ORIGIN-T2D

1. ABSTRACT

Type 2 diabetes mellitus (T2D) is commonly described through a constellation of abnormalities including insulin resistance, β-cell dysfunction, abnormal hepatic glucose regulation, altered glucagon and incretin signaling, ectopic lipid accumulation, obesity, and progressive dysglycaemia.

These abnormalities are biologically real and clinically important. However, an unresolved causal question remains:

What is the earliest event that initiates the biological cascade leading to type 2 diabetes in a particular individual?

The central premise of ORIGIN-T2D is that an abnormality recognized during established insulin resistance, prediabetes, or diabetes may not necessarily represent the initiating event. It may instead be a downstream mediator, compensatory response, consequence, or convergence point generated by an earlier disturbance.

The proposed research therefore reverses the conventional direction of investigation.

Instead of beginning with established diabetes and explaining its metabolic abnormalities, ORIGIN-T2D seeks to identify the earliest reproducible biological divergence between individuals who eventually develop T2D and otherwise comparable individuals who do not.

The study will use prospective longitudinal observation, discordant twin and sibling designs, deep phenotyping, continuous glucose monitoring, metabolic testing, autonomic and endocrine measurements, immune profiling, metabolomics, proteomics, transcriptomics/epigenomics, and other exploratory biological measurements.

The primary objective is not to identify another risk factor.

It is to determine whether a biological event or state transition exists that:

Occurs before conventional metabolic abnormalities.

Predicts subsequent progression.

Is reproducible across independent individuals.

Distinguishes future cases from comparable non-cases.

Plausibly precedes rather than follows insulin resistance and β-cell dysfunction.

Demonstrates causal evidence when experimentally testable.

The study will explicitly test two competing models:

2. THE CENTRAL QUESTION

The conventional question is:

Why does a person with type 2 diabetes have insulin resistance, β-cell dysfunction, and hyperglycaemia?

ORIGIN-T2D asks a different question:

Why did this particular person enter the biological trajectory that produced those abnormalities in the first place?

This distinction is fundamental.

Insulin resistance may contribute directly to hyperglycaemia.

β-cell dysfunction may contribute directly to hyperglycaemia.

Visceral adiposity may contribute to insulin resistance.

Genetic susceptibility may increase probability.

Age may alter biological vulnerability.

Lifestyle may influence several pathways.

But none of these observations, by themselves, establishes that the factor represents the first initiating event in an individual case.

The proposed investigation therefore introduces a distinction between:

Risk factor

A characteristic associated with increased probability of disease.

Susceptibility factor

A characteristic that makes disease more likely under particular circumstances.

Mediator

A biological process through which an upstream exposure or disturbance produces disease.

Consequence

A biological change produced by the disease process.

Trigger

An event capable of initiating a transition into a pathological trajectory.

Initiating event

The earliest causal biological event required for that individual’s disease trajectory.

The purpose of ORIGIN-T2D is to search for the final category without prematurely assuming what biological system contains it.

3. THE CORE HYPOTHESIS

Primary hypothesis

In at least a subset of individuals who develop type 2 diabetes, an upstream biological event or state transition occurs before the conventional metabolic abnormalities of insulin resistance, β-cell dysfunction, and dysglycaemia, and this event contributes causally to the subsequent disease trajectory.

A secondary hypothesis is:

The earliest initiating disturbance may differ between individuals but may converge upon a limited number of downstream biological states that ultimately produce the clinical phenotype of T2D.

This second hypothesis is important because T2D is heterogeneous.

Therefore, the research must be capable of discovering either:

One common upstream mechanism

or

Multiple upstream mechanisms.

4. THE CAUSAL MODEL

The conventional simplified model is:

Genetic susceptibility / lifestyle / obesity

↓

Insulin resistance

↓

β-cell compensation

↓

β-cell failure

↓

Prediabetes

↓

Type 2 diabetes

ORIGIN-T2D proposes that this may be incomplete.

