Alzheimer's Disease Through the Lens of Energy Economics: Adaptive Origins, Comparative Clues & Therapeutic Opportunities

 

Alzheimer's Disease Through the Lens of Energy Economics: Adaptive Origins, Comparative Clues & Therapeutic Opportunities

Jared Edward Reser Ph.D.

 

1. Introduction: From Toxic Debris to Energy Economics

1.1 The metabolic-reduction hypothesis (Reser 2009): core logic

In 2009 I published an article proposing that Alzheimer's pathology may represent a maladaptive extension of an evolutionarily conserved brain-energy reduction mechanism. Central to the hypothesis is the idea that the human brain poses a substantial energetic liability, especially in older hunter-gatherers facing diminished foraging returns and resource availability. The article argued that Alzheimer's would have selectively tempered the most energy demanding areas of the brain to reduce overall energy expenditure without compromising foraging ability. However, when modern longevity lets that same program run far longer than hunter gatherers would have lived in the prehistoric past (55 is the average terminal age of modern hunter-gatherers), it crosses the threshold from adaptive thrift to clinical Alzheimer’s.

 

Reser JE. Alzheimer's disease and natural cognitive aging may represent adaptive metabolism reduction programs. Behav Brain Funct. 2009 Feb 28;5:13.

 

From an evolutionary standpoint, humans historically experienced significant reductions in caloric returns starting around age 45 to 50, when cognitive demands for novel learning typically diminish because of accumulated expertise. I argued that natural selection favored mechanisms enabling late-life individuals to strategically reduce cortical energy expenditure by selectively pruning metabolically expensive neuronal circuits, especially within high-order association cortices and the hippocampus. Such an adaptive "low-power mode" would have conserved valuable metabolic resources during prolonged periods of scarcity without severely compromising essential survival skills, sensorimotor functions, or culturally valuable accumulated knowledge. This dovetails with other efforts made throughout the body to conserve energy with age. It is also in line with the fact that brain metabolic rate diminishes linearly after late childhood.

I further noted striking parallels between the molecular and cellular hallmarks of Alzheimer’s—such as selective regional glucose hypometabolism, synapse elimination, insulin resistance, amyloid accumulation, and tau hyperphosphorylation—and those observed in mammalian brains undergoing adaptive metabolic downshifts. These downshifts include conditions such as starvation, torpor, and hibernation. Intriguingly, these extreme yet reversible physiological states share many molecular mechanisms, suggesting that Alzheimer's pathology might represent a chronic activation of an ancient metabolic conservation strategy. Before we continue, let us look at the key lines of evidence from the 2009 article.

 

 

Evidence category	Core point in the paper
1. Human brain is an extreme metabolic liability	The cortex consumes 20‑25 % of resting energy—more than any other organ—so trimming its cost in late life would have increased survival during food scarcity.
2. Age‑linked cerebral hypometabolism is universal and continuous	Brain glucose use peaks in childhood, declines in adulthood, and keeps falling through old age; AD represents a natural continuation of this trajectory.
3. Selective hit‑map matches an energy‑saving strategy	AD preferentially deactivates high‑cost association cortex & hippocampus while sparing sensorimotor areas; exactly the pattern expected if the goal is to drop “expendable” cognitive load yet preserve basic perception and movement.
4. Parallels with starvation / torpor responses	Hyper‑phosphorylated tau, lowered thyroid & growth hormone, insulin resistance, etc., appear in starving rodents and hibernating animals—and in AD—implying a shared fuel‑deprivation program.
5. Life‑history logic: declining foraging & rising expertise	Hunter‑gatherers’ calorie return is reduced after ~45 y; by then skills are routinised, so costly working‑memory circuits can be pruned without loss of competence.
6. Metabolic‑syndrome & “thrifty” genetics link	AD co‑occurs with low resting metabolic rate, insulin resistance, APOE‑ε4 and other thrifty alleles; their geographic distribution tracks famine‑prone populations.
7. Neuroecology parallels in other species	Birds & mammals down‑regulate hippocampal metabolism when food is scarce; hippocampus is likewise the earliest, most hypometabolic region in AD.
8. Cross‑species ubiquity of plaques & tangles	Multiple mammalian orders show AD‑like lesions, implying deep evolutionary roots and selective value rather than random human pathology.

Table 1: Key lines of evidence in the 2009 Reser article. Because the ageing brain’s energy budget once exceeded what late‑life foragers could reliably earn, natural selection favoured a program that (i) progressively lowers cerebral metabolism, (ii) targets the most expendable cognitive networks, (iii) mirrors documented starvation/torpor responses, and (iv) is embedded in thrifty genes that remain common wherever famine shaped human evolution. 

 

1.2  Purpose and scope of this review

Over the past fifteen years, accumulating data from human neuroimaging studies, comparative biology, genetics, and clinical metabolic interventions have provided support for the 2009 proposition. The purpose of this review is to comprehensively synthesize and critically assess recent evidence that frames Alzheimer's disease as an over-extended brain energy-conservation response rather than purely a toxic amyloid-driven pathology. Vast overlap between Alzheimer's and energy conservation modes will be documented, looking carefully at similarities in molecular pathways with mammalian hibernation, torpor, and starvation. The article will also examine translational therapeutic opportunities informed by this reconceptualization of AD. By leveraging evolutionary insights and comparative biological data, I propose novel therapeutic avenues designed not just to slow the progression of AD, but to potentially reverse pathological changes by re-engaging the brain’s inherent metabolic "wake-up" mechanisms.

Before we continue, let me spell out some of the major similarities between Alzheimer's and adaptive metabolism reduction programs in mammals in a didactic, easy-to-understand way. 

 

As we continue, let me spell out some of the major similarities between Alzheimer's and adaptive metabolism reduction programs in mammals in a didactic, easy-to-understand way. 

 

2. Comparative Review: Starving and Hibernating Mammals Utilize the Molecular Machinery of Alzheimer's 

2.0 Hibernation and Torpor Explained

Many animals face seasonal periods of food scarcity, particularly during winter months when foraging becomes energetically costly or ecologically unfeasible. To survive these conditions, certain species have evolved physiological strategies that dramatically lower their metabolic demands. Hibernation is one such strategy, characterized by prolonged, uninterrupted periods of metabolic suppression. During hibernation, animals such as bears, bats, ground squirrels, and groundhogs exhibit profound reductions in heart rate, body temperature, and neural activity. For instance, hibernating bears may reduce their metabolic rate by as much as 50%, entering a state of sustained energy conservation.

Torpor, by contrast, is a shorter-term, often daily or intermittent form of metabolic depression, observed in species like skunks, chipmunks, hedgehogs, raccoons, and deer mice. Torpor is also found across other taxa, including various birds, reptiles, amphibians, and insects. While hibernation is typically seasonal and extended, torpor allows for rapid, flexible responses to acute environmental stressors such as cold or food shortage. Both hibernation and torpor, along with the broader mammalian response to starvation, serve the adaptive function of minimizing energy expenditure under resource-limited conditions. For context, sleep also offers a mild energy-conserving effect—lowering respiratory rate, body temperature, and metabolic rate—though unlike hibernation and torpor, brain activity during sleep remains relatively high and structured, reflecting its distinct functional role. The relation between AD and brumation, estivation, and lethargy is also ripe for study.

2.1 Hibernators and Starvation: Reversible Tau and Synapse Stripping

During deep hibernation, species such as ground squirrels, hamsters, and bears dramatically reduce their brain's metabolic rate—sometimes by up to 90%—accompanied by extensive tau hyperphosphorylation at the very same epitopes (AT8, Ser396) observed in human Alzheimer's brains (Arendt et al., 2003; Stieler et al., 2011). In parallel, dendritic spine density and synaptic connections are substantially pruned, reducing the metabolic load imposed by inactive circuits. Critically, within hours of rewarming and arousal, this extensive tau phosphorylation is reversed, synapses regenerate rapidly, and cognitive functions are restored without apparent damage (Popov & Bocharova, 1992; von der Ohe et al., 2006).

