Sage — Metabolic Hypothesis of Fuel
This pillar sits on the Shore: Living Water.
The eukaryotic cell and the human organism it constitutes are not free-floating metabolic engines. They are nested symbionts whose energy economy depends on the continuous delivery of substrates that mitochondria can oxidize cleanly and that the gut biota can transform into signaling molecules and short-chain fatty acids. The eukaryotic cell itself arose through an ancient endosymbiosis in which a host cell of Asgard archaeal lineage formed a stable partnership with an alphaproteobacterium that became the mitochondrion. That inner bacterial inheritance still powers oxidative phosphorylation and retains its own circular genome, membrane architecture, and metabolic preferences. At a larger scale, the trillions of bacteria and archaea that colonize the gastrointestinal tract function as an ecological symbiont—an external metabolic organ whose collective genome (the microbiome) encodes the majority of the enzymes required to unlock complex plant polysaccharides, produces the short-chain fatty acids that fuel colonocytes, and generates the epigenetic and signaling molecules that calibrate host immune tone, barrier integrity, and metabolic set-points. Together these two nested bacterial partnerships—the permanent intracellular mitochondrion and the ongoing gut community—constitute the primary architecture of human energy extraction, ordered complexity, and adaptive resilience. Fuel is the substrate that either respects or assaults this architecture: the membrane composition and oxidative preferences of the mitochondrial endosymbiont, the fermentative capacity of the colonic consortium, and the circadian timing that allows both partnerships to recover and realign each night.
The hypothesis is straightforward. Animal-derived fats and proteins constitute the proper primary fuel for the human mitochondrial system because they align with the evolutionary substrate preferences of the alphaproteobacterial ancestor, generate ketones and amino-acid-derived intermediates with low reactive oxygen species burden when oxidized under physiological conditions, and supply the cofactors and structural lipids that maintain mitochondrial membrane integrity. Fermented microbiota-accessible carbohydrates (MACs) and mushrooms constitute the ideal fuel for the gut biota because fermentation both depletes rapidly absorbable simple sugars that would otherwise bypass or overwhelm the partnership and generates the precise metabolites—butyrate, propionate, lactate, bioactive peptides, and fungal polysaccharides—that feed keystone species, reinforce the barrier, and calibrate immune tone. Simple carbohydrates, industrially processed seed oils, and alcohol, by contrast, impose direct mitochondrial toxicity and force compensatory systems into chronic overdrive, fraying both the inner and outer symbioses. Finally, the timing of fuel delivery modulates the cortisol–circadian axis: appropriately timed meals reinforce the daily rhythm of mitochondrial biogenesis, autophagy, and biota recovery; late or continuous feeding disrupts that rhythm and converts food from partner into chronic stressor.
This is not a rejection of plant foods or an absolutist dietary dogma. It is a metabolic hierarchy grounded in the nested bacterial inheritances that made complex life possible and that continue to power and regulate human physiology. The frugal implication is that the highest-leverage interventions are also the most ancestral and accessible: prioritize the substrates mitochondria prefer, feed the biota with pre-digested and fungal partners, avoid the modern toxins that force compensatory cascades, and deliver fuel within the temporal windows that allow both symbionts to rest and realign.
Mitochondrial Substrate Preference: Why Animal Fats and Proteins Are Proper Human Fuel
Mitochondria are not generic energy factories. They retain the metabolic signature of their alphaproteobacterial ancestry. That ancestry specialized in the oxidation of fatty acids and the entry of amino-acid-derived carbon skeletons into the tricarboxylic acid cycle. The eukaryotic host acquired, through the endosymbiotic event, an organelle optimized for beta-oxidation, ketogenesis under low-glucose conditions, and the efficient generation of ATP with relatively controlled electron leak when the proper substrates are supplied in the proper redox context.
