The Nested Causal Architecture
of Human Physical Reconfiguration

§1Introduction: A Fragmented Problem

Over the past three decades, multiple disciplines have independently documented historical declines in human physical parameters. Bioarchaeologists using cross-sectional bone geometry analysis have found that Late Upper Paleolithic Europeans possessed fibular rigidity comparable to that of modern professional ice-hockey players[1], and that the transition from hunting-gathering to agriculture was accompanied by a sustained decline in lower-limb strength and mobility[2]. Paleoanthropologists have documented a 9.9%–17.4% decline in cranial capacity from the Mesolithic peak to the present[3]. Nutritional anthropologists have tracked systemic health deterioration following the replacement of diverse wild game with cereal grains[4,5]. Evolutionary medicine researchers have proposed the “mismatch disease” framework to explain the osteoporosis and muscle atrophy arising from sedentary lifestyles[6].

Yet these findings remain scattered across different disciplines, different temporal scales, and different analytical frameworks. Lieberman’s “mismatch diseases” address culture–biology conflicts on a millennial scale[6]; Falkowski’s oxygen–mammal model addresses geochemical drivers on a hundred-million-year scale[7]; Macintosh’s skeletal biomechanics address behavioral changes on a thousand-year scale[2] — but the three have never been brought into dialogue within a unified causal framework.

More critically, a methodological pitfall has systematically obscured the deepest-level factors. When researchers employ regression models to analyze predictors of human stature — as in Wells et al.’s global analysis of 571 male populations[8] — atmospheric oxygen concentration is automatically excluded because it exhibits virtually no variance within the observational window (declining by only 0.7% over the past 800,000 years of the Pleistocene[9]). But “no variance within the observational window” does not equal “no causal relevance.” Foundations never appear in structural stress analyses of buildings because they impose identical constraints on all floors — yet remove the foundation and the entire building ceases to exist.

This paper proposes a Three-Layer Nested Causal Model (TLNC) that attempts to integrate these disparate findings into a coherent theoretical framework spanning from hundred-million-year to decadal scales, and explicitly sets forth falsifiable predictions.

§2Hypothesis: The Three-Layer Nested Causal Model

We hypothesize that the historical changes in human physical parameters (stature, bone density, muscle mass, and brain volume) are not the linear product of any single factor, but rather the joint output of three nested hierarchical layers. The layers stand in a hierarchical constraint relationship: each layer defines the feasibility domain, efficiency, and upper bound within which the next layer operates — yet this is not an all-or-nothing switch dependency. Lower layers can still partially function when upper layers are compromised; the boundaries simply narrow.

2.1 L1: Atmospheric O₂ — Existential Constraint

Humans are the large animal with the highest per-unit-mass O₂ demand known on Earth. The adult brain constitutes only 2% of body mass yet consumes 20%–25% of the body’s total oxygen and glucose[10,11]. During childhood development, this proportion is even more extreme: at approximately age five, the brain claims 66% of resting metabolic rate[12], causing human children to gain body weight 30- to 100-fold more slowly than non-primate mammals of comparable size[12]. Furthermore, placental reproduction in mammals is itself a high-oxygen strategy — 60% of the oxygen in maternal blood is consumed before it reaches the placenta[7].

Human existence therefore presupposes an atmospheric oxygen concentration sufficient to simultaneously sustain a high-metabolic-rate brain and placental reproduction. Falkowski et al.’s carbon-isotope model shows that atmospheric O₂ approximately doubled over the past 205 million years, reaching a peak of roughly 23% during the Eocene (~50 million years ago), precisely coinciding with the explosive radiation of new orders of large placental mammals[7].

2.2 L2: Ecosystem Structure — Nutritional Supply

On the biological platform afforded by O₂, the human capacity to obtain adequate nutrition depends on ecosystem structure. Data consistently show that stature is determined not by total vegetation (gross NPP) but by the efficiency with which vegetation is converted into animal protein accessible to humans. Wells et al. found that gross NPP is a negative predictor of stature, whereas ecologically relevant NPP (eNPP) with concentrated growing seasons peaks at approximately 60° north and south latitude[8] — precisely corresponding to mid-to-high-latitude grassland ecosystems with the highest density of large ungulates.

