{ "updated": "2026-10-07", "spec_url": "sources/reference_neuron/SPEC.md", "mechanisms": [ { "id": "mgo", "title": "Methylglyoxal in distal skin", "old_page": "05_methylglyoxal_carbonyl_stress.html", "exposure": "Keep free methylglyoxal at 100 µM in the selected skin territory.", "exposure_detail": "At time zero, an ideal noninjurious clamp sets free, unbound extracellular methylglyoxal to 100 µmol/L throughout the baseline arbor's fixed epidermal territory and the underlying 1 mm of dermis. It maintains this concentration continuously to the stated horizon, replenishing loss to metabolism, reaction and transport. This is the unbound methylglyoxal pool, including its rapidly equilibrating aqueous hydrated forms, not protein adducts or a stock-bottle dose. The clamp directly supplies only this skin compartment; circulating MGO, the DRG soma, proximal nerve and central terminals receive no imposed dose. They can respond through ordinary downstream biology. Normal intracellular disposal, skin cells, glia, blood flow and tissue repair remain active. Intracellular MGO is not clamped. The comparison uses otherwise matched healthy reference systems with native local MGO. In the replacement comparison the same arbor removal occurs at time zero in both systems. This is a chosen high local challenge, not a measurement of diabetic skin.", "comparison": "H versus H+M, using the person-weighted, then neuron-weighted, then arbor-weighted population and four exact endpoints in the common specification. The input persists at every horizon. All estimates are q = P(pM - pH >= 0.10), expressed as percent belief, not endpoint risk.", "diabetes_relevance": "The selected challenge is much higher than the circulating concentrations in the original human pain comparison. A 100 µM clamp is 112 times its 0.895 µM painful-group plasma mean and 188 times its 0.531 µM low-pain-group mean. These ratios compare different compartments; they neither prove nor rule out a local skin hotspot. No retrieved study measures this free extracellular concentration continuously before arbor loss in early type 2 diabetes. Tissue adducts, whole-tissue concentrations and culture bath doses do not establish that claim. The separate relevance judgment is very low. A lower natural MGO exposure could still contribute to diabetes through other routes or interactions.", "diabetes_relevance_belief": { "low": 0.1, "central": 1, "high": 8, "claim": "Among 50-year-old men with type 2 diabetes diagnosed within five years and no DPN, at least 10% of people have free extracellular methylglyoxal at or above 100 µM throughout the selected reference arbor's skin territory for at least 30 continuous days before its first arbor loss, using person-weighted selection of the reference neuron and arbor." }, "summary": "Methylglyoxal can change nerve signals and can damage skin fibers. At 100 µM, short experiments also show preserved cells and stronger signals. Keeping that level in skin for months makes loss and slow repair plausible. Death of the distant cell body has much less support. These are judgments about a specified high local exposure; diabetes has not been shown to produce it.", "evidence": [ { "id": "BIER2012", "title": "Bierhaus 2012: Methylglyoxal modification of NaV1.8 facilitates nociceptive neuron firing", "url": "https://doi.org/10.1038/nm.2750", "year": 2012, "system": "Selected human type 2 diabetes pain comparison; supporting mouse and channel experiments", "n": "10 human participants per pain group; original supplement inspected", "result": "Plasma MGO means were 895 and 531 nM in pain and low-pain groups. Sex distributions were 3/7 and 9/1 male/female; average diabetes durations were 11 and 13 years. Channel experiments included 100 µM exposure for 3 h.", "limits": "Association with pain, substantial sex imbalance, established diabetes, no local skin free-MGO trajectory or identified-neuron fate. The corrected article and original table are used." }, { "id": "AND2013", "title": "Andersson 2013: Methylglyoxal evokes pain by stimulating TRPA1", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC3805573/", "year": 2013, "system": "Mouse sensory neurons and GLO1 inhibition in mice", "n": "6 mice per structural group; cells and channel patches are not additional animals", "result": "GLO1 inhibitor 50 mg/kg every other day for two weeks produced TRPA1-dependent hypersensitivity without a statistically established IENF-density reduction. Earlier checked bar reconstruction: approximately 23.5 versus 21.4 fibers/mm, difference -2.1, approximate 95% interval -8.0 to +3.8.", "limits": "The interval is not precise preservation. Local free MGO was not clamped at 100 µM. Pain and crossing density are not the target spike assay or complete arbor destruction." }, { "id": "RADU2012", "title": "Radu 2012: Dual effect of methylglyoxal on sensory-neuron calcium signaling and neurite outgrowth", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11498478/", "year": 2012, "system": "Adult male BALB/c mouse DRG cultures with glial cells", "n": "Outgrowth summaries count 5-13 images per concentration; independent animal/culture replication for these means is not established in the retrieved text", "result": "After a 24 h pulse and washout, outgrowth-area-per-neuron means were 0.69 in controls, 1.36 at 150 µM and 0.35 at 750 µM. The report describes improved adherence/outgrowth at lower doses and cytotoxicity at higher doses.", "limits": "Image counts do not establish animal n. A mixed-culture, survivor-normalized outgrowth measure after washout is not same-neuron repair completion under a sustained clamp. Greater adherence can explain apparent survival gains." }, { "id": "VEL2026", "title": "Velichkova 2026: Methylglyoxal alters C-fibre slowing and heat sensitivity in a sex-dependent manner", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13124350/", "year": 2026, "system": "Juvenile rat isolated dorsal roots; systemic rat behavior", "n": "Male vehicle 9 roots from 6 animals, MGO 9 roots from 5 animals; female vehicle 10 roots from 6 animals, MGO 9 roots from 4 animals", "result": "100 µM for 3 h altered activity-dependent slowing in opposite directions by sex, with no significant change in initial activation threshold, CAP amplitude or average conduction velocity. Acute 10 min application at 100 µM or 1 mM did not alter slowing. Systemic MGO produced heat hyperalgesia in males.", "limits": "Root CAP timing and withdrawal behavior do not measure the specified skin-to-central-axon heat response. Juvenile rats and root exposure differ from adult human skin. No ATP-supply or degeneration rate can be inferred from ADS." }, { "id": "YOU2025", "title": "Yousuf 2025: Diroximel fumarate, MGO-induced nociception and DRG stress", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12015141/", "year": 2025, "system": "Local mouse paw MGO; cultured human and mouse DRG", "n": "Structural experiment: 3 male and 3 female mice per condition. Human culture: ONE 29-year-old male donor, as specified in the supplement; 111-176 cells per arm are not donors.", "result": "A 20 ng intraplantar MGO injection reduced PGP9.5 epidermal crossing density at 24 h; diroximel fumarate pretreatment protected it. Human DRG neurons treated with 1 µM MGO for 24 h increased p-eIF2α; monomethyl fumarate reduced that signal.", "limits": "A bolus has no measured local concentration-time curve. Loss of staining/crossings does not resolve complete destruction, viable retraction or subtype. Human stress-marker induction is not neuron death. Nrf2 activation has multiple actions. Supplement: https://doi.org/10.2337/figshare.28436780" }, { "id": "TOTT2026", "title": "Totta-Griese 2026: The glycolytic metabolite methylglyoxal drives axon degeneration in vitro and in vivo", "url": "https://pubmed.ncbi.nlm.nih.gov/42759686/", "year": 2026, "system": "Primary mouse DRG and adult male mice", "n": "Independent animal/culture n not stated in accessible abstract", "result": "Published online 18 September 2026. Tested 5, 10, 500 and 5000 µM in culture and a single 720 ng injection in mice. Reports axonal degeneration and reduced somata counts at 500 and 5000 µM. Higher-GLO1 substrain cultures were protected; the substrain in-vivo comparison did not show fiber loss.", "limits": "Only the original abstract was recovered; full dose-by-time results and animal n remain unavailable. Do not assign an effect to 5 or 10 µM from this abstract. It extends the same group's conference work and is not an independent replication of it. Injection is not a sustained local clamp; somata counts are not fate-tracked death." }, { "id": "GRIG2021", "title": "Griggs 2021: Methylglyoxal shortens the axon initial segment and changes cortical activity", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC8496204/", "year": 2021, "system": "Mouse cortical cultures with neurons and astrocytes", "n": "3 independent cultures for viability and key structural comparisons", "result": "100 µM for 24 h shortened the AIS without detected cell death; AIS length recovered after medium replacement. Network activity rose briefly, fell about 20% at 0.5-3 h and returned to baseline at 24 h.", "limits": "Cortical networks are not heat-responsive C nociceptors. Nominal media addition is not a maintained free-MGO clamp. The experiment opposes an automatic acute-death or 90%-signal-loss assumption, not every possible long-term effect." }, { "id": "DUR2009", "title": "Duran-Jimenez 2009: Glycated matrix and sensory-nerve outgrowth", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC2780874/", "year": 2009, "system": "Adult rat