Tau is a protein found in neurons, especially in the long axons that carry signals between brain cells. Its main job is to stabilize microtubules, which are like tiny tracks that move nutrients and cell parts around inside the neuron. Think of tau as railroad ties that keep the neuron’s internal tracks straight and working efficiently. In Alzheimer’s disease tau becomes hyperphosphorylated, meaning it gets too many phosphate molecules stuck onto it. This changes its shape and causes it to fall off the microtubules. Without tau, the tracks collapse. Worse, the tau proteins start sticking together and form tangles inside the neurons. These neurofibrillary tangles disrupt communication, block nutrient transport, and eventually kill the neuron. So, in AD, tau goes from being a support structure to being a toxic clog.

Fascinatingly, hibernating, torpid, and starving mammals also show changes in tau—but they don’t develop Alzheimer’s. In hibernators, like ground squirrels and bears, tau also becomes hyperphosphorylated. The tau even detaches from microtubules and there is partial axonal collapse, just as in AD. The railroad tracks are disabled. In other words, this very expensive process of transporting resources from the cell body down the axon to the terminal and synapse is frozen. It is an easy, elegant and reversible way for the brain to save energy. As in AD it reduces the expensive shuttling of cellular products in the most energy-expensive areas of the brain, the cerebral cortex and hippocampus. Fascinatingly, hibernation, prolonged starvation, and Alzheimer’s disease (AD) all flip tau into an energy-saving configuration by tagging exactly the same amino-acid sites, those detected by the AT8 antibody (Ser 202/Thr 205) and by antibodies to Ser 396/Ser 404. Whether it’s a ground squirrel entering torpor, a human fasting, or a neuron sliding into Alzheimer’s disease, the identical phospho-sites and writer/eraser enzymes are engaged.

Normally, many components are shuttled down axons including synaptic vesicles full of neurotransmitters, mitochondria to power the synapse, mRNA and ribosomes for local protein synthesis, as well as growth and repair materials for distant parts of the neuron. Removing the supportive tau freezes activity placing the neurons in a lean, thrifty suspended state. This starts as soon as hibernation begins or after several days of food restriciton. For example, it appears in rat/mouse cortex & hippocampus after 2–5 days of fasting. It causes the neuron to enter a low-power mode—like pausing a supply chain to save fuel during a winter shutdown. When this happens, the synapses temporarily go dark just like a railroad depot in a snowstorm where all the workers have gone home.

However, in hibernating and starving mammals, the tau changes are reversible without lasting repercussions. When hibernating animals wake up the tau phosphorylation reverses and the tau reattaches to the microtubules reinstating axonal transport. However, in Alzheimer’s, tau is hyperphosphorylated, but never de-phosphorylated. The shutdown becomes chronic and the cells never “wakes up.” Eventually, tau forms tangles, which do not happen in hibernation, causing permanent destruction.

GSK-3β is the main tau-phosphorylating enzyme. It adds phosphate groups to tau at many different sites. In hibernation and starvation GSK-3β becomes transiently active, but shuts off when the animal rewarms or eats. In Alzheimer’s, the same enzyme, GSK-3β, is chronically overactive. Something similar is true for the other tau kinases like CDK5. Also, the enzyme that removes phosphate groups from tau, PP2A, is reduced in hibernation/starvation and also reduced in Alzheimer’s.

So, axonal transport is metabolically expensive because it uses ATP-hungry motor proteins (i.e. kinesin, dynein) to carry substances up to 1 meter in long axons. Constant delivery of organelles and materials is required to keep the synapse alive. By detaching tau and halting the supply chain, the brain can shut off one of its most expensive processes. In hibernation and starvation, it’s a reversible suspension of the railroad tracks. In Alzheimer’s, it’s more like a permanent derailment.

2.2 Amyloid-β Production as a Way to Reduce Energy Expenditure

A misfolded protein is a protein that bends into the wrong shape. Instead of floating freely or doing its job, it sticks to other misfolded proteins and starts to form clumps or aggregates. These can block nutrient flow and communication. Amyloid is a type of misfolded protein that forms insoluble, sticky fibers. In Alzheimer’s, the key amyloid protein is called amyloid-beta (Aβ) but it also plays a role in starvation and hibernation.

Multiple studies—especially in hibernating ground squirrels, Syrian hamsters, and bears—have shown that during hibernation levels of amyloid precursor protein (APP) increase in the brain. It comes from the same source as in AD, improperly processed or cleaved APP proteins. However, they do not accumulate to the point of forming amyloid plaques because they are cleared during arousal from hibernation or after refeeding. So, hibernators appear to intentionally use amyloid-beta, but their brains manage it carefully, avoiding the pathology seen in Alzheimer's. Today, researchers believe that Aβ plays an adaptive role in hibernation and starvation, by suppressing synaptic activity and its energetic demands. Further, creating these Amyloid proteins is easy and cheap because it uses preexisting parts (APP).

Synapses (the connection points between neurons) are one of the largest energy sinks in the nervous system. When an animal faces several days without food (starvation) or months without feeding (hibernation), it uses amyloid-beta to slash that budget. Amyloid-β quiets synapses by dampening their energy-expensive mechanisms. This includes acetylcholine receptors which buttress learning and memory. But it also includes NMDA and voltage-gated Ca²⁺ channels which manage excitatory transmission, conserving vast amounts of energy otherwise used in signaling and ion pumping. So in effect, this production of amyloid is like stalling things at the factory, stopping critical junctures for workflow. You could say that amyloid-beta acts like a lockout supervisor, pausing activity at the control panels (receptors) of the loading docks (synapses). But let’s talk more specifically about what it does to acetylcholine receptors.

Acetylcholine (ACh) is one of the brain’s most important neurotransmitters involved in attention, learning and memory. It works by binding to special receptors on neurons, such as the α7 nicotinic acetylcholine receptor (α7nAChR). It’s found all over the brain, especially in the hippocampus and cortex. When acetylcholine binds to α7nAChR, the neuron becomes more excitable, synapses become stronger, and memory circuits become more active and plastic.

Now here’s the twist: the amyloid-beta protein (Aβ)—especially in its small, soluble form—can bind to the α7 nicotinic receptor, just like acetylcholine does. However, amyloid binds much more tightly than acetylcholine and effectively blocks the receptor (competitive binding), preventing acetylcholine from working. The receptor is occupied, but it’s not doing its job. When amyloid-beta clogs up the α7 receptors the brain slips deeper into a low-activity, energy-saving mode. Thus, in hibernation, torpor, and starvation acetylcholine activity is intentionally dialed down as a way to save energy. This supports the idea that Alzheimer’s disease is a pathological version of an ancient energy-saving response, using the same molecules, same receptors, and same circuits, just stuck in the on position too long.

Amyloid formation is not exclusive to the brain—it’s a broader metabolic regulatory phenomenon. In fact, from an evolutionary standpoint, there’s growing support for the idea that protein aggregation may serve an energy-saving function under certain conditions. In type 2 diabetes mellitus (T2DM), amyloid also accumulates pathologically in the pancreas—specifically in the islets of Langerhans. But here, it's not Aβ, it's islet amyloid polypeptide (IAPP), also known as amylin. They form plaque-like deposits that damage or kill beta cells. This accumulation of IAPP reduces insulin secretion—effectively contributing to insulin insufficiency and hyperglycemia and in turn reducing glucose uptake by tissues and accomplishing energy conservation. Amylin fibrils might have evolved as a self-limiting brake on insulin secretion.

What emerges is a picture of amyloid aggregation as part of a broader, conserved strategy to reduce energy expenditure during prolonged energy stress. Thus, rather than being purely toxic junk, amyloid in both the brain and pancreas might reflect a thrifty design that goes awry when unregulated—especially in the context of modern environments (over-nutrition, prolonged lifespan, sedentary living). It also suggests that therapeutic strategies used in T2DM—like insulin sensitizers (GLP-1 agonists, SGLT2 inhibitors)—may have shared relevance in AD, which, as we have discussed, is exactly what recent clinical trials are now exploring.

