Differential Associations of Long-Term Physical Activity Patterns with Executive Function and Brain Functional Connectivity

Objectives: Long-term physical activity patterns may be associated with cognition through distinct mechanisms, yet direct comparisons between accumulated moderate-to-vigorous exercise and daily ambulatory activity remain scarce. This study aimed to examine whether regular exercise time and daily step count show distinct patterns of association with executive function and cortical functional connectivity in healthy young adults, and to characterize the extent to which the two activity indices overlap. Methods: In this cross-sectional study, 48 healthy young adults were stratified by their physical activity over the preceding six months into high versus low regular exercise time groups (cut-off: ≥150 min/week) and high- versus low-daily-step-count groups (cut-off: ≥10,000 steps/day). All participants completed the Stroop task to assess response inhibition and information processing speed, while 63-channel functional near-infrared spectroscopy (fNIRS) was used to monitor cortical functional connectivity. Results: A high regular exercise time (≥150 min/week) was associated with a higher incongruent-to-congruent accuracy ratio in the word test of the Stroop task (t = 2.318, p = 0.0250, η2 = 0.1046), which is consistent with reduced Stroop interference; this association was not accompanied by differences in cortical functional connectivity. Conversely, a high daily step count (≥10,000 steps/day) was associated with shorter reaction time in the Stroop word test (t = 2.048, p = 0.0463, η2 = 0.0836) and with higher connectivity across the sampled cortical regions (HbO: t = 2.104, p = 0.0409, η2 = 0.0878; HbR: t = 2.431, p = 0.0190, η2 = 0.1139), including stronger left precentral–right occipital connectivity at the HbO level (t = 4.837, FDR-corrected q = 0.0295) and right parietal–left inferior frontal connectivity at the HbR level (t = 4.983, FDR-corrected q = 0.0181). Regular exercise time correlated specifically with the word-test accuracy ratio (Spearman ρ = 0.337, p = 0.019), whereas no significant correlations were found between daily step count-related behavioral and brain activity measures. Conclusions: In this cross-sectional sample, physical activity was associated with cognitive performance in a pattern-specific manner: regular exercise was associated primarily with response inhibition, whereas daily walking was associated primarily with processing speed, with the latter accompanied by more extensive connectivity across the sampled cortical regions. These associations are compatible with, but do not demonstrate, dissociable neural mechanisms and should be tested in longitudinal and intervention designs.