sensory neurons on modified laminin/fibronectin; diabetic rat nerve ECM", "n": "At least 4 independent cultures for outgrowth; figure 2 n=4 cultures", "result": "Substrates exposed to 500 µM MGO for 24 h and then washed impaired 18 h sensory-neurite outgrowth. NGF-stimulated longest-neurite length was not significantly reduced although total branching was reduced; GDNF responses and matrix types differed.", "limits": "Pre-glycated, washed matrix is distinct from live skin at 100 µM. Supports a downstream obstacle to repair, not a human completion-time distribution. The assay retains soma survival and does not equal axon destruction." }, { "id": "MOR2017", "title": "Morgenstern 2017: Compensatory carbonyl disposal in GLO1-null Schwann cells", "url": "https://doi.org/10.1074/jbc.M116.760132", "year": 2017, "system": "Murine Schwann-cell clones", "n": "3 independently generated GLO1-null clones; reported viability estimates from at least 6 experiments", "result": "48 h MGO LC50 values: wild type 220±19 µM; GLO1 loss 69±3.7 µM; GLO1 loss plus epalrestat 14.2±8.6 µM, mean±SEM. Baseline MGO could remain near normal through compensatory disposal.", "limits": "Clones/experiments are not human donors. Do not use these LC50 values as a cutaneous-neuron dose-response or estimate of intracellular MGO. The chosen challenge leaves GLO1 and backup disposal intact." }, { "id": "HAN2015", "title": "Hansen 2015: ADDITION-Denmark serum MGO and neuropathy", "url": "https://pubmed.ncbi.nlm.nih.gov/25761542/", "year": 2015, "system": "Well-treated screen-detected type 2 diabetes, approximately 5.8 years duration", "n": "319 vibration-threshold, 543 monofilament, 966 questionnaire, 882 painful-neuropathy assessments; overlapping people", "result": "No association between serum MGO and assessed peripheral neuropathy or painful neuropathy.", "limits": "Cross-sectional serum measurement, not a selective perturbation or local pre-injury exposure. Public abstract lacks effect intervals; a precise upper bound cannot be invented." }, { "id": "BOND2026", "title": "Ziegler 2026: BOND randomized benfotiamine trial", "url": "https://doi.org/10.1136/bmjdrc-2025-005773", "year": 2026, "system": "Established symptomatic type 2 diabetic polyneuropathy", "n": "57 randomized; modified intention-to-treat analysis 51 participants", "result": "600 mg/day for 12 months did not improve structural outcomes. Adjusted IENFD treatment difference -0.486 fibers/mm, 95% CI -1.870 to +0.897; corneal nerve length -0.327 mm/mm², -2.432 to +1.778.", "limits": "Indirect metabolic intervention without verified local free-MGO correction. Established disease, not incident prevention or the isolated high local challenge. Neither statistical null proves absence of an MGO effect." }, { "id": "CHENG2019", "title": "Cheng 2019: MGO, TRPA1, itch and thermal hypoalgesia", "url": "https://www.thno.org/v09p4287.htm", "year": 2019, "system": "Male mice with imposed MGO or STZ diabetes", "n": "Animal groups differ by experiment; not used as a quantitative risk calibration", "result": "Intrathecal/systemic MGO produced thermal hypoalgesia linked to spinal TRPA1, while skin injections evoked itch.", "limits": "Central modulation of withdrawal behavior can change reported heat sensitivity without failure of the peripheral neuron at dorsal-root entry. No direct central MGO dose is included in the selected skin challenge." } ], "calibration": "The source assays constrain plausibility, timing and direction; none supplies a likelihood for the exact human claim. Acute 100 µM root/cortical experiments and lower-dose DRG outgrowth oppose immediate universal death or near-total signal failure. Local mouse structural positives support injury capability within one day, but crossing density cannot be converted into the fraction of complete original-arbor destruction events. Persistent replenishment makes this challenge stronger than a decaying culture dose or single injection; no equal concentration-times-duration law is assumed. Matrix modification and cell stress support slower or failed repair, while low-dose growth enhancement and successful first completion can reduce late non-completion differences. Low judgments emphasize intact clearance, subtype resistance, viable retraction, adaptation and eventual repair. High judgments emphasize continued replenishment, local tissue injury, accumulated adducts and loss of trophic support. Remote soma death receives much less weight than local injury. Long-horizon bounds are deliberately wide because no study follows this exposure for years. Numbers are explicit subjective sensitivity judgments, not a fitted posterior, confidence interval or simulation-derived human risk. The legacy five-year prevention probability is a different claim and was not copied.", "checks": [ { "label": "Challenge/plasma mean ratio", "value": "111.73× painful-group mean; 188.32× low-pain-group mean", "interpretation": "100 µM divided by source means 0.895 and 0.531 µM. A cross-compartment scale comparison, not a local concentration bound." }, { "label": "Mouse local bolus inventory", "value": "20 ng = 0.2775 nmol; at 100 µM this amount occupies 2.775 µL", "interpretation": "Uses MGO molecular weight 72.06 g/mol. This is an instantaneous mass balance assuming all material is free; tissue spread, binding and removal prevent a sustained-exposure equivalence." }, { "label": "Higher injection inventory", "value": "720 ng = 9.9917 nmol; at 100 µM this amount occupies 99.917 µL", "interpretation": "The 2026 source dose is total injected mass, not a measured local or intracellular concentration." }, { "label": "Source outgrowth contrasts", "value": "150 µM/control 1.971; 750 µM/control 0.507", "interpretation": "Recomputed from 1.36/0.69 and 0.35/0.69 in Radu. Nonmonotone nominal dose response and survivor-normalized imaging cannot be translated into repair probabilities." }, { "label": "Repair difference bound", "value": "pM(non-completion,T) - pH(non-completion,T) = FH(T) - FM(T) ≤ FH(T)", "interpretation": "FH/FM are probabilities of first repair completion. A 10-point excess cannot occur before at least 10% of controls complete repair. No measured FH curve is available for the specified human lesion." } ], "estimates": [ { "endpoint": "death", "hours": 1, "low": 0.01, "central": 0.1, "high": 1, "rationale": "The distant soma receives no imposed dose; acute source effects concern excitation. Rapid retrograde lethal injury affecting an extra 10% is very unlikely.", "source_ids": [ "VEL2026", "BIER2012" ] }, { "endpoint": "death", "hours": 24, "low": 0.1, "central": 0.5, "high": 4, "rationale": "Even directly exposed 100 µM cortical cultures lacked detected acute death; high-dose DRG soma loss is at 500 µM or more. Local skin injury could transmit stress, but this is a large compartment gap.", "source_ids": [ "GRIG2021", "TOTT2026", "YOU2025" ] }, { "endpoint": "death", "hours": 168, "low": 0.2, "central": 1, "high": 8, "rationale": "Days allow local degeneration and retrograde responses, but neither mouse skin-fiber loss nor one-donor stress staining establishes loss of connected DRG somata.", "source_ids": [ "YOU2025", "TOTT2026" ] }, { "endpoint": "death", "hours": 720, "low": 0.5, "central": 3, "high": 18, "rationale": "A month of local injury can reduce target support. Preserved proximal axon, other arbors and native carbonyl disposal still make material soma death a minority interpretation.", "source_ids": [ "DUR2009", "MOR2017", "AND2013" ] }, { "endpoint": "death", "hours": 2160, "low": 1, "central": 7, "high": 30, "rationale": "Persistent tissue damage could eventually threaten the parent neuron, but direct soma-fate evidence at this compartment and dose is absent. The high case assumes sustained damaging retrograde signals.", "source_ids": [ "TOTT2026", "MOR2017" ] }, { "endpoint": "death", "hours": 8760, "low": 3, "central": 15, "high": 45, "rationale": "A year increases opportunities for secondary injury while allowing compensation. Culture death at larger doses supplies capability, not an annual hazard; the inference remains remote.", "source_ids": [ "TOTT2026", "MOR2017", "GRIG2021" ] }, { "endpoint": "death", "hours": 43800, "low": 5, "central": 25, "high": 60, "rationale": "Five years of persistent skin stress makes soma loss possible, but loss of one peripheral territory need not kill its parent. No years-long selective exposure or human soma tracking supports a narrow judgment.", "source_ids": [ "TOTT2026", "DUR2009", "BOND2026" ] }, { "endpoint": "function", "hours": 1, "low": 1, "central": 5, "high": 20, "rationale": "Early MGO more often excites than silences nociceptors. Desensitization or local transduction disruption is possible, but a 90% drop in the exact heat signal is a much stronger endpoint than changed slowing.", "source_ids": [ "BIER2012", "VEL2026", "AND2013" ] }, { "endpoint": "function", "hours": 24, "low": 3, "central": 15, "high": 40, "rationale": "Local mouse fiber changes can occur within one day; cortical 100 µM activity recovered by 24 h. These conflicting adjacent observations permit early failure without establishing its prevalence.", "source_ids": [ "YOU2025", "GRIG2021", "RADU2012" ] }, { "endpoint": "function", "hours": 168, "low": 8, "central": 30, "high": 60, "rationale": "A maintained exposure outlasts acute dosing and could damage heat transduction or connected endings. Hyperalgesia, intact propagation and partial injury may still leave more than 10% of baseline signaling.", "source_ids": [ "YOU2025", "TOTT2026", "VEL2026", "AND2013" ] }, { "endpoint": "function", "hours": 720, "low": 15, "central": 