In the evolutionary literature, type 2 diabetes is often interpreted as the pathological extension of a once-adaptive “thrifty” genotype or phenotype—one that evolved to conserve energy during periods of caloric scarcity. Traits such as insulin resistance, fat storage, and reduced glucose uptake would have conferred survival advantages in ancestral environments marked by intermittent famine. The recent characterization of Alzheimer’s disease as “type 3 diabetes” reinforces this perspective, highlighting its shared features with systemic metabolic disorders—particularly insulin resistance, impaired glucose utilization, and amyloid accumulation—further supporting the view that Alzheimer’s, like type 2 diabetes, may represent the maladaptive persistence of an ancestral energy-conservation program.

 

2.3 Kinase/Phosphatase Switch (GSK-3β ↔ PP2A)

The remarkable reversibility in hibernators hinges upon a tightly regulated biochemical switch involving kinase and phosphatase enzymes that control tau phosphorylation. GSK-3β (Glycogen Synthase Kinase 3 beta) is a primary kinase responsible for phosphorylating tau at multiple sites (including AD-relevant ones like Ser396 and Thr231). It is activated in both hibernation and Alzheimer’s. During torpor, GSK-3β activity rises, leading to reversible tau phosphorylation. In AD, persistent GSK-3β activation leads to chronic, irreversible tau aggregation.

PP2A (Protein Phosphatase 2A) is the main tau dephosphorylating enzyme in the brain. It is suppressed during torpor, allowing tau phosphorylation to proceed. In AD, PP2A is chronically downregulated or inhibited, preventing tau clearance and promoting neurofibrillary tangle formation. During arousal from torpor, norepinephrine (from the locus coeruleus) and thyroid hormones suppress GSK-3β and stimulate PP2A. This rapidly removes phosphate groups from tau, thus restoring normal neuronal cytoskeletal integrity (Arendt & Stieler, 2003). In early AD, however, norepinephrine-producing neurons in the locus coeruleus are among the first to degenerate, which may block this arousal-like reversal pathway, leading to sustained tau phosphorylation.

2.4 Norepinephrine, Thyroid, and Brown Adipose Tissue (BAT) Arousal Burst

The rewarming phase in hibernating mammals involves a critical metabolic burst orchestrated largely by norepinephrine (NE) from the locus coeruleus, thyroid hormone release, and brown adipose tissue (BAT) thermogenesis. This coordinated burst re-energizes neurons, rapidly restores ATP availability, and reactivates metabolic enzymes necessary for neuronal recovery and synapse regeneration. Interestingly, this NE-mediated metabolic burst precisely mirrors pathways disrupted early in Alzheimer's disease, providing an additional therapeutic target for reversing Alzheimer's pathology (Cannon & Nedergaard, 2004; Tupone et al., 2013).

2.5 Daily Torpor & Synthetic Torpor in Non-Hibernators

Non-hibernating species, including certain rodents and primates, demonstrate shorter bouts of daily torpor, during which similar transient phosphorylation of tau, metabolic reduction, and synapse pruning occur. Importantly, researchers have experimentally induced synthetic torpor states in laboratory rodents using pharmacological agents like adenosine A₁ receptor agonists. These animals also experience profound yet fully reversible tau phosphorylation, synaptic pruning, and metabolic slowdown, confirming that torpor-like mechanisms are broadly conserved across mammals, including species closely related to humans (Cerri et al., 2013; Tupone et al., 2013).

2.6 Insulin and Leptin Receptors — the Body’s Fuel Sensors

During starvation or hibernation, both insulin and leptin drop, signaling to the brain: “There’s no food coming in—conserve energy!” This message turns off systems that use lots of energy (like thinking and moving) and shifts the body into a low-power mode. In Alzheimer’s, the brain stops responding properly to insulin and leptin—even when they’re present. This is called resistance. The result is similar to starvation: reduced brain energy, memory loss, and cell damage. So the brain ends up acting like it's starving, even though food is available.

When leptin and insulin fall, a part of your brain called the hypothalamus activates certain neurons that release: AgRP (Agouti-related peptide) and NPY (Neuropeptide Y). These are hunger-triggering signals that also lower body temperature, reduce activity, and slow metabolism—traits seen in both hibernation and Alzheimer’s.

2.7 AMPK and mTOR — the Cellular Fuel Gauges

Inside cells, two important proteins track energy use: AMPK and mTOR. AMPK turns on when energy is low; it tells cells to slow down, recycle parts, and survive with less. The other is mTOR which works when energy is high; it tells cells to grow and make new proteins.

When animals hibernate or fast, AMPK increases and mTOR decreases—this slows the brain down and saves energy. In Alzheimer’s, this same balance is off—AMPK stays high, mTOR stays low. That keeps brain cells in a “shutdown mode” for too long, which might contribute to shrinkage and memory loss.

2.8 Adenosine Receptors — the Brain’s Dimmer Switch

Adenosine builds up in the brain the longer you're awake. It tells your brain, “You’re tired, slow things down.” It acts mostly through the A1 and A2A receptors to suppress brain activity and encourage sleep or rest—also helpful during torpor or hibernation. In Alzheimer’s, adenosine signaling becomes too strong or unbalanced, leading to constant suppression of memory circuits and increased inflammation.

2.9 Inflammatory Receptors — IL-1 and TNF

IL-1 and TNF are molecules that trigger inflammation. They bind to receptors that help the brain respond to stress or infection. In short bursts, like at the start of torpor, they help animals cool down and conserve energy. But in Alzheimer’s, these signals become chronic, leading to overactive immune cells in the brain and destruction of healthy neurons.

1.10 Microglial Receptors — the Brain’s Cleanup Crew

Microglia are the brain’s immune cells. They clean up damaged parts, including old synapses. They use receptors like TREM2 and CR3 to find what to remove. In healthy development (and especially in hibernation), microglia trim unneeded connections to make the brain more efficient. But in Alzheimer’s, the cleanup goes too far. These receptors stay turned on, and microglia start removing healthy synapses, leading to memory loss.

Temporary Conclusion

These molecules and receptors aren’t just involved in Alzheimer’s—they’re part of an ancient survival system: When food is scarce or it’s winter, they help animals slow their brains, prune unnecessary circuits, and conserve energy. In Alzheimer’s, these same systems get turned on—but don’t turn off. The brain ends up stuck in a thrifty, low-power state, unable to recover. This may explain why Alzheimer’s looks so much like a hibernation program—it uses precisely the same switches and circuits, but without an exit ramp.

The next two tables document which natural states demonstrate AD-like pathology.

 Alzheimer's Like Features Across Natural and Adaptive Brain States:

3. Mammals with Spontaneous Amyloid + Tau Pathology

3.1 Cetaceans (Dolphins, Whales)

Cetaceans, such as bottlenose dolphins and pilot whales, exhibit naturally occurring Alzheimer's-like pathology, including amyloid-beta plaques and tau tangles. These neuropathological features correlate strongly with extended lifespan, substantial cortical mass, and metabolic stressors inherent to prolonged diving and high cognitive demands (Gunn-Moore et al., 2018; Davis et al., 2019). Importantly, despite prominent pathology, cognitive function in these species remains largely intact, suggesting effective endogenous mechanisms that limit functional impairment.

3.2 Great Apes, Dogs, and Cats

Chimpanzees and other great apes, with lifespans approaching human longevity, display similar amyloid and tau accumulation patterns in aging brains. These AD-like features emerge predominantly in associative cortices and hippocampal regions analogous to those affected early in humans. Yet again, these changes typically appear only at extreme ages beyond typical reproductive periods, supporting the hypothesis that metabolic conservation mechanisms only become pathological when extended unnaturally (Rosen et al., 2008).