2.3 L3: Behavioral Loading — Direct Sculpting

Even when L1 and L2 are both satisfied (adequate oxygen, abundant protein sources), the human musculoskeletal system still requires use to reach its genetic potential. Fetal Akinesia Deformation Sequence (FADS) provides the most extreme counter-evidence: when a fetus is completely immobile, joints contracture, muscles atrophy, and lungs fail to develop despite adequate oxygen and nutrition[13]. Postnatally, systematic reviews of prone positioning (tummy time) have demonstrated a significant positive correlation between use and functional maturation[14].

2.4 The Irreplaceable Nature of the Nesting Relationship

The core claim of the TLNC model is that the nesting relationship among the three layers is irreplaceable — the effectiveness of L3 depends on L2 being satisfied, and the existence of L2 depends on L1 being satisfied. A sedentary agricultural population (L3 degraded) settled in a high-altitude (L1 constrained), tropical dense forest (L2 constrained) environment should theoretically exhibit the full additive superposition of degradation effects from all three layers. Conversely, a highly mobile nomadic population (L3 fully engaged) living on open grasslands (L2 optimal) at sea level (L1 optimal) should approach the upper bound of human genetic physical potential.

§3Evidence: L1 Infrastructure Layer

Evidence supporting O₂ as an existential constraint comes from three temporal scales. It must be made explicit here: L1 explains the feasibility-domain boundary within which human physical potential exists, not the direct drivers of stature and bone density changes over the past 40,000 years. Atmospheric O₂ declined by only 0.7% over 800,000 years of the Pleistocene[9], and the physiological effects of this magnitude of change are very likely far smaller than those of nutrition, disease, physical activity, and genetic drift. The value of L1 lies in answering “why could a species emerge on Earth that devotes 20% of its total oxygen to an organ constituting merely 2% of its body mass,” rather than “why did this species become shorter during the Neolithic.”

Hundred-million-year scale: When atmospheric O₂ was approximately 10% some 205 million years ago, mammalian body mass averaged roughly 2 grams. When O₂ reached its peak of approximately 23% around 50 million years ago, large placental mammalian orders underwent explosive radiation[7]. Earlier Paleozoic evidence is even more dramatic: during the Carboniferous, when O₂ reached 32%, giant dragonflies with wingspans of 70 cm appeared; during the Triassic, when O₂ fell to 13%, giant insects went extinct[15]. Drosophila experiments verified the causal direction of this relationship: in a 40% O₂ environment, body size increased by 15% after 11 generations of selection, yet under 10% O₂, even intense selection for larger body size resulted in significant size reduction — low oxygen imposed a physical ceiling on body size[15].

Ten-thousand-year scale: A Princeton University team reconstructed the atmospheric O₂ record of the past 800,000 years from Antarctic and Greenland ice cores, revealing an O₂ partial-pressure decline of 7‰ (0.7%)[9]. Deeper ice samples (1.5 million years ago) suggest that this decline may have begun during the Mid-Pleistocene Transition, approximately 1.2 to 0.7 million years ago[16]. Although a 0.7% change may be physiologically negligible for humans, it confirms that O₂ is not strictly constant but is slowly declining.

Spatial dimension: At high altitudes (>3,000 m), effective O₂ partial pressure drops by 30%–40%, which co-varies with low megafaunal density, short stature, and broad thoracic morphology (Andean Indigenous peoples, Tibetans). It must be noted, however, that high altitude simultaneously affects L1 (O₂ partial pressure) and L2 (constrained agricultural output, increased ultraviolet radiation, lower temperatures), making it a natural control for L1 + L2 joint constraint rather than a clean isolated test of L1 alone.

§4Evidence: L2 Ecosystem Layer

4.1 Longitudinal Evidence: Megafaunal Disappearance and Synchronous Stature Decline

During the Last Interglacial, European megafaunal biomass averaged 18.9 tonnes/km² (IQR: 6.3–31.4)[17]. This figure has now fallen to 0.6 tonnes/km² — a loss of 94.5%[17]. Over the same broad period, European male stature declined from approximately 179 cm in the Early Upper Paleolithic to 164 cm in the Mesolithic and below 165 cm in the Neolithic[18,19,20]. The synchronicity of these two trajectories is notable within the limits of available temporal resolution.

