Coverage and boundaries of the hypothesis collection

Reviewed 13 September 2026. Index · Probability method.

Scope of the search

This is a broad, mechanism-oriented literature assessment of diabetes-related structural injury or failed replacement of human peptidergic C-fiber terminals. Other neuropathies serve as calibration examples. It includes 45 causal propositions, not 45 mutually exclusive etiologies or a census of every molecular regulator.

The assessment combined the preceding fatty-acid analysis, the repository's topic histories and source corrections, and searches of primary papers indexed through PubMed/PMC and journal sites. Search families combined diabetes/diabetic neuropathy with a mechanism or alias and with sensory neuron, DRG, Schwann cell, axon, skin denervation or intraepidermal fiber density. Follow-up searches checked human evidence, structural endpoints, intervention results and retractions where relevant. Searches were expanded when a separate initiating exposure appeared, as with amylin aggregation.

This was not a database-complete systematic review with a screened denominator. A family absent here has not been proved absent from the literature. The retrieved primary studies support the coverage stated below; a paper mentioning a pathway is not automatically evidence that it damages this fiber class. No claim is made that every citation was available in full text. Local public-data reanalyses are identified as such.

What is grouped together

A separate file is useful when the hypothesis introduces a different exposure, necessary compartment transfer, limiting function, execution mechanism or loss-versus-replacement explanation. Several proteins governing the same causal requirement remain branches inside that file. A file can be broader than one experimentally resolved molecular pathway; its probability then prices that broader union.

Literature term or related family	Where it belongs	Distinction retained
Aldose reductase, sorbitol, SORD, myo-inositol, polyol-associated redox changes	01	Elevated flux, poor disposal and a toxic consequence are separate claims
Glucosamine, hexosamine biosynthesis, O-GlcNAc	02	O-GlcNAc is an optional mediator, not assumed from glucosamine toxicity
Glycated matrix/laminin/collagen, AGE receptors, reactive carbonyls	03, 04, 05	Matrix effects, receptor signaling and direct chemical injury can differ
Glucose oxidation, respiratory reserve, AMPK/PGC-1, NAD balance	06, 07, 27, 28	A narrow overload model is distinct from energy shortage and specific NAD-consuming executors
Palmitate, oleate, acylcarnitines, lipid droplets, oxidized LDL	08, 09, 17	Mixture, unbound exposure and neuronal versus glial routing matter
Serine/alanine balance, PHGDH, SPT, deoxysphingolipids	10	Human inherited toxicity supports capability, not the diabetic exposure bridge
Canonical ceramide species, sphingolipid balance, beneficial fatty-acid products	11, 43	Toxic accumulation differs from a limiting protective product
Insulin/IGF, C-peptide, NGF/TrkA, PTEN and muscarinic growth control	12, 13, 14, 44	Deficiency, resistance, adult dependence and growth restraint differ
Capillary dysfunction, hypoxia, nitric-oxide availability, DAG/PKC, nerve-barrier changes	15, 16	Structural vessel abnormalities need a consequential delivery deficit
Remak-cell support, Schwann mitochondrial/lipid metabolism, satellite glia, organelle transfer	17, 18	C-fibers are unmyelinated; supportive glia still matter
TNF/TLR/NF-kappaB, inflammatory cytokines, protective macrophages, complement	19, 20, 22	Immune activation has no universal harmful sign
NLRP3, gasdermin, TBK1-associated pyroptosis	21	Inflammasome signaling and actual lytic peripheral injury are different
UPR/CHOP, integrated stress signaling, proteostasis, autophagy/lysosomes, mitophagy	23, 24	Adaptive stress responses and static markers do not establish harmful flux
DRP1/MFN/OPA1, mitochondrial fragmentation, transport, HDAC6/tubulin	25, 26	Morphology or motion must change useful local maintenance
SARM1/NMNAT2, PARP, GPX4/iron-dependent lipid injury, caspase-associated soma death	27–30	These are different execution or compartment claims, not interchangeable death markers
Sodium/calcium channels, calcium handling, ion-pump demand	31	Excess excitability must cross a structural injury threshold
Keratinocytes, terminal niche, repair factors, senescence/SASP	32, 33	Failed replacement may lower density without increased destruction
Microbiome, endotoxin, short-chain fatty acids and other microbial products	34	Transferable community effects do not identify one mediator
Treatment-induced neuropathy, rapid HbA1c reduction, severe lows, glucose variability	35, 41, 42	Rate of improvement, absolute fuel deprivation and waveform shape differ
Nerve swelling, entrapment, repetitive mechanical stress	36	A focal lesion is not automatically the cause of symmetric distal DPN
Metformin-associated B12 deficiency, renal failure/uremic exposure	37, 38	These use explicit treatment/comorbidity populations
ROS/RNS, peroxynitrite, antioxidant buffering	39	Damaging chemistry does not require one mitochondrial source
Metabolic memory, chromatin persistence, lasting exposure effects	40	Persistent causal regulation differs from irreversible old damage
hIAPP/amylin oligomers and amyloid-related peptide toxicity	45	Harmful aggregation differs from normal peptide signaling

