{
  "id": "25",
  "title": "Mitochondrial dynamics",
  "summary": "Abnormal mitochondrial fission, fusion, or distribution may compromise the maintenance of sensory terminals.",
  "source_title": "Abnormal mitochondrial fission or fusion impairs terminal maintenance",
  "page": "25_mitochondrial_dynamics.html",
  "source_markdown": "sources/25_mitochondrial_dynamics.md",
  "structured_data": "data/25_mitochondrial_dynamics.json",
  "snapshot_date": "2026-09-14",
  "source_review_date": "2026-09-14",
  "scope_label": "Early type 2 diabetes",
  "scope_exception": false,
  "source_headline_markdown": "**Best judgment: about 16%; skeptical–favorable sensitivity: 0.6%–62%.** These are subjective probabilities of the full human proposition, not treatment effects or confidence limits. Structural MCU evidence and direct injury-model fission experiments support causality in some contexts; neither identifies a selective fission/fusion effect in early human diabetes. Reviewed 14 September 2026. [Probability method](probability_method.md) · [Collection index](README.md).",
  "probability_percent": {
    "skeptical": 0.63,
    "base": 16.26625,
    "favorable": 61.965
  },
  "probability_meaning": "Subjective belief in the complete causal proposition. Scenario endpoints are sensitivity products, not confidence limits.",
  "causal_proposition_markdown": "In adults with type 2 diabetes diagnosed within five years and no clinical DPN at entry, persistent abnormal mitochondrial fission or fusion in peptidergic sensory neurons contributes to physical terminal loss. Selectively normalizing the specified membrane-remodeling machinery from entry would avert at least **20% of five-year net terminal loss**, in a baseline-defined group accounting for at least **10% of reference-care incident DPN**.\n\nThe claim concerns division and joining of mitochondria. Abnormal transport, positioning, biogenesis, degradation, calcium loading and ATP production can be causes or consequences, but their correction is not substituted for a fission/fusion intervention. The target cells are the identified sensory neurons themselves, including their somas, peripheral axons and skin terminals; glial and CNS changes are contextual evidence.\n\n**Population and endpoint.** Reference care is contemporary diabetes care without Q below. Identify adult peptidergic unmyelinated neurons by combined anatomical and molecular classification. Physically track their distal arbors, so that changes in CGRP, PGP9.5 or a fluorescent reporter alone cannot count as disappearance.\n\nFor each person, five-year net terminal loss is physical destruction minus replacement, divided by baseline terminal amount. Newly formed terminals that are later destroyed contribute to both totals. Average people equally and define:\n\n`B = (mean reference net loss − mean loss with Q) / mean reference net loss.`\n\nThe reference mean must be positive. Pain, conduction, mitochondrial shape, somal apoptosis and neurite length in a dish are distinct outcomes.\n\n**Finite baseline groups.** Candidates are the whole eligible population, males, females, baseline BMI ≥30 kg/m², baseline HbA1c ≥7.5%, and a baseline dynamics-positive group. The last has the qualifying rate abnormality below in at least 10% of mapped neurons during a 30-day entry observation. No extra intersections or responder-defined groups qualify.\n\nIncident DPN means new bilateral length-dependent peripheral nerve signs with objective small-fiber loss or abnormal conduction, excluding another cause. Coverage is `C = P(group membership | incident DPN under reference care)`. The subgroup, relevant neurons and qualifying process must remain compatible through the entire chain.\n\n**Direct event measurements.** A fission event is division of one fully tracked mitochondrial matrix and its membranes into two separated daughter organelles. A fusion event is joining of two previously separate organelles, with both membrane continuity and matrix-content mixing. Close contact, crossing trajectories and transient outer-membrane contact without matrix mixing do not count as full fusion. Track complete events in three dimensions and record subsequent rejoining separately.\n\nMeasure completed events per unit mitochondrial protein mass per hour, separately in soma, peripheral axon and terminal arbor. The mass denominator prevents more mitochondria alone from increasing a cell's apparent event rate. Measure both rates; their ratio alone can hide a simultaneous fall or rise.\n\nCompare each 30-day mean with an age-, sex- and anatomical-site-matched nondiabetic reference. The normal median must be positive to apply a fold-change boundary; a zero reference does not establish the stated relative abnormality. An abnormal rate is **at least twice the normal median and above the 95th percentile**, or **at most half the normal median and below the fifth percentile**. At least 10% of mapped neurons, averaged with equal person weight in a candidate group, must sustain a qualifying abnormality for **90 consecutive days during year one**.