Similarly, aged domestic dogs exhibit Alzheimer's-like (neuritic Aβ plaques) plaques, some p-tau but rarely full tangles, and cognitive dysfunction, again emphasizing the role of lifespan extension and metabolic demands in pathology manifestation (Inestrosa et al., 2005; Head, 2013). Aged cats exhibit extracellular Aβ and AT8-positive intraneuronal p-tau accumulate with age; cerebral amyloid angiopathy is also reported. Rhesus macaques show Aβ plaques, tau phosphorylation, gliosis, synapse loss by around 25 years of age.

3.3 Octodon degus

The rodent Octodon degus, has several adaptations to its semi-arid environment in the matorral ecoregion of central Chile. These adaptations, such as a tendency toward the metabolic syndrome, help it save energy. Many individuals spontaneously develop classic AD neuropathology, including amyloid plaques, tau tangles, neuroinflammation, neuron and synapse loss, and cognitive impairment, closely paralleling human Alzheimer's. The rodent lives 7–9 years at least twice as long as most lab rodents indicating that this could be an example of late life runaway thrift. They practice corprophagy, ingesting their feces to extract more nutrients particularly when food is scarce or low in nutrients. They have a specialized digestive system that allows them to efficiently extract nutrients from plant matter including dried vegetation and bark. They have highly reduced basal metabolic rate compared to other rodents. Their metabolism can vary seasonally with lower rates during the summer months (nonbreeding season). This comparative model underlines the evolutionary conserved nature of AD-like responses under metabolic strain.

The presence of Alzheimer’s-like neuropathology in Octodon degus, which does not hibernate, strongly supports the idea that AD may be an evolutionary thrifty program gone awry. In fact, degus are one of the clearest animal cases showing how normal, adaptive energy-saving responses in the brain can slide into chronic degeneration under modern or mismatched conditions. Octodon degus bridge the gap between adaptive plasticity and maladaptive degeneration. Their spontaneous development of AD-like features in old age doesn’t just fit the thrift hypothesis—it embodies it. They show that Alzheimer’s may be less about the “collapse of the brain,” and more about a strategy for conserving it.

3.4 Exception Species (Naked Mole-Rat) and Lessons in Resilience

A notable exception to typical mammalian patterns is the naked mole-rat (Heterocephalus glaber), which despite extraordinary longevity (over 30 years) and high brain amyloid-beta burden, does not develop pathological plaques, tau tangles, or cognitive impairment. This remarkable resilience may reflect unique molecular adaptations to metabolic and oxidative stress, highlighting protective pathways that prevent the pathological progression of Alzheimer’s features. Understanding these mechanisms could inspire novel therapeutic strategies to confer similar protection in humans (Edrey et al., 2013).

3.5 Dehnel’s Phenomenon and Seasonal Brain Size Reduction

Common shrews (Sorex araneus), unable to store significant fat reserves or enter true hibernation, employ a unique seasonal survival strategy known as Dehnel’s phenomenon. Each winter, these shrews undergo substantial brain and skull size reduction (approximately 20%), significantly lowering metabolic demands. Remarkably, brain structures regrow during the warmer months, suggesting innate neural mechanisms capable of reversible brain atrophy and synaptic remodeling. Such extreme, yet reversible, neural adaptations in a non-hibernating mammal further support the plausibility of a programmed, reversible metabolic reduction response in humans. Whether tau participates in Dehnel’s phenomenon remains an open—and testable—question.

3.6 Relation to Neuroecology in Food-Caching Animals

Across a surprising range of taxa, from small mammals to food caching birds, animals trim the most energetically expensive part of their forebrain, the hippocampus, when heightened spatial memory is no longer worth its metabolic cost. These seasonal contractions deliver a measurable saving in ATP demand—exactly when food is scarce or ambient temperature drives up thermoregulatory costs. Alzheimer’s changes begin in the hippocampus. Alzheimer’s disease even targets the same hippocampal subfields first, with FDG-PET showing an early, selective fall in glucose use and synaptic activity. The convergence suggests that the hippocampal atrophy and hypometabolism seen in AD may represent a deeply conserved neuro-ecological program that many birds and mammals activate only temporarily. Understanding how chickadees and shrews switch this circuitry off and back on each year could therefore illuminate new ways to “wake” the human hippocampus from the chronic low-power state characteristic of early Alzheimer’s. Whether tau participates in these phenomena is also an open question.

The widespread presence of Alzheimer-like neuropathological responses across diverse mammalian lineages suggests deep evolutionary roots of metabolic downregulation programs.  

Animal Species that Age into Alzheimer-like Neuropathology Naturally

Taxonomic group	Species / common name	Key Alzheimer-like features reported in aged individuals
Companion mammals	Domestic dog (Canis familiaris)	Diffuse → neuritic Aβ plaques, cerebral amyloid angiopathy (CAA), mild tau-P, Canine Cognitive Dysfunction that maps to lesion load
	Domestic cat (Felis catus)	Cortical Aβ plaques, AT8-positive tau neurites, CAA, spatial & social memory loss
Rodents	Octodon degus	Aβ plaques & CAA, AT8 / PHF-1 tau tangles, microglial activation, object‐ and maze-memory decline
	Guinea-pig (Cavia porcellus)	Human-identical Aβ sequence; cortical plaques and CAA with age or high-cholesterol diet; sparse tau-P
	Gray short-tailed opossum (Monodelphis domestica)	Soluble Aβ accumulation and learning deficits by mid-life (plaques rare)
	Naked mole-rat	High soluble Aβ and tau-P but no plaque or tangle aggregation → a “resilience” comparator
Small primates	Gray mouse lemur (Microcebus murinus)	Aβ plaques, AT8 tau threads + tangles, gliosis, progressive spatial-memory loss by 4–5 y
Large primates	Chimpanzee & occasionally gorilla	Mature Aβ plaques + neurofibrillary tangles in cortex/hippocampus of >35 y specimens
	Rhesus macaque, cynomolgus macaque, vervet (African-green) monkey	Age-dependent cortical plaques, diffuse tau-P, synapse loss, measurable executive-function decline
Marine mammals	Beaked whales, pilot whales, bottlenose & Atlantic spotted dolphins	Dense-core Aβ plaques, CAA, AT8 / PHF-1 tau tangles, microgliosis; strandings often show severe pathology
Other	Sheep (rare reports), deer (sporadic)	Focal cortical plaques ± tau-P — data limited; not yet mainstream models

 

Comparing the living apes provides a living gradient—from mild Aβ accumulation to plaque + tangle formation—helping researchers pinpoint where the adaptive shutdown spirals into irreversible degeneration. All three apes produce human-sequence Aβ and deposit plaques as they age, but tau activation advances only as far as each species’ protection/clearance systems allow:

Extent of Alzheimer-like Neuropathology in the Great Apes

Feature	Orangutan (Pongo spp.)	Gorilla (Gorilla spp.)	Chimpanzee (Pan troglodytes)
Oldest brains examined	Captive adults ≈ 28-36 y (late 30s ≈ human 80s)	Zoo/wild adults ≈ 35-55 y	Sanctuaries & zoos ≈ 35-61 y (late 50s ≈ human 90s)
Amyloid-β (Aβ) plaques	Sparse, diffuse cortical plaques; strong bias toward the less-sticky Aβ40; meningeal & white-matter CAA small.	Numerous plaques in association cortex > hippocampus; extensive CAA (vessel deposits), especially in males.	Abundant dense-core & diffuse plaques in cortex and hippocampus; CAA common; Aβ42 proportion similar to humans.
Tau hyper-phosphorylation (AT8, PHF-1)	Largely absent—no neuronal pretangles, no glial tau.	Present but limited: AT8-positive astrocytic “coiled bodies,” occasional neuritic clusters; neuronal pretangles rare; no mature tangles.	Robust: AT8-positive pretangles, neuritic threads, astrocytic & oligodendroglial tau; true neurofibrillary tangles in hippocampus and temporal cortex of the oldest animals.
Neuroinflammation	Little reported; microglia quiescent.	Moderate microgliosis around plaques and CAA; sex differences noted.	Pronounced microglial activation and astrocytosis around plaques and tau inclusions, mirroring human AD tissue.
Neuron / synapse loss	Not detected; cortical thickness preserved.	No quantified neuron loss; mild synaptic pruning possible but unproven.	Subtle neuron loss in entorhinal cortex and CA1 reported in oldest individuals; synaptic protein reductions parallel plaque load.
Behavioural / cognitive data	No systematic late-life testing (orangutans seldom reach ≥40 y in captivity).	No formal cognitive batteries; anecdotal slowing only after mid-40s.	Documented decline in social decision-making, memory, and problem-solving after ~45 y; parallels lesion burden.
Overall “Braak stage” analogue	Pre-Braak: Aβ present, tau cascade not triggered.	Braak I–II: Aβ plus initial tau-P, but tangles rare.	Braak III–IV: Aβ + tangles in limbic & association cortex—closest non-human match to early/mid human AD.