50, "high": 80, "rationale": "Persistent local chemical and tissue injury makes severe functional loss credible over a month. Lower judgments allow adaptation and functional reserve; no measured human heat-spike curve defines the boundary.", "source_ids": [ "YOU2025", "DUR2009", "MOR2017", "GRIG2021" ] }, { "endpoint": "function", "hours": 2160, "low": 25, "central": 65, "high": 90, "rationale": "Months of replenished 100 µM exposure can maintain damage or impair renewal. Native clearance and preserved branches could sustain useful signals, so near-total heat failure is not assumed inevitable.", "source_ids": [ "YOU2025", "TOTT2026", "DUR2009", "MOR2017" ] }, { "endpoint": "function", "hours": 8760, "low": 30, "central": 70, "high": 95, "rationale": "Continuing local injury makes an extra 10% with severe heat failure likely under the central transfer. Regrowth, adaptation and the difference between crossing counts and spikes keep the low case substantial.", "source_ids": [ "YOU2025", "TOTT2026", "RADU2012", "DUR2009" ] }, { "endpoint": "function", "hours": 43800, "low": 30, "central": 75, "high": 97, "rationale": "The challenge still operates at five years, but this is a current state: recovered or replacement endings can restore signal despite earlier destruction. Chronic adaptation and matched aging remain unmeasured.", "source_ids": [ "TOTT2026", "DUR2009", "MOR2017", "BOND2026" ] }, { "endpoint": "arbor", "hours": 1, "low": 0.03, "central": 0.3, "high": 3, "rationale": "Chemical signaling can be fast, but completed dismantling of every viable part of the original arbor within one hour has little source support at 100 µM.", "source_ids": [ "VEL2026", "AND2013", "GRIG2021" ] }, { "endpoint": "arbor", "hours": 24, "low": 1, "central": 8, "high": 30, "rationale": "Mouse local MGO reduces epidermal crossings at 24 h, supporting rapid injury. Complete physical loss of a chosen baseline arbor is stricter than fewer crossings, marker loss or viable retraction.", "source_ids": [ "YOU2025", "GRIG2021" ] }, { "endpoint": "arbor", "hours": 168, "low": 5, "central": 25, "high": 60, "rationale": "Mouse degeneration evidence makes completed destruction plausible over days; washout growth enhancement and unknown peptidergic susceptibility prevent translating average fiber loss into an event fraction.", "source_ids": [ "TOTT2026", "YOU2025", "RADU2012" ] }, { "endpoint": "arbor", "hours": 720, "low": 12, "central": 45, "high": 80, "rationale": "Continuous delivery gives local injury more time to finish. The low case emphasizes protected residual branches, viable retraction and intact clearance; even dramatic partial loss can fail the strict endpoint.", "source_ids": [ "TOTT2026", "AND2013", "MOR2017" ] }, { "endpoint": "arbor", "hours": 2160, "low": 22, "central": 60, "high": 90, "rationale": "Three months of persistent local stress supports a material cumulative destruction effect. No source directly tracks original arbors, and surviving remnants can sharply separate this endpoint from density decline.", "source_ids": [ "YOU2025", "TOTT2026", "DUR2009" ] }, { "endpoint": "arbor", "hours": 8760, "low": 30, "central": 70, "high": 95, "rationale": "The central judgment expects repeated local insults eventually to remove some whole baseline arbors. Clearance and resistant neuronal subtypes leave a wide lower interpretation; regrowth does not erase destruction.", "source_ids": [ "TOTT2026", "MOR2017", "RADU2012" ] }, { "endpoint": "arbor", "hours": 43800, "low": 30, "central": 75, "high": 97, "rationale": "The original-arbor event accumulates under a continued challenge, but matched control turnover can narrow the excess. Long duration is not a license to assign certainty without control-arbor survival data.", "source_ids": [ "TOTT2026", "DUR2009", "BOND2026" ] }, { "endpoint": "replacement", "hours": 1, "low": 0.01, "central": 0.1, "high": 1, "rationale": "Both arms begin without the selected arbor. A 10-point excess of non-completion requires at least 10% of healthy controls already to meet both repair criteria, which is very unlikely within one hour.", "source_ids": [ "DUR2009", "RADU2012" ] }, { "endpoint": "replacement", "hours": 24, "low": 0.2, "central": 2, "high": 12, "rationale": "Early healthy completion still limits the possible risk difference. Culture sprouting can occur within 18 h, so one day is not forced to zero, but a connected human arbor with half its heat response is more demanding.", "source_ids": [ "DUR2009", "YOU2025" ] }, { "endpoint": "replacement", "hours": 168, "low": 5, "central": 25, "high": 55, "rationale": "During the first week a delay could separate the groups. MGO-modified substrate can restrict branching, while low-dose mixed cultures can grow better after washout; neither is the joint length-and-heat endpoint.", "source_ids": [ "DUR2009", "RADU2012", "YOU2025" ] }, { "endpoint": "replacement", "hours": 720, "low": 15, "central": 55, "high": 85, "rationale": "More healthy arbors should have an opportunity to complete, while continuous MGO can injure growing endings and their substrate. The timing of human control repair and persistence of viable sprouts remain uncertain.", "source_ids": [ "DUR2009", "YOU2025", "TOTT2026" ] }, { "endpoint": "replacement", "hours": 2160, "low": 20, "central": 65, "high": 92, "rationale": "Months can expose a substantial repair delay or persistent local growth obstacle. Matrix and cell-injury findings support the central case; preserved growth in lower-dose cultures supports the low case.", "source_ids": [ "DUR2009", "RADU2012", "MOR2017" ] }, { "endpoint": "replacement", "hours": 8760, "low": 15, "central": 60, "high": 92, "rationale": "Persistent injury could prevent repair for a year, but some exposed neurons may eventually complete once. Successful first completion remains success even if a later arbor is lost, so the risk difference can shrink.", "source_ids": [ "DUR2009", "RADU2012", "YOU2025" ] }, { "endpoint": "replacement", "hours": 43800, "low": 10, "central": 50, "high": 90, "rationale": "Five years allow occasional successful replacement despite ongoing injury; later loss cannot undo first completion. The high case assumes a persistently hostile skin territory or death before completion, while the low case allows eventual success.", "source_ids": [ "DUR2009", "RADU2012", "TOTT2026", "BOND2026" ] } ] }, { "id": "sarm1", "title": "SARM1 activity", "old_page": "27_sarm1_axon_destruction.html", "exposure": "Hold 10% of endogenous SARM1 assemblies active in the selected skin axons.", "exposure_detail": "From time zero, maintain an active TIR configuration in 10% of endogenous full-length SARM1 assemblies in the mapped neuron's processes within the fixed skin territory: epidermis plus the underlying 1 mm of dermis. Distribute the clamped subset throughout all exposed branches and maintain it during turnover and new growth. Keep abundance and localization endogenous. The other 90% retain normal regulation, so total activity can exceed 10%. Catalysis depends on available substrates and products. NAD, ATP, synthesis, signaling and tissue responses remain free to change. The soma and proximal nerve receive no direct intervention. This is an ideal local molecular challenge, not a drug dose or a measured diabetic active fraction.", "comparison": "H retains normal SARM1 regulation. H+M receives the sustained local activity challenge. Other initiating conditions are matched, including normal support of the connected soma and other branches.", "diabetes_relevance": "Diabetic mouse perturbations support SARM1 involvement, but no reviewed human study measured the specified active fraction in this skin compartment before damage. Brief activation episodes may matter without satisfying the continuous 30-day exposure claim. Expression, cADPR concentration and active-assembly fraction are different quantities.", "diabetes_relevance_belief": { "low": 2, "central": 15, "high": 45, "claim": "In at least 10% of person-weighted reference neuron/arbor draws among 50-year-old men within five years of T2D diagnosis and without DPN at entry, at least 10% of local endogenous SARM1 assemblies remain active continuously for 30 days before arbor loss." }, "summary": "Local SARM1 activity has stronger support for axon damage than for death of the distant soma. The uncertain conversion from active fraction to net NAD loss limits every time estimate. Repair is a separate question: deleting SARM1 can also delay nerve clearance and repair.", "calibration": "These are explicit subjective judgments, not a statistically calibrated human risk model. Low judgments emphasize unknown enzyme abundance, compensation and assay-to-adult transfer. High judgments give more weight to conserved human sensory-axon destruction and assume sustained activity exceeds local supply. The central values lie between those readings. Acute forced activation anchors timing; diabetic animal studies support relevance but do not identify this exposure. No source measures the 10-point claim or the complete-arbor endpoint in the reference population. Year-scale estimates extend mechanistic reasoning and retain broad ranges.", "checks": [ { "label": "Conditional NAD balance", "value": "2.67–266.64 hours", "interpretation": "For three chosen enzyme amounts, 10% activity reaches 90% NAD depletion at these times only with zero net supply and SAM-TIR assay kinetics. This is not a measured human time range." }, { "label": "Compensation changes the result", "value": "4.86 micromolar/hour", "interpretation": "At the chosen 0.1 micromolar total enzyme scenario, this net supply prevents