\n\nThe qualifying interval must precede the first **≥5% net physical terminal deficit from entry maintained for 30 days** in the connected arbor under reference care. Ordinary turnover without sustained loss is distinct from this landmark. Rate, duration, unit-prevalence and deficit thresholds are chosen definitions, not measured universal tipping points.\n\n**Three defined interventions.** Q acts on fission, fusion, or both. In the named neurons, it restores out-of-reference membrane-remodeling capacity to the matched nondiabetic median: fission through the DRP1-dependent membrane-scission machinery; fusion through MFN1/MFN2-mediated outer-membrane and OPA1-mediated inner-membrane joining. An out-of-reference capacity is below the fifth or above the 95th percentile in standardized assays with common GTP, membrane substrate and, for fusion, organelle contact opportunities. Normal-range capacity is left alone.\n\nThe ideal intervention changes only these proteins' specified membrane-remodeling actions, including the functional abundance of their active complexes. It does not directly alter MFN2-dependent transport/tethering or OPA1-dependent cristae functions independently of fusion. This requires selective molecular implementations beyond current drugs. Unrelated functions of shared proteins must be verified as directly unchanged.\n\nQ begins at entry and continues for five years. The direction of adjustment is determined by the measured capacity, not by observed benefit or an assumption that less fission is always better. Actual event rates need not become normal if organelle encounters, energy or substrate supply remain limiting. Faithful machinery normalization can therefore have a small or adverse effect.\n\nQ does not directly change calcium entry, neuronal firing, antioxidant activity, mitochondrial biogenesis, lysosomal disposal, general axonal transport or systemic diabetes. All downstream effects on those variables, organelle distribution and replacement remain free to occur. Resource costs and loss of protective segregation or complementation count against B.\n\nFor a combined fission-plus-fusion intervention, both processes must meet R1 in the same qualifying neurons during overlapping pre-deficit intervals. A single-process intervention needs only its corresponding R1 abnormality. The complete proposition requires one same group and one same intervention subset to pass every requirement.",
  "claims": [
    {
      "id": "R1",
      "source_id": "R1",
      "claim": "One fixed group and process subset meet the first-year, 90-day, pre-deficit rate criterion in at least 10% of mapped neurons.",
      "claim_markdown": "One fixed group and process subset meet the first-year, 90-day, pre-deficit rate criterion in at least 10% of mapped neurons.",
      "conditional_percent": {
        "skeptical": 30.0,
        "base": 65.0,
        "favorable": 90.0
      },
      "cumulative_base_percent": 65.0,
      "source_cumulative_display": "65.0%",
      "reason_markdown": "Rodent timing and human organelle changes support occurrence; actual early human rates and direction remain unknown.",
      "question": "Are mitochondrial fission or fusion rates persistently abnormal in sensory neurons before terminal loss?"
    },
    {
      "id": "R2",
      "source_id": "R2",
      "claim": "In one same R1 group/subset, selective Q produces B≥0.05.",
      "claim_markdown": "In one same R1 group/subset, selective Q produces B≥0.05.",
      "conditional_percent": {
        "skeptical": 20.0,
        "base": 55.0,
        "favorable": 85.0
      },
      "cumulative_base_percent": 35.75,
      "source_cumulative_display": "35.8%",
      "reason_markdown": "MCU structure and direct axotomy perturbations support capability, but neither isolates Q in diabetes; protective dynamics and positioning confound attribution.",
      "question": "Would normalizing abnormal mitochondrial fission or fusion machinery avert at least 5% of five-year net terminal loss?"
    },
    {
      "id": "R3",
      "source_id": "R3",
      "claim": "One same group/subset satisfying R2 reaches B≥0.20.",
      "claim_markdown": "One same group/subset satisfying R2 reaches B≥0.20.",
      "conditional_percent": {
        "skeptical": 35.0,
        "base": 70.0,
        "favorable": 90.0
      },
      "cumulative_base_percent": 25.025,
      "source_cumulative_display": "25.0%",
      "reason_markdown": "Large structural contrasts permit materiality, with uncertain mediation, duration and compensation.",
      "question": "Would normalizing abnormal mitochondrial fission or fusion machinery avert at least 20% of five-year net terminal loss?"
