# 24. Slow completion of intracellular autophagic disposal

**Best judgment: about 11%; skeptical–favorable sensitivity: 0.3%–55%.** These are subjective probabilities of the complete human proposition, not treatment effects or statistical confidence limits. Animal structural benefits support a role for cellular maintenance, but completed cargo turnover and selective human prevention remain poorly resolved. Reviewed 14 September 2026. [Probability method](probability_method.md) · [Collection index](README.md).

## The causal claim and its scope

In adults with type 2 diabetes diagnosed within five years and no clinical DPN at entry, **slow disposal of cargo already enclosed in macroautophagosomes**, in connected sensory neurons or their supporting glia, contributes to subsequent terminal loss. Restoring the specified disposal machinery from entry would avert at least **20% of five-year net physical peptidergic C-fiber terminal loss** in a baseline-defined group accounting for at least **10% of reference-care incident DPN**.

This concerns completion of macroautophagy, including disposal of mitochondria. Failure to recognize or sequester damaged cargo, excessive damage production despite normal disposal, proteasomal degradation and chaperone-mediated autophagy are distinct propositions. They can interact with this route, but are not automatically included in its probability. PINK1/Parkin signaling is one possible upstream recognition mechanism, not a requirement for all mitochondrial disposal.

**Population and endpoint.** Reference care is contemporary diabetes care without Q below. Identify adult peptidergic unmyelinated sensory neurons through combined molecular and anatomical classification, then physically track their distal skin arbors. Altered CGRP or PGP9.5 expression alone does not count as loss.

For each person, net terminal loss is physical destruction minus replacement over five years, divided by baseline terminal amount. Newly formed terminals later destroyed enter both totals. Give people equal weight and define:

`B = (mean reference net loss − mean net loss with Q) / mean reference net loss.`

The reference mean must be positive. B describes preservation of net structure; it does not require complete prevention of individual degeneration events.

**Finite baseline groups.** Candidates are the whole eligible population, males, females, baseline BMI ≥30 kg/m², baseline HbA1c ≥7.5%, and a baseline slow-disposal group. The last meets the cargo-completion criterion below in at least 10% of mapped target units during a 30-day entry observation. No additional intersections or groups defined by subsequent treatment response qualify. Incident DPN is new bilateral length-dependent peripheral nerve signs with objective small-fiber loss or abnormal conduction, excluding another cause. Coverage is `C = P(group membership | incident DPN under reference care)`.

**Connected cells and measurement.** A target unit comprises its peptidergic neuron, including soma and axon, its directly associated Remak Schwann cells and satellite glia. Macrophages, vascular cells, myelinating Schwann cells without direct association, circulating cells and CNS cells are outside Q's direct scope.

Measure the interval from verified closure of a cargo-containing, double-membrane autophagosome to lysosomal hydrolysis of at least 90% of its enclosed protein mass into free amino acids or di-/tripeptides. This is the operational completion boundary. Nonperturbing molecular labels must distinguish degradation from fluorophore quenching, dilution, secretion, transport out of the imaging field and transfer to another cell. Transfer is recorded but does not count as intracellular hydrolysis. This is a measured transition with a start and finish, not the abundance of LC3, p62, lysosomes or acidic puncta.

Use the same low tracer load and standardized cargo size/composition in each cell class, with two separately analyzed cargo classes: mitochondria and nonmembranous cytosolic protein aggregates. The assay begins after enclosure, so defective initiation cannot itself create a positive result. Retain the cell's physiological energy and metabolic environment, measure those conditions, and compare with an age-, sex- and site-matched nondiabetic reference. Tracking native unmanipulated cargo verifies that the tracer has not created the bottleneck.

A class qualifies when its median completion time is **at least twice the matched normal median and above the normal 95th percentile**. Its 30-day mean must remain above both boundaries for **90 consecutive days during year one**, in at least **10% of mapped units**, averaged with equal person weight in a candidate group. Either cargo class may qualify; use that same class, cells and group when interpreting the causal comparison.

The qualifying interval must precede the first **≥5% net physical terminal deficit from entry maintained for 30 days** in the connected arbor under reference care. Ordinary turnover without a sustained deficit does not cross this landmark. The thresholds specify the proposition; they are not established biological tipping points.

**The intervention Q.** From entry for five years, selectively restore deficient post-enclosure disposal machinery in the named cells to the matched nondiabetic median functional capacity. The specified processes are transport of closed autophagosomes toward lysosomes, autophagosome–lysosome fusion, lysosomal acidification and hydrolysis of enclosed cargo. For each process, restore functional machinery capacity below the normal fifth percentile to the matched normal median, assessed under a common reference substrate/energy condition; do not raise an already normal or high capacity. In-vivo ATP and substrate supply remain uncorrected. Monitor with standardized process-specific assays and the full completion-time measurement.