Aliases in this table locate a family. They do not confer a separately assessed probability on every named protein, lipid or process.

Evidence deliberately kept outside a direct C-fiber claim

Myelin-specific injury. The MLKL diabetic-mouse study has strong manipulations of myelin decompaction and conduction. Its S441-linked myelin function is distinguished from canonical RIP3-dependent necroptotic phosphorylation. It cannot be copied directly into a claim that unmyelinated C-fibers die by necroptosis. A Remak-cell/support consequence would need the bridge in hypothesis 17; a distinct C-fiber necroptosis price is not justified by that myelin experiment. Primary study.

CNS and pain-only mechanisms. Spinal microglial activation, central sensitization or altered withdrawal behavior can change pain without preserving a peripheral axon. Conversely, a central intervention could conceivably influence peripheral maintenance; that return path must be demonstrated rather than assumed. The TBK1 study's central knockdown and systemic treatment had different structural results, as discussed in 21. The reported diabetes-associated dimethylglyoxal study concerns blood-brain barrier/cognitive outcomes; it does not supply a separate peripheral-fiber cause here. Primary CNS study.

New cell-death labels or regulator lists. A ferroptosis file is included because that specific injury claim has a retrievable experimental nerve literature, but it receives a low price for the human bridge. Cuproptosis, disulfidptosis and a broad “necroptosis” label are not assigned separate human C-fiber probabilities merely from review terminology or marker enrichment. This is a limitation of the assessed evidence, not a declaration that these processes cannot occur. Likewise, miRNAs, lncRNAs and individual kinase targets are grouped by the function their manipulation is proposed to change.

Other causes of neuropathy. Alcohol, chemotherapy, infections, autoimmune disease, inherited neuropathies and nutritional deficiencies unrelated to diabetes are not added as extra diabetes mechanisms. When they establish a useful causal capability, that narrower inference is stated locally. Foot-ulcer trauma is also not treated as an explanation of the earlier transition to incident DPN.

Calibration examples and the inference they buy

Example	Useful inference	Remaining gap
SORD-associated inherited axonal neuropathy	Severe polyol-disposal failure can injure human nerves	Ordinary diabetic exposure; predominantly motor inherited phenotype
HSAN1/deoxysphingolipid disease	The lipid chemistry can damage human sensory nerves	Concentration, duration and prevalence in type 2 diabetes
Human sensory-neuron SARM1 perturbation under other injuries	Human axons possess the execution route	Activation during diabetes and mature fiber subtype
Amyloid polyneuropathy and precursor suppression	A circulating amyloid precursor can causally affect human neuropathy	Amylin is a different precursor with a different local exposure
Nondiabetic advanced CKD with skin denervation	A renal-disease state can accompany human small-fiber loss independently of diabetes	Which exposure is causal and how much it contributes in diabetic CKD
Human randomized DPN intervention results	Challenge an unqualified culture/animal-to-human efficacy inference	A drug trial may not test the exact mediator, stage or structural endpoint

These are selected anchors, not a denominator from which an empirical animal-to-human success rate can be computed. Primary citations and sample-size qualifications are in the linked mechanism files and the method.

Where this leaves the original question

The collection identifies many physically plausible routes and makes their weak links explicit. It does not identify a uniquely proven cause in an individually typed human peptidergic neuron. The strongest confidence often concerns a general capability—energy failure, oxidative injury or axon execution—while the largest uncertainty concerns whether a specific diabetic exposure engages that capability, at the right place and time, enough to account for material human loss.

That separation is the intended use of the prices: decide which missing human bridge carries the uncertainty, rather than mistake a long molecular pathway or an attractive rescue for a completed causal explanation.