Animal fats—particularly the saturated and monounsaturated species abundant in ruminant fat, egg yolk, and certain marine sources, together with cholesterol and the phospholipids of animal membranes—align with this architecture in several reinforcing ways. First, long-chain saturated and monounsaturated fatty acids undergo orderly beta-oxidation, generating acetyl-CoA that feeds the TCA cycle without the obligatory generation of the high NADH/NAD+ ratios that accompany rapid glycolysis. Second, under conditions of moderate carbohydrate restriction or intermittent fasting, hepatic and, to a lesser extent, renal ketogenesis converts excess acetyl-CoA into beta-hydroxybutyrate and acetoacetate. These ketone bodies are not merely alternative fuels; they are preferred substrates for the brain, heart, and skeletal muscle under many physiological states, and they exert signaling effects that include histone deacetylase inhibition, mitochondrial biogenesis via PGC-1α pathways, and reduced inflammatory tone. Third, the structural lipids of animal origin—cholesterol, sphingomyelin, and the relatively saturated phospholipid profiles of terrestrial animal membranes—support the packing and function of mitochondrial inner membranes. Cardiolipin, the signature phospholipid of the inner mitochondrial membrane, requires specific fatty-acyl composition for optimal respiratory-chain supercomplex assembly; chronic displacement by highly unsaturated industrial oils impairs this architecture.
Animal proteins supply the complementary half of the mitochondrial fuel equation. The amino acids released by digestion enter multiple pathways that mitochondria are equipped to handle. Glucogenic amino acids can be converted to TCA intermediates in a controlled, demand-driven manner rather than as a continuous flood of glucose. Branched-chain amino acids support muscle protein synthesis and, via their metabolites, influence mTOR and mitochondrial quality control. Creatine, abundant in animal muscle, is phosphorylated to phosphocreatine, buffering ATP and supporting the high-energy phosphate transfer that mitochondria ultimately recharge. Carnitine, likewise concentrated in animal tissues, is the obligate shuttle for long-chain fatty acids into the mitochondrial matrix; endogenous synthesis is possible but limited, making dietary supply a practical advantage. Taurine, glycine, and other amino acids participate in conjugation reactions, antioxidant defense, and membrane stabilization. In short, animal protein is not merely “building material.” It is a metabolic toolkit that mitochondria and the host genome evolved to utilize in parallel with the fat-derived energy stream.
This preference is visible in human physiology. The brain’s capacity to run on ketones during fasting or carbohydrate restriction is not a pathological adaptation; it is a conserved feature of a species whose ancestral diets were intermittent and often animal-dominant for extended periods. The high mitochondrial density of cardiac myocytes, retinal photoreceptors, and type I skeletal muscle fibers is matched by a preference for fatty-acid and ketone oxidation under ordinary conditions. Even the liver, which can perform de novo lipogenesis from carbohydrate, does so at a metabolic cost and under hormonal conditions that are frequently pathological in the modern context. The evolutionary default, visible in the comparative physiology of hunter-gatherer and pastoralist populations and in the substrate preferences of isolated mitochondria, is that animal fat and protein constitute clean, high-yield fuel when supplied in amounts matched to demand and when the redox and oxygen environments of the cell are not already compromised.
None of this implies that carbohydrate is unnecessary or that plant foods are optional. It does imply a hierarchy. Glucose is a rapid, high-flux fuel best reserved for acute demand or for tissues with limited fatty-acid oxidative capacity under specific conditions. Chronic elevation of glucose and insulin, by contrast, shifts the entire system toward compensatory states—elevated de novo lipogenesis, suppressed ketogenesis, increased mitochondrial ROS generation at complexes I and III, and eventual downregulation of the very oxidative machinery that animal substrates would otherwise support. The proper fuel is therefore the one that keeps mitochondria in their preferred operating mode: high capacity for beta-oxidation and TCA flux, controlled ROS, intact membrane lipids, and the metabolic flexibility to switch into ketosis when exogenous carbohydrate is low.
The Practical Difficulties of Matching These Functions on an Exclusively Plant-Based Diet
It is possible, with careful planning, fortification, and supplementation, to construct an exclusively plant-based diet that meets basic caloric and macronutrient targets. Matching the specific functional profile of animal fats and proteins for mitochondrial and systemic metabolism, however, is substantially more difficult and often incomplete. The challenges are not ideological; they are biochemical and ecological.
