EFFECTS OF INTEGRATED AEROBIC EXERCISE WITH COGNITIVE TRAINING (ACTIV4BRAIN) PROGRAM ON REINFORCEMENT LEARNING IN OLDER ADULTS

Cognitive training has been shown to improve executive functions important for complex task performance, including value-based decision making (reinforcement learning) which involves working memory, processing speed, and sustained attention. Moreover, integrating cognitive training into aerobic exercise sessions might enhance its effects because of aerobic exercise on the brain neurogenesis and neuroplasticity. In this study, we investigated the effect of regular aerobic exercise with cognitive training on cognitive performance in older persons.
Specifically, we were interested in the modulation of behavioral and computational measures of reinforcement learning (RL). Participants (54–75 years) were assigned into three groups: integrated aerobic exercise with cognitive training (Activ4Brain), aerobic exercise only (AE), or passive control (CNT). Participants completed a probabilistic RL task, before and after a 3-month, twice-a-week intervention. The behavioral effects of the intervention were analyzed using both model-agnostic analyses and computational modelling of behavior, in order to probe the underlying cognitive mechanisms. We used a Rescorla–Wagner model with separated learning rates for gain and loss (α_gain, α_loss), and a softmax inverse temperature parameter (β). Response speed improved more in the Activ4Brain group than in the remaining two groups, but only in participants with lower baseline performance. There was no effect of treatment on accuracy. Model-dependent analyses yielded consistent results, as the computational model parameters did not differ between interventions.
These results suggest that integrated cognitive and aerobic exercise training may be beneficial for older adults, enhancing their ability to respond more efficiently during reinforcement learning tasks.