Ben-Dor et al. proposed a framework unifying megafaunal extinction with human evolutionary transitions: humans specialized in hunting large prey because large prey offered higher foraging returns, greater biomass density, higher fat content, and required less complex toolkits[21]. As prey body size progressively diminished, humans had to invest increasing amounts of energy to obtain the same nutritional return. Stable-isotope analysis shows that Late Upper Paleolithic European humans exhibited a high-trophic-level carnivorous dietary profile, but by the Mesolithic the trophic level had markedly declined[22].

4.2 Cross-Sectional Evidence: The Determinative Role of Vegetation Structure

During glacial periods, the Eurasian landmass was dominated by tundra and grassland, with dry steppe covering 2.2 times the area of interglacial periods (4.0 vs. 1.8 Mm²)[23]. These open grasslands supported dense communities of large ungulates — woolly mammoths, wild horses, bison, and giant deer. Post-glacial forest expansion and grassland retreat coincide precisely with the sustained decline in stature.

The intra-Mesolithic east–west European contrast provides quasi-experimental conditions: Eastern Europe retained more open grassland and ungulate communities, and its populations were systematically taller than their contemporaries in the already-forested west[18].

Modern cross-sectional data are even more dramatic: the 32 cm stature difference between the Dinka, who inhabit equatorial open savanna (male average 182 cm), and the Pygmy peoples, who inhabit equatorial closed tropical rainforest at the same latitude (approximately 150 cm), cannot be adequately explained by latitude, temperature, or genetic drift alone, yet aligns strongly with the vast disparity in large-prey availability.

4.3 The Nutritional Degradation Mechanism of Cereal Replacement

When agriculture replaced hunting-gathering approximately 10,000 years ago, the human diet was compressed from hundreds of species of wild plants and animals to a handful of cereal grains. The synthetic conclusion from multiple ethnological and archaeological studies is that the transition to a cereal-based diet led to declines in life expectancy and stature, increases in infant mortality and infectious disease, and multiple nutritional deficiencies[24]. Specific mechanisms include reduced protein quality and quantity; phytic acid in cereals chelating iron, zinc, calcium, and magnesium and thereby reducing their bioavailability[25]; and the appearance in Neolithic populations of high-prevalence skeletal markers of iron-deficiency anemia (cribra orbitalia and porotic hyperostosis) along with an explosion in dental caries[26,4].

A landmark 2022 PNAS study integrated ancient DNA with osteometric data, comparing genetically predicted stature and actual skeletal stature for 167 prehistoric European individuals. The results showed that Neolithic individuals were on average 3.82 cm shorter than their genetically predicted values[27] — direct evidence that this was not a genetic change but rather environmental suppression (nutritional and/or disease-related) of developmental potential.

§5Evidence: L3 Behavioral Layer

5.1 Skeletal Response to Mechanical Loading

Modern human joint trabecular bone density is only one-third to one-quarter that of ancient foragers[28]. This decline began in the Neolithic, not during the Industrial Revolution. Macintosh’s research shows that Central European male farmers around 5,300 BCE possessed lower-limb strength averaging that of today’s cross-country runners, yet a mere 3,000 years later (around 2,300 BCE), this level had dropped to that of sedentary university students[2]. Ruff et al.’s analysis of over 1,800 skeletal samples spanning the past 33,000 years confirmed that the greatest changes in bone strength occurred during the transition from hunting-gathering to agriculture[29].

5.2 A “Use It or Lose It” Evidence Chain from Fetus to Adult

The TLNC model predicts that, even when L1 and L2 are satisfied, L3 (behavioral loading) remains a necessary condition for the body to realize its genetic potential. Evidence spans the entire developmental trajectory: fetal immobility results in joint contractures, muscle atrophy, and pulmonary hypoplasia[13]; infant prone-positioning time correlates significantly and positively with on-time attainment of gross motor milestones[14]; the cross-sectional area of Type II (fast-twitch) muscle fibers in children does not match that of Type I fibers until puberty, reflecting developmental limitations in motor-cortex discharge frequency[30]; and among adults, over just 30 years (1985–2016), average grip strength in 20- to 34-year-old males declined by 20 pounds (~9 kg), with females declining by 10 pounds[31].

