Micronutrient testing: what whole blood shows and the serum value can miss

You are sitting in the consulting room. On the printout there is a column of numbers and next to it a column of reference ranges. Not a single value is flagged. The doctor says the sentence you already know: “Everything is in the green.”

And you walk out and you are still tired.

I know this situation from both sides of the desk. As a physician with a number in front of me that looks unremarkable. And as a person who knows that the body is not much impressed by a reference range.

This text is about what a micronutrient analysis can do and what it cannot. It is not a plea for more laboratory work. It is a plea for more precise laboratory work: fewer numbers, but the right ones.

“Normal” is a statistical term, not a state of your cells

Let us start with what creates this misunderstanding.

A reference range does not come about because somebody worked out which value is good for you. It usually comes about like this: a laboratory measures a parameter in a larger group of people who count as healthy, and defines the range into which the middle 95 percent fall. That is a sensible statistical convention. It is not a statement about an optimum.

Two things follow from this, and both are uncomfortable.

First: five percent of healthy people lie outside the range by definition. If you have twenty parameters determined, you should expect on purely arithmetic grounds that about one of them looks abnormal without anything being meaningful. That is why I think little of indiscriminate large panels. They produce chance findings that create fear and pull follow-up diagnostics behind them.

Second: the reference group lives in the same world as you. If a nutrient is broadly scarce in a population, the reference range migrates downwards with it. The range then describes what is usual, not what would be favourable. This distinction between average and optimum is the core of this whole topic.

Now comes the second level, and that one matters even more. Your body actively holds certain blood values constant. Not so that the number looks nice, but because heart rhythm, nerve excitability and clotting depend on it.

We call this homeostasis. And it is the reason why a blood value is sometimes the last thing to change.

The same logic applies to other nutrients, only with different numbers. Potassium sits almost entirely inside the cell. Zinc is steered through a narrow control loop. By current understanding, selenium is transported preferentially to the organs that need it most urgently.

And from that follows the point that stands above this whole guide: a nutrient is not a switch. It is a player in a network. Whoever turns one screw moves others along with it.

And now you know why “everything is normal” and “I do not feel well” do not have to be a contradiction.

Serum or whole blood: why one blood draw can give two answers

Picture a river running through a city. You want to know how much water is available to the city. So you measure the level in the river.

That is serum.

What you do not see with it: the tanks in the cellars, the tanks on the roofs, the water in the pipes of the houses. That is exactly where everyday life happens. On top of that, the river level is actively regulated so that it does not swing about.

Whole blood is the attempt to count the water in the houses as well. Besides plasma it also contains the cells, above all the red blood cells. For minerals that sit mostly inside the cell, this can be a sensible addition.

Now comes the limitation that is rarely supplied in the advertising for such analyses and that I consider the most important part.

Whole blood is not a gold standard. The reference ranges differ between laboratories. The haematocrit, meaning the share of cells in the blood, influences the result. And for many parameters there are no large studies showing that the whole blood value allows better clinical predictions than the serum value. What we can say with confidence: it is a different angle. What we cannot seriously say: it is always the better one.

And now you know why the question “whole blood or serum” can so rarely be answered with a single word.

Functional markers: measuring the work rather than the store

Here comes the idea that, for me, brought the biggest change in nutrient diagnostics.

Picture a kitchen. You want to know whether there is enough flour. You can look into the pantry and count the sacks. That is the concentration measurement.

Or you look at what comes out of the kitchen. If half-finished doughs are suddenly standing around everywhere, then something is missing somewhere in the process. That is a functional marker.

This is exactly how methylmalonic acid and homocysteine work. Both are intermediates in metabolism that can pile up when an enzyme is not running smoothly because a cofactor is missing.

There is a second piece of work that lights up the point from another direction, and I am showing it to you because it makes the limits visible.

And now you know why in some people I put more weight on methylmalonic acid than on B12 in serum.

Vitamin B12: the value that raises the alarm late

If one topic shows that “measuring instead of guessing” is more than an advertising slogan, it is this one.

Total B12 in serum measures two things at once. It measures the smaller part bound to transcobalamin, which can actually reach the cells. And it measures the larger part that hangs on haptocorrin and is barely available to the cell.

Picture a delivery fleet. One part of the vehicles drives to the houses and delivers. Another part stands in the yard and gets counted. If you count both together, the fleet looks bigger than the delivery performance.

Who is particularly affected? People who eat vegan, because B12 in relevant amounts comes from animal sources. People taking proton pump inhibitors, because stomach acid is needed to release it from food. People on metformin. People after stomach or bowel surgery. And people at an older age, in whom stomach acid production often declines.

This list is not a reason for alarm. It is a reason to look in a targeted way instead of supplementing across the board.

Now comes a point that matters to me, because it shows homeostasis in its purest form.

And one more detail that is often overlooked in practice: if supplementation with B12 has already begun, measuring total B12 is barely interpretable any more. The value is then high, regardless of how the supply looked beforehand. Functional markers are the more helpful choice in this situation, and they too need some distance in time from the last intake.

And now you know why it is worth measuring before the first preparation and not after it.