crossing 30 micromolar NAD. Local human supply and enzyme concentration are unknown." }, { "label": "Naive exponential extrapolation fails", "value": "64.7-fold retention mismatch", "interpretation": "Extending the 15-minute forced-TIR NAD loss as one exponential predicts 0.154% retention at 90 minutes; the paper reports about 10%. Rounded summaries do not provide a statistical test." } ], "evidence": [ { "id": "gerdts2015", "title": "Local SARM1 activation and NAD loss", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC4513950/", "year": 2015, "system": "Engineered mouse DRG cultures; fly axons; injured mouse nerve", "n": "Source-specific experiments; nerve NAD: WT 5, knockout 9 mice. Nerves or culture fields are not donor replication.", "result": "Dimerized TIR domains caused local axon destruction. Global activation could also kill somata. NAD fell by 66% at 15 minutes and 90% at 90 minutes. Increased NAD synthesis could preserve axons.", "limits": "Overexpressed isolated TIR domains differ from 10% of endogenous full-length assemblies. Local distal treatment did not establish distant soma death." }, { "id": "essuman2017", "title": "SARM1 has intrinsic NAD-cleaving activity", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC6284238/", "year": 2017, "system": "Biochemical protein assays and neuronal perturbation", "n": "Assay replication; no human population sample.", "result": "The catalytic TIR domain directly cleaves NAD; catalytic-site disruption separates enzyme activity from mere protein presence.", "limits": "An enzyme reaction does not identify the local adult human enzyme amount, net NAD balance or physical destruction threshold." }, { "id": "angeletti2022", "title": "Human SARM1 substrate kinetics", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC8844822/", "year": 2022, "system": "Recombinant human protein, 25 C", "n": "Four basal NAD series; one fitted series at each positive NMN concentration in Table 1.", "result": "Table 2 gives SAM-TIR NAD kcat 0.448/s and Km 69.5 micromolar. Full-length enzyme has substrate and allosteric regulation.", "limits": "The active fragment is only a conditional rate surrogate. Neither tissue-average NAD nor crude-preparation rates identify a cutaneous axon rate." }, { "id": "chen2021", "title": "SARM1 is required for degeneration in human sensory-neuron cultures", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC8171232/", "year": 2021, "system": "Human iPSC-derived sensory neurons; axotomy and vincristine", "n": "One BJ fibroblast donor background, two knockout clones.", "result": "Knockout preserved axons and NAD after injury. Re-expression and dominant-negative experiments support conserved causal machinery.", "limits": "Two clones are not two donors. Acute cultured injury is not diabetes or an intact 50-year-old human neuron." }, { "id": "loreto2025", "title": "Human SARM1 toxicity can be reversed after dysfunction begins", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC7617922/", "year": 2025, "system": "Human iPSC-derived dopamine neurons; vacor", "n": "Five donor lines overall; at least three lines in most non-knockout experiments.", "result": "At four hours, metabolic and mitochondrial abnormalities preceded visible fragmentation. Removing vacor and adding nicotinamide preserved axons for weeks; adding nicotinamide with continued vacor only delayed degeneration over 72 hours.", "limits": "Dopamine neurons are not skin nociceptors. Rescue changed the exposure. Oxygen consumption was not the specified heat-spike assay." }, { "id": "cheng2019", "title": "Partial preservation in STZ-diabetic mice", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC6804630/", "year": 2019, "system": "Male diabetic mice, lifelong Sarm1 deletion", "n": "Four to six mice per skin arm at 25 weeks.", "result": "Skin fiber density and heat responses were partly preserved despite similar endpoint glucose. Earlier source-image reconstruction gave about 37% reduction of the genotype-normalized fiber deficit.", "limits": "Non-littermate WT controls, global deletion, and density rather than tracked whole-arbor fate. The percentage is deficit preservation, not this view's claim probability." }, { "id": "turkiew2017", "title": "High-fat-diet neuropathy and Sarm1 deletion", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC5585053/", "year": 2017, "system": "High-fat-fed mice", "n": "Skin: four mice per WT arm and eight per knockout arm.", "result": "Knockout mice retained epidermal fibers, but the reported fasting hyperglycemia developed only in WT mice.", "limits": "Changed systemic metabolism prevents clean attribution to axonal SARM1." }, { "id": "brazill2024", "title": "Sex-dependent functional protection in diabetic mice", "url": "https://insight.jci.org/articles/view/175159", "year": 2024, "system": "STZ diabetic mice", "n": "Four to seven mice per functional group.", "result": "Deletion protected male electrophysiology over 15 weeks. Females did not develop that deficit. An initial cohort lacked gross skin loss at 14 weeks.", "limits": "This is not a replicated structural skin-rescue comparison or a single heat-responsive axon readout." }, { "id": "hakim2025", "title": "Modest skin-fiber preservation in diet-induced disease", "url": "https://www.nature.com/articles/s41586-024-08535-1", "year": 2025, "system": "High-fat/high-fructose-fed male mice", "n": "Five animals per genotype for 24-week skin.", "result": "Public source data previously gave 42.36 versus 46.97 fibers/mm: +4.61, Welch interval 0.22 to 9.00. Skin P=.042 becomes .084 across three terminal outcomes.", "limits": "No chow skin arms in the source sheet; no genotype-specific glucose histories. Density is not complete original-arbor destruction." }, { "id": "chen2026", "title": "SARM1 re-expression and K641Q in diabetic mice", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12799148/", "year": 2026, "system": "Neonatal AAV in Sarm1-null mice, followed by HFD/STZ", "n": "Eight animals per skin arm.", "result": "Source means: null healthy 28.55, null diabetic 18.67, diabetic WT re-expression 8.87, diabetic K641Q 18.08 fibers/mm. Null diabetic mice retained a 34.6% deficit.", "limits": "K641Q is an amino-acid replacement, not authentic acetylation. It avoids added injury from WT re-expression; it is not an adult catalytic correction." }, { "id": "schmitd2025", "title": "Sarm1 deletion delays peripheral nerve regeneration", "url": "https://pubmed.ncbi.nlm.nih.gov/41061042/", "year": 2025, "system": "Mouse nerve crush and grafting", "n": "Final article abstract verified; detailed n not recovered. The accessible 2024 preprint reports three biological growth replicates, four to six per graft genotype and seven per toe-spread group; these counts are not asserted for the final version.", "result": "Sarm1 loss preserved distal material but delayed the Schwann-cell repair response and recovery. The final article describes faster injury-associated transcription with deletion but delayed distal nerve repair; hind-paw recovery was delayed, ultimately not impaired.", "limits": "Myelinated nerve-trunk clearance differs from regrowth of an unmyelinated epidermal arbor. Protein deletion is not the inverse of the local activation challenge." }, { "id": "czech2023", "title": "TIR-1/SARM1 regulation of regeneration differs from degeneration", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC10121217/", "year": 2023, "system": "Invertebrate axon injury with human SARM1 expression", "n": "Worm axons/animals, not human donors; assay-dependent n.", "result": "TIR-1 inhibited regeneration through signaling separable from its NADase activity; human SARM1 also inhibited regeneration in that system.", "limits": "A human protein in worms is not a human neuron experiment. This cannot calibrate a catalytic activity effect on human replacement." }, { "id": "shi2026", "title": "Caspase cleavage links SARM1 and apoptosis", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12846840/", "year": 2026, "system": "Biochemical and mouse-cell experiments, including embryonic DRG trophic withdrawal", "n": "Three reported biological replicates in key DRG panels; embryo/litter independence not specified.", "result": "SARM1 can participate in apoptosis after caspase cleavage. Trophic-deprived neurons also retain a cleavage-independent route.", "limits": "Embryonic whole-culture withdrawal cannot set an adult soma-death probability after a small local skin perturbation." }, { "id": "kim2026", "title": "Metastable axons can retain ATP despite SARM1 activation", "url": "https://www.nature.com/articles/s41419-026-09329-1", "year": 2026, "system": "RAB7A-L129F DRG model and sensory-neuron AAV in mice", "n": "Independent n not recovered from the accessible early article abstract.", "result": "The October 2 accepted report describes NAD loss without ATP depletion or fragmentation in culture, with autophagy limiting progression. Mouse skin fibers were affected.", "limits": "Abstract-level use only. CMT2B is not diabetes, and no dose conversion to a 10% active fraction is available." }, { "id": "lundback2026", "title": "Base-exchange inhibitors can activate SARM1 at low dose", "url": "https://www.nature.com/articles/s42004-026-02074-8", "year": 2026, "system": "Biochemical, rodent and human-neuron assays", "n": "Assay-specific replication; no treated human DPN cohort.", "result": "Sub-inhibitory concentrations of some inhibitors activated SARM1 and accelerated neurite damage; duration affected preservation.", "limits": "A compound label does not establish a continuously inactive enzyme. These exposures