Chronological age is an accident of the calendar. Chrona reads blood chemistry, DNA methylation and wearable telemetry to compute the age your tissues are actually living at — organ by organ, with confidence intervals a clinician can defend.
Deployed inside longevity clinics, health systems and underwriting teams
One has the arterial stiffness of a 61-year-old. The other has the immune repertoire of someone 12 years younger. Same birthday, same census box, radically different risk. Chronological age explains roughly a third of the variance in mortality — the rest is written in the tissue.
Chrona quantifies that remainder. Not as a wellness score, but as a regulated clinical measurement with intervals, provenance and a paper trail.
Read the validation recordof mortality variance explained by chronological age alone, in the Chrona reference cohort.
median spread in biological age between top and bottom deciles at chronological 50.
hazard ratio for all-cause mortality per decade of biological age excess.
of that excess is attributable to modifiable inputs across an 18-month window.
Every Chrona output traces back to the same calibrated model family, so the composite index, the organ clocks and the intervention forecast never contradict each other in front of a patient.
The Chrona Index compresses 118 features — metabolic panel, inflammatory markers, methylation beta-values and 90-day wearable telemetry — into one 0–100 score, delivered with a 95% credible interval and a full feature-attribution trace.
A patient can carry a 38-year-old immune system and a 52-year-old liver at the same time. Chrona resolves nine organ systems independently, each with its own assay basis, so intervention lands where the deficit actually is.
A biological age of 41 means little without direction. Pace of Aging measures how many biological years a patient accrues per calendar year — the earliest signal that a protocol is working, detectable within two quarters.
Chrona runs counterfactuals against the fitted model: adjust a modifiable input, see the projected 18-month effect on biological age with uncertainty bands. Effect sizes are drawn from intervention arms in the reference cohort, never extrapolated beyond observed support.
Nine systems, nine independent estimators, one chart. Hover any segment to inspect its assay basis and deviation from the patient’s chronological age. This is the view most clinicians open first.
Dashed ring = chronological age 47 · all nine systems currently resolve inside it
Nine systems resolved independently from a single draw. Two are flagged for lagging the patient’s own composite index by more than 2.5 years.
A stripped-down public build of the Chrona engine — six modifiable inputs, the same effect directions as the clinical model, computed entirely in your browser. Nothing is transmitted, stored or logged.
Illustrative model for demonstration. Effect directions mirror the clinical engine; magnitudes are simplified and this is not a diagnostic device or medical advice.
Four stages, fully auditable. Every result carries the assay lot, model version and calibration date that produced it — so a number from March is still interpretable in November.
Venous draw processed through CLIA/ISO-15189 partner labs, plus optional methylation array and a 90-day wearable export. Chrona never handles the phlebotomy — only the data contract.
Batch effects, assay drift and demographic confounds are regressed out against the reference cohort. This is the stage most consumer tests skip, and the reason their results move when nothing about the patient has.
An ensemble of elastic-net and gradient-boosted estimators, one per organ system, trained on 214,880 adults with a median 8.4 years of follow-up. Outputs are Bayesian-calibrated rather than point-predicted.
Results land as a signed API response, a clinician dashboard and a patient-facing PDF that a non-specialist can read without being frightened by it. Every figure links back to its input provenance.
Chrona is deployed in regulated settings, so the validation record is the product. Prospective replication, external cohorts, and every negative result we found on the way.
Compared against chronological age alone in a held-out validation split (n = 42,976), adjusted for sex, smoking status and ancestry stratum.
Duplicate draws, same patient, 14 days apart, across 3 partner labs and 2 assay lots (n = 1,204).
Zürich, Rotterdam, Kyoto and Baltimore. Effect direction held in all four; magnitude attenuated 9–14% out of sample.
Including two that failed to replicate our original dermal clock, which we retired in v3.8.