3.7 Ancient Torpor Machinery in Placental Mammals: Why Humans Lost Behavioral Hibernation but Retained Cellular Toolkit

Humans probably descend from hibernating ancestors. Evidence suggests early mammals could enter torpor or hibernation-like states. Many modern mammals—especially small insectivores like shrews, hedgehogs, and tenrecs—exhibit daily torpor or seasonal hibernation, and are considered evolutionary relics of early mammalian forms. The molecular machinery required for torpor (e.g. tau phosphorylation, synaptic pruning, PP2A modulation, thyroid hormone control) is conserved across nearly all mammals, including humans—even though we don’t behaviorally hibernate. Certain strepsirrhine primates, like the fat-tailed dwarf lemur (Cheirogaleus), do hibernate—sometimes for months at a time—with all the classical features: metabolic suppression, tau phosphorylation, reversible synapse loss. They have “fat tails” because they store fat in their tails.

Humans, though not true hibernators, retain a remarkably conserved cellular and molecular toolkit associated with these states, including the kinase/phosphatase regulatory mechanisms that modulate tau phosphorylation, adenosine-based metabolic control pathways, and potent proteostatic clearance systems (Tupone et al., 2013; Cerri et al., 2013). Understanding these evolutionary genetic trade-offs and the ancient origin of torpor-related cellular pathways not only provides insight into Alzheimer's pathogenesis but also opens new therapeutic avenues focused explicitly on reactivating protective metabolic mechanisms and reversing pathological processes.

4. The High Cost of Thinking: Life-History Trade-offs and the Concept of Late Life Thrift

The human brain is notoriously expensive, metabolically speaking. Despite constituting only about 2% of total body mass, it consumes roughly 20–25% of the resting metabolic rate (RMR)—a caloric commitment far exceeding any other organ in the body, and substantially higher than the cerebral metabolic demand of other primates or mammals (Raichle & Gusnard, 2002). This energetic cost is predominantly due to cortical neurons, particularly those within associative areas like the default-mode network, where continuous baseline activity maintains readiness for cognitive processing (Buckner et al., 2008). Of course, these are the same areas effected by AD. Crucially, this metabolic investment is disproportionately allocated to higher-order processing areas of the cerebral cortex critical for memory, decision-making, and planning—precisely those areas compromised early in Alzheimer’s Disease (AD) (Sokoloff, 1999).

From an evolutionary perspective, the allocation of energy to cognitive functions must balance against the caloric returns obtained from environmental exploitation. Ethnographic and anthropological data from modern hunter-gatherer populations such as the Hadza and Tsimané show a consistent pattern: caloric productivity in humans typically peaks between ages 30–40 and subsequently declines sharply, dropping significantly after age 45–50 (Kaplan et al., 2000; Gurven et al., 2006). In evolutionary terms, the metabolic cost-benefit ratio of supporting energetically demanding neural networks declines with age, thereby potentially selecting for mechanisms to reduce unnecessary energetic expenditures in the aging brain. AD first throttles and deactivates the highest cost, plastic networks (prefrontal cortex, posterior cingulate, medial temporal lobes etc.) while sparing sensorimotor regions essential for day to day survival, mirroring a form of precision triage rather than an indiscriminate degenerative map. 

Life-history theory predicts that organisms allocate energy strategically across their lifespan, optimizing reproductive success and survival. Given the intense metabolic demands of the human brain and the declining returns on cognitive investments in later life, there is strong evolutionary rationale for a mechanism to selectively reduce cerebral metabolism after reproductive maturity. Strategic pruning of neural networks and reallocation of limited caloric resources toward essential physiological functions could have prolonged survival and enhanced indirect fitness benefits such as care-giving, knowledge transfer, and social roles within the community (Hawkes & Coxworth, 2013).

5. Human Evidence for a Built-In Low-Power Mode

5.1 FDG-PET chronology: hypometabolism precedes amyloid and tau pathology

A growing body of longitudinal neuroimaging evidence reveals that glucose hypometabolism, detectable via fluorodeoxyglucose positron emission tomography (FDG-PET), emerges significantly earlier than the appearance of classical Alzheimer's biomarkers such as amyloid plaques or tau tangles. Studies from the Alzheimer's Disease Neuroimaging Initiative (ADNI) and the Wisconsin Registry for Alzheimer's Prevention (WRAP) demonstrate a clear timeline: reduced cerebral glucose metabolism in specific cortical regions appears up to 10–15 years before cognitive symptoms or detectable amyloid or tau accumulation (Jagust & Landau, 2021; Mosconi et al., 2008). This observation strongly suggests that hypometabolism is not merely a consequence of neuronal injury but rather a potential initiating factor, consistent with the metabolic-reduction hypothesis.

5.2 Fuel specificity: preserved ketone uptake, insulin resistance ("type 3 diabetes")

Despite significant glucose hypometabolism, emerging dual-tracer PET studies show preserved brain uptake of alternative fuels, particularly ketone bodies. Clinical research demonstrates that while glucose metabolism significantly diminishes in early Alzheimer’s and mild cognitive impairment (MCI), the brain’s ability to utilize ketones remains largely intact (Cunnane et al., 2016). Such fuel specificity implies that the observed metabolic reduction is not due to generalized mitochondrial dysfunction or neuron death but instead reflects a selective downregulation of glucose uptake and processing pathways, akin to the adaptive metabolic shift seen during fasting or hibernation.

This selective glucose impairment parallels the widely recognized insulin resistance in Alzheimer's, often termed "type 3 diabetes" (de la Monte & Wands, 2008). Reduced insulin sensitivity and signaling within the brain further underscore an evolutionary context wherein neurons selectively restrict glucose use—possibly an adaptation to prolonged caloric insufficiency or energetically demanding conditions. This shift resembles the “fuel switch” that occurs during starvation, torpor, or hibernation, where the brain begins favoring alternative substrates (primarily ketones) to reduce reliance on glucose, which becomes scarce or deliberately restricted.

Multiple randomized controlled trials (e.g. BENEFIC, Henderson et al.) show that providing ketones exogenously through medium-chain triglycerides (MCTs) or ketone esters can raise brain ATP production and metabolism in association cortices, improve memory and executive function, and restore functional connectivity on PET and fMRI (Croteau et al., 2018; Fortier et al., 2021). If cognition and metabolic function can be restored with simple fuel repletion, it supports the idea that many neurons in early AD are not lost, but rather functionally silenced or idling—consistent with an evolutionarily conserved energy-saving mechanism.