differ from the ideal active-fraction intervention." }, { "id": "ctis2026", "title": "TNV108 first-in-human trial record", "url": "https://euclinicaltrials.eu/ctis-public/view/2025-524258-32-00?lang=en", "year": 2026, "system": "Clinical trial registry", "n": "Ongoing recruiting record; no efficacy result used.", "result": "The registry lists a phase 1 study in healthy adults and patients with DPN.", "limits": "Trial registration and dosing announcements are not evidence of human structural benefit." } ], "estimates": [ { "endpoint": "death", "hours": 1, "low": 0, "central": 0.05, "high": 0.2, "rationale": "The intervention is confined to skin axons. Completed distant soma loss within one hour is very unlikely; no source supports that transfer.", "source_ids": [ "gerdts2015", "shi2026", "loreto2025" ] }, { "endpoint": "death", "hours": 24, "low": 0.01, "central": 0.2, "high": 1, "rationale": "Whole-neuron forced activation can kill cultured somata in a day. Local distal activation is a different intervention, so that result earns little weight here.", "source_ids": [ "gerdts2015", "shi2026", "loreto2025" ] }, { "endpoint": "death", "hours": 168, "low": 0.05, "central": 0.5, "high": 3, "rationale": "Repeated local degeneration could start retrograde stress, but the connected soma has not been directly exposed.", "source_ids": [ "gerdts2015", "shi2026", "loreto2025" ] }, { "endpoint": "death", "hours": 720, "low": 0.1, "central": 1, "high": 6, "rationale": "Loss of one cutaneous arbor is not death of its parent neuron. Evidence for a ten-point excess soma-death risk is weak.", "source_ids": [ "gerdts2015", "shi2026", "loreto2025" ] }, { "endpoint": "death", "hours": 2160, "low": 0.2, "central": 2, "high": 12, "rationale": "Longer deprivation can matter, but other branches and support remain intact. Adult target-population evidence is absent.", "source_ids": [ "gerdts2015", "shi2026", "loreto2025" ] }, { "endpoint": "death", "hours": 8760, "low": 1, "central": 5, "high": 25, "rationale": "Persistent local injury allows indirect routes to soma loss; their size is unmeasured. This raises belief slightly, with wide transfer sensitivity.", "source_ids": [ "gerdts2015", "shi2026", "loreto2025" ] }, { "endpoint": "death", "hours": 43800, "low": 2, "central": 10, "high": 40, "rationale": "Years of local injury could affect soma survival, but no target-specific data establish a ten-point excess. Healthy aging also changes the comparator.", "source_ids": [ "gerdts2015", "shi2026", "loreto2025" ] }, { "endpoint": "function", "hours": 1, "low": 5, "central": 20, "high": 50, "rationale": "Engineered SARM1 can change NAD within minutes. Whether 10% of endogenous assemblies suppress the defined heat signal by 90% within an hour depends on unknown local capacity and supply.", "source_ids": [ "gerdts2015", "chen2021", "loreto2025", "kim2026" ] }, { "endpoint": "function", "hours": 24, "low": 15, "central": 45, "high": 80, "rationale": "Acute sensory-neuron injury supports loss of axon function on this timescale, but the exact active fraction and heat assay have not been tested.", "source_ids": [ "gerdts2015", "chen2021", "loreto2025", "kim2026" ] }, { "endpoint": "function", "hours": 168, "low": 25, "central": 60, "high": 90, "rationale": "Sustained local activity can impair transduction, propagation or terminal integrity. Human sensory-cell evidence supports machinery, not this magnitude.", "source_ids": [ "gerdts2015", "chen2021", "loreto2025", "kim2026" ] }, { "endpoint": "function", "hours": 720, "low": 30, "central": 65, "high": 95, "rationale": "Ongoing activation makes major loss of heat signaling plausible. Compensation and partial arbor survival leave a substantial alternative.", "source_ids": [ "gerdts2015", "chen2021", "loreto2025", "kim2026" ] }, { "endpoint": "function", "hours": 2160, "low": 30, "central": 70, "high": 95, "rationale": "The exposure persists in new branches. This supports sustained functional loss, although regeneration and adaptation can preserve signaling.", "source_ids": [ "gerdts2015", "chen2021", "loreto2025", "kim2026" ] }, { "endpoint": "function", "hours": 8760, "low": 25, "central": 70, "high": 95, "rationale": "Continued exposure keeps the claim plausible; no additional longitudinal human evidence justifies narrowing the range.", "source_ids": [ "gerdts2015", "chen2021", "loreto2025", "kim2026" ] }, { "endpoint": "function", "hours": 43800, "low": 20, "central": 65, "high": 95, "rationale": "Current function can recover through adaptation or replacement. Comparator aging and uncertain long-term compensation slightly lower central belief.", "source_ids": [ "gerdts2015", "chen2021", "loreto2025", "kim2026" ] }, { "endpoint": "arbor", "hours": 1, "low": 0.5, "central": 5, "high": 20, "rationale": "Biochemical disturbance can start quickly, but completed dismantling of an entire adult arbor takes more than an enzyme reaction. The one-hour structural claim has low belief.", "source_ids": [ "gerdts2015", "chen2021", "angeletti2022", "kim2026" ] }, { "endpoint": "arbor", "hours": 24, "low": 10, "central": 35, "high": 70, "rationale": "Local engineered activation destroys axon segments within a day. Unknown endogenous dose and the stronger whole-arbor endpoint lower transfer confidence.", "source_ids": [ "gerdts2015", "chen2021", "angeletti2022", "kim2026" ] }, { "endpoint": "arbor", "hours": 168, "low": 20, "central": 55, "high": 85, "rationale": "Multiple acute studies support local destruction. The 10% assembly challenge is weaker and less characterized than forced TIR expression or toxin dosing.", "source_ids": [ "gerdts2015", "chen2021", "angeletti2022", "kim2026" ] }, { "endpoint": "arbor", "hours": 720, "low": 25, "central": 65, "high": 90, "rationale": "Repeated local activity gives time for complete destruction. NAD replacement and resistant branches still provide plausible alternatives.", "source_ids": [ "gerdts2015", "chen2021", "angeletti2022", "kim2026" ] }, { "endpoint": "arbor", "hours": 2160, "low": 30, "central": 70, "high": 95, "rationale": "The cumulative endpoint permits delayed destruction after compensation initially succeeds. The effect size remains unmeasured in the reference population.", "source_ids": [ "gerdts2015", "chen2021", "angeletti2022", "kim2026" ] }, { "endpoint": "arbor", "hours": 8760, "low": 30, "central": 75, "high": 95, "rationale": "Continued activation supports a substantial cumulative effect. The broad range retains enzyme-abundance, compensation and species-transfer uncertainty.", "source_ids": [ "gerdts2015", "chen2021", "angeletti2022", "kim2026" ] }, { "endpoint": "arbor", "hours": 43800, "low": 25, "central": 75, "high": 95, "rationale": "A large cumulative effect remains plausible, but healthy control arbor turnover can reduce the difference. No five-year human trajectory is observed.", "source_ids": [ "gerdts2015", "chen2021", "angeletti2022", "kim2026" ] }, { "endpoint": "replacement", "hours": 1, "low": 0, "central": 0.1, "high": 1, "rationale": "Both arms start with the same removed arbor. Unless many healthy arbors already complete repair, exposure cannot add ten points of non-completion risk.", "source_ids": [ "schmitd2025", "czech2023", "gerdts2015" ] }, { "endpoint": "replacement", "hours": 24, "low": 0.1, "central": 1, "high": 5, "rationale": "Early non-completion is common in both arms. This floor on healthy repair, rather than lack of molecular activity, keeps the excess-risk claim weak.", "source_ids": [ "schmitd2025", "czech2023", "gerdts2015" ] }, { "endpoint": "replacement", "hours": 168, "low": 2, "central": 10, "high": 30, "rationale": "Some healthy local repair becomes plausible. Re-entering branches encounter the sustained active fraction, but the exact repair timing is not known.", "source_ids": [ "schmitd2025", "czech2023", "gerdts2015" ] }, { "endpoint": "replacement", "hours": 720, "low": 15, "central": 40, "high": 75, "rationale": "Continued activation can damage new branches before they restore both length and heat signaling. Clearance and support effects are not equivalent to protein knockout.", "source_ids": [ "schmitd2025", "czech2023", "gerdts2015" ] }, { "endpoint": "replacement", "hours": 2160, "low": 20, "central": 55, "high": 85, "rationale": "The contrast is more informative after healthy arbors have time to recover. Persistent catalytic stress plausibly prevents a substantial fraction of completions.", "source_ids": [ "schmitd2025", "czech2023", "gerdts2015" ] }, { "endpoint": "replacement", "hours": 8760, "low": 20, "central": 60, "high": 90, "rationale": "Repeated exposure may keep replacement below both thresholds, though an intact parent neuron and compensatory growth can permit success.", "source_ids": [ "schmitd2025", "czech2023", "gerdts2015" ] }, { "endpoint": "replacement", "hours": 43800, "low": 10, "central": 50, "high": 90, "rationale": "This is first completion, not current innervation. A branch that briefly completes repair counts as completed thereafter, so central belief falls despite continued exposure.", "source_ids": [ "schmitd2025", "czech2023", "gerdts2015" ] } ] }, { "id": "ngf", "title": "Loss of NGF support", "old_page": "14_ngf_trophic_support.html", "exposure": "Keep mature NGF available to one selected skin arbor at 10% of its healthy local level.", "exposure_detail": "From time zero until the stated horizon, extracellular mature NGF that can bind neuronal receptors throughout the fixed epidermal territory of the selected arbor is clamped to 10% of the matched healthy local concentration history, including at regrown terminals in the repair