Sit still in a quiet room, think about nothing in particular, and the organ doing the thinking is still the most expensive thing you own. The brain is roughly 2 percent of body mass and consumes around 20 percent of the body's total energy supply, and about 80 percent of that goes to neurons alone: not to insight or memory as such, but to the dull municipal work of pumping ions back across membranes, clearing and recycling neurotransmitters, and paying for synaptic transmission. Those figures open a new review by Carmen Lahr, Elle Wilson, Melanie Sengupta, Michela Bernini and Alexander Skupin, published online in Neurobiology of Disease on 23 September 2026 as article 107615 and currently in press as a journal pre-proof. The authors use them to set up a question this section keeps circling back to: if the budget is that tight, what happens to the household when one member stops contributing?
The review's organising move is to stop treating brain metabolism as one pathway and start treating it as a division of labour. Neurons are the big spenders. Astrocytes handle glucose uptake, glycogen storage, lactate production, neurotransmitter recycling, neurovascular coupling and lipid metabolism, supplying neurons rather than competing with them. Oligodendrocytes, best known for myelin, also feed axons directly, passing across lactate and TCA cycle intermediates; the authors note that when that coupling fails, axonal dysfunction follows. Microglia shift their own metabolism as they switch roles, through what the review calls immunometabolic reprogramming, inflammatory signalling and the clearance of debris. And the vasculature is not plumbing but a participant: endothelial cells, pericytes, astrocytic endfeet and neurons together form a neurovascular unit that reads metabolic signals and widens vessels where the work is being done, so that blood flow and glucose arrive at the regions that are busy.
What makes this feel less like a textbook diagram and more like a living system is how fast the accounting moves. In the sequence the review illustrates, a neuron fires, potassium and neurotransmitter spill into the extracellular space, astrocytes take them up, and the sodium-dependent transport that does the taking up switches on the astrocyte's own sodium-potassium pump. That pump costs ATP, the astrocyte ramps up glycolysis to pay for it, and lactate goes out to the neuron. The act of cleaning up after a signal is itself what triggers the next delivery of fuel. The review also points to glycogen as the brain's small local reserve: astrocytes break it down and hand over lactate when glucose runs low, and glycogen builds back up when neuronal activity drops, an observation that dates in the cited literature to work from 1972 on anaesthetised brains.
The pivot of the paper is a claim about direction of travel. Disturbances in energy production and use, the authors write, are now recognised as "early and potentially causal contributors to neurodegeneration, rather than simply downstream consequences of neuronal loss". That is the difference between metabolism as a symptom and metabolism as a cause, and it is the reason this review belongs in a section about shared grammar rather than separate diseases. Using Alzheimer's and Parkinson's as their principal models, Lahr and colleagues pull together mitochondrial dynamics, glial immunometabolism, lipid droplet biology, patient-derived cell models, metabolic imaging biomarkers and emerging metabolic therapies, and argue that cell-type specialisation, mitochondrial function and the coupling between cells together decide which neurons are vulnerable and which are not. Some of the material they survey is frankly preclinical: the mitochondrial fission inhibitor peptide P110, for instance, is described as delaying disease progression in mouse models of amyotrophic lateral sclerosis, which is a statement about mice and not about people.
For a family reading this after a diagnosis, the honest translation is narrow and still worth having. It does not change a prescription or a care plan this week. What it changes is the picture of what is going wrong: not one rogue protein acting alone in one cell type, but a supply chain in which astrocytes, oligodendrocytes, microglia and small blood vessels all have invoices to pay, and in which the loss of neurons may be the last line of an account that went wrong much earlier and somewhere else. That framing is why metabolic imaging is being studied as a way of seeing trouble before it is visible as atrophy, and why metabolic treatments are a research direction at all. None of that is settled, and the review says as much by ending on challenges and future perspectives rather than on answers.