The alternative model is:

UNKNOWN UPSTREAM EVENT

↓

PRIMARY BIOLOGICAL DISTURBANCE

↓

NEUROENDOCRINE / IMMUNE / CELLULAR / METABOLIC / BIOELECTRICAL / OTHER DYSREGULATION

↓

INSULIN RESISTANCE / β-CELL DYSFUNCTION / ALTERED GLUCAGON / ECTOPIC LIPID / OTHER METABOLIC ABNORMALITIES

↓

DYSGLYCAEMIA

↓

TYPE 2 DIABETES

The crucial research objective is to determine whether the box labelled:

UNKNOWN UPSTREAM EVENT

actually exists.

5. WHY THE QUESTION IS SCIENTIFICALLY IMPORTANT

T2D does not behave as a simple deterministic disease.

Genetic susceptibility is substantial but does not determine an individual’s outcome. Twin studies demonstrate substantial heritability while also identifying twin pairs who remain discordant for T2D.

Similarly, monozygotic twin studies have identified differences in gene expression, DNA methylation, and other molecular characteristics between diabetic and non-diabetic twins.

However, an important limitation remains:

A biological difference observed after diabetes develops may be a cause, a mediator, a compensatory response, or a consequence.

Therefore, finding a difference is insufficient.

ORIGIN-T2D focuses on temporal ordering.

6. THE MOST IMPORTANT CONCEPT: THE ZERO POINT

The study will define three different temporal landmarks.

Z0 — Causal zero point

The actual initiating biological event.

This may be invisible with current technology.

Z1 — First detectable biological abnormality

The earliest measurable deviation from the individual’s previous biological state.

Z2 — First conventional metabolic abnormality

For example:

measurable insulin resistance,

abnormal β-cell response,

altered hepatic glucose regulation,

abnormal glucose tolerance,

elevated fasting glucose,

elevated postprandial glucose.

Current medicine may frequently identify Z2.

The objective of ORIGIN-T2D is to determine whether Z1 can be identified and whether Z1 is a measurable approximation of Z0.

7. THE 9,999 VS 1,000 PRINCIPLE

The study adopts a central analytical principle:

Non-cases are not merely controls. They are biological evidence.

If 10,000 individuals possess a particular risk factor but only a proportion develop T2D, the non-progressors may contain mechanisms of resistance or protection.

Therefore:

Instead of asking only:

What do future diabetic individuals have?

we ask:

What separates future diabetic individuals from future non-diabetic individuals despite apparently similar circumstances?

This turns disease resistance into an object of investigation.

8. STUDY POPULATIONS

The proposed program will use several complementary populations.

9. SPECIAL COHORT B — HIGH-RISK NON-PROGRESSORS

Individuals with substantial conventional diabetes risk who remain normoglycaemic.

These people are extremely important.

They allow the question:

What is protecting them?

Measurements should include:

insulin sensitivity,

insulin secretion,

hepatic metabolism,

autonomic function,

inflammatory markers,

endocrine profiles,

metabolomics,

microbiome where feasible,

sleep/circadian measures,

physical activity,

environmental exposures,

genetic susceptibility,

epigenetic signatures.

10. SPECIAL COHORT C — LOW-CONVENTIONAL-RISK PROGRESSORS

Individuals who develop T2D despite lacking the conventional phenotype expected to explain their disease.

Examples could include individuals who are:

relatively lean,

physically active,

without marked visceral obesity,

without major conventional metabolic abnormalities.

The purpose is not to label them “mysterious.”

It is to deliberately examine cases that challenge the explanatory power of conventional models.

11. SPECIAL COHORT D — DISCORDANT SIBLINGS

Biological siblings in whom:

one develops T2D,

another does not.

This partially controls for:

shared genetics,

childhood environment,

family background,

socioeconomic factors.

12. SPECIAL COHORT E — DISCORDANT MONOZYGOTIC TWINS

This is one of the most powerful natural experiments available.