 

Table 1. Therapeutic metabolic trials and cognitive outcomes

Intervention	Duration	Metabolic outcome	Cognitive outcome	Reference
MCT ketogenic supplementation	6 months	↑ Brain ketone uptake (PET)	Improved episodic memory, executive function	Fortier et al., 2021
Ketone ester (single dose)	Acute	↑ Parietal metabolism/connectivity	Improved attention, reaction time	Cunnane et al., 2016
Intranasal insulin (40 IU/day)	4 months	↑ FDG uptake, especially in APOE ε4	Improved memory, cognitive performance	Craft et al., 2017
GLP-1 agonist (liraglutide)	12 months	Stabilized glucose metabolism (FDG-PET)	Slowed cognitive decline	Gejl et al., 2016
Sodium selenate (PP2A activator)	3 months	↓ Phospho-tau (CSF)	Trend towards cognitive improvement	Malpas e

 

5.3 Ancient Torpor Machinery in Placental Mammals: Why Humans Lost Behavioral Hibernation but Retained Cellular Toolkit

Humans probably descend from hibernating ancestors. Evidence suggests early mammals could enter torpor or hibernation-like states. Many modern mammals—especially small insectivores like shrews, hedgehogs, and tenrecs—exhibit daily torpor or seasonal hibernation, and are considered evolutionary relics of early mammalian forms. The molecular machinery required for torpor (e.g. tau phosphorylation, synaptic pruning, PP2A modulation, thyroid hormone control) is conserved across nearly all mammals, including humans—even though we don’t behaviorally hibernate. Certain strepsirrhine primates, like the fat-tailed dwarf lemur (Cheirogaleus), do hibernate—sometimes for months at a time—with all the classical features: metabolic suppression, tau phosphorylation, reversible synapse loss. They have “fat tails” because they store fat in their tails.

Humans, though not true hibernators, retain a remarkably conserved cellular and molecular toolkit associated with these states, including the kinase/phosphatase regulatory mechanisms that modulate tau phosphorylation, adenosine-based metabolic control pathways, and potent proteostatic clearance systems (Tupone et al., 2013; Cerri et al., 2013). Understanding these evolutionary genetic trade-offs and the ancient origin of torpor-related cellular pathways not only provides insight into Alzheimer's pathogenesis but also opens new therapeutic avenues focused explicitly on reactivating protective metabolic mechanisms and reversing pathological processes.

6. Therapeutic Opportunities Informed by Hibernation Biology

6.1 Re-fuel Strategies (Ketones, Insulin, GLP-1/GIP, SGLT-2)

If Alzheimer's disease represents an overextended metabolic conservation response, then therapeutic strategies should first aim to restore balanced fuel supply to the brain. As discussed, recent evidence strongly supports the use of ketogenic strategies, intranasal insulin, GLP-1 receptor agonists, and sodium-glucose co-transporter-2 (SGLT-2) inhibitors to re-establish metabolic flexibility in early Alzheimer’s (Cunnane et al., 2016; Craft et al., 2017).

Ketone supplementation, such as medium-chain triglycerides (MCT) or ketone esters, bypass impaired glucose metabolism by providing alternative fuel to neurons, rapidly restoring cognitive functions, synaptic plasticity, and metabolic activity in compromised regions. Intranasal insulin and GLP-1 receptor agonists (e.g., liraglutide, semaglutide) directly counteract the cerebral insulin resistance characteristic of Alzheimer's, re-sensitizing neurons to glucose utilization pathways. Similarly, SGLT-2 inhibitors can modulate systemic glucose homeostasis, potentially optimizing cerebral metabolism.

6.2 Proteostatic Mechanisms that Allow Safe Reversal in Hibernators

One remarkable difference between AD and adaptive states such as hibernation lies in the robust proteostatic mechanisms that permit safe and complete reversal of pathological changes in the latter. Hibernating mammals effectively manage protein aggregates through enhanced protein quality-control systems, including molecular chaperones and autophagic clearance pathways. Cold-shock proteins such as RNA-binding motif protein 3 (RBM3) are significantly upregulated during hibernation and cold stress, directly promoting synaptic integrity, tau dephosphorylation, and neuronal recovery upon rewarming (Peretti et al., 2015).

Moreover, the dramatic arousal-induced reactivation of protein phosphatases such as PP2A rapidly reverses tau phosphorylation, enabling swift restoration of normal neuronal structure and function (Arendt & Stieler, 2003). These protective mechanisms appear to be muted or impaired in the human Alzheimer’s brain, suggesting therapeutic opportunities aimed at restoring these endogenous repair pathways. Namely, understanding that tau hyperphosphorylation, synaptic loss, and metabolic depression are naturally reversed during hibernation arousal offers crucial insights for Alzheimer's therapy, highlighting potential targets to mimic the hibernator’s natural "OFF-switch."

6.2.1 Norepinephrine / LC stimulation

Norepinephrine (NE) released from the locus coeruleus (LC) during arousal plays a pivotal role in rapidly reversing tau phosphorylation, reactivating metabolic processes, and restoring cognitive function in hibernators. Alzheimer's pathology is characterized by early LC degeneration and reduced NE signaling. Pharmacological strategies aimed at boosting LC-NE signaling (e.g., atomoxetine, selective norepinephrine reuptake inhibitors, vagal nerve stimulation) might effectively emulate this natural arousal mechanism, facilitating metabolic and synaptic recovery (Mather & Harley, 2016).

6.2.2 PP2A activators (sodium selenate, SAMe)

Protein phosphatase 2A (PP2A) activation is critical for tau dephosphorylation during the natural arousal phase in hibernators. Clinical trials employing sodium selenate, a PP2A activator, have already demonstrated potential benefits, significantly reducing phosphorylated tau levels and stabilizing cognitive function in early Alzheimer's patients (Malpas et al., 2023). Similarly, S-adenosylmethionine (SAMe), another PP2A-enhancing compound, may represent an additional therapeutic avenue to reverse pathological tau phosphorylation, mirroring the hibernation reversal mechanisms.

6.2.3 Thermogenic and thyroid mimetics

Hibernators exit torpor via a robust metabolic burst involving thyroid hormone release and brown adipose tissue (BAT) thermogenesis. Translating this biological insight, therapeutic strategies incorporating mild thermogenic stimuli (e.g., infrared warming, thyroid hormone analogs) may reactivate dormant neural pathways, reestablish ATP availability, and stimulate neuronal recovery, providing a novel and biologically coherent approach to Alzheimer's treatment (Cannon & Nedergaard, 2004).

6.3 Pulse-torpor Approaches and Synthetic Dormancy Compounds

Given the success of induced torpor in animal models, researchers could explore synthetic dormancy protocols for Alzheimer’s treatment. Such "pulse-torpor" approaches might combine pharmacological induction of temporary metabolic suppression (e.g., via A₁ adenosine receptor agonists) with controlled rewarming and metabolic stimulation phases (Tupone et al., 2013). Short, therapeutic cycles of induced dormancy followed by controlled arousal could facilitate clearance of pathological tau, restoration of neuronal energy states, and synaptic remodeling—effectively replicating the reversible cycle naturally used by hibernators.

6.4 Lifestyle Translations: Intermittent Fasting, Cold-hot Conditioning, HIIT

Lifestyle interventions that replicate elements of ancestral metabolic stress responses may offer practical, non-pharmacological methods to prevent or reverse early Alzheimer's pathology. Intermittent fasting and calorie restriction reliably activate AMPK/mTOR signaling, reduce insulin resistance, and mimic starvation-induced metabolic adaptations that could re-establish neuronal fuel flexibility and resilience (Mattson et al., 2018).

Additionally, temperature-based interventions, such as alternating mild cold and heat exposure ("cold-hot conditioning"), can activate brown adipose tissue, release norepinephrine, stimulate thyroid pathways, and promote neuronal rejuvenation mechanisms observed during natural arousal from torpor. High-intensity interval training (HIIT), similarly, induces systemic metabolic shifts and improves insulin sensitivity, potentially restoring cerebral metabolic balance (Northey et al., 2018).

