Brain glucose dynamics map a layer of complexity that weakens with age
A new preprint from Monash University argues that the brain’s metabolic activity is not just a background fuel signal but a dynamic system with measurable complexity. Using simultaneous functional PET with 18F-fluoroedeoxyglucose and functional MRI in 84 adults, investigators report that metabolic entropy captures a form of non-linear, high-frequency complexity in cerebral glucose signaling. The same signal appears to track how flexibly the brain moves between network states, especially across unimodal and higher-order associative systems.
The message is simple and clinically relevant: cerebral metabolism is not static, and its temporal richness declines in normative ageing.
PET-derived entropy behaved differently from haemodynamic entropy
The study set out to answer a basic question in systems neuroscience: whether entropy measures applied to brain signals carry the same biological meaning in metabolism and hemodynamics. They do not. PET multiscale entropy, derived from dynamic glucose uptake, reflected non-linear and high-frequency metabolic complexity. In contrast, haemodynamic multiscale entropy appeared to represent non-linearly constrained neurovascular phase dynamics rather than the same kind of metabolic variability.
That separation matters because it suggests that PET and fMRI are not interchangeable views of the same phenomenon. The metabolic signal seems to carry a deeper readout of energetic coordination, while haemodynamic entropy may be shaped more tightly by vascular and coupling constraints. In other words, complexity in glucose processing and complexity in blood-oxygen dynamics overlap only partially, and the difference is biologically meaningful.
Metabolic complexity tracked metastable network behavior
Greater metabolic complexity was associated with greater network metastability, particularly in unimodal and higher-order associative networks. Metastability here refers to the brain’s ability to remain flexible without collapsing into rigid synchrony, a property that supports efficient switching between functional configurations. A metabolically complex system can sustain this kind of controlled instability because energy delivery and utilization can adapt to shifting network demands.
One sentence captures the central result: higher metabolic entropy aligned with a more flexible brain.
Regional metabolic complexity also correlated with greater cortical thickness, higher cerebral metabolic rates of glucose, and greater cerebral blood flow. That triad makes biological sense. Thicker cortex, stronger glucose utilization, and richer perfusion all point to tissue with preserved structural and energetic capacity. In practical terms, the signal is not merely an abstract statistical feature; it appears to sit on top of a recognizable substrate of brain health.
Ageing reduced the brain’s metabolic richness
Metabolic complexity was lower in older adults than in younger adults. The age effect extended beyond a minor shift in signal amplitude and suggests that normative ageing may reduce the brain’s capacity to sustain diverse metabolic states. As that entropy falls, network behavior may become less adaptable, with less room for the metastable transitions that support efficient cognition.
The authors also found that higher global metabolic complexity was associated with better cognitive performance. At the voxel level, entropy clusters in both imaging modalities were linked to cognition, although the size and spatial location of those clusters varied. The pattern is important but not overinterpreted: these associations were largely accounted for by age variance. Age therefore appears to sit upstream of much of the relationship between entropy and performance, acting as a major driver of both metabolic complexity and cognitive outcome.
Why the glucose signal may be more informative than the haemodynamic one
The divergence between modalities hints at a mechanistic hierarchy. Glucose metabolism is the substrate that powers synaptic transmission, ion pumping, and large-scale network coordination. If the metabolic signal becomes less complex with age, the brain may lose some ability to redistribute resources rapidly across competing circuits. That would naturally erode metastability, especially in association networks that require broad integration and frequent reconfiguration.
By comparison, haemodynamic complexity may be too downstream to fully capture those energetic changes. Vascular responses are important, but they are coupled to neuronal demand rather than representing demand itself. The preprint therefore positions metabolic entropy as a more neurobiologically grounded marker of brain ageing than haemodynamic entropy alone.
What this means for brain ageing research
The study reframes entropy as more than a generic measure of signal irregularity. In this dataset, metabolic entropy behaved like a readout of the brain’s capacity to maintain rich and adaptable network dynamics. Its association with metastability, cortical thickness, glucose metabolism, blood flow, and cognitive performance gives the measure a coherent physiological context.
A second sentence bears repeating: the ageing brain showed less metabolic complexity.
Because the work is a preprint, it remains preliminary and should be interpreted with that boundary in mind. The abstract does not provide enough detail to evaluate every analytic choice, effect size, or potential source of bias, and peer review has not yet tested the full methods and interpretation. Even so, the central signal is internally consistent: entropy in cerebral glucose dynamics declines with age and appears to support the unstable-but-functional network behavior that characterizes a responsive brain.
If these findings hold up, metabolic multiscale entropy could become a useful biomarker in future work on healthy ageing, neurodegeneration, and cognitive resilience. For now, the study adds a mechanistic layer to a long-standing idea in neuroscience: the brain works best when its activity remains complex enough to adapt, but ordered enough to stay efficient.