STAGE 01
Healthy Beta Cell
Islet β-cells maintain glucose homeostasis through tightly regulated, biphasic insulin secretion. Cellular architecture and mitochondrial function are intact.
Kihealth Labs · Scientific Foundation
The molecular biology of disease interception.
Disease begins in cells, not in symptoms. Kihealth Labs measures the molecular signals that pancreatic β-cells release when they are stressed, injured, and dying — years before glucose, HbA1c, or clinical presentation reveal the disease.
Pancreatic β-Cell BiologyCell-Free DNATissue-Specific MethylationDigital Droplet PCRLiquid Biopsy
The Window of Opportunity
Beta-cell death is measurable before HbA1c ever changes.
Traditional biomarkers move only after substantial, irreversible β-cell loss. Kihealth detects the molecular signature of active β-cell death years earlier — inside the window where intervention still preserves function.
% Demethylated INS cfDNA
Rises first — peaking inside the window — then declines as the at-risk reserve is depleted.
Functional Beta-Cell Mass
Largely preserved through the window of opportunity, then declines once clinical dysfunction begins.
Insulin
Compensatory secretion masks early β-cell loss; first-phase release blunts before fasting insulin falls.
C-peptide
Tracks endogenous insulin output; declines as functional β-cell reserve is exhausted.
HbA1c
Stable early, rising sharply only once significant beta-cell loss has occurred.
Traditional Biomarkers Remain Stable
HbA1c and glucose often stay within the normal range while silent beta-cell injury is already underway.
Kihealth Detects Active Beta-Cell Death
Demethylated INS cfDNA rises as beta cells undergo apoptosis — a direct, earlier molecular signal.
Intervene Before Irreversible Loss
Earlier detection opens the window to preserve functional beta-cell mass before substantial loss has occurred.
Interactive Disease Progression
From the first molecular signal to clinical diagnosis.
Diabetes doesn't begin on the day it's diagnosed. Step through each phase to compare Kihealth's molecular signal against insulin, C-peptide, and traditional glycemic markers.
Phase 02 · Injury~5 years before diagnosis
Active β-Cell Death
Apoptotic β-cells release tissue-specific, demethylated INS cfDNA fragments into circulation. Functional mass is still largely intact and both insulin and C-peptide remain within reference ranges.
Biological Signals
- ↑ Demethylated INS cfDNA in plasma
- Autoantibody seroconversion (T1D)
- Islet inflammation intensifies
Kihealth Signal
Peak detection window — Beta Intercept™ signal is maximal.
Molecular signal precedes clinical diagnosis by years.
Relative Biomarker Signal at this Phase
Normalized to peak · illustrative
Kihealth cfDNA95%
C-peptide88%
Insulin90%
HbA1c8%
Fasting Glucose12%
The Biology of Disease Onset
Diabetes does not begin with high blood sugar.
Both Type 1 and Type 2 diabetes begin as a cellular disease of the pancreatic β-cell. Long before fasting glucose or HbA1c cross clinical thresholds, β-cells are exposed to autoimmune attack, metabolic overload, chronic inflammation, and lipotoxicity that impair insulin secretion and eventually trigger apoptosis.
By the time hyperglycemia is measurable, the biological cascade has been active for years — and up to 50% of functional β-cell mass may already be lost. Conventional laboratory markers describe the endpoint of that process, not its onset.
Kihealth Labs was founded to measure the earlier chapters of that story: molecular evidence of active β-cell stress, injury, and death — the biology that precedes clinical disease.
Figure 1
Pancreatic islet of Langerhans — β-cells (insulin), α-cells (glucagon), and intra-islet capillaries.
Kihealth Detection window
The Life Cycle of a Beta Cell.
Diabetes is diagnosed at the end of a long, silent decline. By the time glucose crosses the clinical threshold, most beta cells are already lost.
Stage 01
Healthy
100%
Islets produce insulin in balance with demand.
Ki Detects
Stage 02
Stressed
78%
Metabolic load and inflammation strain beta cells.
Ki Detects
Stage 03
Silent Loss
40%
Beta cells die and release cfDNA — years before symptoms.
Stage 04
Diabetes
10%
Glucose crosses clinical thresholds. Most beta cells are gone.
Why Kihealth exists
We detect beta-cell death while it’s happening — not years later, when the diagnosis is already too late.