    },
    {
      "id": "R4",
      "source_id": "R4",
      "claim": "One same group/subset satisfying R3 has C≥0.10.",
      "claim_markdown": "One same group/subset satisfying R3 has C≥0.10.",
      "conditional_percent": {
        "skeptical": 30.0,
        "base": 65.0,
        "favorable": 90.0
      },
      "cumulative_base_percent": 16.26625,
      "source_cumulative_display": "16.3%",
      "reason_markdown": "A common metabolic setting makes coverage plausible, but the specified early state and benefit are not jointly observed.",
      "question": "Does the benefiting subgroup account for at least 10% of incident DPN cases under reference care?"
    }
  ],
  "sections": [
    {
      "heading": "The causal claim and its scope",
      "markdown": "In adults with type 2 diabetes diagnosed within five years and no clinical DPN at entry, persistent abnormal mitochondrial fission or fusion in peptidergic sensory neurons contributes to physical terminal loss. Selectively normalizing the specified membrane-remodeling machinery from entry would avert at least **20% of five-year net terminal loss**, in a baseline-defined group accounting for at least **10% of reference-care incident DPN**.\n\nThe claim concerns division and joining of mitochondria. Abnormal transport, positioning, biogenesis, degradation, calcium loading and ATP production can be causes or consequences, but their correction is not substituted for a fission/fusion intervention. The target cells are the identified sensory neurons themselves, including their somas, peripheral axons and skin terminals; glial and CNS changes are contextual evidence.\n\n**Population and endpoint.** Reference care is contemporary diabetes care without Q below. Identify adult peptidergic unmyelinated neurons by combined anatomical and molecular classification. Physically track their distal arbors, so that changes in CGRP, PGP9.5 or a fluorescent reporter alone cannot count as disappearance.\n\nFor each person, five-year net terminal loss is physical destruction minus replacement, divided by baseline terminal amount. Newly formed terminals that are later destroyed contribute to both totals. Average people equally and define:\n\n`B = (mean reference net loss − mean loss with Q) / mean reference net loss.`\n\nThe reference mean must be positive. Pain, conduction, mitochondrial shape, somal apoptosis and neurite length in a dish are distinct outcomes.\n\n**Finite baseline groups.** Candidates are the whole eligible population, males, females, baseline BMI ≥30 kg/m², baseline HbA1c ≥7.5%, and a baseline dynamics-positive group. The last has the qualifying rate abnormality below in at least 10% of mapped neurons during a 30-day entry observation. No extra intersections or responder-defined groups qualify.\n\nIncident DPN means new bilateral length-dependent peripheral nerve signs with objective small-fiber loss or abnormal conduction, excluding another cause. Coverage is `C = P(group membership | incident DPN under reference care)`. The subgroup, relevant neurons and qualifying process must remain compatible through the entire chain.\n\n**Direct event measurements.** A fission event is division of one fully tracked mitochondrial matrix and its membranes into two separated daughter organelles. A fusion event is joining of two previously separate organelles, with both membrane continuity and matrix-content mixing. Close contact, crossing trajectories and transient outer-membrane contact without matrix mixing do not count as full fusion. Track complete events in three dimensions and record subsequent rejoining separately.\n\nMeasure completed events per unit mitochondrial protein mass per hour, separately in soma, peripheral axon and terminal arbor. The mass denominator prevents more mitochondria alone from increasing a cell's apparent event rate. Measure both rates; their ratio alone can hide a simultaneous fall or rise.\n\nCompare each 30-day mean with an age-, sex- and anatomical-site-matched nondiabetic reference. The normal median must be positive to apply a fold-change boundary; a zero reference does not establish the stated relative abnormality. An abnormal rate is **at least twice the normal median and above the 95th percentile**, or **at most half the normal median and below the fifth percentile**. At least 10% of mapped neurons, averaged with equal person weight in a candidate group, must sustain a qualifying abnormality for **90 consecutive days during year one**.\n\nThe qualifying interval must precede the first **≥5% net physical terminal deficit from entry maintained for 30 days** in the connected arbor under reference care. Ordinary turnover without sustained loss is distinct from this landmark. Rate, duration, unit-prevalence and deficit thresholds are chosen definitions, not measured universal tipping points.\n\n**Three defined interventions.** Q acts on fission, fusion, or both. In the named neurons, it restores out-of-reference membrane-remodeling capacity to the matched nondiabetic median: fission through the DRP1-dependent membrane-scission machinery; fusion through MFN1/MFN2-mediated outer-membrane and OPA1-mediated inner-membrane joining. An out-of-reference capacity is below the fifth or above the 95th percentile in standardized assays with common GTP, membrane substrate and, for fusion, organelle contact opportunities. Normal-range capacity is left alone.