Yasnost
Document · v1.0 · April 2026~4 min read

Methodology.

How Yasnost turns a lab PDF into a Health Score, biological age and an action plan. Here are the formulas, the sources, and the limits of what we do.

§ 02 · Health scale

Health Score — one number instead of thirty

Health Score is a 0–100 number that compresses your labs into one metric. It is not a medical diagnosis and not a "health grade". It's a composite index that's useful for tracking change: did your own number move up or down compared to your previous draw.

The score is a weighted sum of six components:

ComponentWeightWhat's in it
Biological age
25 %
Levine PhenoAge formula — see below
Metabolism
20 %
Glucose, insulin, HOMA-IR, lipid panel, HbA1c
Hormones
15 %
TSH, free T4, testosterone, cortisol, DHEA-S
Inflammation
15 %
hs-CRP, ferritin (as inflammation marker), ESR
Wearable data
15 %
Sleep, activity, resting heart rate — if a wearable is connected
Nutrients
10 %
Vitamin D, B12, folate, iron, magnesium, zinc
Σ Weight total100 %

If a component is missing, its weight redistributes proportionally across the rest. A user with no hormone panel and no wearable still gets an honest score from what's available — no penalty for missing fields.

High
80+

Components sit inside the functional band.

Mid
60–80

Some shifts that are worth a look.

Needs attention
< 60

Several systems need attention.

Per-marker thresholds are explained in the action plan.

§ 03 · Where the reference ranges come from

Which "normal" Yasnost shows — and why there isn't just one

A classic lab report carries one reference range per marker, and each lab sets it itself — by its own assay, instruments and population. That's why the same panel at Invitro, Hemotest and Helix can come back with different "normals". This isn't a lab error — it's how clinical diagnostics actually work.

We don't pretend there's a single universal "correct" range. The source for each reference is chosen per marker, by which source is most reliable for it.

Marker categoryReference sourceExample
Stable lab markersClinical lab textbook (Dolgov / Menshikov, 2025 ed.)Albumin, creatinine, glucose
Covered by clinical guidelinesRussian Ministry of Health guideline with direct citationTSH, lipid panel, HbA1c
Method-dependentThe reference printed on your PDF — no reliable alternativeInsulin, cortisol, testosterone, vitamin D
Qualitative testsThreshold defined by the assay manufacturerAntibodies, infection markers

For method-dependent markers, no single "correct" range exists in principle: different methods (Roche, Abbott, Siemens) produce results that differ by 30–137%. A textbook that ignores method is harmful here. So Yasnost honestly shows the reference from the lab where you tested — and says so plainly. Each marker card in the app shows which source was used and links to it.

The ±5% edge band — our only "extra layer" on top of the range

A standard lab report is binary: in range or out. If your marker sits 1% below the upper limit it's "normal" and you'll never know. A month later it might be 1% above and now it's "out of range". Yasnost adds one in-between band:

  • ±5%
    Edge of range
    A value within 5% of either limit gets an "edge" highlight.
  • →
    Out of range
    A value past the limit gets the standard out-of-range highlight.
§ 04 · Biological age

Biological age via the Levine formula (2018)

Biological age is an estimate of how physiologically "worn" the body is compared to your calendar age. We use the validated PhenoAge formula, published by Dr Morgan Levine in 2018 in Aging. It was trained on the NHANES III cohort (~9,900 people) and validated on NHANES IV.

The calculation needs 9 biomarkers + calendar age:

  • Albuming/L
  • Creatinineµmol/L
  • Glucosemmol/L
  • C-reactive protein (hs-CRP)mg/L
  • Lymphocytes%
  • MCV — mean corpuscular volumefL
  • RDW — red cell distribution width%
  • Alkaline phosphatase (ALP)U/L
  • White blood cell count (WBC)10⁹/L

If even one marker is missing, bio age isn't computed. We don't impute averages and we don't guess. The report shows which markers are missing and where to test them.

Levine formula · 2018
xb = −19.9067
   − 0.0336 · albumin
   + 0.0095 · creatinine
   + 0.1953 · glucose
   + 0.0954 · ln(CRP)
   − 0.0120 · lymphocytes%
   + 0.0268 · MCV
   + 0.3306 · RDW
   + 0.00188 · ALP
   + 0.0554 · WBC
   + 0.0804 · age
PhenoAge = 141.50225 + ln(−0.00553 · ln(1 − MC)) / 0.090165

where  MC = 1 − exp(−exp(xb) · (exp(120·γ) − 1) / γ),   γ = 0.0076927

A positive gap (bio age > calendar) signals that metabolic, immune or kidney function is running like that of older people. These 9 markers aren't the full picture of ageing, and the formula doesn't predict an individual lifespan. It's a change tracker: it makes sense to compare your bio age with your previous draw, not with a "norm".

§ 05 · Analytics

How Yasnost uses AI

AI is used for two things only:

Used for
  1. PDF parsing. Extracting marker name, value, unit and date from any lab's format.
  2. Plain-language summary. Turning already-computed numbers into a readable note: what shifted, what to watch, what to flag with your doctor.
Not used for
  • Health Score — deterministic formula, see above.
  • Biological age — Levine formula, see above.
  • "In range / out of range" calls — fixed thresholds from the catalogue.
  • Supplement and dose suggestions — bound to the marker and its delta, not to a model's free interpretation.

Which means the numbers on the dashboard are reproducible: feed in the same labs twice and you get the same score and the same bio age.

§ 06 · Limits

What we don't do

  • — 01
    We don't diagnose.
    "High CRP" in the report is a signal, not a diagnosis.
  • — 02
    We don't prescribe and we don't override what your doctor has prescribed.
    Any change to a regimen goes through a clinician.
  • — 03
    We don't replace a clinic visit.
    If the report shows red flags, the next step is a doctor — not a tweak to your supplements.
  • — 04
    We don't train models on your labs.
    Data sits on Russian servers in line with Federal Law 152-FZ.
  • — 05
    Every marker carries its own disclaimer.
    Each biomarker card in the app spells out the limits of what we can say — especially for method-dependent markers, where no reliable universal range exists.
§ 07 · Sources

Sources

  1. Dolgov V.V., Menshikov V.V. — Clinical Laboratory Analytics.
    Vol. 1–4, 2025 edition (Russian)
    Primary source for stable lab markers.
  2. Russian Ministry of Health clinical guidelines.
    cr.minzdrav.gov.ru
    Priority source for markers covered by official guidelines — legal priority since 01.01.2025 under Federal Law 489.
  3. Levine ME et al. An epigenetic biomarker of aging for lifespan and healthspan.
    Aging (Albany NY). 2018;10(4):573–591 · DOI: 10.18632/aging.101414 · PMC5940111
    Original paper on PhenoAge.
  4. NHANES — National Health and Nutrition Examination Survey.
    cdc.gov/nchs/nhanes
    Training and validation cohort for PhenoAge.
§ 08 · Contact

Questions about the methodology

If you're a clinician, researcher or journalist and want to discuss the methodology — write to us. We reply.

support@yasnosty.ru →
← Back home