The ideal manipulation replaces inactive machinery or supplies functional copies dedicated to these autophagic processes. Shared motors or signaling proteins must be altered only in their autophagic interactions; transport of unrelated vesicles is not directly changed. It preserves endogenous upstream cargo recognition and sequestration. It is applied to all enclosed cargo, **without selecting material according to whether removing it later benefits the axon**.

This idealization requires molecular selectivity unavailable in current drugs. It specifies what is changed rather than assuming a beneficial flux. Q does not directly supply ATP, repair mitochondria, inhibit SARM1, stimulate biogenesis, increase autophagosome initiation, remove extracellular debris or correct diabetes. Resource use, cargo availability and downstream feedback remain free to change. Restoring machinery may fail to shorten completion time when energy is limiting; that possibility counts against the causal effect. More successful disposal may consume useful organelles faster than replacement and can worsen maintenance.

One same baseline group must satisfy the exposure, structural effect, magnitude and coverage conditions. If several groups qualify initially, at least one must continue satisfying the entire chain.

## Reasoning from the biology

A cell's cargo burden is the balance of generation, recognition, sequestration, delivery, hydrolysis, export and replacement. An increase in autophagosome number can result from increased entry or decreased exit. A decrease can result from faster completion or reduced formation. p62 is both produced and removed. A single static measurement cannot recover these rates.

The proposed route is:

`slow post-enclosure completion → changed retained cargo and recycled substrates → destruction minus replacement of terminals.`

Retained organelles can produce oxidants or inflammatory signals, occupy transport capacity, or supply inadequate ATP. Successful disposal can recover substrates and permit replacement. These are alternative or interacting consequences, not mandatory additional discounts. The claim does not require somal apoptosis or prove that one particular mitochondrial product mediates all benefit.

Conversely, autophagy can slow because energy supply has already deteriorated. Repairing disposal machinery then may have little effect. Increased cargo production can overwhelm an otherwise normal system, while another disposal route or mitochondrial biogenesis can compensate. More fission can facilitate sequestration, but fragmentation alone neither demonstrates completed mitophagy nor determines whether increased removal will be useful.

The intervention's net effect therefore cannot be inferred from a regulator's name or from one favorable marker. SIRT3 changes both mitochondrial maintenance and SARM1-related chemistry; AMPK changes glucose and lipid metabolism as well as autophagy. Their effects support possible pathways but do not identify Q.

## Evidence that moves the judgment

**AICAR has structural benefits, with important limits to their size and attribution.** A 2025 study administered 500 mg/kg AICAR to HFD-fed mice, beginning either with the diet for four months or after two months of functional neuropathy. Skin IENFD improved in both experiments. However, the text's reported means imply 114% and 108% closure of the untreated deficit, whereas the actual Figure 5E and 6F means imply approximately **59% and 58%**. Two-pixel reading sensitivity gives about 51–67% and 49–67%, not sampling confidence intervals. The study also gives conflicting n: 6–8 versus 6 for prevention, and 6 versus 8 for delayed treatment.

Table 1 reports glucose **178→138 mg/dL** and HbA1c **6.8→5.8%**, with changes in weight, lipids and insulin resistance. DRP1, phospho-ULK1 and LC3-II measurements do not establish completed turnover. Thus a sizeable treatment-associated structural effect remains plausible, but neither complete restoration nor clearance-specific mediation is secure. Delayed-treatment improvement informs treatment of established dysfunction; the study lacks baseline skin counts establishing repair of a measured structural deficit. [Chandrasekaran et al., 2025](https://pmc.ncbi.nlm.nih.gov/articles/PMC11720447/).

**Human follow-up exists, but does not measure this transition.** A sural-nerve study began with 30 men, ten each with normal glucose tolerance, impaired tolerance or T2 diabetes. Only **13** returned about 11 years later; seven had developed T2 diabetes. Among those seven, median myelinated fiber density changed from **6,070 to 2,973/mm²**, while unmyelinated axon density did not significantly change. Seven transitions to diabetes are not seven established incident-DPN events.

Autophagic structures became less evident in myelinated axons and more prominent in Schwann cells. Detailed autophagy comparisons used **7/10/9 donors at baseline and 3/9 at follow-up**, after unsuitable tissue was excluded. Static structures cannot distinguish slow completion from reduced formation, altered cellular composition or late injury. This is valuable longitudinal human morphology, but not a direct early peptidergic-terminal clearance experiment. The unmyelinated null is also too small and indirect to establish preserved turnover. [Mohseni et al., 2017](https://pmc.ncbi.nlm.nih.gov/articles/PMC5561322/).