\n\nThe ideal intervention changes only these proteins' specified membrane-remodeling actions, including the functional abundance of their active complexes. It does not directly alter MFN2-dependent transport/tethering or OPA1-dependent cristae functions independently of fusion. This requires selective molecular implementations beyond current drugs. Unrelated functions of shared proteins must be verified as directly unchanged.\n\nQ begins at entry and continues for five years. The direction of adjustment is determined by the measured capacity, not by observed benefit or an assumption that less fission is always better. Actual event rates need not become normal if organelle encounters, energy or substrate supply remain limiting. Faithful machinery normalization can therefore have a small or adverse effect.\n\nQ does not directly change calcium entry, neuronal firing, antioxidant activity, mitochondrial biogenesis, lysosomal disposal, general axonal transport or systemic diabetes. All downstream effects on those variables, organelle distribution and replacement remain free to occur. Resource costs and loss of protective segregation or complementation count against B.\n\nFor a combined fission-plus-fusion intervention, both processes must meet R1 in the same qualifying neurons during overlapping pre-deficit intervals. A single-process intervention needs only its corresponding R1 abnormality. The complete proposition requires one same group and one same intervention subset to pass every requirement."
    },
    {
      "heading": "Reasoning from the biology",
      "markdown": "Fission redistributes existing material between daughters; it does not create mitochondrial mass. It can provide smaller transportable units or separate dysfunctional material for disposal. Fusion can exchange contents, complement local defects and alter how damage propagates. Both processes can support maintenance, and either extreme can impair it.\n\nThe relevant structure is:\n\n`fission/fusion events → organelle complementation, segregation and distribution → destruction minus replacement of terminals.`\n\nEnergy supply, calcium buffering, oxidants and disposal are alternative or interacting intermediates. A particular intermediary is not mandatory merely because one study measures it. Somal death is also unnecessary for loss of a long terminal arbor.\n\nA static population of smaller organelles can arise from increased fission, reduced fusion, selective removal of large mitochondria, altered growth, transport or imaging resolution. A larger fluorescent object may represent one enlarged organelle or an unresolved cluster. Neither appearance determines the two event rates.\n\nThe often-used “interconnectivity” measure area/perimeter does not solve this problem. For a circle, `A/P = r/2`: it falls simply because the circle becomes smaller. Without direct events or a true network measure, it cannot distinguish fragmentation from size.\n\nLikewise, MFN2 mutations and MCU deletion are not pure fission/fusion experiments. MFN2 participates in positioning and transport; MCU changes mitochondrial calcium entry. A beneficial downstream shape change can accompany protection caused by another output. Selective Q and its net terminal effect are the additional propositions being priced."
    },
    {
      "heading": "Evidence that moves the judgment",
      "markdown": "**MCU deletion preserves diabetic mouse skin fibers, but does not isolate dynamics.** George and colleagues studied a ten-week HFD model with constitutive MCU deletion in the Nav1.8 lineage. Figure 7 uses **3/3/5/5 animals** for regular-diet heterozygotes/homozygotes and HFD heterozygotes/homozygotes. Digitized densities are approximately **34.3/37.7/17.2/37.0 fibers/mm**. The HFD genotype contrast is **+19.75 fibers/mm**, with an approximate Welch 95% interval of **12.00–27.51**; an exploratory exact n=5/5 label comparison gives p=0.00794.\n\nNormalizing to each genotype's regular-diet group gives about **96% preservation of the deficit**. This is a large animal structural result, but the comparison changes calcium uptake from development onward. It measures mixed Nav1.8-lineage fibers rather than only peptidergic terminals. Both HFD genotypes become obese and glucose-intolerant; that classification does not establish identical metabolic histories. [George et al., 2022](https://pmc.ncbi.nlm.nih.gov/articles/PMC8720329/).\n\n**The same study's timing and shape evidence are less specific.** Smaller somal mitochondria appear at two weeks of HFD, before the allodynia and skin-fiber loss timing established in earlier experiments. These are not repeated observations of the same terminals. Morphology statistics count 60 or 120 mitochondria, without a stated independent-animal n for those comparisons.