The Exact Meaning of 'Your Tests Are Normal' — How Reference Ranges Are Made

"But all my tests are normal."

Many people come to my consultation room after hearing this. And after it, almost the same sentence follows.

"So am I just being sensitive?"

To answer this question, today I want to tell a slightly different story. Not a story about disease, but a story about tests. More precisely, a story about what the word 'normal' means when it's used on a test sheet.

Once you know this, you'll see that "normal, yet in pain" is not a contradiction.

The reference range is not a 'condition of health' but a 'distribution of people'

Let me start with the most important fact.

The normal range (reference range) printed on a test sheet was not created as a number meaning that being inside it makes you healthy.

The way it's made is this. You gather many people judged to be healthy. You measure all their test values and lay them out in a row. And you cut out the interval that holds the middle 95%. You discard 2.5% each at the top and bottom.

That cut-out interval is the normal range printed on the test sheet.

This method carries one unavoidable consequence.

By definition, 1 in 20 completely healthy people falls outside the reference for that item.

The discarded top and bottom 2.5% are those people. Not because they have a disease. Because we decided to cut it that way in the first place.

So the more tests you do, the more 'abnormalities' appear

Let me take one more step here.

In a health screening, it's never the case that only one test item comes out. Usually dozens come out at once.

If the probability of falling within normal on one item is roughly 95%, then the probability that ten items are all within normal is a bit over 60%. Put the other way around —

If a healthy person with no disease takes ten tests, the probability that one or more comes out beyond the reference is around 40%.

You've probably seen a red arrow or two on a health screening result. A good portion of those arrows are not signals of disease, but shadows produced by the way the numbers are cut.

This is not my interpretation. It is a textbook fact of laboratory medicine. And it's only half of this story.

The truly important thing is the opposite direction

So far I've told you about "cases that are outside normal but not disease."

But what troubles me far more in the consultation room is the opposite side.

"Cases that are inside normal, but not normal for that person."

To explain this, one concept is needed.

You have your own usual value

The body's numbers have their own set-point for each person. And they move within a narrow band around that set-point.

Here two spreads must be distinguished.

	What it is
Within-individual spread	The range within which one person's value rises and falls in daily life
Between-people spread	The spread arising because each person has a different set-point

The reference range is made by embracing all of the second spread too. Since it must hold the differing set-points of many people, the range inevitably widens.

But some items have a very narrow within-individual spread and a very wide between-people spread. In such items, a strange thing happens.

My value changes to double its usual, and yet it still registers as 'within the normal range.'

Let me take a kidney number as an example

The clearest example is creatinine, which reflects kidney function.

Creatinine is a waste product coming from muscle. So people with a lot of muscle have a high usual value, and people with little muscle have a low usual value. Someone small and lean often stays around 0.6.

Now suppose this person's kidney function drops by half. Creatinine roughly doubles. From 0.6 it becomes 1.2.

But the upper end of the normal range on a test sheet is usually around 1.2–1.3.

Kidney function has halved, yet the test sheet registers 'normal.'

This is not because the test is bogus. It's because we're measuring an individual by a group's standard. Because medicine knows this limitation, kidney function is not judged by the single creatinine number but is calculated together with a value corrected by inputting age and sex (the glomerular filtration rate). It is a countermeasure made by people who know the limitation.

Thyroid numbers have a similar nature. The set-point clearly differs from person to person, and within that person it moves narrowly. Even while moving inside the group reference, it can be a large change for that person.

A test is 'one point at one moment'

Here another thing overlaps.

The body doesn't stay still through the day.

• The stress hormone is highest in the morning and lowest in the dead of night. That difference reaches several fold.
• Blood pressure differs each time you measure it. Posture, whether you just walked, even the mere fact of being in a clinic changes it. (If your blood pressure differs each time you measure)
• Meals, sleep, posture, and the day's tension shake many numbers.

A test is the act of marking a single point on that swaying curve. Depending on where that point happens to land, 'normal' and 'borderline' are divided.

So I try not to judge from a single number. Rather than a single point, I try to see the direction the points draw.

And there are things tests don't see well

This is the core of what I want to say today.

Most tests look at substance and structure. How much of what is in the blood. How the tissue is shaped. At seeing these, modern testing is enormously powerful.

But the body has, besides substance and structure, a layer called regulation.

• How quickly it responds
• Whether it rises and falls in an appropriate band
• After being shaken, how long it takes to return to place

An echocardiogram looks at the structure of the heart. The heart's ability to regulate its beat to fit the situation is seen by another test. A single blood-pressure reading looks at the pressure at that moment. In what band the blood pressure swings over a day is seen only by 24-hour measurement.

Even when the structure is fine, regulation can be disturbed. And in a state of disturbed regulation, a person definitely struggles.

When a test says "normal," that word usually means this.

