The future of healthy brain aging will be metabolic: a predictive lesson from data-driven modeling

Advances in modern medicine, public health, and living standards over the last century have significantly extended life expectancy. However, with the increase in longevity, the prevalence of neurological diseases linked with aging-including dementia and simple cognitive decline-has tremendously increased and will continue to rise with the projected aging of the population. We need urgently to identify the critical causative factors and find ways to counteract or reverse their effects to promote healthy brain aging and further extend healthy life expectancy. In their lead article, Shichkova and colleagues (Shichkova et al. 2025) used data-driven modeling to identify the major metabolic alterations that can impact neuronal activity in the aging brain and to suggest how these could be reversed.A reduction in glucose consumption (or hypometabolism), as identified in specific brain regions using 18 fluorodeoxyglucose positron emission tomography ( 18 FDG-PET), has become a hallmark of both normal and pathological brain aging. Indeed, clear evidence of hypometabolism has been reported in neurodegenerative diseases such as Parkinson's disease and Alzheimer's disease (AD) (Zilberter and Zilberter 2017). Moreover, these metabolic alterations have also been observed in asymptomatic, at-risk individuals-sometimes years before any detectable sign of disease-suggesting that such metabolic deficits are not a mere consequence of neurodegeneration. Hypometabolism has been detected in normal brain aging, and associated as a risk factor for cognitive decline. The exact causes of such hypometabolism, and the link with the eventual cognitive deficits, remain unclear for the moment. Elucidating these relationships and the underlying pathological processes affecting brain metabolism might allow the design of adequate interventions to delay or even prevent cognitive decline.Based solely on histological observations, 19 th century neuroanatomists already suggested that glia, and more specifically astrocytes, could regulate the supply of essential metabolites to neurons. However, only in the last 30 years has clear evidence implicated astrocytes in this critical function. Two seminal observations sparked the investigations that led to these findings. First, brain activation triggered by sensory stimulation or a cognitive task was shown to produce a concomitant increase in blood flow over the associated cortical area. This "neurovascular coupling" was proposed to ensure a sufficient supply of oxygen and nutrients to neurons in the active area. Its cellular and molecular bases remained elusive for decades despite intense research efforts. Only recently has the role of astrocytes in the regulation of cerebral blood flow upon neuronal activation been uncovered, revealing intricate metabolic interactions between neurons, astrocytes, and blood vessels (Lia et al. 2023). The second key finding followed the development of a method to measure and visualize localized glucose consumption by the brain. The 2-deoxyglucose method revealed the tight relationship between activation of a specific brain area and localized glucose consumption, giving rise to the concept of neurometabolic coupling. Again, the prominent role of astrocytes and the description of the associated mechanism-the so-called astrocyte-neuron lactate shuttleforegrounded the tight metabolic interactions between astrocytes and neurons (Bonvento and Bolaños 2022). These findings also raised the question of the importance of these interactions for cognition. Indeed, interference with the transfer of energy substrates from astrocytes to neurons led not only to alterations in neuronal responses upon brain activation, but also to deficit in behavioral performances requiring intact activation of the associated brain areas (Roumes et al. 2023). These observations highlighted the fact that neurons are not energetically autonomous; their activity (and the cerebral functions they subserve) critically depends on interactions with their astrocytic partners. Suddenly, we must consider a new factor contributing both to cognition and to pathological processes affecting cognition. Effectively, astrocytes are emerging as a novel therapeutic target for neurological diseases affecting cognition, such as AD (Rodriguez-Giraldo et al. 2022). Considering their role in neuroenergetic regulation, astrocytes might constitute a priority target for interventions aimed at restoring or preserving metabolic proficiency in the central nervous system.Interestingly, conditions affecting peripheral energy homeostasis, such as diabetes, can directly cause cognitive decline or are a risk factor for developing dementia, notably via alterations in brain energy metabolism (Yanai et al. 2024). Indeed, some investigators consider AD as a "type 3" form of diabetes. Brain insulin resistance and dysfunction of insulin signaling pathways are suspected to be responsible for cognitive impairments and drugs that treat these pathologies have been proposed as potentially effective solutions.Insulin and other hypoglycemic agents such as metformin have shown beneficial effects in animal models and patients with AD. Insulin sensitizers such as rosiglitazone and pioglitazone, which are agonists of nuclear receptor peroxisome proliferator-activated receptor- (PPAR), have reversed behavioral deficits in preclinical AD models and in patients. Similarly, analogs of the gut hormone glucagon-like peptide-1 (GLP-1), such as liraglutide and semaglutide, have demonstrated neuroprotective activity against the toxicity of A oligomers (produced in abnormal quantities in AD) and the decline in behavioral performances normally exhibited by genetic mouse models of AD. In addition, fibroblast growth factor 21 (FGF21), a hormone produced predominantly by the liver involved in glucose and lipid homeostasis, was shown to improve behavioral performance in a diabetesinduced cognitive decline mouse model and a genetic mouse model of AD, by modulating components of the astrocyte-neuron lactate shuttle. Pharmacological agents are not the sole therapeutic option with the potential to prevent aging-related cognitive decline. There is accumulating evidence that physical exercise exerts cognitive benefits in both young and aging subjects, although the precise mechanisms involved are certainly pleiotropic and not yet well-defined. One hypothesis suggests that physical exercise might benefit cognition by improving brain energy metabolism. Effects on brain mitochondrial activity are among those observed, for example. But, interestingly, exercise also improves cognitive functions by modulating various components of the astrocyte-neuron lactate shuttle, and by allowing muscle-derived lactate to enter the brain and sustain brain activity (Xue et al. 2022). Further investigations are necessary to decipher the critical impact of exercise on cerebral metabolic processes and how this might improve cognition, but the improvement of astrocyte-neuron metabolic interactions already appears important.The brain is a complex organ involving several cell types and multiple interactions between them, especially in the domain of neuroenergetics. To elucidate how such interactions are important for cognition we need to use tools that account for this complexity while extracting the key elements to guide our quest for interventions to prevent aging-associated cognitive decline. Some successful examples show how mathematical modeling could help identify mechanisms underlying complex biological data or to predict the impact of an external stimuli or treatment on a particular biological response. The brain represents a particular challenge in this respect but recent efforts, such as those of Shichkova and colleagues (Shichkova et al. 2025), offers the promise to take us one step further. Considering the laborintensive and costly biological investigations, and the ethical aspects of animal experimentation, computational modeling should provide a useful approach to rapidly develop and evaluate anti-aging strategies aimed at preserving metabolic proficiency and thereby prevent cognitive decline.A paradigm shift is taking place in the field of aging, with efforts dedicated to promoting healthy aging. The increased incidence of neurodegenerative diseases associated with longer life expectancy is mobilizing substantial resources to uncover the underlying causes and to develop therapeutic solutions. Along with several clinical trial failures, the realization that genetics-driven hypotheses alone are not providing successful options to prevent agingrelated cognitive decline has led to a reappraisal. Mounting biological evidence, together with data-driven modeling, points to metabolic dysfunctions during aging as fundamental events and potential therapeutic targets. We should pay attention to these lessons now if we want to prepare ourselves for a long and bright future. Artificial intelligence was not used to generate any part of the text or correct this manuscript.