Animal proteins supply all essential amino acids in ratios that closely match human tissue requirements, with high digestibility and a leucine content sufficient to robustly trigger muscle protein synthesis. Most individual plant proteins are incomplete or limited in one or more essential amino acids (commonly lysine, methionine, or leucine). Combining complementary plant sources can approximate a complete profile, yet the overall digestibility and leucine density typically remain lower, requiring larger total protein intakes and careful meal composition to achieve comparable anabolic signaling. Several mitochondria-relevant compounds are either absent or present only in negligible amounts in plants: creatine is essentially exclusive to animal muscle; carnitine is concentrated in animal tissues and synthesized endogenously only in limited quantities; taurine is likewise scarce in the plant kingdom. These are not optional additives for many high-demand tissues; they participate directly in high-energy phosphate buffering, fatty-acid transport into mitochondria, and membrane and antioxidant defense.
Healthy lipid matching is equally constrained. Cholesterol, a structural component of animal membranes and a precursor for steroid hormones and bile acids, is absent from plants. The long-chain saturated and monounsaturated fatty-acid profiles characteristic of ruminant fat and egg yolk support mitochondrial membrane packing and cardiolipin composition more readily than the highly polyunsaturated profiles of many seed and nut oils. Conversion of plant-derived alpha-linolenic acid (ALA) to the long-chain omega-3s EPA and DHA is inefficient in humans, often below 5–10 percent and further reduced by high linoleic-acid intake. Consequently, exclusive plant diets frequently produce lower circulating and tissue EPA/DHA unless algae-derived supplements are added. Many common plant oils simultaneously deliver large quantities of linoleic acid, the same omega-6 PUFA whose excess incorporation into membranes promotes peroxidation and the generation of reactive aldehydes already described.
Certain vitamins and related cofactors present further gaps. Preformed vitamin A (retinol) is found in animal foods; plant provitamin A carotenoids require conversion whose efficiency varies widely among individuals. Vitamin B12 is produced by bacteria and concentrated in animal tissues; reliable plant sources do not exist without fortification or supplementation. Vitamin K2 (menaquinones), particularly the longer-chain forms, is far more abundant in animal foods and certain fermented products than in unfermented plants. Choline, essential for membrane phospholipids and one-carbon metabolism, is more concentrated and bioavailable in egg yolks and liver than in most plant foods. Heme iron is absorbed more efficiently than non-heme plant iron and is less affected by phytate and polyphenol inhibitors.
None of these difficulties render an exclusively plant-based diet impossible. Individuals who meticulously combine complementary proteins, emphasize lower-linoleic plant fats, consume fermented foods, and supplement strategically (B12, algae DHA/EPA, creatine, and often vitamin D, K2, and choline) can achieve nutritional adequacy and, in some cases, excellent health markers. The metabolic and practical cost, however, is higher: greater attention to food combination, higher total food volume for equivalent protein density, dependence on industrial fortification or isolated supplements for several critical nutrients, and residual gaps in compounds such as creatine and carnitine that the body must then synthesize under greater biosynthetic demand. From the standpoint of the nested symbionts, the animal-derived package of complete protein, preferred fatty-acid species, cholesterol, creatine, carnitine, and associated vitamins arrives pre-assembled in forms that mitochondria and host tissues are evolved to utilize directly. Replicating that package from plants alone is an engineering problem that can be solved, but it is not the default solution the holobiont was designed to run on.
Fermented MACs and Mushrooms: Ideal Fuel for the Gut Biota
The outer symbiont—the gut community—has its own substrate preferences, and those preferences are not identical to the host’s. Humans possess roughly 20,000 protein-coding genes; the collective microbiome encodes an estimated 3 to 10 million. The functional consequence is that the hologenome—the combined genetic repertoire of host plus microbes—vastly expands metabolic capacity. The human genome encodes only a handful of enzymes capable of cleaving the diverse glycosidic bonds in plant cell-wall polysaccharides and resistant starches. Gut bacteria, especially Bacteroidetes and certain Firmicutes, deploy dozens to hundreds of carbohydrate-active enzymes per species, organized in polysaccharide-utilization loci. These enzymes release mono- and oligosaccharides that feed both the microbes and the host via short-chain fatty acids (SCFAs). Without this microbial genetic repertoire, the caloric yield from a diverse plant-based diet would collapse, and the signaling molecules derived from fiber fermentation would disappear.