Post a biomarker payload, receive a fully attributed age profile. Median response 340 ms, versioned models, and a sandbox that returns synthetic patients so you can build before your lab contract is signed.
v4.2 — old versions stay callable for 24 months.index.recomputed and pace.threshold.# request a full age profile
curl https://api.chrona.bio/v4/index \
-H "Authorization: Bearer $CHRONA_KEY" \
-H "Chrona-Version: 4.2" \
-d '{
"subject_ref": "8842-K",
"chronological_age": 47,
"panel": {
"hs_crp": 0.4,
"hba1c": 5.9,
"cystatin_c": 0.82,
"alt": 41,
"vo2_max": 48
},
"methylation": "epic_v2:s3://…/8842K.idat",
"include": ["organ_clocks", "pace", "shap"]
}'import { Chrona } from "@chrona/sdk";
const chrona = new Chrona(process.env.CHRONA_KEY, {
version: "4.2"
});
const profile = await chrona.index.create({
subjectRef: "8842-K",
chronologicalAge: 47,
panel: {
hsCrp: 0.4,
hba1c: 5.9,
cystatinC: 0.82,
alt: 41,
vo2Max: 48
},
include: ["organ_clocks", "pace", "shap"]
});
console.log(profile.chronaIndex); // 41.2from chrona import Chrona
client = Chrona(api_key=os.environ["CHRONA_KEY"], version="4.2")
profile = client.index.create(
subject_ref="8842-K",
chronological_age=47,
panel={
"hs_crp": 0.4,
"hba1c": 5.9,
"cystatin_c": 0.82,
"alt": 41,
"vo2_max": 48,
},
include=["organ_clocks", "pace", "shap"],
)
print(profile.chrona_index) # 41.2{
"chrona_index": 41.2,
"ci_95": [39.8, 42.6],
"delta_years": -5.8,
"pace_of_aging": 0.81,
"organ_clocks": {
"immune": 38.4,
"cardiovascular": 39.7,
"hepatic": 44.9
},
"flags": ["hepatic_lag", "msk_lag"],
"model_version": "4.2.1"
}“The organ panel changed how I sequence care. I stopped treating ‘aging’ as one problem and started treating the hepatic clock that was eight years ahead of everything else. The patient understood the chart in about forty seconds, which is the part I did not expect.”
“We piloted three biological age vendors. Chrona was the only one whose test–retest variance was small enough that a six-month follow-up meant anything at all.”
“The API shipped in a week. The clinical validation pack took our compliance team four months to get through — which is exactly the right ratio, and the opposite of what we normally see.”
“For underwriting, a 0.847 C-statistic on all-cause mortality is not a marketing number. It moved our pricing bands. That has not happened from a new data source in nine years.”
For single-site longevity and precision-health practices.
Includes 120 results / month
Start a pilotFor multi-site groups and digital health platforms.
Includes 600 results / month
Request accessFor health systems, insurers and research cohorts.
Volume from 10,000 results / year
Talk to usAssay and phlebotomy costs are billed by your lab partner, not by Chrona. All plans include the reference cohort recalibration at no additional cost.
The four we get on every first call, and the two that matter most to compliance teams.
Ask something elseNo. Chrona is a clinical decision-support tool, not a diagnostic under EU MDR or FDA classification. It is used by licensed clinicians alongside standard workup — it does not diagnose disease, and we are explicit about that in every report footer.
Three things: normalisation, interval reporting and stability. Consumer tests typically report a point estimate from a single assay with no batch correction, which is why results can swing several years between draws with no underlying change. Our test–retest SD is 0.31 years across labs and assay lots.
A standard metabolic and inflammatory panel gets you the Chrona Index and six of the nine organ clocks. Methylation array unlocks the remaining three and tightens the interval by roughly 30%. Wearable telemetry is optional but materially improves the pace-of-aging estimate.
EU customers are pinned to Zürich and Frankfurt; US customers to us-east and us-west. Data never crosses regions. We process under a BAA for HIPAA and as a processor under GDPR Art. 28, and we do not train models on customer data without a separate, revocable research agreement.
Six months is the shortest interval where a change in pace of aging is reliably distinguishable from assay noise. Quarterly testing produces prettier charts and no additional clinical signal, so we do not recommend it.
Old versions stay callable for 24 months, and every stored result records the version that produced it. When we retire a clock — as we did with the v3.8 dermal model — we publish the failure analysis and re-score affected patients at no cost.
Clinical pilots start with a 40-result validation batch against your existing panel data. Most teams see their first Chrona Index inside three weeks.
SOC 2 Type II · ISO 27001 · HIPAA BAA available · EU data residency