Monozygotic twins provide extremely similar genomes, yet T2D can remain discordant.

Existing international datasets have identified discordant monozygotic twin pairs, demonstrating that such discordance can be studied systematically.

Existing research has also shown molecular differences in adipose tissue and skeletal muscle among T2D-discordant twins.

ORIGIN-T2D would take the question one step further:

Which differences appeared BEFORE the diabetic phenotype?

That temporal distinction is critical.

13. SPECIAL COHORT F — REVERSERS

Individuals who:

develop prediabetes and return to normal glucose regulation,

or:

develop T2D and subsequently achieve sustained remission.

These individuals are biologically valuable because the trajectory has changed direction.

We should ask:

What changed when disease progression stopped?

The existence of diabetes prevention and remission trajectories demonstrates that early metabolic deterioration is not necessarily a one-way process.

This provides another natural experiment for identifying mechanisms of progression versus recovery.

14. THE LONGITUDINAL DESIGN

Participants should not simply receive one annual HbA1c.

That would miss the central question.

The study should repeatedly measure multiple biological layers.

Core measurements

Conventional metabolic layer

fasting glucose,

HbA1c,

fasting insulin,

C-peptide,

oral glucose tolerance testing,

insulin sensitivity,

insulin secretion,

glucagon,

lipids,

liver enzymes,

body composition.

Dynamic glucose layer

Continuous glucose monitoring during predefined intervals.

Rather than simply asking:

Is glucose high?

we ask:

How does glucose regulation change over time?

15. NEUROENDOCRINE LAYER

Because the upstream disturbance may involve regulatory systems outside classical metabolic tissues, potential measurements include:

autonomic function,

heart-rate variability,

sympathetic/parasympathetic balance,

cortisol rhythm,

catecholamine patterns,

hypothalamic-pituitary-adrenal axis markers,

sleep/circadian measures,

appetite regulation,

selected neuroendocrine hormones.

This does not assume that the nervous system is the answer.

It keeps the nervous system inside the hypothesis space.

16. IMMUNE / INFLAMMATORY LAYER

Measurements could include:

inflammatory cytokine panels,

immune-cell phenotyping,

innate immune activation markers,

adaptive immune signatures,

complement pathways,

inflammatory lipid mediators.

The important question would be:

Does a reproducible immune disturbance appear before conventional insulin resistance?

If yes, it becomes a candidate upstream pathway.

If it appears only afterward, it becomes more likely to be downstream.

17. MOLECULAR LAYER

A subset of participants would undergo:

genomics,

epigenomics,

transcriptomics,

proteomics,

metabolomics,

lipidomics,

miRNA profiling.

The purpose is not to collect enormous quantities of data simply because technology allows it.

The purpose is to reconstruct the sequence of biological change.

Existing twin studies demonstrate that molecular differences can be detected, but they also illustrate the central problem: associations can reflect either causal or consequential pathways.

ORIGIN-T2D therefore prioritizes repeated sampling.

18. CELLULAR / BIOELECTRICAL EXPLORATORY LAYER

This is deliberately exploratory.

Potential areas include:

ion-channel biology,

membrane potential,

calcium signaling,

cellular electrical activity,

mitochondrial function,

redox state,

intracellular signaling dynamics.

The protocol would explicitly state:

No assumption is made that bioelectrical dysfunction causes T2D.

It is included because the hypothesis is intended to remain open to previously underappreciated biological domains.

19. ENVIRONMENTAL LAYER

Potential exposures include:

medications,

pollutants,

endocrine-disrupting chemicals,

occupational exposures,

sleep disruption,

circadian disruption,

psychosocial stress,

infectious history,

diet,

physical activity,

socioeconomic changes,

major life events.

Again, these are candidates—not predetermined causes.

20. THE MOST IMPORTANT FEATURE: HIGH-FREQUENCY SAMPLING AROUND TRANSITION

Annual measurement is inadequate if the initiating event occurs rapidly.