challenge. This chosen 90% ligand reduction is not a measured 90% reduction in receptor signaling. Other arbor territories, soma and axons receive no independent clamp. Connected skin, glia, vessels, immune responses, endogenous compensation and downstream receptor signaling remain free to respond. ProNGF, total skin NGF and skin RNA are not the input.", "comparison": "Healthy reference H versus the same reference with this sustained local mature-NGF reduction H+M. The separate repair comparison starts with the same selected-arbor removal in both arms. All reference-neuron and endpoint definitions follow the common specification.", "diabetes_relevance": "Diabetes can alter NGF biology, but the exact severe exposure is not established. Human precursor-like skin NGF fell 97.1% in one small study while skin NGF RNA rose 3.26-fold in another; neither measures accessible mature ligand. Recent mature-NGF blood measurements concern established neuropathy/foot lesions. Early obese/diabetic mouse skin can instead have increased NGF signaling. Capacity under this imposed challenge, diabetes reaching it, and benefit from correcting NGF in diabetes are separate claims. NGF-R100W mouse repair results support capacity to improve damaged tissue; the 1,019-person rhNGF trial did not show broad clinical benefit in established DPN. None establishes incident-DPN prevention.", "diabetes_relevance_belief": { "low": 2, "central": 10, "high": 30, "claim": "Among men aged 50 with type 2 diabetes diagnosed within five years and no DPN at entry, at least 10% of person-weighted reference arbors undergo at least 90% depletion of accessible extracellular mature NGF continuously for at least 30 days before first arbor loss. These prevalence, dose and duration cutoffs are chosen conventions. The numbers are subjective belief in that claim, not measured exposure prevalence." }, "summary": "Low NGF can weaken heat signaling and reduce skin innervation without killing the adult sensory neuron. Whether this challenge completely destroys an original arbor or prevents its eventual replacement is less certain. Adult survival, human NGF-blockade data and NGF-independent injury repair are important counterevidence.", "evidence": [ { "id": "schwartz1982", "title": "Adult sensory neuron number after NGF autoimmunization", "url": "https://pubmed.ncbi.nlm.nih.gov/6180809/", "year": 1982, "system": "Adult rats and guinea pigs; in vivo NGF autoimmunization", "n": "Animal counts and duration not recovered from primary abstract.", "result": "No change in sensory ganglion neuron number or size distribution, despite reduced substance P in ganglia, cord and skin.", "limits": "Primary abstract read. Adult survival counterevidence; no single-arbor tracking, local mature NGF measurement or human five-year follow-up." }, { "id": "lindsay1988", "title": "NGF and BDNF enhance regeneration but are not required for adult sensory neuron survival", "url": "https://pubmed.ncbi.nlm.nih.gov/3249232/", "year": 1988, "system": "Adult rat DRG culture", "n": "Independent culture/animal count not recovered.", "result": "Adult sensory neurons survived without added NGF; growth-factor addition enhanced axonal regeneration.", "limits": "Primary abstract and available article text inspected. Dissociated cultures do not measure fate of a connected human distal-leg neuron; survival and axon growth differ." }, { "id": "bennett1998", "title": "Endogenous NGF regulates adult rat nociceptor sensitivity", "url": "https://pubmed.ncbi.nlm.nih.gov/9749782/", "year": 1998, "system": "Adult rat saphenous skin; local TrkA-IgG for 10–12 days", "n": "152 nociceptor recordings; independent animal counts and structural arm n not recovered.", "result": "Heat-responsive nociceptors decreased from 57% to 32%; epidermal PGP9.5 innervation density decreased 44%. Heat stimulus-response curve shifted right and became shallower; mechanical responses were unchanged.", "limits": "Primary abstract read; full numerical tables not recovered. Sequestration was not a measured 90% mature-ligand clamp. Different units were sampled, and neither baseline-normalized 90% spike loss nor complete loss of identified arbors was measured." }, { "id": "diamond1992", "title": "Adult rat sensory axons regenerate and recover skin function during anti-NGF treatment", "url": "https://pubmed.ncbi.nlm.nih.gov/1313494/", "year": 1992, "system": "Adult rat cutaneous nerve crush and excision; systemic anti-NGF", "n": "Independent animal counts not recovered from accessible primary abstract.", "result": "Post-crush regrowth and recovery of heat, touch and mechanical nociception were unaffected at doses almost eight times those blocking collateral sprouting. Regeneration traversed 4-mm gaps; anti-sprouting activity persisted for at least eight weeks.", "limits": "Primary abstract read; full scanned article could not be recovered. Strong distinction between injury-induced regeneration and collateral sprouting. Nerve-trunk injury and field-level behavioral recovery differ from same-neuron epidermal-arbor repair." }, { "id": "hoffman2011", "title": "Chronic NGF autoimmunization: limited density and assay-dependent heat effects", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC3203207/", "year": 2011, "system": "Rats immunized at six weeks, boosted at ten weeks, assessed through fourteen weeks", "n": "Main contrast: 6 cytochrome-C controls and 10 NGF-immunized rats; NGF rats split after treatment into high/low titre groups of 5 each. Separate pilot had 15 rats.", "result": "Median epidermal length-density was 0.486 in controls, 0.443 with low titres and 0.438 with high titres; density group differences were not significant. High-titre hot-plate responses were about twice as slow, while plantar thermal latency and tail-flick tests were unchanged.", "limits": "Full primary text read. Ratio of high-titre/control medians is 0.9012; no precision for that ratio can be reconstructed. Printed absolute length-density unit scale is unclear. Titres do not quantify free mature NGF. Hot-plate latency is not a linear measure of one neuron's spikes." }, { "id": "brown2014", "title": "Human NGF blockade: distal-leg fiber density in tanezumab trial", "url": "https://clinicaltrials.gov/study/NCT00863772?tab=results", "year": 2014, "system": "Painful osteoarthritis; tanezumab 5 or 10 mg IV every eight weeks; planned 24-week assessment", "n": "220 randomized, 219 treated. Week-24 density-change analysis: 52 at 5 mg, 53 at 10 mg, 53 placebo, with LOCF.", "result": "At 10 cm above the lateral malleolus, mean density change was +0.31 (SD 4.17), −0.91 (SD 3.86) and −0.52 (SD 4.15) fibers/mm for 5 mg, 10 mg and placebo. Placebo baseline mean was 12.43 fibers/mm.", "limits": "Sponsor-posted numerical results and primary article abstract/sections read (PMID 25073573). Early clinical hold: most patients received one or two doses. No verified local 90% depletion, subtype isolation or identified-arbor fate. Missing values use last observation carried forward. Clinical efficacy confirms activity, not the specified local dose." }, { "id": "bramson2015", "title": "NGF blockade in painful DPN: pain and skin fiber outcomes", "url": "https://clinicaltrials.gov/study/NCT01087203?tab=results", "year": 2015, "system": "Established painful DPN; tanezumab 20 mg SC at day 1 and week 8", "n": "38 tanezumab and 35 placebo participants in registry density analysis.", "result": "Week-16 distal-calf least-squares density changes were −0.15 (SE 0.21) versus +0.40 (SE 0.22) fibers/mm. Pain improved in the published study.", "limits": "Registry result and primary article abstract read (DOI 10.1111/pme.12677). LOCF and early hold apply. Covariance of adjusted means is not supplied, so no reconstructed adjusted confidence interval is claimed. Established DPN is not the healthy reference or incidence prevention." }, { "id": "apfel1998", "title": "Phase II rhNGF treatment of symptomatic diabetic polyneuropathy", "url": "https://pubmed.ncbi.nlm.nih.gov/9748012/", "year": 1998, "system": "Randomized human rhNGF supplementation for six months", "n": "250 patients.", "result": "Three prospectively specified multiple-endpoint analyses favored NGF (p=0.032, 0.008, 0.005), including sensory examination and quantitative sensory tests.", "limits": "Primary abstract read. Does not establish terminal destruction, same-neuron repair or correction of measured endogenous mature NGF deficiency." }, { "id": "apfel2000", "title": "Phase III rhNGF trial did not confirm broad clinical benefit", "url": "https://pubmed.ncbi.nlm.nih.gov/11056593/", "year": 2000, "system": "Established diabetic polyneuropathy; rhNGF 0.1 microgram/kg SC three times weekly for 48 weeks", "n": "504 rhNGF and 515 placebo participants.", "result": "Primary lower-limb neuropathy score comparison did not show benefit (p=0.25); 418 versus 461 completed the regimen. Most secondary outcomes also did not show benefit.", "limits": "Primary abstract rechecked; prior repository full-table calibration retained separately. This tests a treatment regimen in established disease, not selective restoration of measured terminal NGF or incident DPN prevention. No identified-arbor structural endpoint." }, { "id": "anand1996", "title": "Human diabetic skin NGF and sensory dysfunction", "url": "https://pubmed.ncbi.nlm.nih.gov/8640566/", "year": 1996, "system": "Human early diabetic polyneuropathy; skin measurements and sensory/autonomic testing", "n": "Participant counts and diabetes-type breakdown not recovered from accessible primary abstract.", "result": "Lower skin NGF and substance P accompanied length-dependent