Butyrate, the most consequential of these SCFAs for epithelial health, is the preferred energy substrate for colonocytes, which derive approximately 60–70 percent of their ATP from its beta-oxidation rather than from glucose or glutamine. Butyrate maintains the hypoxic environment of the crypt base that favors stem-cell quiescence and limits expansion of facultative anaerobes and pathobionts. It upregulates tight-junction proteins (ZO-1, occludin, claudins), stimulates goblet-cell differentiation and mucus production, induces antimicrobial peptides, and promotes differentiation of regulatory T cells in the lamina propria. When butyrate production falls, colonocytes shift toward less-efficient glycolysis, the barrier loosens, mucus thins, and bacterial products such as lipopolysaccharide can translocate, driving metabolic endotoxemia and low-grade systemic inflammation. Keystone butyrate producers such as Faecalibacterium prausnitzii, Roseburia spp., and Eubacterium rectale, together with mucin-specialists such as Akkermansia muciniphila, therefore exert outsized influence on barrier integrity and immune tone. The ideal fuel for this community is therefore not the simple sugars that the host can absorb in the small intestine, nor the refined starches that flood the proximal gut, but the complex, partially pre-digested, and often fungal substrates that reach the colon in forms that keystone mutualists can process into beneficial outputs while minimizing the expansion of pathobionts.
Microbiota-accessible carbohydrates—resistant starches, non-starch polysaccharides, oligosaccharides, and certain fibers—are the primary feedstock. When these MACs arrive intact, Bacteroidetes and specific Firmicutes deploy their polysaccharide-utilization loci, release oligosaccharides, and ferment them to acetate, propionate, and butyrate. The problem with modern diets is not the presence of plant carbohydrates per se but their form and context. Highly refined flours and sugars are absorbed before they can serve the biota; the residual fraction that reaches the colon is often insufficient in diversity and quantity to sustain the full functional repertoire of a resilient community.
Fermentation solves this problem at the source. Traditional lactic-acid fermentation of vegetables and dairy, as well as the fungal fermentations that produce tempeh, miso, and certain mushroom preparations, accomplishes several simultaneous metabolic improvements. First, the microbes of the fermentation process consume a substantial fraction of the simple sugars and rapidly fermentable oligosaccharides that would otherwise spike host glycemia or feed opportunistic organisms in the small intestine. Lactose in milk is converted to lactic acid; the residual lactose content of properly fermented yogurt or kefir is markedly reduced. Glucose, fructose, and sucrose in raw cabbage or other vegetables are largely depleted during sauerkraut or kimchi production, leaving behind organic acids, bacteriocins, and a matrix of residual complex carbohydrates and polyphenols. Second, fermentation generates additional bioactive compounds—short-chain organic acids, bioactive peptides, B vitamins, and in some cases vitamin K2—that further support both host and microbial metabolism. Third, the live organisms or their postbiotic residues can help re-seed or functionally support a depleted gut community, although the primary benefit is the pre-digestion and metabolic transformation of the substrate itself.
Mushrooms occupy a special niche within this framework. As the fruiting bodies of fungi, they supply beta-glucans, chitin, and a suite of complex polysaccharides that are poorly digested by human enzymes yet fermentable by specific bacterial taxa. These fungal fibers act as selective prebiotics, often favoring butyrate producers and organisms associated with barrier integrity. Beyond substrate, mushrooms deliver immunomodulatory compounds that interact with pattern-recognition receptors in ways that can promote tolerance rather than inflammation when the overall microbial ecology is healthy. In the context of the holobiont, mushrooms are neither pure host fuel nor pure microbial fuel; they are a bridging substrate that the outer symbiont processes into signals the host immune system has co-evolved to interpret.
The net effect of prioritizing fermented MACs and mushrooms is that the gut biota receives the substrates it is equipped to handle, produces the metabolites the host epithelium and immune system require, and is less frequently confronted with the simple sugars and additives that favor pathobiont expansion. This is the ecological counterpart of supplying mitochondria with animal fat and protein: each symbiont is given the fuel that matches its evolutionary design, and the partnership as a whole remains coherent.
