Reading your own blood work

The range on your lab report was not built for you. It was built from everybody else.

Two labs can print the same testosterone result and disagree about what it means. The reason is in how that band gets drawn.

Updated August 14, 2026 · 9 minute read

The results come back and the word at the end of the line is normal. Nothing is flagged. The visit lasts four minutes, the paper goes in a drawer, and that is the end of it. For most men that is the right outcome. What almost nobody gets told is what that word was actually measuring.

It was not measuring you. It was measuring the distance between your number and the middle of a reference group that the laboratory chose, on criteria that have nothing to do with you. That is not a criticism of the laboratory. It is the definition of the thing, and it is written down in the standards the laboratory follows.

Below is how that group gets picked, why two respected laboratories can draw the line in two different places, what a second number does that the first one cannot, and what any of it is worth to a man who feels fine on paper and flat in real life.

A reference range is a percentile, not a verdict

Start with the definition, because the definition is doing all the work here. From a 2024 paper in the Journal of Clinical Medicine on how laboratories verify their own limits:

By definition, reference intervals include the central 95% of results measured in non-diseased reference individuals. Journal of Clinical Medicine, 2024 · full text

The same paper describes how the two numbers get produced: laboratory results are collected from apparently healthy individuals, and the 2.5th and 97.5th percentiles of that distribution become the lower and upper limits. Everything between them is printed as normal.

Read that once more with a calculator. If the band is the middle 95 percent of a reference group, then one person in forty from that same healthy group falls below the lower limit, and one in forty falls above the upper one, purely by construction. Nothing has gone wrong. The band was built to leave them out.

This is why a laboratory report can only ever answer one question: is this result unusual for the group we compared it with. It is a good question. It is not the same question as whether the number is a good one for the man holding the paper.

Change the group, change the number

If the limits are percentiles of a group, then the choice of group is the whole ball game. Two respectable answers to the same question, both published, both current:

Quest Diagnostics, clinical education center, testosterone FAQ. Travison et al., Harmonized Reference Ranges for Circulating Testosterone Levels in Men of Four Cohort Studies in the United States and Europe, cohorts: Framingham Heart Study, European Male Aging Study, Osteoporotic Fractures in Men Study, Male Sibling Study of Osteoporosis.

Quest does not hide the problem. On the same page, in the answer about where the cutoff for testosterone deficiency should sit, the laboratory writes it plainly:

The population defining the TT reference range, and thus the T-deficiency cutoff, has been debated. Quest Diagnostics, clinical education center

So take a man of 38 with a result of 300 ng/dL. It is inside the Quest band, because that band was drawn from men up to the age of ninety. It is inside the harmonized band, because that band starts at 264. And it sits just below the 5th percentile of healthy nonobese men his own age, whose middle value is 531. All three statements are arithmetic. Only the first one is printed on the sheet, and only the first one gets read out loud in the appointment.

The tube is the same. The method is not.

There is a second source of movement in that number, and it has nothing to do with statistics. It is the assay. The Endocrine Society recommends measuring total testosterone by liquid chromatography with tandem mass spectrometry. Most routine testing has historically used an immunoassay instead, and Quest describes the difference in one sentence:

Immunoassay produces reliable results for normal or elevated levels of total testosterone but lacks sensitivity, specificity, and precision at lower concentrations. Quest Diagnostics, clinical education center · condensed from the published answer

Read the shape of that sentence. The cheaper method is at its weakest exactly where the decision gets made. Quest states that in men with low levels, total testosterone should be measured only by mass spectrometry, and that its own mass spectrometry assay is certified by the CDC Hormone Standardization Program. Two men can hand over identical blood and walk out with different numbers, and neither of them will ever know which method was run unless somebody looks.

One number is really three

Total testosterone is a sum. Quest breaks the circulating total into three parts:

SHBG is the variable in that picture. When it moves, the split moves with it, and the total on the sheet can stay exactly where it was. Quest is explicit about when the second measurement earns its place: free testosterone is helpful when the total sits near the decision level, or when a shift in SHBG is likely. It also notes that free and bioavailable testosterone cannot be measured directly at all in routine practice, so the figure is calculated.

None of this is exotic. It is the standard content of a laboratory education page. It just does not fit in a four minute appointment, and it is not printed next to the word normal.

What Marek Health does with the same sheet

Marek Health is a telehealth platform built around this exact gap. Not around the drug, around the reading. It is worth quoting their own answer on hormone therapy eligibility, because it says the quiet part in public:

The ranges we have for hormone health changes between labs but they're also largely arbitrary and don't paint the whole picture of what's going on. Marek Health, frequently asked questions

On their testosterone page, one of the four things they position themselves against is not a competitor. It is a document: oversimplified lab value ranges. The other three are disinterested doctors, outdated treatment protocols and stubborn insurance companies.

What their panels look at

Their published panel categories, in their words:

Panel categories as published on marekhealth.com/get-started, read August 14, 2026.

The four steps, as they describe them

Marek Health is careful about what it is, and the distinction is on their own FAQ page: they are not a clinic. They describe themselves as a platform that connects clients with specialized healthcare providers, and the providers do the prescribing. That is a narrower promise than most of the category makes, and it is the honest one.

See how Marek Health reads a panel Guided Optimization on marekhealth.com · United States residents only

This page is an independent explainer. The link goes to Marek Health.

Ten weeks, in their clients' words

The photographs below are published by Marek Health on its home page under the heading real client results in just 10 weeks. The reviews are reproduced from their own site with the names and dates they display, and every cut inside a quote is marked.

Reviews reproduced from marekhealth.com as published on August 14, 2026. Individual experiences are individual. Marek Health displays a 5.0 Trustpilot score from 820 or more reviewers on its own get-started page.

What it costs, and who it is not for

Worth knowing before the call, because the model is unusual and the numbers are published:

That is a real cost, and it buys a specific thing: somebody whose job is to read the whole sheet with you rather than the flagged lines. Whether that is worth it depends entirely on whether you have ever left an appointment with the word normal and the feeling that it did not answer the question you came in with.

• A reference range is the middle 95 percent of a group. One healthy person in forty sits below it before anyone is examined.
• The group is a choice. Men up to 90 in one published range, healthy nonobese men aged 19 to 39 in another.
• The method moves the number too. Immunoassay loses precision at exactly the concentrations where the decision is made.
• The total hides three fractions, and the free one is 1 to 4 percent of it.
• None of that is on the sheet. It takes a person and 25 minutes.

Start with the intake call Marek Health Guided Optimization · $299 assessment · direct pay, no insurance billing