small-fiber dysfunction and reduced axon-reflex vasodilation.", "limits": "Primary abstract read. Cross-sectional association does not establish temporal order, free mature ligand at terminals or a sustained 90% reduction." }, { "id": "diemel1999", "title": "Human diabetic skin NGF mRNA was increased", "url": "https://pubmed.ncbi.nlm.nih.gov/10229303/", "year": 1999, "system": "Lateral calf skin biopsies", "n": "19 diabetic participants and 8 controls for RNA; separate thermal comparison used 24 controls.", "result": "Mean NGF mRNA was 5.12 (SD 3.88) in diabetes versus 1.57 (SD 0.95) in controls: a 3.26-fold ratio. No correlation with thermal measures was found.", "limits": "Primary abstract read. RNA does not measure release, processing or local mature ligand. Does not refute a processing defect, but refutes using all NGF assays as equivalent." }, { "id": "yiangou2002", "title": "Human diabetic skin showed lower precursor-like NGF immunoreactivity", "url": "https://pubmed.ncbi.nlm.nih.gov/12365567/", "year": 2002, "system": "Human skin from subclinical diabetic neuropathy and controls", "n": "6 people per group.", "result": "Median optical intensity of a 53-kDa precursor-like band was 1.5 in diabetes versus 52.0 in controls, a 97.1% reduction in the ratio of medians.", "limits": "Primary abstract read. The band is not extracellular mature NGF. This superficially similar percentage cannot establish the 90% mature-ligand challenge or its duration." }, { "id": "christianson2003", "title": "NGF increased branching without restoring diabetic cutaneous innervation", "url": "https://pubmed.ncbi.nlm.nih.gov/12618126/", "year": 2003, "system": "STZ-diabetic mice; two weeks of intrathecal NGF, GDNF or neurturin", "n": "Animal counts and arm-level numerical uncertainty not recovered.", "result": "NGF increased axon branching but did not restore cutaneous innervation density; GDNF and neurturin increased both.", "limits": "Primary abstract read. A null without arm uncertainty is not equivalence. Different delivery routes and trophic factors matter. Branching, density and repair to the specified function/length thresholds are distinct." }, { "id": "cheng2012", "title": "Early diabetic mouse skin can show increased NGF-dependent innervation", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC3225563/", "year": 2012, "system": "Young db/db mice; anti-NGF from six to eight weeks", "n": "4 mice per group; sections were not independent animals.", "result": "TrkA-positive epidermal fibers rose about 1.5–2-fold during early allodynia. Anti-NGF reduced this increase in db/db mice but did not change control db/+ density.", "limits": "Full primary text checked. Young genetic obesity/diabetes is not a 50-year-old human. Changes include marker expression and total density. No later preventive benefit from prolonged blockade was established." }, { "id": "koui2025", "title": "Obese mouse skin NGF sensitization can precede later axon loss", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC11952797/", "year": 2025, "system": "Diet-induced-obesity mouse ear skin; ex vivo nociceptor calcium imaging", "n": "Late density figure reports n=4 per group; acute inhibitor panels report n=6. These panel counts are not inflated by imaging fields.", "result": "Keratinocyte NGF/TrkA/PI3K signaling increased sensory responses before later axon loss. Brief anti-NGF exposure reduced calcium responses.", "limits": "Full primary text inspected. Ear skin and capsaicin-evoked calcium differ from the target. Acute neutralization did not test whether sustained blockade prevented or caused later structural loss." }, { "id": "ngfr100w2025", "title": "Local NGF-R100W/VEGFA mRNA treatment increased diabetic mouse skin innervation", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12278795/", "year": 2025, "system": "Male db/db mice treated from age 16 to 22 weeks with plantar mRNA-LNP", "n": "Figure 5 structural comparisons report n=6 per group; blood-flow comparisons n=10. Culture cell counts are not donor n.", "result": "NGF-R100W monotherapy increased innervation; combination with VEGFA gave stronger structural and functional effects. The text reports GAP43-positive density of 23.4 fibers/mm with combination versus about 11 in the control group.", "limits": "Full text/captions read; total-PGP9.5 figure coordinates were not recovered. GAP43 is a repair-associated marker, not total arbor survival. Engineered ligand and combination treatment do not identify physiological native-NGF correction, quantitative synergy or incident prevention." }, { "id": "danos2025", "title": "Acute NGF stimulation reduced axon-maintenance proteins and accelerated post-cut fragmentation", "url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC12005098/", "year": 2025, "system": "E13.5 mouse sensory cultures, DIV7–8; deprivation/addback and axotomy", "n": "Figure 8 protein and fragmentation panels each use 4 independent cultures from different litters; SARM1 panel uses 3.", "result": "At one/two hours after cutting, NMNAT2 remaining was approximately 50%/20% without NGF and 25%/15% with NGF. Acute deprivation delayed fragmentation; NGF addback accelerated it.", "limits": "Final full primary article checked, not preprint. Bcl-xL overexpression preserved survival during prolonged deprivation. A severed embryonic-derived axon is not the intact adult target. The result defeats a universal NGF-equals-maintenance rule; it does not establish that chronic withdrawal protects intact human arbors." }, { "id": "sambataro2025", "title": "Mature and precursor NGF differed in established diabetic foot disease", "url": "https://pubmed.ncbi.nlm.nih.gov/41056698/", "year": 2025, "system": "Cross-sectional established T2D neuropathy with and without foot lesions", "n": "374 adults with T2D neuropathy (106 without lesions, 119 with lesions without macrovascular disease, 149 revascularized with lesions), plus 53 healthy controls.", "result": "Advanced foot-lesion groups had lower circulating mature NGF and higher proNGF than less advanced lesion/no-lesion groups.", "limits": "Primary abstract read. Blood concentration, advanced ulceration/ischemia and prevalent neuropathy do not identify accessible mature NGF at intact pre-loss skin terminals; no 90%/30-day exposure prevalence is supplied." }, { "id": "lankford2013", "title": "Adult rat nerve regeneration during NGF antibody treatment", "url": "https://pubmed.ncbi.nlm.nih.gov/23531437/", "year": 2013, "system": "Adult rat unilateral sciatic crush; weekly muMab911 for eight weeks from day minus one", "n": "Thirty animals across separate experiments: sixteen behavior/regeneration, six pharmacokinetics, eight retrograde labeling; not thirty per endpoint.", "result": "No detected impairment of gait recovery or nerve regeneration; a small reduction in soma size occurred. Plasma antibody exposure was measured.", "limits": "Independent reviewer recovered primary abstract and available source sections. No measured 90% local mature-NGF clamp, no tracked epidermal arbor, and no human transfer rate." } ], "calibration": "Every plotted value is a subjective q: belief that H+M adds at least ten percentage points to the specified endpoint probability. Low/central/high are alternative evidence-transfer judgments, not measured human risk, confidence limits or a fitted posterior. Low judgments emphasize residual ligand, adult compensation, human blockade and injury-induced regrowth; high judgments emphasize selective vulnerability of NGF-responsive heat neurons and sustained near-depletion. The central functional judgment after weeks is only slightly above even odds because rat heat-responsive units can disappear, but the exact 90%-spike-loss criterion is unmeasured. A twofold hot-plate latency is not converted into spike loss. Death remains low because adult sensory survival differs from developmental dependence. Complete-arbor loss is less certain than reduced density. Long-horizon arbor and replacement q can decline as healthy arbors turn over or delayed repair catches up. No injury coefficients or human fate rates were fitted.", "checks": [ { "label": "New human density constraint", "value": "10-mg tanezumab minus placebo: −0.39 fibers/mm; descriptive unadjusted 95% interval −1.916 to +1.136.", "interpretation": "Calculated from mean changes, SDs and n=53/53 at the exact distal-leg site. This is −3.14% of placebo baseline density (interval −15.41% to +9.14%). It does not exclude a 10% density loss and does not measure complete-arbor events or the imposed exposure." }, { "label": "Dose-arm cross-check", "value": "5-mg tanezumab minus placebo: +0.83 fibers/mm; descriptive unadjusted 95% interval −0.762 to +2.422.", "interpretation": "n=52/53; no consistent large density decline across dose arms. Trial truncation and LOCF constrain inference." }, { "label": "Chronic rat source check", "value": "High-titre/control median epidermal length-density ratio 0.438/0.486 = 0.9012; 9.88% lower.", "interpretation": "n=5/6, group difference was not significant. The source's reduced-density wording does not demonstrate a large common complete-arbor effect." }, { "label": "Endpoint identification check", "value": "A 44% density decrease can arise from 44% shortening of every equal-sized arbor or complete loss of 44% of them.", "interpretation": "Constructed algebraic example, not biological data. The same density leaves complete destruction at 0% or 44%; source density cannot identify the target event." }, { "label": "Exposure-form check", "value": "Precursor-like band reduction 97.12%; skin RNA mean ratio 3.261.", "interpretation": "Both