\n\nAt ten weeks, the reported area change alone predicts an area/perimeter ratio of **0.784** under a simple same-shape size model; the observed ratio is **0.778**. At two weeks the corresponding ratios are 0.686 and 0.727. These calculations do not fit a biological mechanism; they show that the index adds no independent demonstration of changed fission/fusion. The source's live calcium and genetic results remain useful even though morphology does not measure event rates. [George et al., 2022](https://doi.org/10.1097/j.pain.0000000000002391).\n\n**Early glucose experiments support an injury mechanism, with a weaker endpoint than terminal survival.** Edwards and colleagues found smaller, more numerous DRG mitochondria in db/db mice, using **two animals per group** for ultrastructure. Their intervention experiment used embryonic rat DRG cultures exposed to **25 versus 45 mM glucose**. DRP1 knockdown reduced a cleaved-caspase-3 blot signal. The paper describes increased survival, but that experiment does not provide a directly counted terminal-preservation effect or a live fission-event rate. Mdivi-1 corroboration is described as data not shown. This supports cellular susceptibility rather than quantitative adult human prevention. [Edwards et al., 2010](https://doi.org/10.1007/s00125-009-1553-y).\n\n**Human skin supplies anatomical relevance, not a matching fragmentation direction.** Hamid and colleagues studied **seven donors each** with no diabetes, diabetes without DPN and DPN. Mitochondrial fluorescent signals in skin fibers shifted toward **larger** volumes in the diabetic groups at the thigh. Normal distal-leg fibers also had larger signals than normal thigh fibers. Imaging magnification changed estimated volumes, and clustered objects could not be equated with single mitochondria.\n\nAverage mitochondrial number per nerve volume and mitochondrial volume fraction did not significantly differ between groups. These small cross-sectional comparisons do not prove equivalent function, nor do they measure fission, fusion or pre-loss timing. Diabetes duration averaged about six years, and DPN was already established in the affected group. [Hamid et al., 2014](https://pmc.ncbi.nlm.nih.gov/articles/PMC4241807/).\n\n**Human DRG phosphorylation does not provide a persistent positive anchor.** Reanalysis of public donor-level reporters identifies DRP1-S616 in 17 confidently localized spectra, with five painful-DPN and four nondiabetic donors. The DPN/control phospho-S616/total-DRP1 ratio is **0.968-fold**, with AUROC **0.35** for the proposed increased-in-DPN direction. All DPN donors occupy one TMT channel series and controls the other, so disease and channel effects are inseparable. This descriptive null neither supports a large persistent bulk-DRG elevation nor tests transient axonal events or incident disease. The assay measures a phosphosite, not fission flux. [Public PXD039344 source](https://proteomecentral.proteomexchange.org/cgi/GetDataset?ID=PXD039344).\n\n**Direct fission-dependent degeneration exists in another human-neuron context.** After laser axotomy in induced human neurons, live imaging identifies rapid fission preceding a retrograde wave of degeneration. DRP1 knockdown and an independent DRP1-interaction inhibitor suppress the fission response. Reanalysis of the source workbook, averaging **three differentiations** equally, gives a **0.366 reduction in the 24-hour increase in axon-degeneration index** with DRP1 knockdown. This is an actual morphology-based axon effect; it is not 69 independent human donors or an intact diabetic-terminal experiment. [Gómez-Deza et al., 2024](https://doi.org/10.1038/s41467-024-54982-9).\n\nThe same neuronal platform supplies a counterexample. In vincristine injury, roughly 80% DRP1 transcript knockdown enlarges mitochondria by about 60%, yet does not reduce axon degeneration. Mdivi-1 remains protective after DRP1 knockdown. Thus mdivi-1 benefit and mitochondrial size cannot be used as interchangeable evidence for fission causation. This limits a universal model; it does not establish that the diabetic context behaves like vincristine. [Gómez-Deza et al., 2023](https://doi.org/10.1038/s41419-023-06227-8).\n\n**Inherited neuropathy also separates fusion from positioning.** In embryonic sensory neurons expressing disease-associated MFN2 mutants, disrupted positioning accompanies segmental axonal degeneration. OPA1 depletion impairs fusion while sparing transport and does not produce the same degeneration in the tested interval. MFN1 expression can rescue the MFN2-mutant phenotype. These results support the importance of organelle maintenance but prevent inherited MFN2 neuropathy from uniquely validating a diabetic fusion defect. [Misko et al., 2012](https://pmc.ncbi.nlm.nih.gov/articles/PMC3319368/).\n\nFinally, AICAR treatment in HFD mice improves skin fibers while increasing DRP1 and autophagy-associated proteins. Because systemic metabolism also changes and actual rates are not measured, this is not evidence that more fission causes protection. It nevertheless cautions against assigning a fixed adverse sign to increased DRP1 abundance. [Chandrasekaran et al., 2025](https://pmc.ncbi.nlm.nih.gov/articles/PMC11720447/).\n\nNo accessed human study directly measures the specified early event-rate abnormality together with selective Q, five-year structural benefit and incident-case coverage."