Ferritin without an inflammation marker gives only half an answer

Hardly any value is misread as often as this one. And hardly any has so much to do with quality of life.

Ferritin is the storage protein for iron. A low value speaks for empty stores, that is undisputed. And here comes the sentence I leave out of no consultation: a newly found iron deficiency is not a finding you simply top up. It is a question. Namely: where is the iron being lost?

In men and in women after the menopause, answering it usually includes a medical work-up of the gastrointestinal tract, because a silent bleed can sit behind it, including one arising from a polyp or a tumour. In women before the menopause, menstrual bleeding is the most common reason, but that too deserves to be looked at once. Coeliac disease or impaired absorption in the gut can also be behind it. Taking iron without having asked this question can make the value look good and cover up the cause.

The problem also begins at the other end, though.

Ferritin is at the same time an acute phase protein. With inflammation it rises, regardless of how much iron is actually available. A cold last week, a silent inflammation in the abdomen, excess weight, an autoimmune disease: all of that can lift the value.

The result is treacherous. The ferritin value looks decent, and the iron supply to blood formation is not. There is a term of its own for this state, functional iron deficiency.

And then there is the question of the cut-off itself. It is more uncomfortable than it sounds.

What follows from this in practice? A ferritin value without an accompanying inflammation marker is hard to interpret. Alongside it, a blood count, transferrin saturation and, depending on the question, markers of the red cell series such as the haemoglobin content of reticulocytes or zinc protoporphyrin make sense.

And now you know why with exhaustion I almost never order a ferritin on its own.

Omega-3 index and 25-OH vitamin D: two values that are well defined

After so much caution it is time for the other side. There are parameters where the measurement is pleasingly clear. I use two of them regularly.

The omega-3 index

Here it is not the plasma that is measured but the membrane of the red blood cells. The reason is elegant: an erythrocyte membrane is built up over weeks. So it does not tell you what you ate yesterday, it tells you how your supply looked on average.

The index states what share EPA and DHA make up of all the fatty acids in the membrane. And unlike with many nutrient markers, here there are data linking the value to a hard endpoint.

A smaller study from Erlangen fits the picture. In 71 people with symptoms who had a computed tomography of the coronary arteries, the omega-3 index was lower in those with strongly pronounced calcification than in the rest. The authors write themselves that this finding has to be confirmed in larger groups. I am showing it to you as a piece of the puzzle, not as proof (Bittner 2020).

25-OH vitamin D

With vitamin D people often ask why the active form is not measured. The answer has to do with homeostasis again: the active form is tightly regulated through parathyroid hormone and can even look high normal in a deficiency. The storage form 25-OH vitamin D has a clearly longer half-life and reflects the supply better.

But modesty applies here too, for a reason that hardly anyone has on their radar.

With selenium there is a similar subtlety that I find interesting. The concentration in serum is a pointer. As a more functional marker, the transport protein selenoprotein P is discussed, which can mirror the supply to the organs and can run into a plateau once intake is sufficient. A review from Berlin describes this in detail (Schomburg 2022). For broad practice, however, this marker is not yet standard.

And now you know why I follow some values over time and others only once for orientation.

What an analysis cannot do: hair mineral analysis and IgG tests

Now the part where I risk disappointing you. I am writing it anyway, because you are offered both of these on every corner of the internet.

And first of all: if you have had one of these tests done, that is not a reproach. They are marketed professionally, they look scientific, and the longing for an answer is entirely understandable.

The hair mineral analysis

The idea sounds good. Hair grows over months and should therefore contain a kind of long term record. For individual substances in strictly standardised toxicological procedures, that is indeed established.

For commercial mineral profiles the data look different.

This study is old, that is true. To this day I know of no comparably designed work showing that reproducibility between the laboratories has clearly improved. As long as that remains so, I hold to this appraisal, and I will gladly change it if new data suggest it.

IgG and IgG4 tests on foods

Here the matter is immunologically even clearer, and in a certain way it is comforting.

And now you know why I prefer to test less and to take seriously what has been tested.

Pre-analytics: time of day, haemolysis and the half story in the tube

Now comes the least spectacular section of this text. And perhaps the one with the most practical effect.

Pre-analytics is everything that happens before the actual measurement: the time of day, whether you are fasting, how long the tourniquet was on your arm, how the tube was transported, how quickly it was centrifuged. And this part decides the result more often than most people suspect.

One example every practice knows: haemolysis. When red blood cells burst during the draw or during transport, they release their content into the serum. And because potassium sits mostly inside the cell, the measured value then rises without anything having changed in the body. For other intracellular parameters the same effect can play a role.

With homocysteine there is a related problem that is even more subtle. Red blood cells keep releasing homocysteine after the blood draw. If the sample stands too long before plasma and cells are separated, the value rises. An international expert panel has published a detailed guidance document on this (Refsum 2004), covering exactly such practical questions: sample collection, handling, biological influences, reference ranges and the variation within one and the same person.

With zinc it is the time of day and the last meal. An international expert panel describes in its report that the plasma concentration is influenced by meals, time of day, inflammation and certain medicines and hormones. The same panel also records that the data for hair, urine, nail and blood cell values are so far not sufficient to recommend them for judging zinc status.

