At BloodTrack, people upload their own pathology reports and we turn them into something they can actually read and track over time. That gives us an unusual vantage point: a growing, fully anonymised library of real-world blood work. We went through 7,411 individual marker results from 333 reports uploaded by 115 people to answer one question — what patterns actually show up when ordinary people start tracking their own bloods?
Everything below is aggregated. No individual is identifiable, and we only report figures backed by a meaningful sample — most numbers here rest on dozens to hundreds of results, and we show the sample size (n) next to every finding. This is not a population study: our users skew toward health-conscious adults, many of them men optimising their hormones. Read it as a snapshot of people who track their blood work, not of the general public.
The headline: 7 in 10 reports had something out of range
Across 264 reports with a full panel (five or more markers), 70% contained at least one out-of-range result, averaging 2.9 flags per report. Looking at every result individually, 10.8% fell outside the lab's reference range. A perfectly clean panel is the exception, not the rule — even among people healthy enough to be tracking proactively.
1. The relationships hiding in your hormones
The most interesting findings in our data weren't flag rates — they were how markers move together. For every pair of hormone markers people tested in the same report, we calculated the correlation. Here is the full picture:
A few relationships stand out:
- Free and total testosterone track tightly (r = 0.89, n = 41) — expected, but it confirms we are reading both consistently.
- LH and FSH move in lockstep (r = 0.69, n = 51) — the two pituitary signals that drive the testes rise and fall together, which is why they tend to be suppressed together on testosterone therapy.
- Free testosterone is linked to thicker blood (r = 0.50, n = 23) — while total testosterone is not (r = −0.17). It is the free, usable fraction that tracks with a higher haematocrit.
- More free testosterone, more estradiol (r = 0.50, n = 40) — the signature of aromatisation, where the body converts testosterone into estrogen.
- SHBG pulls the other way — inversely linked to both free testosterone (r = −0.30) and haematocrit (r = −0.52): more binding protein, less free hormone, thinner blood.
2. The hidden cost of optimisation: thicker blood
That free-testosterone-to-haematocrit link matters. Reports carrying the hallmark of testosterone therapy (suppressed LH alongside maintained testosterone) ran a median haematocrit of 0.49, versus 0.43 for everyone else (n = 9 vs 15) — a shift toward the range where doctors start recommending blood donation to manage thickness. Raising red-cell volume is a well-recognised effect of testosterone therapy; our data shows it tracking with the free hormone specifically, not the total number on the page. It is easy to ignore and easy to monitor.
3. The “your testosterone is normal” trap
Most doctors quote your total testosterone. But much of it is bound up and biologically unavailable; what your body can actually use is free testosterone. Of 63 reports that measured both, 47 had a perfectly normal total reading — yet 8 of those (about 1 in 6) had flagged free testosterone. A “normal” headline number can quietly hide an abnormal usable one. The culprit is usually SHBG: in our data total testosterone barely tracked SHBG (r = 0.03, n = 33), but SHBG was inversely tied to free testosterone (r = −0.30, n = 24) — the marker almost no one asks for, silently setting how much of your testosterone actually counts.
4. When blood breaks, it usually breaks in one direction
Grouping markers into panels reveals something clean: most systems fail predictably in a single direction. Each row below is hundreds of results, not a handful.
| Panel | Results (n) | % flagged | Direction of the flags |
|---|---|---|---|
| Lipids (cholesterol, LDL, HDL, triglycerides) | 428 | 21% | 86% too high |
| Liver enzymes (ALT, AST, GGT, ALP) | 390 | 14% | 100% too high |
| Glucose / HbA1c | 200 | 15% | 96% too high |
| Testosterone (total + free) | 150 | 25% | Mixed — slightly more high than low |
| Gonadotropins (LH, FSH) | 103 | 27% | 64% suppressed (low) |
Liver enzymes, glucose and lipids were flagged high essentially every time they were flagged at all. The reproductive hormones did the opposite — suppressed. And testosterone broke both ways: in this cohort, an abnormal testosterone result was slightly more likely to be too high than too low. By body system, hormones (16.5% flagged) and lipids (17.2%) topped the list, while electrolytes (1.9%) and minerals (2.7%) almost never strayed — your body defends sodium and calcium far more tightly than it defends your cholesterol.
5. The fingerprint of testosterone therapy
Those patterns connect. Of the reports measuring both testosterone and LH, 28% (9 of 32) showed suppressed LH alongside normal-or-high testosterone — the classic signature of exogenous (supplemented) testosterone, where added testosterone switches off the body's own production signal. Combined with the haematocrit finding, it paints a clear picture of who tracks their bloods most closely: people managing testosterone, watching the downstream effects their GP visit alone might miss.
6. Lean on the outside, flagged on the inside
The triglyceride-to-HDL ratio is a useful proxy for insulin resistance. The median in our data was a healthy 0.59 — but 18% (n = 40) sat above the 1.5 warning line: early metabolic risk hiding in people who often look fit. Among reports testing them, roughly a third had elevated glucose or liver enzymes. Looking well and being metabolically well are not the same thing.
7. Same blood, different number — depending on your country
One of the most common sources of confusion isn't a result at all — it's units. Australia and most of Europe report in SI units; the United States uses conventional units. The same sample can read very differently:
- Testosterone: 21 nmol/L (AU / UK / EU) ≈ 600 ng/dL (US) — multiply nmol/L by 28.8
- Cholesterol: 5.0 mmol/L (AU / EU) ≈ 193 mg/dL (US) — multiply mmol/L by 38.7
- Glucose: 5.5 mmol/L (AU / EU) ≈ 99 mg/dL (US) — multiply mmol/L by 18
- Estradiol: 100 pmol/L (AU / EU) ≈ 27 pg/mL (US) — divide pmol/L by 3.67
If you are comparing your results to advice written for another country, convert first — otherwise a normal number can look alarming, or an abnormal one reassuring.
What this does — and doesn't — mean
BloodTrack users upload their own lab reports, which we read and structure automatically. For this report we aggregated every result, stripped all identifying information, and only published figures backed by a meaningful sample (most marker stats draw on 30–430 results; correlations on 20–57 paired results, which we report as “in our data,” not as proven population effects).
Our users are a small, self-selected, engaged group — skewing toward men optimising hormones and metabolic health, tested overwhelmingly through Australian pathology labs. The numbers describe the people who track their blood work on BloodTrack, not Australians or the public at large. None of this is medical advice; always interpret your results with your doctor.
One last pattern worth noting: 36% of our users uploaded more than one report (the most dedicated tracked 38). The people who look once rarely learn much. The people who look repeatedly are the ones who catch the trends above before they become problems. If you have a stack of old pathology PDFs, that is where your own story is hiding — start with a free check.