source calculations are correct, but neither calibrates the proposed mature extracellular NGF depletion." } ], "estimates": [ { "endpoint": "death", "hours": 1, "low": 0, "central": 0.01, "high": 0.1, "rationale": "A local ligand reduction is unlikely to complete irreversible death of more than an additional tenth of mature connected somata within one hour. This is a biological judgment, not a measured event rate.", "source_ids": [ "schwartz1982", "lindsay1988" ] }, { "endpoint": "death", "hours": 24, "low": 0, "central": 0.05, "high": 1, "rationale": "One day permits signaling changes, but the connected adult soma retains other support. Adult NGF deprivation does not reproduce the rapid death seen in immature cultures.", "source_ids": [ "schwartz1982", "lindsay1988" ] }, { "endpoint": "death", "hours": 168, "low": 0.1, "central": 0.5, "high": 3, "rationale": "Adult withdrawal studies show phenotype and terminal changes without corresponding sensory-soma loss. A severe local terminal deficit alone remains an unlikely cause of a ten-point death excess.", "source_ids": [ "schwartz1982", "bennett1998" ] }, { "endpoint": "death", "hours": 720, "low": 0.2, "central": 1, "high": 5, "rationale": "Several weeks of deprivation can change skin fibers and heat behavior, while adult sensory survival is comparatively independent of NGF. Localizing the clamp to one arbor further limits soma deprivation.", "source_ids": [ "schwartz1982", "hoffman2011" ] }, { "endpoint": "death", "hours": 2160, "low": 0.5, "central": 2, "high": 10, "rationale": "Prolonged terminal stress could affect the soma, but adult animal survival and human blockade evidence argue against common completed death. No human soma-count study measures this contrast.", "source_ids": [ "schwartz1982", "brown2014" ] }, { "endpoint": "death", "hours": 8760, "low": 1, "central": 5, "high": 20, "rationale": "Long exposure leaves time for secondary injury. Other arbors, the central axon and intrinsic adult survival programs remain available; clinical blockade does not establish a large death effect.", "source_ids": [ "lindsay1988", "brown2014" ] }, { "endpoint": "death", "hours": 43800, "low": 2, "central": 10, "high": 35, "rationale": "Five-year adult-human local withdrawal is unmeasured. A cumulative secondary-death effect is possible, but adult independence and the local exposure scope keep the central judgment low.", "source_ids": [ "schwartz1982", "lindsay1988", "brown2014" ] }, { "endpoint": "function", "hours": 1, "low": 0.1, "central": 1, "high": 5, "rationale": "NGF signaling can change excitability rapidly, but losing at least 90% of the baseline heat-evoked spikes in enough reference neurons within one hour is a much stronger event.", "source_ids": [ "bennett1998", "koui2025" ] }, { "endpoint": "function", "hours": 24, "low": 2, "central": 10, "high": 30, "rationale": "Channel regulation and sensitization can change over a day. Evidence for near-total loss of the specified central-axon heat response at this time is limited.", "source_ids": [ "bennett1998", "hoffman2011" ] }, { "endpoint": "function", "hours": 168, "low": 15, "central": 45, "high": 75, "rationale": "The local adult-rat withdrawal study approaches this time and loses many heat-responsive units. Transfer to the chosen 47 C assay, human subtype and 10% residual-spike cutoff remains uncertain.", "source_ids": [ "bennett1998", "hoffman2011" ] }, { "endpoint": "function", "hours": 720, "low": 20, "central": 55, "high": 85, "rationale": "Bennett's 57% to 32% heat-responsive change makes a material functional effect plausible after weeks. Only a modest fraction of the selected baseline-responsive population must cross the strict cutoff, but neither that fraction nor the exact dose transfer is measured.", "source_ids": [ "bennett1998", "hoffman2011", "brown2014" ] }, { "endpoint": "function", "hours": 2160, "low": 20, "central": 55, "high": 85, "rationale": "Persistent withdrawal can maintain a less heat-responsive state. Mixed behavioral assays and human blockade results leave substantial room for residual function and compensation.", "source_ids": [ "bennett1998", "hoffman2011", "brown2014" ] }, { "endpoint": "function", "hours": 8760, "low": 20, "central": 55, "high": 85, "rationale": "The same plausible loss of heat transduction persists, but no matched year-long spike study resolves whether compensation restores more than 10% of baseline firing.", "source_ids": [ "bennett1998", "brown2014" ] }, { "endpoint": "function", "hours": 43800, "low": 15, "central": 50, "high": 85, "rationale": "A long-lived low-response state remains plausible. Years of adaptation and loss of function in the healthy comparator may reduce the excess; there is no five-year target assay.", "source_ids": [ "bennett1998", "brown2014", "bramson2015" ] }, { "endpoint": "arbor", "hours": 1, "low": 0, "central": 0.01, "high": 0.1, "rationale": "Complete physical dismantling of an initially healthy arbor within an hour is very unlikely after local NGF reduction alone; immediate signaling change is not destruction.", "source_ids": [ "bennett1998", "danos2025" ] }, { "endpoint": "arbor", "hours": 24, "low": 0.1, "central": 0.5, "high": 3, "rationale": "One day could initiate branch changes, but there is no adult evidence that this local challenge completely removes at least an extra tenth of baseline arbors this quickly.", "source_ids": [ "bennett1998", "danos2025" ] }, { "endpoint": "arbor", "hours": 168, "low": 3, "central": 15, "high": 40, "rationale": "Adult withdrawal can reduce epidermal innervation within about two weeks. Partial shortening and whole-arbor disappearance are unresolved, so the seven-day complete-loss claim remains uncertain.", "source_ids": [ "bennett1998", "hoffman2011" ] }, { "endpoint": "arbor", "hours": 720, "low": 8, "central": 30, "high": 60, "rationale": "The 44% rat density decrease supports structural capacity, but could consist largely of partial branch loss or reduced renewal. Human blockade and the weaker chronic-rat result oppose automatic conversion to complete destruction.", "source_ids": [ "bennett1998", "hoffman2011", "brown2014" ] }, { "endpoint": "arbor", "hours": 2160, "low": 10, "central": 40, "high": 70, "rationale": "Continuing withdrawal gives incomplete damage more time to reach full arbor destruction. Direct human density data temper this inference, and none follows the original arbor.", "source_ids": [ "bennett1998", "hoffman2011", "brown2014" ] }, { "endpoint": "arbor", "hours": 8760, "low": 10, "central": 40, "high": 75, "rationale": "Sustained exposure could cumulatively eliminate arbors. Unknown native turnover and replacement can produce the same density, so a ten-point excess of original-arbor events remains unresolved.", "source_ids": [ "bennett1998", "brown2014" ] }, { "endpoint": "arbor", "hours": 43800, "low": 5, "central": 30, "high": 75, "rationale": "Both arms can lose original arbors over five years. Healthy turnover could erase much of the excess cumulative event risk even if deprivation accelerates loss; the target's turnover history is unmeasured.", "source_ids": [ "bennett1998", "brown2014" ] }, { "endpoint": "replacement", "hours": 1, "low": 0, "central": 0.01, "high": 0.1, "rationale": "Both arms begin with the same arbor removal. Within one hour, completion of both half-length and half-response criteria is unlikely even in the healthy arm, leaving little excess non-completion.", "source_ids": [ "diamond1992" ] }, { "endpoint": "replacement", "hours": 24, "low": 0.1, "central": 1, "high": 5, "rationale": "Very early growth differences need not create a ten-point difference in completed functional repair. The healthy completion rate at one day is unknown.", "source_ids": [ "diamond1992", "lindsay1988" ] }, { "endpoint": "replacement", "hours": 168, "low": 5, "central": 20, "high": 50, "rationale": "NGF promotes branching and growth, so a delay is plausible. Injury-triggered regeneration can proceed during anti-NGF treatment, and the exact healthy repair schedule is unknown.", "source_ids": [ "diamond1992", "lindsay1988", "christianson2003", "lankford2013" ] }, { "endpoint": "replacement", "hours": 720, "low": 10, "central": 40, "high": 70, "rationale": "During a plausible repair interval, less branching or heat recovery could delay completion. Adult anti-NGF regeneration counterevidence prevents treating the delay as obligatory.", "source_ids": [ "diamond1992", "christianson2003", "ngfr100w2025", "lankford2013" ] }, { "endpoint": "replacement", "hours": 2160, "low": 10, "central": 35, "high": 70, "rationale": "A remaining delay may persist, but injury-induced regrowth and other growth factors can allow the original neuron to complete repair. Density rescue alone does not determine both required thresholds.", "source_ids": [ "diamond1992", "christianson2003", "ngfr100w2025", "lankford2013" ] }, { "endpoint": "replacement", "hours": 8760, "low": 5, "central": 25, "high": 60, "rationale": "Many delayed arbors may eventually satisfy the first-completion criteria despite continued low NGF. The source regeneration result weighs against assuming a permanent block.", "source_ids": [ "diamond1992", "christianson2003", "lankford2013" ] }, { "endpoint": "replacement", "hours": 43800, "low": 3, "central": 20, "high": 60, "rationale": "This is failure ever to complete repair by five years, not current arbor quality. Long recovery time makes persistent ten-point excess less likely than a month-scale delay, with no direct long-term human evidence.", "source_ids": [ "diamond1992", "brown2014", "christianson2003", "lankford2013" ] } ] } ] }