Therefore, participants approaching a biological transition should enter an intensified observation phase.

Stable phase

Assessment every 6–12 months.

Early abnormality phase

Assessment every 1–3 months.

Transition phase

Shorter intervals where feasible.

The objective is to capture:

Before → transition → after

rather than simply:

Healthy → diabetic.

21. PRIMARY ENDPOINT

The primary endpoint should not simply be “development of diabetes.”

The primary endpoint should be:

Identification of a reproducible biological signal or state transition that precedes the emergence of conventional metabolic abnormalities and is independently associated with subsequent progression to T2D.

This keeps the study focused on the actual scientific question.

22. SECONDARY ENDPOINTS

Identification of distinct upstream trajectories.

Identification of protective signatures in non-progressors.

Identification of biological differences in discordant twins and siblings.

Identification of changes preceding insulin resistance.

Identification of changes preceding β-cell dysfunction.

Identification of changes preceding dysglycaemia.

Determination of whether candidate signals remain predictive after adjustment for:

BMI,

age,

family history,

genetic risk,

physical activity,

diet,

conventional metabolic measures.

Identification of biomarkers capable of predicting the transition.

Determination of whether candidate mechanisms are reversible.

Generation of experimentally testable causal hypotheses.

23. THE CRITICAL ANALYTICAL QUESTION

For every candidate signal X:

We ask:

Does X occur before:

insulin resistance?

hyperinsulinaemia?

impaired insulin secretion?

ectopic fat?

rising glucose?

HbA1c elevation?

If:

X → insulin resistance → diabetes

that is interesting.

If:

insulin resistance → X → diabetes

then X is unlikely to be the initiating event.

This temporal analysis is fundamental.

24. CAUSALITY FRAMEWORK

Every candidate must pass progressively stronger tests.

25. THE “BROKEN SHIP” TEST

Every proposed mechanism must answer:

Is this the broken ship, or what caused the ship to become broken?

For example:

Insulin resistance

Could be a mediator.

Question:

What caused the insulin resistance?

β-cell dysfunction

Could be a mediator.

Question:

What caused the β-cell dysfunction?

Visceral fat

Could contribute.

Question:

Why did this individual develop that particular fat distribution?

Genetic susceptibility

Could explain vulnerability.

Question:

What biological event converted susceptibility into disease?

Age

Could modify vulnerability.

Question:

What biological change associated with aging actually initiates the transition?

This does not deny the causal role of these factors.

It simply refuses to stop the investigation prematurely.

26. PROTECTIVE MECHANISM ANALYSIS

This is one of the innovative components.

Instead of examining only:

Future diabetic vs healthy

we calculate:

What biological mechanism allows an individual exposed to substantial metabolic stress to remain metabolically normal?

For example:

Two individuals may have:

similar BMI,

similar insulin resistance,

similar family history,

but only one develops diabetes.

The non-progressor may possess:

superior β-cell compensation,

different autonomic regulation,

different hepatic signaling,

altered immune regulation,

different mitochondrial function,

different endocrine signaling,

or another protective mechanism.

That protective mechanism may reveal the missing upstream event indirectly.

27. DISCORDANT-TWIN SUBSTUDY

This should be a major component.

Population

MZ twin pairs discordant for:

T2D,

prediabetes progression,

or metabolic deterioration.

Measurements

Both twins undergo identical:

metabolic testing,

CGM,

body composition assessment,

autonomic assessment,

endocrine testing,

immune profiling,

metabolomics,

proteomics,

epigenomics,

transcriptomics where feasible.

Primary question

What biological difference separates the trajectories after genetic similarity has been maximally controlled?

Existing research demonstrates that discordant twin studies can reveal molecular and epigenetic differences.

ORIGIN-T2D would add a critical longitudinal dimension wherever historical samples or prospective follow-up permit.