---

Table 2: Candidate Interventions Mapped to Hibernator Arousal Cascade

Stage of Arousal Cascade	Intervention Strategy	Potential Therapeutic Agents
Metabolic Fuel Restoration	Ketogenic supplementation; Insulin sensitizers; GLP-1 agonists; SGLT-2 inhibitors	MCTs, ketone esters, intranasal insulin, liraglutide, semaglutide
Tau Dephosphorylation	PP2A activation; kinase inhibition	Sodium selenate, SAMe, GSK-3β inhibitors
NE-BAT Metabolic Reactivation	NE enhancement; BAT stimulation; thyroid mimetics	Atomoxetine, selective norepinephrine reuptake inhibitors, thyroid analogs
Synaptic Restoration	Cold-shock proteins; mild hypothermia; thermogenic activation	RBM3 upregulation, cold-hot conditioning, infrared warming

(Abbreviations: MCT, medium-chain triglyceride; GLP-1, glucagon-like peptide-1; SGLT-2, sodium-glucose co-transporter-2; PP2A, protein phosphatase 2A; SAMe, S-adenosylmethionine; BAT, brown adipose tissue; NE, norepinephrine; RBM3, RNA-binding motif protein 3.)

---

These insights from hibernation biology substantially expand our therapeutic toolkit, reframing Alzheimer's not simply as neurodegeneration but as reversible metabolic dormancy. Leveraging this evolutionary wisdom provides unique, promising, and physiologically coherent opportunities for Alzheimer’s intervention, focusing explicitly on metabolic recalibration, neuronal reactivation, and synaptic regeneration.

 

7. Outstanding Questions and Future Research

The hypothesis that Alzheimer's disease represents an ancestral metabolic down-regulation program, pathologically overextended in contemporary humans, opens numerous intriguing avenues for future investigation. Here are four critical research directions needed to further substantiate this theory, advance clinical translation, and bridge disciplinary gaps.

7.1 Quantifying caloric savings of the cortical downshift

A foundational assumption of the metabolic-reduction hypothesis is that selective neuronal downscaling provides meaningful energy savings. While general metabolic shifts have been demonstrated via neuroimaging studies, no precise quantification of the caloric benefits associated with cortical downregulation in Alzheimer’s has been systematically conducted. Future studies should aim to integrate detailed metabolic imaging (FDG-PET, ketone PET, arterial spin labeling MRI) with metabolic chamber experiments and calorimetry methods to measure exactly how much energy the brain conserves during progressive neurodegeneration.

A detailed model quantifying these metabolic savings in relation to real-world caloric demands would allow researchers to rigorously test evolutionary predictions and more accurately assess whether the observed energetic reductions provide a meaningful adaptive advantage. Such quantification could fundamentally strengthen the metabolic-reduction model's explanatory power.

7.2 Identifying biomarkers of reversible vs. irreversible tau states

The concept of Alzheimer’s pathology as a potentially reversible metabolic state raises the critical issue of distinguishing reversible tau phosphorylation states from irreversible tau aggregation. The ability to reliably identify individuals with early-stage, reversible changes would allow timely metabolic interventions before irreversible damage occurs. Current biomarkers (e.g., tau PET, CSF phospho-tau, plasma markers) do not effectively discriminate between reversible, hyperphosphorylated tau and irreversible, aggregated tau deposits.

Thus, future research should focus on developing advanced biomarker panels—combining phospho-tau species, PET imaging tracers sensitive to aggregated versus soluble tau states, and markers of synaptic integrity (e.g., neurogranin, SNAP-25)—to reliably differentiate reversible tau phosphorylation associated with metabolic states from irreversible pathological aggregation. Such biomarkers would be indispensable for optimizing intervention timing and patient selection for metabolic-based therapies.

7.3 Safety and feasibility of torpor-mimetic therapies in humans

Experimental evidence in animal models demonstrates that controlled synthetic torpor or transient hypothermia can safely induce tau hyperphosphorylation and subsequent de-phosphorylation upon arousal. Translating these concepts safely to humans, however, remains largely unexplored. Crucial safety questions include the tolerability of mild therapeutic hypothermia or metabolic suppression, potential cardiovascular and immune implications, and the reversibility of induced metabolic downshifts in older adults.

Preliminary feasibility studies, beginning with controlled, short-duration torpor-mimicking protocols in healthy volunteers, are necessary to evaluate these therapies. Early-phase clinical trials should incorporate rigorous metabolic monitoring, cardiovascular and cognitive assessments, as well as detailed biomarker analyses, to comprehensively establish safety profiles. Demonstrating human feasibility would mark a critical step toward using evolutionary-informed metabolic recalibration strategies to halt or reverse Alzheimer's pathology.

7.4 Cross-disciplinary collaborations: neurology × comparative physiology × anthropology

The integrative nature of the metabolic-reduction hypothesis inherently calls for interdisciplinary collaboration across traditionally isolated research fields. Effective exploration and translation of this hypothesis require that neurologists, comparative physiologists, evolutionary anthropologists, and metabolic specialists come together to bridge knowledge gaps. Neurologists can offer expertise in clinical disease processes and therapeutic trial design. Comparative physiologists contribute essential insights regarding reversible tau biology and metabolic regulation from animal models. Anthropologists, comparative physiologists and neuroecologists provide data and context concerning evolutionary life-history traits, aging patterns, and metabolic adaptations.

Fostering these collaborative networks would allow for rich cross-fertilization of ideas, methods, and perspectives—accelerating discovery and facilitating the development of novel therapeutic paradigms grounded in evolutionary biology. Strategic funding and institutional support for multidisciplinary research initiatives could significantly accelerate this promising avenue toward a comprehensive understanding of Alzheimer's disease. By addressing these outstanding questions, researchers will substantially clarify the role of metabolic conservation in Alzheimer's pathogenesis, determine its therapeutic potential, and potentially reshape our fundamental approach to dementia prevention and treatment.

7.5 Implication for other Neurodegenerative Diseases

If we view Alzheimer’s disease as the pathological prolongation of an ancient energy-conservation program, this framework could potentially be extended to help explain other neurodegenerative diseases. Each of these conditions appears to selectively affect brain regions or circuits with exceptionally high energy demands, suggesting that they too may reflect maladaptive activation of conserved “thrift” responses to metabolic stress.

In vascular dementia (VaD), the cortical and subcortical regions most affected—such as the watershed cortex and deep white matter—are those most vulnerable to chronic hypoperfusion. Rather than sudden infarction, these areas may enter a state of prolonged fuel shortage that drives them into a hypometabolic, low-functioning state. This pattern is strikingly similar to the early hypometabolism seen in Alzheimer’s, but here triggered by vascular insufficiency. From this perspective, VaD may represent a vascularly-induced thrift program, where circuits enter dormancy to survive low perfusion. The implication is that therapeutic strategies aimed at metabolic re-fueling—such as ketones, improved blood flow, or insulin-sensitizing agents—might revive partially idling circuits before they undergo irreversible degeneration.

In dementia with Lewy bodies (DLB), pathology centers on regions like the posterior cingulate and visual association cortex, and includes disruptions in the cholinergic brainstem system. Alpha-synuclein, the principal aggregating protein in DLB, is known to inhibit synaptic vesicle recycling—an energy-intensive process. This suggests that alpha-synuclein may function as a synaptic “brake,” reducing the energy cost of constant synaptic activity. Behaviorally, DLB includes vivid hallucinations and REM sleep disturbances, echoing the dream-rich, low-arousal physiology of torpor. If DLB represents a form of partial, dysregulated metabolic suppression, then interventions that restore mitochondrial throughput or mimic arousal neuromodulators such as norepinephrine and acetylcholine could help restore function even without fully removing alpha-synuclein aggregates.

Frontotemporal dementia (FTD) presents a somewhat different profile, targeting anterior brain regions such as the anterior cingulate, orbitofrontal cortex, and anterior temporal lobes. These areas support social cognition, emotional regulation, and executive function—all metabolically expensive and reliant on spontaneous firing and integration. FTD may represent an alternative “special-purpose” thrift response, where under conditions of chronic stress or energetic crisis, the brain selectively sheds high-cost social-cognitive circuits rather than memory-related ones. The behavioral symptoms—disinhibition, apathy, and loss of empathy—may mirror what happens in animal models when frontal circuits are pruned to conserve energy. If so, targeted metabolic interventions that support prefrontal circuits could slow early FTD progression.

