Contact us
Pick whichever is easiest; we reply quickly.
Leave a message
Scan to add us on WeChat and chat with us
How 13 hormones and signalling molecules rise and fall during and after a run. Every curve is graded for evidence, and you can upload your own running records for a personal reading generated in your browser.
How 13 hormones and signalling molecules rise and fall during a run and in the hours after it, drawn on one chart.The line style is the evidence: solid lines mean human measurements under similar conditions; dashed lines are estimates. Pick an intensity and a duration; further down you can also upload your own running records to get a personal reading in your browser.
“Relative response”: each substance is scaled so that its largest response across this page’s scenarios is 100, so substances cannot be compared with each other; to compare sizes, switch to “Fold change”. Falls in insulin and testosterone are drawn below the baseline, where the scale is tighter than above it.
Follows the scenario chosen above. The horizontal axis is time after the run (log scale). Solid bars mean human measurements, outlined bars are estimates, hatching means unknown.
What each substance is, how it changes during a run and what to watch for. The letter is the evidence grade: A means several human studies or meta-analyses agree, B means fewer or mixed human studies, C means mainly animal studies, conference abstracts or inference.
Many of the “arrows” on a runner’s check-up report are left by the runs of the previous days. First tell the shadows of training apart, then decide what is worth re-testing.
| Test | Typical deviation | Why | What to do |
|---|---|---|---|
| Creatine kinase (CK) | Raised, up to several times the upper reference limit | Muscle micro-damage: rises within 12 hours, peaks at 1–3 days and settles 3–5 days after it stops | No long or hard runs in the 72 hours before the test; a very low CK-MB share points to skeletal rather than heart muscle |
| Urea, uric acid | Raised | May be linked to dehydration, fasting or recent exercise | Re-test after normal drinking and rest |
| Creatinine and eGFR | Creatinine high, eGFR low | Muscle mass, recent exercise and dehydration all push creatinine up | Re-test under standard conditions; if needed, add a urine albumin-to-creatinine ratio |
| Urine specific gravity, urine ketones | Higher specific gravity, ketones positive | A long fast and no water in the morning concentrate the urine | Fasting is not the same as not drinking: you may drink plain water before the blood test |
| Urine protein | Weakly positive | Transient proteinuria after exercise is common | Re-test with a first-morning sample on a rest day |
| Total bilirubin | Slightly raised | May be linked to long fasting and similar | Look at the ratio of direct to indirect bilirubin; let a doctor judge |
| hs-CRP | Raised after a marathon | Post-race inflammatory response, peaking at 24–72 hours and settling in about 3 days | Re-test at least a week after a big race |
The above helps you read a lab report; it does not replace a doctor’s judgement.
Nearly ten years of phone-app records from one amateur runner: about 1,500 runs, about 14,000 km, including more than ten marathons and longer races.
These records have no heart rate, so the chart on the right shows only monthly distance and 48-hour coverage: how many days in a month fell within 48 hours after a run, the window of raised muscle insulin sensitivity.
Names, age, race results and other personally identifying details have been removed from the case, and it contains no health data.
Choose your running records and the page estimates your intensity structure, 48-hour coverage and distance “spikes”, then uses the same model as above to draw your typical run and the curve of any single run.Files are parsed only in your browser: nothing is uploaded or saved, and closing the page clears everything.GPS coordinates are used only to work out distance, then discarded.
Supported: FIT, GPX and TCX (single activities); activity-list CSV exported from Garmin, Strava and others; JSON with date, distance and duration fields (for example a Codoon export); and ZIP archives holding these files. When a single-activity file has heart rate, the curves use minute-by-minute heart rate.
The defaults are a rough conversion of the common three-zone model. If you have had a lactate or ventilatory threshold test, enter your measured values.
Each substance is described by a simple kinetic model: during the run it moves towards a target set by intensity and duration, and after the run it falls back at its own clearance rate. Cortisol has an extra delay, so its peak comes after the finish. The model parameters are anchored to the human studies in the table below; anything beyond the anchors is drawn dashed.
Intensity is expressed as a percentage of VO₂max: heart rate is converted with the Swain 1994 regression and pace with Daniels’ VDOT formula. The three intensity levels correspond to about 40%, 60% and 80% of VO₂max, exactly the three intensities at which Hill 2008 measured cortisol.
| Substance | Anchoring studies | What it constrains |
|---|
Pick whichever is easiest; we reply quickly.
Scan to add us on WeChat and chat with us