5.3 Energy Allocation Redistribution

A surprising finding provides a deeper mechanistic explanation for L3: Pontzer et al. discovered that modern hunter-gatherers (the Hadza) and modern Westerners show no statistically significant difference in total energy expenditure (TEE) after controlling for lean body mass and fat mass[32]. This implies that the human total metabolic rate may be a species-level constraint, with the difference lying in how energy is allocated: hunter-gatherers devote more energy to muscular activity and immune repair, while modern humans store more energy as fat. Obesity-rate comparisons (hunter-gatherers <5% vs. modern Western nations ~40%[33]) and body-fat-ratio comparisons (males 9–18% vs. 20–30%) support this interpretation.

5.4 The Immune Energy Sink: Forced Metabolic Reweighting Under Constant TEE

Pontzer’s constant-TEE finding implies a deeper mechanism: if the total human metabolic budget is a species-level constraint, then any increase in the metabolic load of one system must “borrow” energy from others. Agricultural sedentism caused population densities to surge and brought humans into close proximity with livestock, introducing an entirely new infectious-disease burden — pathogens such as smallpox, measles, and rubella, which require hundreds to thousands of hosts for sustained transmission, became endemic only after Neolithic settlement[38]. Urlacher et al.’s study of Shuar children in Ecuador provided direct evidence: chronic immune activation associated with soil-transmitted parasites reduced child growth by as much as 49%[39]. Research in the Bolivian Amazon similarly showed that adaptive immune markers are negatively correlated with child stature — in high-pathogen environments, energy is preferentially allocated to immune defense rather than skeletal growth[40].

Under the hard constraint of constant TEE, this means that the skeletal and stature declines in agricultural populations may have been not merely the result of L2 (reduced nutritional quality) or L3 (altered mechanical loading), but also of a forced metabolic reweighting triggered by the sedentary mode of life — the immune system, an “energy black hole,” crowded out the energy that would otherwise have been devoted to bone mineralization and muscle protein synthesis during the developmental period. This mechanism constitutes an important mediating pathway between L2 and L3, providing a more precise thermodynamic explanation for the effects of the transition from hunting-gathering to agriculture at the level of energy allocation.

§6Cross-Layer Interactions and Positive Feedback Loops

The TLNC model not only describes three independent layers but, more importantly, reveals positive feedback loops between them that lend the process of parametric change a self-accelerating character.

L2 internal loop: Megafaunal extinction → release of herbivory pressure → forest canopy closure → further reduction of large-prey habitat → even lower residual prey density. Paleovegetation reconstructions confirm that “mosaic open woodland — landscapes combining open, sparse-canopy, and closed-canopy patches — was prevalent prior to the Late Pleistocene megafaunal collapse. As large herbivore populations declined, canopy expanded under reduced browsing pressure”[34].

L2→L3 cross-layer amplification: Cereal replacement of animal protein → phytic-acid chelation of calcium, iron, and zinc → insufficient bone mineralization + depleted raw materials for muscle synthesis → reduced bone density → increased fracture risk → further activity restriction.

L3 internal degradation spiral: Sedentism → bone density decline → osteoporosis → fear of falling/fracture → even greater sedentism. This loop has been thoroughly documented in contemporary elderly populations.

6.1 Modern Humans Are Not Declining on All Metrics

It must be made explicit here that the TLNC model describes the directional changes in specific physical parameters (bone density, muscular strength, body fat ratio, and other metrics maintained by high levels of physical activity) following alterations in ecological and behavioral conditions, not a blanket value judgment on the “human condition.” Modern humans significantly outperform any historical period across multiple dimensions: stature has substantially recovered and in some cases exceeded pre-industrial levels following nutritional improvement (Dutch male average: 183 cm); infant and child mortality has fallen from pre-industrial rates of 30–50% to below 0.5% in developed nations; average lifespan has extended from approximately 30–35 years in the hunter-gatherer era (including infant mortality) to over 70–80 years; and cognitive ability, education, and technological cognitive offloading continue to advance. Specialized modern athletes under training conditions can match or even exceed ancient human levels on specific metrics[37]. Agriculture, while reducing individual physical robustness, dramatically increased population carrying capacity. A more accurate description, therefore, is not “humanity is degenerating” but rather: under the historical transformation of ecological and behavioral conditions, human physical parameters have undergone structural redistribution — some metrics declined, others rose, and overall population-level fitness arguably increased.