    },
    {
      "heading": "Probabilities of the logical claims",
      "markdown": "Each conditional assumes every preceding row in one compatible group and intervention subset. Multiple qualifying alternatives are handled as a joint OR claim, not by adding their probabilities.\n\n| # | Claim | Skeptical / best / favorable | Best cumulative | Basis |\n|---|---|---:|---:|---|\n| R1 | One fixed group and process subset meet the first-year, 90-day, pre-deficit rate criterion in at least 10% of mapped neurons. | 30 / **65** / 90% | 65.0% | Rodent timing and human organelle changes support occurrence; actual early human rates and direction remain unknown. |\n| R2 | In one same R1 group/subset, selective Q produces B≥0.05. | 20 / **55** / 85% | 35.8% | MCU structure and direct axotomy perturbations support capability, but neither isolates Q in diabetes; protective dynamics and positioning confound attribution. |\n| R3 | One same group/subset satisfying R2 reaches B≥0.20. | 35 / **70** / 90% | 25.0% | Large structural contrasts permit materiality, with uncertain mediation, duration and compensation. |\n| R4 | One same group/subset satisfying R3 has C≥0.10. | 30 / **65** / 90% | **16.3%** | A common metabolic setting makes coverage plausible, but the specified early state and benefit are not jointly observed. |\n\nR1 does not define an abnormal rate as harmful. R2 asks whether correcting the specified machinery changes net structure. A 10% benefit passes R2 and fails R3; a large benefit restricted to a group contributing 3% of cases passes R3 and fails R4. A vague “deficient functional organelle population” is not charged again as a separate overlapping causal claim."
    },
    {
      "heading": "Ideal experiments that would resolve each claim",
      "markdown": "Use representative adults in the defined early-T2 population, fixed baseline groups, matched nondiabetic reference measurements and physically identified target arbors. The idealizations are long-term nonperturbing event tracking and molecular manipulations that separate membrane remodeling from other functions of shared proteins.\n\n**R1 — event rates, with no predecessor assumed.** Track mitochondrial membranes and matrix continuity in three dimensions through complete scission and joining events. Quantify fission and full fusion separately per protein mass and time in the named compartments. Measure the 30-day means, 90-day persistence, fraction of target neurons and timing relative to sustained physical terminal decline. Verify that differences do not arise from unresolved clusters, altered mass, motion or selective disappearance from view.\n\n**YES:** one fixed group/process subset meets all defined boundaries, with overlapping abnormalities for a combined subset. **NO:** complete observations show that every candidate combination fails at least one criterion. Changed shape, DRP1 RNA or an isolated phosphosite is not the defined measurement; missing event tracking leaves the claim **unresolved**.\n\n**R2 — structural causation conditional on R1.** Randomize compatible participants from entry to reference care or each qualifying Q subset for five years. Verify normal-range fission/fusion machinery capacity, directly preserved nonremodeling protein functions, and comparable initiating glycemic, lipid, nutritional and mechanical exposures. Measure actual event rates and allow calcium, ATP, disposal, positioning and replacement to change downstream. Use an independent molecular implementation and reversal of the capacity adjustment to distinguish remodeling from an intervention artifact.\n\nCount physical terminal destruction and replacement. **YES:** one compatible group/subset has a positive reference-loss denominator and B≥0.05. **NO:** every faithfully implemented comparison has B<0.05, including worsening or inability of normalized machinery to help in the existing cellular environment. Failed target delivery or unverified molecular specificity is unresolved. MCU deletion, systemic AICAR, pain relief or a caspase blot alone leaves this exact comparison unresolved.\n\n**R3 — magnitude conditional on R1–R2.** Estimate the same five-year B in the same qualifying group/subset, counting all maintenance benefits and costs. **YES:** one reaches B≥0.20. **NO:** all R2 combinations have 0.05≤B<0.20. A transient benefit, or a finite uncertainty interval crossing 0.20, does not settle this row.\n\n**R4 — coverage conditional on R1–R3.** Ascertain baseline membership and incident DPN under reference care throughout the representative population. **YES:** at least one R3 group has C≥0.10. **NO:** every such group is below that boundary. Count people rather than mitochondria, images or neuronal differentiations. Missing incident ascertainment or constructing a new responder group leaves coverage unresolved.\n\nThese ideal experiments define the propositions. They are not requests for a current laboratory program."
