Global functional connectivity of cognitive control networks predicts task-switching performance in older adults.

Older adults performed more slowly and less accurately in the Overlap version of the task-switching paradigm compared to the Separate version.

• Participants aged 55-80 years old performed both Separate and Overlap versions of the task-switching paradigm
• Behavioral analyses included n = 118 participants
• The Overlap version places greater working memory demands on participants by requiring maintenance and coordination of multiple task rules simultaneously
• The Overlap version showed sensitivity to age-related declines in executive function

Greater CCN global functional connectivity was associated with smaller mixing costs in the Overlap version of the task-switching paradigm.

• Brain-behavior analyses were conducted in n = 112 participants
• Six 8-minute resting-state fMRI sessions were collected over two days
• Whole grey-matter GFC was calculated, followed by average GFC extraction from the CCN, DMN, and SMN
• Mixing cost reflects the ability to maintain and coordinate multiple task rules in working memory and is sensitive to age-related declines in EF
• The association was specific to the Overlap version, which has higher working memory demands

Greater Default Mode Network global functional connectivity was associated with smaller mixing costs in the Overlap version.

• DMN GFC showed a significant association with mixing costs in the Overlap version
• The DMN association was not observed for the Separate version mixing costs
• This finding suggests the DMN plays a role in task-switching performance under higher working memory demands
• GFC was extracted as an average from the DMN region of interest following whole grey-matter GFC calculation

Greater Somatomotor Network global functional connectivity was associated with smaller mixing costs in the Overlap version but larger mixing costs in the Separate version.

• SMN GFC showed opposing associations with mixing costs depending on task version
• SMN GFC was linked to larger mixing costs in the Separate version
• SMN GFC was also linked to smaller switching costs in the Separate version
• Switching cost is more closely linked to age-related general slowing in processing speed rather than executive function declines
• This pattern suggests the SMN's role in task-switching performance differs depending on working memory load of the task

Mixing cost is more sensitive to age-related declines in executive function, while switching cost is more closely linked to age-related general slowing in processing speed.

• Mixing cost reflects the ability to maintain and coordinate multiple task rules in working memory
• Switching cost reflects the specific cost of switching between tasks on switch trials versus repeat trials
• These two components of task-switching performance are differentially sensitive to different aspects of aging
• This distinction motivated the use of two task versions (Separate and Overlap) with differing working memory demands

The study used a high-reliability resting-state fMRI protocol consisting of six 8-minute sessions collected over two days.

• Brain-behavior analyses were conducted in n = 112 participants aged 55-80 years old
• Six sessions of resting-state fMRI were collected, totaling 48 minutes of resting-state data
• Sessions were distributed across two days
• Whole grey-matter GFC was calculated from these sessions followed by network-specific GFC extraction for the CCN, DMN, and SMN