Illustrative model of beta-cell decline. Values are directional, not patient-specific. Not for diagnostic use.
The Beta-Cell Progression
From healthy islet to clinical diabetes.
Seven molecular stages define the trajectory from β-cell homeostasis to overt diabetes. Kihealth's diagnostics are designed to detect the earliest of these stages — when intervention remains most effective.
STAGE 02
Ki detects
Metabolic Stress
Chronic hyperglycemia, hyperlipidemia, and increased secretory demand drive endoplasmic reticulum stress and mitochondrial dysfunction within the β-cell.
STAGE 03
Ki detects
Inflammation
Islet-resident macrophages release IL-1β and TNF-α; in T1D, autoreactive T cells infiltrate the islet. Local cytokine signaling amplifies β-cell injury.
STAGE 04
Ki detects
Apoptosis
Injured β-cells undergo caspase-mediated programmed cell death, releasing fragmented, tissue-specific DNA into the peripheral circulation.
STAGE 05
Ki detects
Loss of Function
β-cell mass falls and remaining cells de-differentiate. First-phase insulin release is blunted before fasting glucose becomes abnormal.
STAGE 06
Prediabetes
Impaired fasting glucose and impaired glucose tolerance emerge. HbA1c drifts into the 5.7–6.4% range; ~50% of β-cell function may already be lost.
STAGE 07
Diabetes
Clinical diagnosis by HbA1c ≥ 6.5%, fasting glucose ≥ 126 mg/dL, or symptomatic hyperglycemia. Microvascular complications have often begun.
Kihealth window of molecular detection (stages 2–5)
Clinical diagnosis threshold (stages 6–7)
Figure 2
Cell-free DNA fragments carrying tissue-specific cytosine methylation marks.
Cell-Free DNA · Methylation
A molecular record of cell death, circulating in blood.
When any cell in the body dies, it releases short, nucleosome-protected DNA fragments into the bloodstream. This cell-free DNA (cfDNA) circulates transiently in plasma, providing a near–real-time readout of tissue turnover across the organism.
Every cell type carries a distinct cytosine methylation signature — an epigenetic barcode established during differentiation. By interrogating methylation at β-cell–specific loci (INS, GCK), we can measure the fraction of cfDNA that originated specifically from pancreatic β-cells.
The result is a tissue-specific, minimally-invasive assay for active β-cell death — biology that is otherwise invisible without pancreatic biopsy.
Fragment size
~160–200 bp
Nucleosomal footprint of apoptotic release.
Plasma half-life
~15–120 min
Rapid turnover reflects real-time biology.
Tissue signature
Methylation
Cytosine methylation encodes cell-of-origin.
Assay input
1–2 mL plasma
Standard EDTA venous draw.
Liquid Biopsy
Blood as a diagnostic window into internal organ biology.
Liquid biopsy is the analysis of circulating molecular material — cfDNA, cell-free RNA, proteins, and extracellular vesicles — from a routine blood draw. Pioneered in oncology, it eliminates the anatomical and safety limits of tissue biopsy while providing a systemic view of disease biology.
For metabolic and autoimmune disease, liquid biopsy is uniquely powerful: pancreatic islets are inaccessible for routine sampling, and β-cell biology is otherwise measurable only by proxy. A plasma sample brings that biology into the diagnostic workflow.
Clinical applications
- Earlier disease detection
- Longitudinal disease monitoring
- Therapeutic response tracking
- Pharmaceutical research
- Companion diagnostic development
- Preventative medicine applications
Figure 3
Post-centrifugation EDTA tubes — plasma (upper) contains cell-free DNA; cellular fraction (lower) is discarded.
Figure 4
Digital droplet PCR — nanoliter emulsion partitioning enables absolute quantification of rare molecular targets.
Digital Droplet PCR
Quantifying rare molecular signal, droplet by droplet.
β-cell–derived cfDNA is exceptionally rare — often less than 0.1% of total plasma cfDNA. Detecting it reliably requires a technology with single-molecule sensitivity. Digital droplet PCR (ddPCR) partitions each reaction into tens of thousands of nanoliter droplets, enabling absolute quantification by Poisson statistics — without a standard curve, and with far greater precision than conventional qPCR.
- 1Cell-free DNA extractionTotal cfDNA isolated from EDTA plasma using magnetic-bead chemistry with automated liquid handling.
