Amylin aggregation

Amyloidogenic human islet amyloid polypeptide (hIAPP/amylin) forms harmful species that reach peripheral sensory compartments and materially damage peptidergic C-fibers in a subgroup of type 2 diabetes.

Scope: Default early type 2 DPN, with a subgroup identifiable by hIAPP aggregation/exposure independently of treatment response. The hypothesis concerns harmful aggregation states, not all physiological amylin signaling.

Causal role: systemic peptide exposure with local aggregation toxicity. Unless the scope specifies an exception, materiality means at least 20% less five-year net terminal loss in an identifiable subgroup representing at least 10% of incident DPN cases in the stated population. The target remains human peptidergic C-fibers when a study measures a broader endpoint.

Source assessment

Base judgment: 20.3%; skeptical–optimistic sensitivity: 1.0%–55.1%. Subjective probabilities of the causal proposition below, not effect sizes. Decimal places only show arithmetic. Reviewed 13 September 2026. Probability method · Collection index.

Reasoning from the biological requirements

A secreted peptide becomes an axonal toxin only if production, delivery, aggregation and clearance yield a harmful local species. A high circulating total concentration is neither necessary nor sufficient for a high local oligomer burden. The species must contact the axon or a connected support compartment at an effective concentration and duration. Membrane injury, organelle stress and failed clearance are alternative downstream routes. Aggregates observed after denervation could be retained debris or parallel diabetes pathology; timing and a causal effect of preventing aggregation are therefore essential.

Probabilities of the necessary claims

Conditional prices are skeptical / base / optimistic. Each price after the first assumes all earlier requirements are true. Their order is an order for assessing joint belief, not necessarily a linear biological pathway. The cumulative column uses the base scenario. Temporal order, connected compartments and the causal prevention comparison are included in these requirements, without extra generic multipliers.

See the claims and probabilities table above.

R1

Human tissue association and hIAPP models support the possibility, but longitudinal human aggregation state is uncertain.

R2

Bulk plasma and skin deposits do not directly measure exposure at the terminal; acute injection doses are a major extrapolation.

R3

Aggregation controls and animal fiber counts strengthen causality, while human subtype and toxic threshold remain unresolved.

R4

A small case-control tissue study does not determine incidence, attributable loss or five-year preservation.

Evidence that moves the judgment

• Albariqi et al. found more IAPP-positive oligomers and lower IENFD in skin from six people with T2D neuropathy versus nine non-T2D controls. There was no diabetes-without-neuropathy comparison establishing DPN specificity. hIAPP-transgenic mice developed fiber loss without hyperglycemia; non-aggregating comparators and inhibition of aggregation supported an aggregation-dependent effect in exposure models. 2023 primary study.

• The same paper calculates that an effective intravenous bolus produced a maximal plasma concentration about 1,000 times that in its transgenic model or human T2D. The chronic transgenic findings partly address that limitation, but neither converts an acute injected dose into a human toxic threshold. Dermal deposits also require a bridge to epidermal terminal injury. Exposure and anatomical details.

• Other amyloid neuropathies provide a human capability anchor: the 225-participant APOLLO randomized trial showed improved neuropathy outcomes when transthyretin production was suppressed. This demonstrates that an amyloid precursor can be causally relevant to human neuropathy; it does not establish that amylin has the same dose, tissue distribution or effect in diabetes. Adams et al., 2018.

Why these odds and how uncertain they are

This is more than an analogy to amyloidosis: there is a diabetes-relevant peptide, human tissue localization and controlled animal structural evidence. Those features keep the probability well above an unsupported possibility. The largest uncertainty is the local human exposure bridge. Aggregation can concentrate material, but invoking concentration without measuring it cannot resolve a dose mismatch. The small human comparison also cannot distinguish a cause of neuropathy from a feature of diabetes generally.

The skeptical scenario assigns the human deposits to parallel or downstream pathology and the animal injury to a special exposure regime. The optimistic scenario assumes persistent local oligomers expose a vulnerable subgroup at concentrations not captured by bulk plasma assays. Human amylin's aggregation behavior also means an ordinary mouse diabetes model can miss this candidate; absence in that model is not a clean human null.

Conditional inputs: base [75, 60, 75, 60]%; skeptical [40, 25, 40, 25]%; optimistic [90, 85, 90, 80]%. Products give the headline prices. The span measures sensitivity to these interpretations; it is not a statistical confidence interval.

Numerical update example: Public human evidence relating local harmful hIAPP species to subsequent fiber loss, with diabetic controls and a credible exposure-response relationship, would raise requirement 2. Evidence that the observed deposits are biologically inert at human exposure would lower it. Changing only requirement 2 from 60% to 85% raises the whole price to 28.7%; reducing it to 20% lowers the price to 6.8%. These are hypothetical evidence updates. Even resolving that one requirement with certainty would leave a complete price of 33.8% if the others stayed unchanged.

Overlap and boundary

Can converge on oxidative injury, organelle maintenance or calcium imbalance. This is distinct from C-peptide deficiency and insulin trophic signaling. Other amyloid neuropathies are calibration examples, not extra diabetes hypotheses.