    },
    {
      "heading": "Why these odds and how uncertain they are",
      "markdown": "The strongest evidence is that targeted neuronal perturbation can preserve skin fibers in diabetic mice, and direct fission perturbation can preserve axons after another insult. The principal gap is whether fission/fusion normalization itself carries a material part of the diabetic structural effect in the intended human cells. Static morphology, phosphorylation and gene expression cannot settle that question.\n\nBest conditionals **[65,55,70,65]%** give **16.26625%**. Skeptical **[30,20,35,30]%** gives **0.63%**; favorable **[90,85,90,90]%** gives **61.965%**. Decimal places reproduce arithmetic, not precision of biological belief. These coordinated scenarios have no assigned weights and are neither independent marginal distributions nor empirical confidence limits.\n\nIf morphology mostly reflects size, positioning or compensation and the diabetic MCU benefit is mainly calcium-mediated, **[40,30,45,45]%** gives **2.4%**. If selective dynamics correction preserves terminals in a frequent early human state, **[85,80,85,80]%** gives **46.2%**. Related uncertainties move together.\n\nChanging only R2 from 55% to 80% gives **23.7%**; reducing it to 25% gives **7.4%**. Even R2 known true would leave 29.6% with the other judgments unchanged. These are arithmetic illustrations, not fitted updates; a strong selective experiment could change several rows. Merging R1 and R2 must retain their **35.75%** joint probability. Merely naming more molecular stages must not multiply additional discounts.\n\nThe twofold/half-normal, 90-day and 10%-neuron boundaries are definition choices. A milder or shorter abnormality might matter but fail the stated R1. The 20% structural and 10% case thresholds similarly determine the scope. Existing evidence does not support interpolating probabilities across alternative thresholds."
    },
    {
      "heading": "Overlap and boundary",
      "markdown": "This route overlaps [post-enclosure disposal](24_autophagy_mitophagy.md), [axonal transport](26_axonal_transport.md), [calcium injury](31_hyperexcitability_calcium.md), [energy insufficiency](07_bioenergetic_insufficiency.md) and [SARM1](27_sarm1_axon_destruction.md). Distribution changes are included as downstream consequences, not as a substitute for the defining dynamics lesion. Their probabilities cannot be summed.\n\nEstablished-DPN human morphology, constitutive mouse prevention, traumatic axotomy and inherited MFN2 neuropathy inform different parts of the reasoning. None independently establishes prevention of incident human DPN by selective fission/fusion normalization."
    }
  ],
  "source_claim_table_markdown": "| # | Claim | Skeptical / best / favorable | Best cumulative | Basis |\n|---|---|---:|---:|---|\n| R1 | One fixed group and process subset meet the first-year, 90-day, pre-deficit rate criterion in at least 10% of mapped neurons. | 30 / **65** / 90% | 65.0% | Rodent timing and human organelle changes support occurrence; actual early human rates and direction remain unknown. |\n| R2 | In one same R1 group/subset, selective Q produces B≥0.05. | 20 / **55** / 85% | 35.8% | MCU structure and direct axotomy perturbations support capability, but neither isolates Q in diabetes; protective dynamics and positioning confound attribution. |\n| R3 | One same group/subset satisfying R2 reaches B≥0.20. | 35 / **70** / 90% | 25.0% | Large structural contrasts permit materiality, with uncertain mediation, duration and compensation. |\n| R4 | One same group/subset satisfying R3 has C≥0.10. | 30 / **65** / 90% | **16.3%** | A common metabolic setting makes coverage plausible, but the specified early state and benefit are not jointly observed. |"
}