28. A PARTICULARLY POWERFUL DESIGN

The strongest design would be:

Twin A

Currently healthy.

Twin B

Currently healthy.

Follow both prospectively.

If Twin A develops the first measurable abnormality:

Immediately intensify observation of both.

Then compare:

What changed in A?

Did B show the same change?

If not:

What appeared in A first?

This could provide a remarkably clean natural experiment.

29. STATISTICAL STRATEGY

The analysis should avoid simple correlation hunting.

Methods may include:

longitudinal mixed-effects models,

time-to-event analysis,

joint longitudinal-survival models,

latent trajectory analysis,

change-point detection,

time-varying covariates,

causal mediation analysis,

dynamic Bayesian networks,

structural equation models,

machine-learning discovery models,

targeted causal inference,

replication cohorts.

But machine learning must remain subordinate to biological reasoning.

The algorithm should not simply find:

“X predicts diabetes.”

It should help answer:

“X changed before Y, which changed before Z.”

30. THE “CHANGE-POINT” IDEA

One particularly interesting analytical strategy is to identify the moment at which an individual’s biology changes from one stable state to another.

For example:

State 1

Stable glucose regulation.

↓

Transition

Unknown biological change.

↓

State 2

Compensation.

↓

State 3

Metabolic deterioration.

↓

State 4

Diabetes.

If we can statistically identify the transition point and then inspect what changed immediately around it, we may get much closer to the initiating event.

31. CAUSAL GRAPH

A simplified causal structure would initially be represented as:

Genetics → susceptibility

Environment → ?

Age → ?

Unknown trigger → primary disturbance → metabolic mediators → hyperglycaemia → T2D

with:

Genetics → susceptibility to the unknown trigger or downstream effects

and multiple feedback loops potentially operating downstream.

The analysis must distinguish:

cause,

confounder,

mediator,

consequence.

32. WHAT WOULD COUNT AS A MAJOR DISCOVERY?

The ideal result would look something like:

A biological signal appears consistently years before conventional dysglycaemia, precedes measurable insulin resistance, is independent of established risk factors, is reproducible across cohorts, is present in future cases but not matched non-progressors, and mechanistic experiments demonstrate that altering the pathway modifies downstream metabolic deterioration.

That would constitute strong evidence for an upstream mechanism.

But even a weaker finding could be important.

For example:

Different T2D patients show three distinct upstream trajectories rather than one universal trigger.

That would also substantially change the conceptual model.

33. WHAT WOULD FALSIFY OUR HYPOTHESIS?

This is essential.

A serious research hypothesis must be capable of being wrong.

Our hypothesis would be weakened or falsified if:

34. WHAT IF THERE IS NO SINGLE TRIGGER?

This is an important possibility.

The final model could become:

Trigger A → pathway A

Trigger B → pathway B

Trigger C → pathway C

↓

Common metabolic network

↓

T2D

In that situation, the research would still have succeeded.

We would have discovered that the apparent “single disease” contains multiple upstream causal pathways converging on a common phenotype.

That may ultimately be more useful than finding one universal trigger.

35. THE NEWBORN QUESTION

The earlier idea about identifying the initiating signal in a newborn is intellectually important but must come much later.

We should not screen newborns for an unvalidated marker and intervene simply because the marker appears.

First:

Discovery

↓

Validation

↓

Causal confirmation

↓

Risk/benefit assessment

↓

Intervention trial

↓

Only then potentially preventive screening

If an upstream mechanism were eventually proven causal, early-life detection could become an extraordinary preventive possibility.

But that would require evidence far beyond association.

36. THE RESEARCH PROGRAM IN FOUR PHASES

37. PHASE 1 — RETROSPECTIVE DISCOVERY

Use existing cohorts and biobanks.

Search for individuals who:

were initially normoglycaemic,

subsequently developed T2D.

Compare them with:

matched individuals who remained normoglycaemic.

Then examine stored biological samples collected years before diagnosis.