6.2 Epigenetic Transmission: The Missing L2.5 Layer

TLNC V3 jumped directly from L2 (ecological nutrition) and L3 (behavioral loading) to phenotypic output (stature, bone density), lacking an intergenerational transmission mechanism on the appropriate temporal scale. Epigenetics is precisely the “L2.5 layer” connecting environmental load (macro-level) to gene expression (micro-level). Recent evidence indicates that 82.8% of stature-associated genes contain CpG islands, and the proportion of DNA hypermethylation modules detected in stature-associated genes is significantly higher than in control genes (42% vs. 1.5%, P = 8.0 × 10⁻¹⁷)[41]. More critically, nutritional-stress-induced DNA methylation changes can be transmitted across generations — in animal models, phenotypic and epigenetic changes triggered by a high-energy dietary transition show progressive accumulation from the F0 through the F3 generation[42]. Human data similarly show that pregnancy-period nutritional changes induce epigenetic alterations correlated with offspring childhood obesity risk[43].

This implies that the nutritional stresses imposed by the agricultural transition (cereal replacement, phytic-acid chelation, micronutrient deficiency) and the increased immune burden brought by sedentism did not act solely through contemporaneous mechanical responses (use it or lose it) and contemporaneous nutritional insufficiency — they may have accumulated across generations via DNA methylation and histone modifications, forming a progressive epigenetic “lock-in.” This provides a supplementary explanation, beyond contemporaneous environmental effects, for why Neolithic stature was persistently below genetically predicted values (by 3.82 cm[27]). Future TLNC validation should incorporate epigenetic transmission as a key mediating pathway from L2 to phenotype.

§7Brain Volume Change: The Most Sensitive Cross-Layer Signal (Contested Section)

Brain volume is the most alluring — and the most cautiously handled — variable in the TLNC framework. We dedicate a separate contested section to it, explicitly distinguishing “observational facts,” “what TLNC can explain,” and “what TLNC cannot explain.”

7.1 Observational Facts

Henneberg’s global study of more than 14,000 crania showed a Holocene brain-volume decline of approximately 100–150 mL[3]. DeSilva et al.’s change-point analysis found that the rate of brain-volume decline over the past 10,000 years was 36 times the rate of increase over the preceding 800,000 years[35]. Hawks’s analysis demonstrated that body-mass changes can account for only one-fifth to one-seventh of the observed brain-volume decline[36]. However, other researchers (e.g., Villmoare and Grabowski 2022) have challenged sampling methodologies, arguing that the magnitude of the Holocene decline may have been exaggerated by selective sampling.

7.2 Explanatory Pathways TLNC Can Offer

TLNC offers a possible cross-layer explanation for brain-volume change: L2 degradation (reduced animal protein and long-chain fatty acids → insufficient brain nutrition during development) and L3 degradation (reduced sensorimotor richness → altered synaptic pruning patterns) may have jointly affected brain developmental trajectories. The human brain at age five consumes up to 66% of resting metabolism[12], making it acutely sensitive to nutritional supply.

7.3 What TLNC Cannot Explain

At least the following competing hypotheses exist, which TLNC cannot rule out: (a) The self-domestication hypothesis — following increases in social complexity, reduced selection pressure against aggression led to overall gracilization, with brain-volume reduction being part of whole-body downsizing; (b) The cultural offloading hypothesis — language, writing, and social division of labor rendered it unnecessary for individuals to independently accomplish all cognitive tasks, reducing selective pressure for large brain volume; (c) The neural efficiency hypothesis — brain-volume reduction does not indicate functional decline but may instead reflect improved neural connectivity efficiency and energy optimization (fewer neurons achieving equal or greater computational capability); (d) Measurement and sampling biases — differences in cranial morphology, sex ratios, and regional composition across time periods may affect mean values.