"No trace of the dangerous diseases we look for was found.
And your value is inside the middle 95% of the group."This is enormously important information. But this word does not mean"your body is well regulated." It was not a test that asked that in the first place.

Let me show you how this plays out in one person

Since words can sound abstract, let me follow to the end one chain I actually meet in the consultation room.

Suppose someone has had diarrhea for a long time. It begins with a state in which the immunity of the gut mucosa is excessively excited.

What happens in the body	And the test then
① The gut mucosal immunity is over-excited and absorption drops	Colonoscopy — normal (no visible ulcer or tumor)
② Since absorption fails, energy metabolism is cut down	Blood sugar, liver values — normal (still within reference)
③ Since energy is lacking, peripheral blood flow is reduced (hands and feet turn cold)	Vascular ultrasound — normal (the vessels aren't blocked)
④ From enduring long, the adrenal stress hormone is depleted	Cortisol — lower end of the normal range (all the more if measured in the morning)
⑤ Since the hormone is cut down, the power to hold on to sodium and potassium wavers	Electrolytes — inner edge of the normal range

All five stages have progressed, and yet the tests are normal from start to finish.

Why is that? Because each stage hasn't broken down enough to be called a disease. Each one is 'within reference.' What broke down is not each item, but the chain formed by stringing the items together.

And the tests look at items. They don't look at the chain.

Telling this person "it's normal, so you're fine" is like, in a situation where each floor is fine but the building is tilting, inspecting the bricks one by one and saying "the bricks are all normal." The bricks really are normal. What is tilting is not the bricks.

This chain is what I try to see. Which axis wavered first, and in which direction it spread. If you misjudge the starting point here — for instance, seeing cold hands and feet and trying only to warm the blood flow — the upstream cause stays as it is, so it comes back. (Why long-standing diarrhea drains your strength too · The misunderstanding of saying "the body is cold")

That doesn't mean ignoring the tests

Let me make this clear here. I know this story can be used dangerously.

If this piece is read as saying "tests can't be trusted," I'd rather not write it.

Tests save people. Cancer, infection, anemia, thyroid disease, diabetes, diseases of the heart and kidney — these cannot be sorted out by symptoms alone, and only tests can sort them out. And missing these diseases kills people.

What I do is not to replace tests. It is to deal with the problem that remains in the place after tests have filtered out the dangerous diseases. The order must not be reversed.

So I ask you to bring your test results. If you don't have them, I recommend getting them. Because that is how I divide whether or not it's a problem I should be dealing with.

Then what do I look at?

Before a person whose tests are normal, I look at these things.

First, I read again by the standard of that person's usual. Rather than "within the normal range," I ask "how does it compare with your usual." This is why I ask you to bring old test sheets. When you place the value from five years ago beside the value now, even if both are within the normal range, the direction shows.

Second, I look at the pattern of time. When in the day is it worst? Does it follow the seasons? After doing what does it worsen, and after doing what does it ease? How long does it take to recover? Problems of regulation reveal themselves through pattern more than through numbers. (What time does in a long-standing illness)

Third, I look at whether the symptoms move together. If several symptoms that seem unrelated to each other worsen together and improve together, that may mean not separate diseases but that a single regulation is wavering. (When symptoms that seem unrelated come together)

Fourth, I set the point for testing again. Even if it's normal now, if the direction troubles me, I tell you when to look again. One 'normal' is not a permanent pardon.

When you should go to the hospital first

Below are signals that must not be seen as "problems of regulation." Even if the tests were normal, if the following are present, you should go back to the hospital.

• Unintended weight loss (a clear drop over a few months)
• Pain that wakes you at night, pain that doesn't ease with rest
• Fever that continues, night sweats that soak your clothes
• Blood-tinged phlegm, stool, or urine, black stool
• Suddenly arising neurological symptoms — weakness on one side, slurred speech, visual disturbance
• Symptoms that only keep worsening (going in one direction, not up and down)
• Being told the tests were normal, but it's been a long time since that test

The last one is especially important. "It was normal last year" is not today's diagnosis.

Speaking honestly

Among what I've said so far — that the reference range is made from the middle 95% of healthy people, so a certain proportion of healthy people fall outside it; that because set-points differ from person to person, a group standard can miss an individual's change — these are established facts of laboratory medicine. They are not a story I made up.

From there on is my interpretation.

That "in people who struggle even though their tests are normal, something is happening at the layer of regulation," and that "that regulation is formed by several axes entangled together" — this is a framework I've built in clinical practice, and it is not yet a proven fact. I believe this framework is useful, but believing and being proven are different. I think it's right to speak of them distinctly.

---

The words "your tests are normal" are good news. Because they mean nothing dangerous was found.

But those words are not a synonym for "you're fine." And above all, they are absolutely not a synonym for "you're being sensitive."

A test finds what it was made to find. About what it was not made to find, it has simply said nothing.

What troubles you is a matter outside the test sheet. And outside the test sheet, too, there is a body.