This is potentially much cheaper than immediately starting a 10,000-person prospective study.

It also lets us ask:

Was something already happening years before diabetes became visible?

38. PHASE 2 — PROSPECTIVE ORIGIN COHORT

Recruit a large cohort without diabetes.

Repeated measurements.

Deep phenotyping.

Nested intensive sampling around biological transitions.

This phase establishes temporal order.

39. PHASE 3 — CAUSAL VALIDATION

Candidate mechanisms identified in Phases 1–2 are tested experimentally.

Depending on the mechanism:

cellular models,

organoids,

animal models,

human physiological intervention,

natural experiments,

pharmacological perturbation,

genetic approaches.

The objective:

Does changing the candidate mechanism alter the downstream disease trajectory?

40. PHASE 4 — PREVENTION

Only after causal validation:

Develop an intervention targeting the earliest mechanism.

The ultimate conceptual pathway becomes:

Detection of upstream trigger

↓

Interruption before metabolic damage

↓

Preservation of normal physiology

↓

Prevention of T2D

That is the preventive-medicine vision behind the entire project.

41. WHY WE SHOULD NOT BEGIN WITH “ONE TRIGGER”

There is a subtle danger in the original idea.

If we become convinced that there must be one lightning strike, we may accidentally blind ourselves to a distributed system failure.

The biological reality may resemble:

Many small disturbances

↓

Loss of regulatory reserve

↓

Threshold crossing

↓

Metabolic cascade

rather than:

One dramatic event

↓

Diabetes

Therefore, the protocol must allow both possibilities.

42. THE “THRESHOLD” MODEL

An alternative model is:

Several individually insufficient disturbances accumulate.

For example:

Disturbance A = 20%

Disturbance B = 15%

Disturbance C = 25%

Disturbance D = 20%

↓

Biological reserve falls below threshold

↓

Metabolic system changes state

↓

T2D trajectory begins

In that case, there is still an upstream causal process, but no single event.

This is why state-transition analysis is more scientifically robust than assuming a single trigger.

43. WHAT MAKES THIS DIFFERENT FROM ORDINARY RISK-PREDICTION RESEARCH?

Ordinary prediction asks:

Who will get diabetes?

ORIGIN-T2D asks:

What biological event caused the trajectory to change?

A model that predicts diabetes with high accuracy but cannot explain the initiating mechanism would be useful clinically, but it would not answer our central question.

We are therefore interested in:

Prediction + temporal ordering + mechanism + causality.

44. THE MOST IMPORTANT EXPERIMENTAL QUESTION

For every candidate:

If we could go back to before the disease began, would removing this factor prevent the entire downstream cascade?

That is the ultimate test.

Not:

Does it correlate with diabetes?

Not:

Is it higher in diabetics?

Not:

Does treating it improve glucose?

But:

Does it sit upstream of the disease-generating cascade?

45. PROPOSED PRIMARY RESEARCH QUESTION

Among individuals initially free of type 2 diabetes, what biological event or state transition most consistently precedes and predicts the divergence toward type 2 diabetes, after accounting for established genetic, metabolic, behavioral, and environmental risk factors?

46. PROPOSED SECONDARY QUESTIONS

Does a common upstream event exist across T2D cases?

If not, how many distinct upstream trajectories exist?

Does the earliest abnormality occur in metabolism, endocrine regulation, nervous-system regulation, immunity, cellular signaling, mitochondrial biology, environmental exposure, or another domain?

Can the earliest transition be detected before insulin resistance?

Can it be detected before β-cell dysfunction?

Can it be detected before abnormal glucose?

What distinguishes future non-progressors from future progressors?

What biological mechanisms protect high-risk individuals?

Do discordant twins demonstrate reproducible pre-disease biological divergence?

Which candidate abnormalities are causal rather than consequential?

47. THE FIRST SCIENTIFIC PRODUCT

Before attempting the enormous prospective study, the most realistic first step would be to produce a scientific manuscript from this concept.