One last example that convinced me particularly, because it recreates the serum versus whole blood question directly in the laboratory.

And now you know why with follow-up checks I insist that they happen under conditions as similar as possible.

The four lenses: what a laboratory value tells you about your system

In clinical psychoneuroimmunology we look at a symptom through four lenses. With nutrient diagnostics that is not a philosophical detour but a very practical sorting aid. It stops us from looking at a value in isolation.

And because exhaustion is the complaint that brings most people to the laboratory, a clear order belongs here. Before I think about micronutrients, other things belong checked: an underactive thyroid, an anaemia together with the question of its source, disturbed sleep including pauses in breathing at night, a diabetes, a depression, a heart or kidney disease, coeliac disease, a recent infection. This is not a box-ticking exercise. Whoever skips this level and orders the larger nutrient panel straight away can lose months.

These four lenses explain why I ask the same question before every analysis: what exactly do I want to know, and what changes if the answer comes out one way or the other?

Covering requirements and therapeutic high dose are two different worlds

At this point a distinction belongs that almost always blurs in public discussion.

Food supplementation in the narrower sense fills gaps. Those are the doses you buy in the drugstore or online. They orient themselves on reference values from bodies such as the German Nutrition Society or authorities such as EFSA. No therapeutic effect is to be expected from them, and that is not their job either.

Orthomolecular medicine means something else: clearly higher doses, used in a targeted way, limited in time, under medical supervision and with laboratory monitoring. Only in this range can a nutrient behave like a medicine, with everything that comes with it, side effects included.

Why real food comes first

I say this in every conversation, and I mean it exactly like this: from real food of good origin, nutrient supply is the first choice. Not out of nostalgia, but because a food delivers nutrients in a matrix of cofactors that a capsule does not rebuild.

No supplement replaces nutrition, sleep, movement, sunlight and relationship. If someone comes to me with fifteen preparations and four hours of sleep, the analysis is rarely the most urgent building site.

And why topping up is nevertheless more often justified today

At the same time it would be dishonest to write off supplements across the board. There are reasons why targeted topping up can make sense more often today than fifty years ago.

First, the foods themselves. There are comparisons across decades that suggest lower contents of individual nutrients in fruit and vegetables. These comparisons have clear weaknesses: old and new analytical methods are hard to set against each other, and part of the effect may be explained by higher yielding varieties in which the same mineral is spread over more mass. I regard this as a signal that can be taken seriously. As proof it does not hold, and from it alone I would not derive any supplementation.

Second, the composition of what lands on the plate. Heavily processed foods as a rule deliver plenty of energy and little micronutrient density. Dietary surveys in Western countries describe a high share of energy intake for them. That too is a way of sorting the picture and not evidence that a deficiency follows from it in you.

Third, consumption. Environmental exposures such as heavy metals, mould toxins, plasticisers and pesticide residues, along with chronic stress, can raise the need for protective substances such as zinc, selenium, magnesium and glutathione precursors. This is a consideration I find clinically plausible. Solid studies that put a number on this extra need in humans are, for most of these factors, not known to me.

Fourth, the medicines. For some regularly taken substances an influence on the nutrient household is described. It is best studied for proton pump inhibitors, that is acid blockers with substances such as omeprazole or pantoprazole, where an influence on the uptake of vitamin B12 and magnesium is described, and for metformin, where an influence on B12 uptake is described.

For diuretics, the water tablets, an influence on potassium and magnesium is discussed. For statins a drop in the coenzyme Q10 level in blood is described, while it remains open whether a benefit of supplementation follows from it. For hormonal contraception the data on B6, B12, folate and zinc are inconsistent.

These substances are mostly prescription-only, only single acid blockers are available in low strengths without a prescription at the pharmacy. And now comes the most important sentence of this paragraph: do not stop any of these medicines on your own and do not change a dose on your own initiative. An acid blocker stopped abruptly can lead to a fierce return of the symptoms. A diabetes medicine that is stopped can send blood sugar off the rails. A statin or a diuretic that is stopped can raise again exactly the risk it was prescribed against.

So these are not arguments against these medicines. They are arguments for taking a look together with the doctor who prescribed them.

What this costs in practice in Germany

Money deserves plain speech, but not a price list. Statutory health insurance covers laboratory values when there is a concrete medical indication, meaning a justified suspicion or a follow-up check. A broad screening without symptoms usually does not fall under this and is billed as a self-pay service.

Within the self-pay world a simple rule applies: the more elaborate the procedure, the more expensive. A serum value is cheap. Intracellular measurements, methylmalonic acid, holo-transcobalamin or a fatty acid profile in erythrocytes lie clearly above that. A broad package adds up quickly.

That is why I do not think the decisive question is “what does it cost?” but “which decision depends on it?”. A value that changes nothing is too expensive even for very little money.

Three levers you can put into practice today

No panel, no shopping list, no brands. Just three things that lie in your hands.

And now you know why I prefer to determine three values in a targeted way rather than thirty on suspicion.

Frequent questions about micronutrient testing

How this topic connects with the others

Diagnostics is never an end in itself. These paths lead onwards, depending on which question brought you here.