Abstract
There is an increasing recognition that impairments in everyday functional skills as well as cognitive changes are seen in older people with mild cognitive impairment (MCI). While pharmacological interventions have not been remarkably successful, computerized cognitive training (CCT) has been reported to have efficacy in several meta-analyses for MCI and healthy aging. However, CCT alone does lead to the spontaneous learning of novel functional skills. Many older people, both with MCI and normal cognition (NC), are challenged in the performance of everyday functional skills that require utilization of technology. Hence, the computerized Functional Skills Assessment and Training system (FUNSAT) has been developed to train an array of health-relevant technology focused skills for participants with NC and with MCI. The current FUNSAT software supports fully remote training, can also accommodate CCT on the same platform, and allows for remote monitoring of training engagement and success. In this study, all participants trained fully remotely.
In the current study, 50 participants with MCI (Jak-Bondi Criteria) were randomized to receive remote FUNSAT training alone or in combination with a lead-in booster session of 6-8 hours of CCT. An in-person version of this same training system was reported on previously (Czaja et al., 2020; Harvey et al., 2021) and demonstrated efficacy for acquisition of all 6 skills at endpoint in participants with MCI. Training took place for up to 24 1-hour sessions over 12 weeks, with six different functional skills trained (Ticket Purchase, ATM Banking, Medication Management, Telephone Voice Menu, Internet Banking, and Internet shopping and delivery. Participants graduated from training on a task if they performed it with one or fewer errors on two successive training sessions. 50 healthy controls received home-based FUNSAT training only in the same time period. At the beginning and end of the study and then 30 days after training, the participants were assessed with different forms of fixed difficulty versions of each of the six tasks. Performance was indexed by time to completion and error rates. Other outcomes measures were collected and will be reported later, including performance on a computerized neuropsychlogical assessment (The Brief Assessment of Cognition; BAC), and ecological momentary assessment (EMA) with device-based paging. Participants were paged on a daily basis and asked if they had performed any of the trained tasks (banking, refilling prescriptions), as well as answering other questions about other untrained uses of technology, such as sending text messages and taking digital pictures. These results will be reported separately
All 6 tasks were performed more poorly by the MCI participants at baseline, with effect sizes over d=1.2. For all six tasks, the training effect size from first assessment to final training session for both completion time and errors ranged from 0.96, to 2.00 for the healthy participants, with the 30-day post-training assessment indicating considerable maintenance of these gains, with effect sizes ranging from 0.6 to 1.44 compared to baseline performance on these same 6 skills. For the participants with MCI, effects sizes for gains during training ranged from d=0.96 to 2.08 and maintenance to the 30 day post training assessment ranged from 0.4 to 1.8 SD improvements compared to baseline. There were no significant differences in training gains associated with combined CCT plus FUNSAT training vs FUNSAT only training, despite 20% fewer FUNSAT raining sessions in the combined treatment group.
This study replicated previous findings of substantial gains in FUNSAT performance with training in MCI, while demonstrating persistence post training and proving that remote training is feasible and efficacious. Burst CCT training manifested the same benefits as concurrent training in the previous studies, with reduced drop-out compared to our previous trial. These data indicate that participants with significant cognitive challenges at baseline can still engage in and benefit from both functionally targeted and cognitively targeted training. The effect sizes for training gains were very similar to our previous study. In those studies, combined training with CCT and FUNSAT also demonstrated a synergistic benefit on performance on the cognitive assessment battery. The current study, demonstrating excellent efficacy and persistence of gains, suggests that remote training may also be able to bypass the considerable logistical challenges of in person training interventions.
This study was funded by National Institute of Aging Grant Number 2R44 AG057238 to Mr. Kallestrup.