Lost Dynamics and the Dynamics of Loss: Longitudinal Compression of Brain Signal Variability is Coupled with Declines in Functional Integration and Cognitive Performance

Abstract

Reduced moment-to-moment blood oxygen level-dependent (BOLD) signal variability has been consistently linked to advanced age and poorer cognitive performance, showing potential as a functional marker of brain aging. To date, however, this promise has rested exclusively on cross-sectional comparisons. In a sample of 74 healthy adults, we provide the first longitudinal evidence linking individual differences in BOLD variability, age, and performance across multiple cognitive domains over an average period of 2.5 years. As expected, those expressing greater loss of BOLD variability also exhibited greater decline in cognition. The fronto-striato-thalamic system emerged as a core neural substrate for these change–change associations. Preservation of signal variability within regions of the fronto-striato-thalamic system also cohered with preservation of functional integration across regions of this system, suggesting that longitudinal maintenance of “local” dynamics may require across-region communication. We therefore propose this neural system as a primary target in future longitudinal studies on the neural substrates of cognitive aging. Given that longitudinal change–change associations between brain and cognition are notoriously difficult to detect, the presence of such an association within a relatively short follow-up period bolsters the promise of brain signal variability as a viable, experimentally sensitive probe for studying individual differences in human cognitive aging.

Dynamic regulation of neural variability during working memory reflects dopamine, functional integration, and decision-making

“…We also found that working memory load-based upregulation of "local" signal variability was tightly coupled with increased functional integration within every network examined. This provides first human evidence that local dynamics and functional integration are closely aligned outside of the resting-state 20,21 . Crucially, striato-thalamic functional integration uniquely accounted for a sizable proportion of ΔSDBOLD in every network, but functional integration in non-striato-thalamic networks was not independently associated with modulation of striato-thalamic SDBOLD.…”

“…In the present study, we test for unique associations between brain signal variability regulation and our three research targets (D2 capacity, functional integration, and decisionmaking; Figure 1), specifically within the domain of working memory. Neural variability has proven highly responsive to working memory demands 5,7,10,32,33 , dopamine levels 5,10,12 , and the striato-thalamic system in general 5,20,21,32 . However, it is not yet known whether one's capacity to regulate neural variability in the face of varying working memory load is associated with any of our research targets.…”

“…Our recent resting-state fMRI work in humans has shown that higher temporal variability within a region reflects greater (i.e., lower dimensional) functional integration between regions across moments 20 . Critically, higher temporal variability in thalamus and striatum were among the strongest markers of functional integration across the entire brain, both cross-sectionally and longitudinally across the adult lifespan 20,21 . In simple terms, as striato-thalamic activity fluctuates during resting-state, the brain functionally integrates across moments.…”

“…Among other functions, the indirect D 2 striato-thalamic pathway allows for flexible shifting between ongoing response modes and the processing of new or salient cognitive demands 17,18 . Given that the striato-thalamic system also appears to be central for understanding how local (region level) variability reflects brain function overall 20,21 , the D 2 system may provide an essential window into the origin and functional importance of temporal variability in the striato-thalamic system, especially as cognitive demand varies. We presume that the ability to upregulate striato-thalamic variability under load requires D 2 capacity to meet demand.…”

Signal Variability and Cognitive Function in Older Long-Term Survivors of Breast Cancer with Exposure to Chemotherapy: A Prospective Longitudinal Resting-State fMRI Study

“…Although FC and SD measures are different yet complimentary features for brain functional characterization, SD BOLD has been shown to be a more sensitive parameter for detecting subtle brain changes in aging studies, and it is therefore more appropriate for characterization of the CRCI in long-term older cancer survivors. Garrett et al showed that SD BOLD was coupled with dynamic functional integration and cognitive performance in aging [ 18 ]. They found that the SD BOLD trajectory followed an inverted U-shape across the lifespan, i.e., from low variability in infancy, to high variability in young adulthood to lower variability in older adulthood [ 19 ].…”