An even more damaging paradox is this: the reduction in brain volume occurred not only during the nutritionally impoverished Neolithic but continued into the industrial era, when nutrition was abundantly available and body size had already begun to rebound. This directly undermines the explanatory power of L2 (nutritional degradation → stunted brain development) — if stature recovered after nutritional improvement but brain volume did not, then nutrition may not be the primary driver of brain-volume decline. The self-domestication and neural-efficiency hypotheses may be more competitive in explaining this paradox. TLNC’s explanatory pathway for brain volume must be strictly delimited as “one of several possible mechanisms.”

Critical caveat: brain-volume decline is not equivalent to cognitive decline, nor does it mean “humans are getting dumber.” TLNC merely flags brain-volume change as a “potential cross-layer degradation signal,” not as a “confirmed degradation outcome.” Equating brain volume with intelligence is a reductionism this paper explicitly rejects.

§8Falsifiable Predictions

A valuable hypothesis must generate predictions that can be refuted by empirical data. The TLNC model yields the following three core predictions:

§9Discussion

9.1 Methodological Contribution

The core methodological contribution of this paper is identifying a systematic bias in cross-scale causal analysis: when researchers analyze variable relationships within a specific temporal window, variables that remain constant within that window are automatically excluded — regardless of whether those variables are existential preconditions for the entire system on a longer temporal scale. This is not a flaw in statistical methods but rather an inherent limitation of the statistical framework: regression models can only detect covariance among variables that vary, yet the defining characteristic of existential preconditions is precisely that they do not vary.

We recommend introducing a “Hierarchical Constraint Audit” step in causal inference within evolutionary biology: before constructing any regression model, first identify which variables in the current analytical window serve as existential preconditions on deeper temporal scales, and explicitly label them as “infrastructure layer” rather than “irrelevant variables.” This methodological principle can be generalized to the clearing layer in financial systems, the compute and power layer in AI systems, the infrastructure layer in urban systems, and analogous domains.

9.2 Limitations

This paper has several important limitations. First, our data are drawn from heterogeneous sources across different studies, methods, and samples, rather than from a unified prospective cohort or standardized database. Second, there is a lack of dedicated research on the direct quantitative relationship between O₂ and intra-Pleistocene changes in human physical parameters — our L1 argument relies primarily on hundred-million-year-scale correlations and high-altitude spatial controls; on the ten-thousand-year scale, it is indirect inference, and its explanatory power for changes over the past 40,000 years is far weaker than that of L2 and L3. Third, causal attribution of brain-volume change remains fiercely debated (see §7), and our cross-layer explanation is merely one of several competing hypotheses. Fourth, the data in this paper are heavily biased toward Europe — systematic stature and skeletal-biomechanics data from the Paleolithic through Bronze Age come predominantly from European sites, with comparable data from Africa, East Asia, and the Americas being extremely scarce, limiting global validation of the model. More precisely, the current version of this paper is “an initial TLNC argument centered on European and selected ethnographic evidence” rather than a globally validated framework. Fifth, L2’s inter-population lateral comparisons (e.g., Dinka vs. Pygmy) are heuristically valuable but involve numerous confounders (genetic, disease, parasitic, socioeconomic); more rigorously matched designs are needed in the future.

9.3 Self-Reflection on the Risk of Fractal Forcing

This paper harbors a structural risk that may be termed “fractal forcing”: in order to maintain the symmetrical elegance and logical completeness of the three-layer nested model, variables that operate at entirely different physical magnitudes and temporal scales — the hundred-million-year doubling of atmospheric O₂ from 10% to 23%, versus the century-scale micro-decline of 0.1%; the ten-thousand-year catastrophic collapse of megafaunal communities, versus the decadal decline in grip strength — are stitched together within a single causal framework. The allure of such cross-scale suturing is real — it produces an aesthetically pleasing symmetrical structure and a compelling narrative — but it also carries a genuine risk of overfitting. Readers should note that the three layers of TLNC are not equivalent in physical robustness. The evidence for L3 comes from experiments and quasi-experiments (RCTs, astronaut data, skeletal biomechanics); the evidence for L2 comes from observational associations and natural experiments; L1 has hard evidence on the hundred-million-year scale but provides only a conceptual framework on the ten-thousand-year scale. If one strips L1’s causal claims for the most recent 10,000 years, retaining only the L2 × L3 interaction, TLNC remains fully viable as a theoretical synthesis of “how agriculture and modern life have restructured the human musculoskeletal system” — and this is likely the most robust core of this paper.