Parkinson’s disease (PD) affects dopaminergic neurons in the substantia nigra, which are uniquely vulnerable because they engage in continuous autonomous pacemaking and calcium signaling—processes that impose extreme mitochondrial stress. The presence of alpha-synuclein aggregates and mitochondrial downregulation in these cells may reflect a cell-autonomous attempt to reduce energy consumption. However, this thrift response ultimately leads to insufficient ATP production and cellular degeneration. If PD represents a chronic misfiring of this protective strategy, then therapeutic efforts could focus on supporting ATP generation through ketones or NAD⁺ precursors, buffering calcium loads, or introducing structured “rest periods” for pacemaking neurons.

Amyotrophic lateral sclerosis (ALS) primarily affects motor neurons with extremely long axons, which must sustain high-speed axonal transport and constant energy delivery across distances greater than one meter. These neurons are profoundly dependent on ribosomal translation and mitochondrial transport. In ALS, TDP-43 pathology suppresses protein synthesis and ribosome biogenesis—an energy-saving adaptation that may become catastrophic over time as motor axons run out of supplies. From this standpoint, ALS reflects an attempt to downscale energetically unsustainable axonal function. Restoring mitochondrial trafficking and supplying alternative fuels like ketones directly to axons may offer therapeutic benefit by converting maladaptive chronic thrift into a manageable, potentially reversible state.

Across these disorders, certain unifying predictions emerge. Early hypometabolism likely precedes protein aggregation in each case, suggesting that interventions targeting metabolism could delay or prevent structural damage. Neurons in affected circuits may respond positively to alternative fuels like β-hydroxybutyrate or lactate, even if aggregates are still present. The role of neuromodulators such as adenosine, norepinephrine, and thyroid hormones appears central in regulating transitions between reversible dormancy and irreversible degeneration. Therapeutic strategies that mimic or modulate these arousal signals, or that support cyclic metabolic recalibration rather than linear decline, may prove useful across Alzheimer’s, vascular dementia, Lewy body dementia, FTD, Parkinson’s, and ALS.

Ultimately, this energy-based framing shifts the focus away from protein removal and toward restoring the brain’s ability to enter and exit metabolic conservation states in a controlled manner. These diseases may not reflect total failure, but rather ancient neural responses to stress—programs meant to protect and preserve neurons under energetic duress, now running open-loop in modern human lifespans. The challenge and opportunity lie in understanding how to guide these circuits out of chronic thrift and back toward functional equilibrium.

Alzheimer's Comparison with other Neurodegenerative forms of Dementia:

8. Conclusion: Waking the Brain from an Over-Extended Thrift Mode

Integrating data across multiple biological scales—ranging from human neuroimaging studies to comparative animal physiology, molecular signaling pathways, and evolutionary genetics—we find compelling support for this metabolic-reduction perspective. Human imaging studies consistently demonstrate early cortical glucose hypometabolism preceding clinical symptoms and protein deposition. Comparative biology further reveals striking parallels between Alzheimer’s neuropathology and the fully reversible brain changes observed during torpor, hibernation, and acute starvation in numerous mammalian species. Critically, these adaptive states deploy shared molecular pathways, involving reversible tau phosphorylation, selective synaptic pruning, amyloid accumulation, and controlled metabolic downshifts—offering a conceptual and practical blueprint for therapeutic intervention.

The central shift advocated by this evolutionary-informed perspective is thus one of therapeutic strategy. Rather than merely attempting to clear amyloid plaques or tangles after the fact, a promising alternative involves recalibrating cerebral metabolism and reactivating the brain’s intrinsic metabolic "wake-up" pathways. Leveraging hibernation biology, synthetic torpor induction, ketone-based metabolic interventions, and insulin-sensitizing therapies represent a new generation of Alzheimer’s treatments, designed not just to slow or stabilize progression, but potentially reverse pathology by restoring dormant neural circuits.

This reframing necessitates collaborative research programs bridging neurology, comparative physiology, molecular neuroscience, evolutionary anthropology, and metabolic medicine. Addressing critical questions—such as quantifying energy savings from neural downshifts, distinguishing reversible from irreversible tau states, and safely translating torpor-mimetic therapies to humans—holds promise to transform our clinical approach. Indeed, viewing Alzheimer's through the evolutionary lens of energy economics not only advances our fundamental understanding but presents immediate opportunities for novel, targeted, and potentially transformative interventions aimed at waking the aging brain from its over-extended thrift mode.

 

Conclusory Table: New evidence—beyond the 2009 paper—for framing Alzheimer’s disease (AD) as an over‑extended, brain‑energy‑reduction program

Evidence pillar	New highlights from our 2024‑25 review	Why it strengthens the “metabolic‑program” model
1  Early, fuel‑specific hypometabolism precedes plaques/tangles	Longitudinal FDG‑PET shows cortex‑wide glucose use falling years before tau or amyloid PET turn positive and before cognitive decline. Hypometabolism predicts who will accumulate tau next.	Positions reduced energy use as a trigger, not a by‑product, of pathology.
2  Restoring fuel flow improves cognition	6‑month MCT (ketogenic) BENEFIC trial → ↑ ketone uptake in attention network, ↑ functional connectivity, better memory in MCI ; 15‑month open MCT study stabilised MMSE in 80 % of mild‑AD participants.	Shows that supplying alternative “torpor fuel” reverses part of the deficit—exactly what would help a brain stuck in a thrift mode.
3  Intranasal or sensitiser insulin boosts brain metabolism	Multiple RCTs of intranasal insulin report short‑term memory gains and higher regional FDG uptake in APOE4 carriers.	Treats the insulin‑resistant, low‑glucose state as a reversible throttle—matching a programmed reduction.
4  Reversible AD‑type tau in hibernators and “synthetic torpor”	Ground‑squirrels, hamsters, bears, and pharmacologically‑torpid rats show massive tau phosphorylation, synapse loss, hypometabolism during torpor that clears within hours of arousal via PP2A and NE bursts.	Direct demonstration that the molecular signature of AD is a normal, energy‑saving state that can be switched off.
5  Cross‑species occurrence tracks longevity & metabolic stress	Spontaneous amyloid + tau pathology verified in aged dolphins, chimpanzees, dogs, degus—all long‑lived or prone to insulin resistance; naked‑mole‑rats accumulate soluble Aβ but never aggregate or degenerate.	Shows the program is ancient, turns pathological only under certain life‑history or metabolic contexts; also proves plaques/tangles can be benign or reversible.
6  Selective “hit‑map” matches energy triage	Human and animal data confirm association‑cortex/hippocampus (highest idle glucose burn) are first to decline, primary sensory–motor areas are last.	Aligns perfectly with a strategy that trims the most energetically costly yet least survival‑critical tissue.
7  Neuromodulator OFF‑switch circuitry is already perturbed in prodromal AD	A₂A‑adenosine receptor density is up‑regulated, locus‑coeruleus NE neurons are lost, orexin and thyroid signals are dysregulated before dementia.	Matches the same levers that flip torpor on/off—suggesting AD brains are stuck mid‑cycle.
8  Thrifty genotypes and metabolic syndrome overlap	APOE‑ε4 and other famine‑advantage alleles raise AD risk but also correlate with enhanced fertility or early‑life survival under infection/malnutrition.	Confirms the trade‑off logic: genes that helped conserve energy under scarcity now predispose to late‑life thrift overshoot.
9  Calorie restriction & intermittent fasting blunt pathology	CR / fasting‑mimicking diets in AD mice lower Aβ load, reduce P‑tau, improve cognition—paralleling natural torpor benefits.	Manipulating the same metabolic pathways can push the program back toward adaptive territory.
10  Synthetic torpor rescues AD model mice	Single pharmacologically‑induced torpor cycle (A₁‑adenosine agonist) restored LTP and memory in APP/PS1 mice while clearing P‑tau.	Proof‑of‑concept that intentionally cycling into—and out of—thrift mode can repair diseased circuitry.