A note from the founder

Bars span first to last dose of each compound; the number is doses taken. Two eras: retatrutide alone from Feb 14, dosed every few days by design, then a daily stack from Mar 23. Across that daily stretch I dosed on 161 of 163 days (98.8%). Across the whole 200 days it is 170 days — a lower number that mostly reflects the every-few-days titration at the start, which is why both are here.

The doses I did not take

121 logged skips, kept in the record rather than deleted: Semax 58 · Selank 58 · Kisspeptin-10 2 · CJC-1295 (No DAC) + Ipamorelin 2 · Tesamorelin 1. Adherence numbers that only count the doses you took are not adherence numbers.

Weight

17 self-reported weigh-ins, 285 to 220 lb (-65 lb). Dots sit at their real dates, so the gaps are real — the largest is 48 days with no reading. No trend line is drawn, because 17 uneven points do not support one. A home scale is not a controlled measurement, and diet and training changed alongside everything else.

Bloodwork

Read the denominator first. 303 results across 81 markers, but they are not equally comparable. 62 markers were measured at both Jan 23 26 and Jul 7 26, so those carry a real before-and-after. 19 do not: 10 were measured once, and the rest were added or dropped between panels. Every one of them is below anyway, with the reason it has no delta printed on the row. The Mar 8 24 draw predates the protocol by 22 months and is shown as a detached ring, never averaged in.

The six that actually changed category

Ranked by the band they moved between, not by percentage — a big percentage inside one band is noise, and ranking by it would have put a TSH wobble above a liver enzyme leaving the high range. 4 improved a category, 2 lost one.

All 81 markers

The five findings I'd rather not have published

Of the 62 markers measured at both ends, 9 moved into the optimal band and 3 moved out of it. But a count is not a finding. These are the five results that actually changed how I think about what I did, and four of them are unflattering. A page that leads with its wins is marketing; the wins are in the register above and you can read them there.

Finding 01 · a safety signal

My hematocrit was high before I started, and I raised it anyway.

46.4% in 2024. 50.4% at baseline — already flagged high, three weeks before my first dose. Then 51.6, then 48.3 on the partial May panel, then 51.5% at the endpoint. It is the one marker on this page that was out of range when I started, stayed out of range throughout, and finished higher than it began.

Rising haematocrit is a known effect of androgens and of growth-hormone secretagogues, and I was running both. Thicker blood is not a cosmetic result; it is the mechanism behind the cardiovascular risk that gets people taken off testosterone therapy. The honest reading is that I ran a protocol with a known haematologic effect on top of a marker that was already elevated, for six months, and only noticed the shape of it when I built the tool that draws this chart.

The 48.3 in May is almost certainly hydration, not improvement — it is a single partial panel between two readings of 51.6 and 51.5. I have left it in rather than smoothing it out, because that is the kind of point that makes a chart lie in your favour.

Finding 02 · the axis

My testosterone went down, not up.

Total testosterone 614 → 461 → 520 ng/dL: down 15% from baseline, and out of the optimal band into merely-in-range. Free testosterone 133.3 → 74.7 → 103. LH dipped to 2.2 at the April draw before recovering to 3.7. SHBG went the other way, 18 → 24, which is what a large weight loss usually does and which by itself pushes free testosterone down.

I am publishing all four numbers together because publishing only SHBG — which moved into the optimal band and looks like a win — would be a lie of composition. An axis has to be read as an axis. The April trough is the part I cannot explain away: total T, free T and LH all bottomed on the same draw, which is the shape of transient suppression rather than of noise in one assay.

Finding 03 · the null result

129 doses of growth-hormone secretagogues moved IGF-1 by 2%.

77 doses of Tesamorelin and 52 of CJC-1295 / Ipamorelin over the run. IGF-1 is the standard downstream readout for that entire class — if the mechanism is doing what the category claims, this is the marker where you see it.

255 → 228 → 261 ng/mL. Net +2%, with the middle draw lower than baseline. On my own body, at the doses I ran, that is a null result, and it is the single most useful thing on this page for anyone deciding whether to spend money on that category. I would not have found it without drawing labs, and I would not have published it if this page were a sales asset.

Finding 04 · the immovable one

Lp(a) was 146 at the start and 145 at the end.

146 → 141 → 145 nmol/L against a cutoff of 75. Roughly double, at every draw, for the whole run. Lp(a) is largely set by genotype and is famously indifferent to diet, training and weight loss, so this is the number behaving exactly as the literature says it will.

I include it because it is the counterweight to everything above it. I lost 65 lbs and moved a lot of markers; the marker that carries the most independent cardiovascular risk in my panel did not move at all, and was never going to. Any page that shows you a wall of improving numbers without showing you the one that is fixed is teaching you the wrong lesson about what these tools do.

Finding 05 · no verdict available

My thyroid antibodies tripled, and I am not going to score that for you.

TPO antibodies 3 → 10 → 11 IU/mL. Our catalog publishes no scoring band for antibody titres, so the register above shows the values and the word not scored rather than a colour and a verdict — and it deliberately shows no percentage, because "+267%" on a titre with no published range reads as an emergency and means very little on its own.

What I can say is the direction and the size, which is why the numbers are printed. A rise like that inside a reference range is the sort of thing a clinician wants to see the trend on, and it is on my list for the next draw. Saying "we don't know" in public is the whole point of the exercise.

Why you should not attribute any of this to peptides

This is an N of 1, uncontrolled, unblinded, and confounded on purpose — I was not running an experiment, I was trying to get healthy. Everything below changed inside the same 200 days:

So: I cannot tell you which change did what, and I will not pretend to. What this page is good for is the method — measure first, log every dose, draw again, publish the whole panel including the parts that argue against you. That is the thing I would want from someone else, and it is what the app was built to make ordinary.

Who I am, and what I am not

I am not a physician, a pharmacist, a nurse, or a scientist. I hold no clinical credential of any kind. I am one person who got tired of guessing, started measuring, and then built software to make the measuring easier. Nothing on this page is medical advice, and nothing on it should be read as a recommendation to run what I ran — several of these compounds are research chemicals with no approved human indication, and I made decisions about them that a doctor would likely have argued with. Talk to a clinician who knows your history. What I can offer you is my full record and an honest account of its limits, which is more than I could find anywhere when I started.

Where the numbers come from

Exported from my own account with the app's export button, then computed by tools/build-n1.mjs — nothing on this page is typed by hand. Three things you should know about the raw file. The log held 1060 rows, but 238 were duplicates from re-importing the same CSV twice, leaving 822 real entries and 701 doses actually taken. Two rows carried a 2016 date, a transposition of 2026, corrected rather than dropped. And per-vial tracking only began Jul 9, about five months in, so the 16 vials here are the later part of the run, not all of it. Bloodwork is scored against the same goodtides marker catalog the app uses.

I built what I wished existed.

One person. 200 days. Every number.

What I ran, and when