Working title

“Beyond Metabolic Mediators: A Framework for Identifying the Earliest Upstream Initiating Event in Type 2 Diabetes.”

It would:

define the causal problem,

review current models,

distinguish risk factors from initiating mechanisms,

discuss discordant individuals,

discuss temporal ordering,

identify gaps,

propose the ORIGIN-T2D framework,

define falsification criteria.

That is a realistic first scientific product.

48. THE SECOND SCIENTIFIC PRODUCT

After literature and database work:

Working title

“The Earliest Biological Divergence Before Type 2 Diabetes: A Longitudinal Multi-Omic Framework.”

This would establish the actual study protocol.

49. THE THIRD SCIENTIFIC PRODUCT

After data collection:

Working title

“Identification of Upstream Biological States Preceding Type 2 Diabetes.”

This would report the discovery phase.

50. THE FOURTH SCIENTIFIC PRODUCT

If a candidate mechanism survives causal testing:

Working title

“Causal Validation of an Upstream Mechanism in the Initiation of Type 2 Diabetes.”

That is where the hypothesis becomes potentially transformative.

51. THE CENTRAL PHILOSOPHY

The entire program can be reduced to one sentence:

Do not stop at the first abnormality we can explain; continue upstream until we identify what initiated it.

And another:

Do not study only why the 1,000 became diabetic; study why the 9,999 did not.

Those two principles are the intellectual foundation of ORIGIN-T2D.

52. THE ULTIMATE MODEL

The final scientific model could eventually become one of three possibilities.

53. FINAL RESEARCH STATEMENT

Type 2 diabetes should not necessarily be conceptualized only as a disease whose recognizable metabolic abnormalities constitute its beginning.

The abnormalities currently used to describe T2D may represent different points within a longer causal trajectory.

The fundamental research challenge is therefore to reconstruct that trajectory backward and identify the earliest biological divergence that distinguishes an individual who will develop disease from one who will not.

ORIGIN-T2D proposes that the most informative scientific comparison may not be simply:

Diabetic vs healthy

but:

Future diabetic vs future non-diabetic, particularly when conventional risk factors are similar.

Discordant twins, discordant siblings, high-risk non-progressors, low-risk progressors, and individuals who reverse metabolic deterioration provide natural experiments through which this question can be investigated.

The ultimate objective is not merely earlier prediction.

It is:

Identification of the earliest causal event sufficiently upstream that interrupting it could prevent the downstream metabolic cascade from ever developing.

That is the distinction between predicting diabetes and potentially preventing its initiation.

54. THE NEXT RESEARCH STEP

The next stage should be ORIGIN-T2D Part I: The Scientific Evidence Map.

The literature should be systematically examined across:

Earliest insulin resistance.

Earliest β-cell dysfunction.

Central nervous system and hypothalamic regulation.

Autonomic nervous system.

Endocrine signaling.

Immune and inflammatory mechanisms.

Pancreatic islet biology.

Liver, adipose, and muscle signaling.

Mitochondrial biology.

Ion channels and bioelectrical signaling.

Gut and microbiome.

Infectious mechanisms.

Environmental exposures.

Epigenetic and developmental programming.

Psychological, stress, and circadian pathways.

Metabolic resilience and protective mechanisms.

Discordant twins.

Individuals who develop T2D despite low conventional risk.

Individuals with high conventional risk who never develop T2D.

Individuals who reverse prediabetes or T2D.

For every domain, we ask exactly five questions:

Did it appear before diabetes?

Did it appear before insulin resistance?

Does it explain the transition rather than merely correlate with it?

Can it be experimentally manipulated?

Has anyone demonstrated that manipulating it prevents the entire downstream cascade?

That last question is the killer question.

The purpose of ORIGIN-T2D is not to prove that the conventional model is wrong.

It is to determine whether we have been stopping the causal investigation one level too late.