cognitive outcome: Working memory delayed recognition task (No Distractor, Ignore Distractor, Attend Distractor conditions). - Exploratory attention measure: Complex visual discrimination task with variable distraction (Filter task), focusing on response time variability (RTV). - Motor control check: Basic Response Time (BRT) task to rule out basic motor speed confounds. - Physical outcomes: Diastolic blood pressure (mmHg) and Limits of Stability (LoS; balance, movement velocity) on a balance manager system. Young adult comparison: An untrained cohort of young adults (n=51; ~20-year-olds) completed the CPT and EEG in a single visit for age-comparison benchmarks. Follow-up: Both groups invited for 1-year post-training assessment to evaluate maintenance of effects (primarily cognitive/EEG outcomes). Blinding and expectancy: Multi-layered blinding (data collection, analysis). Expectancy matching established in separate samples and verified within the study after initial training exposure; no significant expectancy differences between BBT and MBT on targeted outcomes. EEG acquisition and preprocessing: 64-channel BioSemi ActiveTwo, digitized at 1024 Hz, 0.01–100 Hz bandpass during acquisition; preprocessing in EEGLAB with artifact handling (noisy channel interpolation, downsampling, high-pass filtering at 1 Hz, ICA for ocular artifacts, average reference). Epochs −1000 to +1000 ms; power computed via FFT, baseline-corrected (dB) with focus on midline frontal theta. Statistical analyses: Repeated measures ANOVAs tested Session (Pre vs Post) by Group (BBT vs MBT) interactions; post-hoc paired and independent t-tests where appropriate. ANCOVA used where baseline trends existed. Effect sizes reported as Cohen’s d; significance threshold p<0.05. Maintenance assessed by comparing post-training to 1-year follow-up within groups.

- Sustained attention (CPT, tau): Significant Session×Group interaction, F(1,47)=7.49, p=0.009, Cohen’s d=0.79. BBT improved (Δ=−15.73 ms; t(23)=2.29, p=0.03); MBT showed a non-significant decline (Δ=+6.46 ms; t(23)=−1.60, p=0.113). At baseline, BBT older adults performed equivalently to young adults (t(23)=0.42, p=0.67); after training, BBT performed significantly better than young adults (t(24)=2.40, p=0.03). At 1-year, no significant change from post-training in either group. - Neural marker (midline frontal theta power during CPT): Significant Group×Time interaction, F(1,43)=6.61, p=0.014, d=0.74. BBT increased theta power post-training (t(23)=2.77, p=0.011); control did not (t(20)=0.79, p=0.447). Relative to young adults, BBT showed age-related reductions at pre (t(25)=3.10, p=0.003) that rose to equivalence post (t(20)=0.11, p=0.91). Similar patterns observed for a complex visual discrimination task (exploratory). - Physical fitness: BBT showed improvements beyond control in diastolic blood pressure and balance (Limits of Stability). Table values indicate diastolic BP decreased from 78.7 to 73.0 mmHg (BBT, with post significantly different from pre) and LoS movement velocity increased from 3.88 to 4.59 m/s (BBT, significant), whereas MBT showed no meaningful change. - Working memory: Exploratory analyses indicated that improvements did not consistently emerge across working memory outcome measures, unlike some prior work. - Basic motor speed: No differential changes in mean RT or RTV on the Basic Response Time task between groups (mean RT interaction F(1,40)=2.16, p=0.15; post-training between-group t(42)=0.44, p=0.67; RTV interaction F(1,40)=0.51, p=0.85), suggesting attention gains were not due to simple motoric speeding. - Maintenance: At 1-year, behavioral attention performance and neural measures did not differ significantly from immediate post-training levels in either group, indicating potential maintenance, though interpretations are cautious due to sample size and non-significant tests.

The integrated, closed-loop BBT intervention produced significant gains in sustained attention and in a neural signature of attention (midline frontal theta), alongside improvements in physical health indicators (blood pressure, balance), outperforming an expectancy-matched active control. These findings support the hypothesis that simultaneously adapting cognitive and physical demands within a cohesive training experience can enhance attentional control in older adults beyond placebo-like engagement. Notably, BBT elevated midline frontal theta power from an age-reduced baseline to young adult-equivalent levels and improved CPT performance to better-than-young levels post-training, suggesting compensatory or enhancement effects on neural systems underlying attention. The absence of differential changes in basic motor speed supports a cognitive locus for the observed attention benefits. Compared to many prior combined interventions requiring 12–30 weeks, BBT achieved benefits after 8 weeks, highlighting efficiency and feasibility. Preliminary 1-year assessments suggest potential maintenance of attention gains, although the stability of these effects warrants cautious interpretation and may benefit from booster sessions. Overall, the results reinforce the promise of integrated, adaptive cognitive-physical training to bolster both neural and behavioral aspects of attention in aging.

An 8-week, integrated, closed-loop cognitive-physical exergame (BBT) improved older adults’ sustained attention performance and midline frontal theta power beyond an expectancy-matched active control, while also enhancing diastolic blood pressure and balance. Post-training attention performance surpassed that of untrained young adults, and neural activity reached age-equivalent levels. Preliminary 1-year data suggest maintained attention performance. These findings support integrated adaptive training as a viable, efficient approach to enhance cognitive and physical function in older adults. Future work should directly compare combined versus single-domain components, include mechanistic probes to parse contributions of cognitive vs physical elements, assess broader cognitive transfer (e.g., working memory), optimize dosing and booster schedules, and examine long-term durability with larger samples.

- Component attribution: The design does not isolate contributions of cognitive vs physical training; thus, synergistic or additive effects cannot be determined. - Synergy evidence: Prior meta-analyses suggest combined training does not always exceed cognitive-only effects; synergy was not directly tested here. - Working memory: Improvements were not consistently observed across all working memory measures. - Contextual differences: Training contexts differed (in-lab supervised BBT vs at-home MBT), potentially introducing non-specific factors (e.g., supervision, environment). - Expectancy control: Although expectancy was matched and verified, unmeasured experiential differences could still influence outcomes. - Follow-up: 1-year interpretations rely on non-significant differences with relatively small returning samples; measure sensitivity over long intervals is uncertain. - Generalizability: Healthy, motivated older adults with specific inclusion criteria may limit generalization to broader populations.