**The 2024 SIRT3 study includes nerve structure, but not selective completion flux.** Intrathecal SIRT3 overexpression at day 14 after STZ improved pain, conduction, mitochondrial measures and supplementary myelinated/unmyelinated axon counts in HFD/STZ rats. FoxO3a knockdown counteracted these effects. Structural TEM sampled **three rats per group**; 11–20 fields do not increase the biological n. The structure was DRG-associated axons, not epidermal terminals. Mitochondrial colocalization, mitophagosomes and LC3/p62 blots do not establish post-enclosure hydrolysis rates. FoxO3a has functions beyond autophagy. [Yang et al., 2024](https://pmc.ncbi.nlm.nih.gov/articles/PMC10993345/).

In contrast, a 2026 study found that **Sirt3 deletion** preserved diabetic mouse skin fibers through a SARM1-related mechanism. The same DRG endpoints improved with opposing SIRT3 manipulations: source means give ATP increases of approximately **1.63-fold after overexpression in rats and 1.83-fold after knockout in mice**, with ROS about 0.71-fold in both. Timing, species and target outputs differ. This rules out a universal SIRT3 treatment direction, not a beneficial effect of selective completion correction. A K641Q mutant is not a direct measurement of endogenous SARM1 acetylation occupancy. [Chen et al., 2026](https://pubmed.ncbi.nlm.nih.gov/41512034/).

**Recent culture studies do not close the rate-measurement gap.** The 2025 PINK1 study used ND7/23 hybrid cells, including **50 mM glucose for 48 hours**, rapamycin and PINK1 knockdown. It supports glucose-sensitive mitochondrial injury and an upstream recognition route in a culture system. The accessible account does not establish adult distal-terminal preservation or the post-enclosure rate defect priced here. [2025 ND7/23 study](https://doi.org/10.1016/j.neuroscience.2025.03.052).

A 2026 paper specifies bafilomycin treatment in Methods, but its results and accessible supplement do not provide the paired inhibitor comparison. The supplement repeats summary tables without biological-replicate rows. “Mitophagy rate” is tabulated for undefined mild/moderate/severe impairment conditions, while LC3-II/I and its reciprocal are inconsistently described. No corresponding skin terminal intervention result is presented. Its stated flux reduction therefore cannot be taken as a resolved quantitative completion defect. [Li et al., 2026](https://doi.org/10.1371/journal.pone.0344082).

Existing human cross-sectional proteomics and small diabetic-mouse protein screens supply context, not direct disposal rates. Neither a marker increase nor a corrected omics null resolves this functional claim. No accessed study combines measured early human completion times, selective Q, physical terminal accounting and incident-DPN coverage.

## Probabilities of the logical claims

Each row after R1 assumes all preceding requirements for the same group and compatible connected cells. These are conditional judgments about one joint proposition.

| # | Claim | Skeptical / best / favorable | Best cumulative | Basis |
|---|---|---:|---:|---|
| R1 | At least one fixed group and cargo class meet the specified first-year, 90-day, pre-deficit slow-completion criterion in at least 10% of units. | 25 / **55** / 85% | 55.0% | Clearance biology is plausible, but most DPN evidence measures abundance or initiation rather than completion in early human cells. |
| R2 | In one same R1 group, selective Q produces B≥0.05. | 15 / **50** / 85% | 27.5% | Structural drug and regulator results support capability; metabolic effects, upstream actions and resource limitation weaken attribution to Q. |
| R3 | One same group satisfying R2 has B≥0.20. | 30 / **65** / 90% | 17.9% | Large animal structural effects leave room for material preservation, but their fraction mediated by this route is unknown. |
| R4 | One same group satisfying R3 accounts for C≥0.10. | 30 / **60** / 85% | **10.7%** | Common metabolic stress makes appreciable coverage plausible; the defined early rate defect and benefit have not been jointly measured. |

R1 does not call the delayed cargo harmful. R2 resolves the net structural consequence of correcting the machinery, including adverse loss of adaptation or substrate costs. A 10% preservation effect passes R2 and fails R3; a strong effect in a very rare baseline group passes R3 and fails R4. Retained material and its particular toxic products are mechanistic explanations, not duplicated causal requirements.

## Ideal experiments that would resolve each claim

The ideal capabilities are nonperturbing cell-resolved cargo tracking, autophagy-specific manipulation of otherwise shared machinery, and complete physical tracking of human terminals for five years. Use representative early-T2 participants and the fixed baseline groups, with matched nondiabetic measurements defining the normal references. Another animal or immortalized-cell model cannot by itself resolve a human row.