Simple Carbohydrates: Direct Mitochondrial Stress and Compensatory Cascades
Seed Oils: Membrane Incorporation, Peroxidation, and Electron-Transport Impairment
Alcohol: Redox Imbalance, Acetaldehyde Toxicity, and Multi-Level Compensation
Fermentation as Pre-Digestion: Depleting Simple Carbohydrates in Dairy and Vegetables
The practical corollary of the above toxicities is that the form in which plant and dairy carbohydrates are consumed matters as much as their presence. Raw milk contains lactose, a disaccharide that in lactose-tolerant adults is cleaved to glucose and galactose and absorbed, but that in many individuals reaches the colon incompletely and can feed less desirable organisms or produce osmotic symptoms. Traditional fermentation by lactic-acid bacteria converts a large fraction of that lactose into lactic acid and other organic acids, simultaneously lowering the simple-sugar load, generating bioactive peptides and B vitamins, and introducing or supporting organisms that can colonize or influence the gut. The resulting yogurt, kefir, or cheese is no longer a simple-carbohydrate delivery vehicle; it is a transformed matrix whose residual sugars are minimized and whose microbial and metabolic outputs support the partnership.
The same principle applies to vegetables. Fresh cabbage, carrots, cucumbers, and other produce contain free sugars and rapidly fermentable oligosaccharides. When these are subjected to lactic-acid fermentation under anaerobic conditions, the resident or inoculated bacteria consume the simple sugars preferentially, producing lactic and acetic acids that lower pH, inhibit pathogens, and leave behind a residual fiber matrix enriched in organic acids, polyphenol metabolites, and, in many cases, live cultures. The MAC profile is altered but not eliminated; the critical change is the depletion of the rapid-sugar fraction that would otherwise challenge small-intestinal handling or favor pathobionts. The fermented product therefore arrives in the colon already partially processed, with a lower risk of generating the gas, osmotic load, or ecological imbalance associated with large doses of unfermented simple carbohydrates, while still supplying the complex substrates that butyrate producers require.
This is why fermented MACs occupy a privileged position in the fuel hierarchy. They are not merely “probiotic foods.” They are pre-digested, sugar-depleted, metabolite-enriched substrates that respect both the host’s limited capacity for rapid carbohydrate disposal and the biota’s need for complex fermentable material. The same logic extends, with appropriate caveats, to certain fungal fermentations and to the inclusion of whole mushrooms whose structural polysaccharides serve analogous prebiotic roles.
Food Timing, Intermittent Fasting, and the Four-Hour Pre-Sleep Window
Fuel is not only chemical; it is temporal. The cortisol–circadian system and the mitochondrial and microbial systems it coordinates operate on a roughly 24-hour cycle. Cortisol rises in the early morning, mobilizing glucose, supporting alertness, and preparing the organism for daytime activity and feeding. Insulin sensitivity is correspondingly higher in the earlier part of the day. As the day progresses, the system shifts toward repair, growth-hormone release, and the nocturnal processes of autophagy, mitophagy, and microbial community stabilization that occur most effectively in the absence of continuous nutrient influx.
Late eating—particularly within the four hours before habitual bedtime—disrupts this architecture on multiple levels. Gastric emptying and intestinal absorption continue into the night, elevating nocturnal glucose and insulin at a time when the system is primed for fasting physiology. The resulting metabolic noise impairs the depth of the cortisol decline that should accompany sleep onset, fragments sleep architecture, and reduces the window for growth-hormone-mediated repair. Mitochondrial quality-control processes that are upregulated during fasting and sleep are blunted. In the gut, continuous substrate availability prevents the periodic reductions in luminal load that allow butyrate producers competitive advantage and that give the epithelium time to restore barrier integrity under lower inflammatory pressure. The cumulative effect is a flattening of the circadian metabolic amplitude: less robust daytime insulin sensitivity, less effective nocturnal recovery, and a gradual upward drift in inflammatory and adiposity set-points.
Intermittent fasting, or more precisely time-restricted eating within a consistent daily window that ends at least four hours before sleep, restores the temporal structure. A feeding window of eight to twelve hours, aligned with daylight and ending in the late afternoon or early evening, allows a prolonged overnight fast that promotes ketone production, autophagy, and microbial recovery without requiring extreme caloric restriction. The avoidance of food in the pre-sleep window is not an arbitrary rule; it is the practical expression of allowing the cortisol decline, the growth-hormone pulse, and the mitochondrial and epithelial repair programs to proceed without nutrient interference. When the food that is consumed within the daytime window is itself composed of animal fats and proteins plus fermented MACs and mushrooms, the chemical and temporal signals reinforce one another: clean mitochondrial substrates by day, biota-supporting fermentation products, and a clear fasting interval by night.
