Reference Ranges Don't Tell Your Story

When you get a blood test back, you typically see two things: your value and a reference range. It might say "TSH: 2.0 mIU/L (reference: 0.4–4.0 mIU/L)". Your value sits within the range, the result looks unremarkable, and nothing more is said about it. But that single data point, evaluated in isolation, can mask a change that is anything but unremarkable.

I know this from personal experience.

A Number That Meant Nothing, Until It Meant Everything

For most of my adult life, my TSH, the hormone that regulates thyroid function, consistently measured around 1.1 mIU/L. That was my baseline, year after year. Then, gradually, it started to creep upward. Over four years it moved from 1.1 to 2.0, and at each checkup the result came back with no comment. It was within range, so it was fine.

Then it spiked to above 20, marking the debut of hypothyroidism.

Looking at any single measurement in isolation, none of the intermediate values would have raised a flag. A TSH of 1.4 is normal. A TSH of 1.7 is normal. A TSH of 2.0 is normal. But when you look at the full picture, my TSH had increased by roughly 90% from my personal baseline over four years, drifting steadily in one direction before eventually crossing into clinical territory. That drift was a signal. It just wasn't visible to anyone who was only ever looking at one data point at a time.

The window for cheap, early intervention had been open for years. Nobody noticed it closing.

How Reference Ranges Are Actually Calculated

To understand why this happens so often, it helps to understand how reference ranges are constructed in the first place.

The process is straightforward. Researchers measure a biomarker in a large group of healthy individuals, sort the results, and remove the lowest 2.5 percent and the highest 2.5 percent. The middle 95 percent defines the reference range. By definition, this means that 5 percent of entirely healthy people will fall outside the range at any given time, and conversely, people with meaningful changes underway can remain well within it for years.

Reference ranges are a statistical description of a population, not a prescription for your personal health. They are designed to catch significant outliers, which they do reasonably well, but they are not designed to detect gradual, directional change in an individual over time.

The Difference Between Population Normal and Your Normal

This distinction matters more than most people realize. Reference ranges answer the question: is this value unusual compared to most other people? They do not answer the question: is this value unusual for you?

Those are fundamentally different questions, and for many biomarkers the second one is far more useful.

Your body operates within its own set of ranges, narrower than the population reference and specific to your physiology, your age, your habits and your history. A value that sits comfortably within the population reference range can still represent a significant departure from your personal baseline, and it is precisely those departures that tend to be meaningful early on, before they become impossible to ignore.

There are also structural limitations to how reference ranges are built. The populations used to derive them are not always representative across age, sex or ethnicity. Many reference ranges do not distinguish between a 28-year-old and a 58-year-old despite the fact that the body changes substantially over those decades. Biomarkers can also fluctuate based on time of day, recent exercise, sleep quality or whether you have eaten, none of which are consistently controlled for in the real world.

These are well-known methodological limitations within laboratory medicine, and they do not make reference ranges useless. They make them insufficient as the sole lens for evaluating your health.

What You Miss When You Only Look at One Data Point

The practical consequence of relying only on reference ranges is that you end up evaluating each blood test as if it exists in a vacuum, with no memory of what came before and no way to see direction or velocity.

A ferritin of 22 µg/L looks fine on paper. But if your ferritin has been 95, then 68, then 41, then 22 over the past three years, that trajectory tells a completely different story than a stable value of 22. The same applies to lipids, fasting glucose, inflammatory markers, thyroid hormones and dozens of other biomarkers that shift gradually before they shift dramatically.

The people most likely to catch these signals early are the ones who have access to their own history and look at it as a series rather than as isolated snapshots. Most people never get that view, not because the data does not exist, but because nobody has assembled it for them and no one has told them to look for it.

The Cost of Catching Things Late

Early intervention is almost always cheaper than late intervention, in every sense of the word. Catching a thyroid trend before it becomes overt hypothyroidism means more options, lower treatment burden and better outcomes. The same logic applies across most chronic conditions where blood biomarkers give advance warning.

One challenge many healthcare systems have is that they often first operate in snapshots. You come in, a value is checked against a range, and if nothing is flagged you leave without a broader picture of where you are heading. This model works well for acute problems but it works less well for the slow, directional changes that precede most chronic disease.

The good news, though, is that the information needed to catch drifting baselines, and changes earlier is often already being generated. It just is not being looked at in a way that makes the pattern visible. This is why I created mittblod.se, to empower people to better understand their health.