What Conventional Lab Ranges Miss

A conventional reference range is a statistical description of the people who got tested, not a statement about health. Knowing how it is built changes what you do with a result that comes back "normal" in a patient who clearly is not.

Where the range comes from

A reference interval is the middle 95% of results from a reference population. That is the whole definition, and everything that bothers us about it follows from those two words.

Middle 95% means the top and bottom 2.5% are cut off. It is a description of spread, not a judgement about health. By construction, one in twenty perfectly well people falls outside it — and, far more consequentially, it makes no claim at all about whether the 95% inside are doing well.

Reference population is where it gets uncomfortable. In practice that population is often assembled from people who had the test run at that laboratory. People get blood drawn because something prompted it. So the comparison group skews toward the unwell, the medicated and the metabolically struggling — not toward a cohort of thriving adults.

Put those together and "normal" means statistically unremarkable among people who got tested here. That is genuinely useful information. It is not the same question as is this value consistent with this patient feeling well, and the second question is the one your patient came in with.

The drift matters too. As the underlying population becomes more insulin resistant and more subclinically hypothyroid, the range built from that population moves with it. The bar does not hold still while the population changes; it follows.

The B12 that looks fine

Serum B12 is the clearest illustration, because the marker that is reassuring and the marker that is accurate are two different tests.

Serum B12 measures what is circulating. It does not measure what has made it into cells and is doing the job. A patient can sit comfortably inside the serum reference range and be functionally deficient at the tissue level, which is why the fatigue, the paraesthesia and the cognitive fog do not resolve when you tell them their B12 is normal.

Methylmalonic acid is the test that answers the real question. B12 is the cofactor that converts methylmalonyl-CoA to succinyl-CoA. Take the cofactor away at the cellular level and the substrate backs up, so MMA rises. It rises because the cell is short, regardless of what the serum says.

	Conventional	Functional
Methylmalonic acid	< 318 nmol/L	< 260 nmol/L

That gap between 260 and 318 is the entire point of this article. It is not a rounding error or a matter of taste. It is a stretch of results where the report says normal, the patient says otherwise, and the biochemistry is on the patient's side.

TPO antibodies sitting under 34

The second example is sharper, because here the conventional threshold is answering a question nobody asked.

	Conventional	Functional
TPO antibodies	< 34 IU/mL	< 10 IU/mL

The conventional cutoff is built to answer does this patient have autoimmune thyroid disease, diagnosably, today. That is a reasonable question and the threshold answers it reasonably well.

It is not the question in front of you. A patient at 28 IU/mL does not have a diagnosis. They do have antibodies — measurable, elevated above what a well thyroid produces, and sitting in a range where the autoimmune process has started and the tissue damage has not yet accumulated enough to show up on a TSH.

That patient is the one you can actually help. By the time they cross 34 and earn the diagnosis, you are managing established disease rather than intercepting it. The conventional range was never wrong; it was answering a diagnostic question in a setting where your job is a clinical one.

The same marker, a different range

Move a patient from one laboratory to another and their reference range changes without anything happening to the patient.

Different analysers, different assay methods and different reference populations all produce different intervals for the same marker. This is ordinary and expected, and it has two consequences worth holding onto.

The first is practical: when you track a patient over years, a result that appears to have moved may only have changed laboratory. Before you interpret a trend, check that the numbers came from the same assay.

The second is conceptual, and it is the more important one. If the range were a property of human physiology, it would not vary by vendor. It varies because it is a property of a measurement and a population — which is exactly what this article has been saying. Nobody has to be wrong for two laboratories to disagree.

The part that gets overclaimed

Functional ranges are not all built on the same quality of evidence, and pretending otherwise is how this field loses credibility with the people it most needs to convince.

Some are on firm ground. MMA as a marker of tissue B12 status has solid biochemistry behind it. Homocysteine, HbA1c and fasting insulin thresholds are supported by outcome data. Others are narrower and rest more on clinical consensus and accumulated experience than on trials, and you will find different numbers in different textbooks and from different teachers.

So the honest position is this: a functional range is a tool for deciding where to look, not a diagnostic standard. A result outside one is a reason to investigate, correlate with symptoms and often to retest. It is not a diagnosis, and it should never be handed to a patient as one.

And the conventional range is not the enemy. It is built to detect frank pathology and flag what needs urgent attention, and it does that job well. A practitioner who dismisses it entirely will eventually miss something that mattered.

Using both numbers

The two ranges answer different questions, so you do not have to choose between them.

Read the conventional range first and ask whether anything here needs urgent attention or a referral. That is what it is built for, and it is the question with the shortest fuse.

Then read the functional range and ask where to intervene. That is where the work is for most of the patients sitting in front of you — the ones who are not diagnosable and are not well.

The third number is better than either, and it is the one most people skip: the patient compared with themselves. A ferritin that has fallen from 90 to 45 over two years is a more useful finding than either range can give you, and both readings are "normal." Trends surface problems while they are still cheap to fix.

And none of it means anything without the symptom picture. A number outside a functional range in someone who feels well is a note to watch. The same number in someone exhausted is a place to start.

The GHT Academy Blood Chemistry Guide carries this out across 85 commonly ordered markers — conventional and functional ranges side by side, with what drives each one high or low. CBC, CMP, electrolytes, liver and lipid panels, the full thyroid panel, cortisol, DHEA and sex hormones. It is the reference I built for my own team.