Studies have shown that Tai Chi Chuan (TCC) training has benefits on task-switching ability. However, the neural correlates underlying the effects of TCC training on task-switching ability remain unclear. Using task-related functional magnetic resonance imaging (fMRI) with a numerical Stroop paradigm, we investigated changes of prefrontal brain activation and behavioral performance during task-switching before and after TCC training and examined the relationships between changes in brain activation and task-switching behavioral performance. Cognitively normal older adults were randomly assigned to either the TCC or control (CON) group. Over a 12-week period, the TCC group received three 60-min sessions of Yang-style TCC training weekly, whereas the CON group only received one telephone consultation biweekly and did not alter their life style. All participants underwent assessments of physical functions and neuropsychological functions of task-switching, and fMRI scans, before and after the intervention. Twenty-six (TCC, N = 16; CON, N = 10) participants completed the entire experimental procedure. We found significant group by time interaction effects on behavioral and brain activation measures. Specifically, the TCC group showed improved physical function, decreased errors on task-switching performance, and increased left superior frontal activation for Switch > Non-switch contrast from pre- to post-intervention, that were not seen in the CON group. Intriguingly, TCC participants with greater prefrontal activation increases in the switch condition from pre- to post-intervention presented greater reductions in task-switching errors. These findings suggest that TCC training could potentially provide benefits to some, although not all, older adults to enhance the function of their prefrontal activations during task-switching.

Tai Chi Chuan (TCC) exercise has been shown to improve cognitive task-switching performance in older adults, but the extent of this positive effect varies among individuals. Past research also shows that brain white matter integrity could predict behavioral gains of cognitive and motor learning. Therefore, in this randomized controlled trial (NCT02270320), we examined whether baseline integrity of three target white matter tract groups was predictive of task-switching improvement after 12-week TCC training in middle-aged and older adults. Thirty-eight eligible participants were randomly assigned to a TCC group (n = 19) and a control group (n = 19). Cognitive task-switching and physical performances were collected before and after training. Brain diffusion spectrum MR images were acquired before training and the general fractional anisotropy (GFA) of each target white matter tract group was calculated to indicate baseline white matter integrity of that group. Correlation and regression analyses between these GFAs and post-training task-switching improvement were analyzed using adjusted p-values. After 12 weeks, significant task-switching and physical performance improvements were found only in the TCC group. Moreover, higher baseline GFA of the prefronto-striato-thalamo-prefrontal loop fibers (r = −0.63, p = 0.009), but not of the prefronto-parietal/occipital (r = −0.55, p = 0.026) and callosal (r = −0.35, p = 0.189) fiber groups, was associated with greater reductions of task-switching errors after the TCC training. Multiple regression analysis revealed that baseline GFA of the prefronto-striato-thalamo-prefrontal loop fibers was the only independent white matter integrity predictor of task-switching error reductions after TCC training (β = −0.620, adjusted R2 change = 0.265, p = 0.009). These findings not only highlight the important role of baseline integrity of the prefronto-striatal circuits in influencing the extent of positive cognitive task-switching effects from short-term TCC training, but also implicate that preserving good white matter integrity in the aging process may be crucial in order to gain the best cognitive effects of exercise interventions.