9.4 Relationship to Existing Frameworks

The TLNC model does not replace existing frameworks such as “mismatch diseases”[6], the “expensive-tissue hypothesis”[10], or the “prey-size decline hypothesis”[21]; rather, it embeds them within a deeper hierarchical structure. Mismatch diseases are manifestations of L3 degradation; the expensive-tissue hypothesis describes energy trade-offs under L1 constraints; prey-size decline is a specific mechanism of L2 degradation. TLNC’s distinctive contribution lies in revealing the nested dependency relationships among these separately described phenomena.

9.5 Logical Extrapolation (Non-Core Prediction)

The following extrapolation is not a core prediction of TLNC but a logical extension of the framework, with extremely low confidence, recorded solely as a long-term monitoring hypothesis. Fossil-fuel combustion over the past century has reduced atmospheric O₂ by approximately 0.1%[9]. However, the sigmoidal oxygen-dissociation curve of human hemoglobin enables the body to achieve instantaneous compensation through minor adjustments in respiratory rate and heart rate until O₂ concentration drops to approximately 19%, maintaining arterial blood oxygen saturation above 97%. Consequently, the current 0.1% decline produces no physiologically detectable effect; its impact is orders of magnitude smaller than those of air pollution, elevated CO₂, sleep deprivation, and physical inactivity. Only if O₂ continues to decline for thousands of years and cumulatively exceeds the hemoglobin compensatory range (below approximately 19%) would a detectable signal potentially emerge — a condition that, at current rates, would require tens of thousands of years to reach.

§10Conclusion

The history of human physical parameters is not a unidirectional curve but a multi-phase arc. From the 130 cm of Australopithecus to the 170 cm of Homo erectus, stature accomplished a 40 cm leap over two million years — this is the phase in which L1 (atmospheric O₂ accumulation to its 21% platform) truly exerted causal force, making a large, high-metabolic, bipedal body possible. This was followed by the Middle Pleistocene plateau (Homo heidelbergensis 163.6 cm, Neanderthals 160.6 cm), after which early Homo sapiens pushed stature to a peak of 177–179 cm — at which point L2 (open grassland × megafaunal prey × high-protein diet) and L3 (high-intensity migratory hunting) were jointly in their optimal state. Then came a sustained decline of approximately 20,000 years (179 → <165 cm), corresponding to megafaunal extinction (L2 collapse) and agricultural sedentism (L3 degradation + increased immune burden). Finally, a partial recovery since industrialization (170–183 cm) corresponds to the compensatory repair of L2 through dairy and industrialized protein — yet bone density (reduced to one-quarter of ancient levels), muscular strength (grip strength declining 16–22% within 30 years), and body fat percentage (rising from 9–18% to 20–30%) remain unrestored, because L3 (sedentism) has never been repaired.

This complete arc — rise → plateau → peak → decline → partial recovery — is precisely the best illustration of the TLNC three-layer model: L1 defines the physically feasible domain for human physique (the leap from 130 to 170 cm required the O₂ platform); L2 sets the nutritional ceiling within that feasible domain (the peak corresponds to optimal ecology, the decline to ecological collapse); L3 determines whether bone and muscle actually approach that ceiling (modern stature recovery without bone-muscle recovery is exactly the result of L2 repair without L3 repair).

The core contribution of the TLNC model is not to prove that “humans are degenerating” — such a judgment is oversimplified and neglects the enormous modern advances in lifespan, infant survival, and cognitive offloading. Its contribution lies in proposing a deeper analytical framework: the morphological expression of any living organism is not the product of a single variable but the joint output of three nested layers of constraint — foundational environment, ecological supply, and behavioral use. And the methodological reminder that “constant ≠ irrelevant” in cross-scale causal analysis holds general significance for evolutionary biology, ecology, and indeed any discipline involving hierarchical systems.