- 2Bisulfite conversionUnmethylated cytosines are converted to uracil, preserving the methylation fingerprint of the tissue of origin.
- 3Droplet partitioningEach sample is partitioned into ~20,000 nanoliter emulsion droplets, isolating individual DNA molecules.
- 4Target amplificationMethylation-specific probes anneal to β-cell–derived loci and are amplified within positive droplets.
- 5Absolute quantificationPoisson statistics on positive droplet counts yield an absolute copies-per-mL β-cell cfDNA measurement.
High-Complexity Laboratory
Reproducibility engineered into every well.
Kihealth's CLIA / COLA-certified laboratory runs cfDNA extraction, bisulfite conversion, and ddPCR assays on validated robotic liquid-handling platforms. Automation reduces operator variability, standardizes assay conditions, and enables the throughput required for population-scale longitudinal monitoring.
±0.5%
Inter-assay CV
20k
Droplets per reaction
96
Samples per plate
The Intercept IQ™ Platform
Translating molecular biology into actionable diagnostics.
Intercept IQ™ is the diagnostic layer built on the biology described above. It integrates β-cell cfDNA measurement, autoimmune and metabolic biomarkers, laboratory automation, longitudinal patient data, and AI-enabled analytics into a unified precision-diagnostics platform that can be extended across disease areas.
Molecular biomarker discovery
Liquid biopsy testing
Laboratory-developed diagnostics
Longitudinal patient data
AI-enabled predictive analytics
Physician-facing reporting
Patient monitoring infrastructure
Pharmaceutical & therapeutic monitoring
Extensible across disease areas
Metabolic Disease
Oncology
Neurodegenerative
Therapeutic Monitoring
Preventative Medicine
Current Diagnostic Programs
Precision diagnostics for β-cell disease.
Type 1 Diabetes
BetaIntercept™ T1D
Combines β-cell–specific cfDNA methylation signal with islet autoantibody panels (GADA, IA-2A, ZnT8A, IAA) to detect active autoimmune β-cell destruction before clinical onset.
Type 2 Diabetes
BetaIntercept™ T2D
Integrates β-cell cfDNA turnover with C-peptide, insulin, and glycemic markers to quantify metabolic β-cell stress in individuals at risk for T2D and progression.
Scientific Validation
Built on decades of peer-reviewed islet science.
Kihealth Labs advances its science through analytical validation, prospective and retrospective clinical studies, academic partnerships, and access to some of the most important longitudinal biobanks in diabetes research.
Yale University technology license
Foundational β-cell methylation IP developed and licensed from Yale School of Medicine.
High-complexity LDT validation
Full analytical validation package covering LoD, linearity, precision, specificity, and clinical concordance.
CLIA / COLA laboratory infrastructure
Kihealth-operated high-complexity clinical laboratory with automated ddPCR workflows.
Nemours pediatric longitudinal T1D study
Multi-year cohort tracking β-cell cfDNA dynamics in at-risk pediatric patients.
DAISY cohort analysis
Retrospective analysis of the Diabetes AutoImmunity Study in the Young biobank.
TEDDY / TrialNet collaboration
Access to prospectively-collected samples from The Environmental Determinants of Diabetes in the Young.
Yale pancreatic transplant research
Reference cohort validating β-cell cfDNA release from graft injury.
ADA scientific presentations
Poster and abstract presentations at the American Diabetes Association Scientific Sessions.
Peer-reviewed publications
Manuscripts, white papers, and preprints spanning three decades of foundational islet science.
Why It Matters
From reactive diagnosis to proactive interception.
Kihealth Labs is building diagnostics for a future where disease is detected at its molecular origin, monitored across the trajectory of the underlying biology, and treated before irreversible tissue loss occurs. By combining pancreatic biology, cfDNA methylation, ddPCR, laboratory automation, and longitudinal data, Kihealth aims to shift medicine from reactive diagnosis to proactive disease interception.
Earlier
Detect biology at the cellular stage, years before symptoms.
More predictive
Model trajectory across the disease continuum.
More personalized
Tuned to each patient's molecular fingerprint.
Kihealth Labs is the scientific foundation of the Kihealth ecosystem.
Combining molecular diagnostics, β-cell biology, liquid biopsy, AI-enabled analytics, and longitudinal biomarker data to build the next generation of precision disease interception.