**R1 — completion deficit, with no predecessor assumed.** Follow the two defined cargo classes from closed autophagosome to verified intracellular hydrolysis in connected neurons, Remak cells and satellite glia. Standardize tracer amount, size and composition; separately measure formation, export, energy state and native-cargo behavior. Determine rolling completion times, persistence, unit prevalence and time relative to the sustained terminal-deficit landmark. **YES:** one group/class meets every threshold. **NO:** complete observation shows that every candidate group/class fails at least one. A normal completion time despite reduced initiation fails this particular row. Missing hydrolysis measurements, late-only tissue or a static marker result leave it **unresolved**.

**R2 — structural causation conditional on R1.** Randomize compatible participants from entry to reference care or Q for five years. Verify normal-range restoration of each deficient post-enclosure process, unchanged direct action on initiation and unrelated transport, and comparable initiating glycemic, lipid, nutritional and mechanical exposures. ATP use, cargo burden, replacement and downstream signaling must be allowed to respond. Use an independent molecular implementation and reversal of the restoration to distinguish the specified machinery from an intervention artifact.

Measure actual completed hydrolysis alongside physical terminal destruction and replacement. Machinery engagement is distinct from guaranteed flux rescue: if energy limitation prevents a normal machinery complement from helping, the faithful Q effect can be small or zero. **YES:** one compatible group has a positive reference-loss denominator and B≥0.05. **NO:** all faithfully implemented comparisons have B<0.05, including worsening. A failed delivery or nonselective manipulation leaves the row unresolved. A beneficial systemic drug, pain relief or increased LC3 alone is also unresolved.

**R3 — magnitude conditional on R1–R2.** Estimate the same five-year B in the same qualifying groups, retaining every downstream benefit and cost. **YES:** at least one reaches B≥0.20. **NO:** all R2 groups remain between 0.05 and 0.20. A transient early benefit that disappears by year five fails the five-year criterion; a finite uncertainty interval crossing the boundary leaves it unresolved.

**R4 — coverage conditional on R1–R3.** Measure baseline group membership and reference-care incident DPN in the representative population. **YES:** one R3 group accounts for at least 10% of those cases. **NO:** every R3 group accounts for less. Count people, not cells or terminals. Do not construct a new responder union or substitute prevalence among established-DPN biopsy patients. Incomplete incident ascertainment leaves coverage unresolved.

These are experiments that define the claims' truth conditions, not requirements for a current laboratory program.

## Why these odds and how uncertain they are

The main uncertainty is not whether lysosomes can degrade organelles. It is whether slow completion is present early in the intended cells and whether correcting the specified machinery, without changing diabetes or upstream tagging, materially preserves terminals. Large animal drug effects keep the favorable scenario open. Their nonselectivity, source inconsistencies and the human rate-measurement gap prevent a high best judgment.

Best conditional inputs are **[55,50,65,60]%**, giving **10.725%**. Skeptical **[25,15,30,30]%** gives **0.3375%**; favorable **[85,85,90,85]%** gives **55.27125%**. Decimal places show multiplication, not elicitation precision. The scenarios are coordinated interpretations with no assigned weights; they are not independent probability distributions or an empirically calibrated interval.

If marker changes mostly reflect formation, systemic metabolism or already failing cells, **[40,25,45,45]%** gives **2.0%**. If a frequent early completion bottleneck and a large selective structural effect are established, **[75,75,80,80]%** gives **36.0%**. These scenarios move related requirements together.

As an arithmetic illustration, raising only R1 from 55% to 80% changes the product to **15.6%**; lowering it to 25% gives **4.9%**. Resolving R1 as true would still leave 19.5% if later judgments were unchanged. Real flux evidence could update several rows, so this is not a forecast. Merging R1 and R2 must retain their **27.5%** joint probability; splitting disposal into more named stages alone must not reduce the headline.

The half-speed, 90-day and 10%-unit definitions are consequential. Shorter or milder deficits could still matter but fall outside R1. The 20% structural and 10% case thresholds similarly define materiality and coverage. Changing them changes the proposition and requires fresh judgment; available data do not support numerical interpolation.

## Overlap and boundary

This route overlaps [ER responses](23_er_stress.md), [mitochondrial dynamics](25_mitochondrial_dynamics.md), [axonal transport](26_axonal_transport.md), [energy insufficiency](07_bioenergetic_insufficiency.md) and [SARM1](27_sarm1_axon_destruction.md). Autophagosome transport can be a shared bottleneck, while energy failure can both cause and follow slow disposal. These probabilities cannot be added.

[Macrophage clearance](20_loss_of_protective_macrophages.md) concerns already dead extracellular material; this file concerns intracellular completion in living neurons and their direct glia. A benefit of transplanting healthy organelles, stimulating upstream tagging or increasing all autophagy does not establish that post-enclosure disposal was the limiting step.