Food itself modulates the cortisol–circadian axis. Protein- and fat-dominant meals produce relatively stable postprandial responses and do not provoke the sharp glucose–insulin spikes that can secondarily stress the hypothalamic–pituitary–adrenal axis. Large simple-carbohydrate loads, by contrast, can produce reactive hypoglycemia and subsequent cortisol counter-regulation, adding noise to the daily rhythm. Alcohol further disrupts sleep architecture and nocturnal metabolic recovery. Thus the same substrates that are toxic or beneficial at the mitochondrial and microbial levels also shape the endocrine timing system that coordinates those levels across the day.
Implications and Frugal Restoration Pathways
If the nested symbionts of the human organism—mitochondria within cells and the gut biota within the gut—have distinct yet complementary substrate preferences, and if modern dietary patterns systematically violate those preferences while also violating the temporal windows required for recovery, then the chronic-disease epidemic is not a mysterious collection of independent failures. It is the predictable consequence of feeding the wrong fuels at the wrong times to a partnership that evolved under very different conditions.
The restorative pathway is correspondingly clear and frugal. Prioritize animal fats and proteins as the primary energy and structural substrates for the mitochondrial system. Include fermented vegetables, fermented dairy where tolerated, and mushrooms as the primary carbohydrate and prebiotic inputs for the gut biota, thereby depleting simple sugars at the source and supplying the metabolites that maintain barrier and immune coherence. Minimize or eliminate industrially processed seed oils, minimize refined simple carbohydrates, and treat alcohol as the mitochondrial and barrier toxin it is. Structure eating within a consistent daytime window that ends at least four hours before sleep, allowing the cortisol decline, the overnight fast, and the associated mitochondrial and microbial recovery processes to occur without interference.
These practices do not require exotic supplements, continuous monitoring devices, or expensive proprietary formulations as first-line measures. They require the consistent application of ancestral substrate hierarchies and circadian timing—resources that remain available at ordinary grocery and market costs. Targeted adjuncts (creatine, specific mitochondrial cofactors, or carefully chosen fermented products) may become rational once the foundational pattern is established and individual response can be observed. The testable prediction is that individuals and populations who adopt this fuel hierarchy and timing will show measurable improvements in markers of mitochondrial function, fecal SCFA profiles, barrier integrity, insulin sensitivity, inflammatory tone, and the multi-system outcomes currently labeled as separate chronic diseases.
The primary partnership of mitochondrial endosymbiont and gut biota is the upstream determinant of metabolic coherence. That partnership can only function optimally when it is supplied with the substrates and the temporal structure it evolved to handle. Animal fats and proteins power the inner symbiont cleanly. Fermented MACs and mushrooms feed the outer symbiont productively. Simple carbohydrates, seed oils, and alcohol impose direct toxicity and force compensatory cascades that fray both partnerships. Food timing either reinforces or dismantles the circadian coordination that allows nightly recovery. Honor the fuel preferences and the temporal windows of the nested bacterial inheritances that constitute the human organism, and the downstream expressions of metabolic disease become less inevitable. Treat fuel as optional or as a continuous, poorly timed stream of industrial substrates, and the compensatory systems will continue to convert an upstream mismatch into the chronic conditions that dominate modern morbidity.
The partnership remains available. The proper fuels remain available. The daily rhythm remains available. What is required is the consistent decision to stop treating the ancient metabolic architecture as infinitely plastic and to begin feeding it according to the design that evolution and endosymbiosis actually produced.
A working hypothesis, open to correction and revised as new evidence arrives. Last revised October 17, 2026.
Related Strength tool (not Fuel): Sage — Methylene Blue · Coach · Surf Shop.
See the notice at the top of this page. Working hypothesis, not medical advice.
Related: HydroEye on Surf Shop (sold through MDVite, EnSurf LLC’s store; not manufactured by EnSurf). More on HydroEye: Sage · Surfer.
Related: Sage — A Metabolic Hypothesis of Ocular Surface Disease (Wipeouts · Ophthalmology).