A result inside the reference range is often described as normal, and normal is often heard as ideal. Those are not the same claim.
For some markers the gap between them matters a great deal; for others it does not exist. This page explains how reference ranges are built, where the distinction applies, and why we do not publish a list of optimal numbers.
Reviewed by Dr. David Oliak, board-certified surgeon
Updated August 2026
A laboratory establishes a reference range by measuring a marker across a reference population and reporting the interval containing the middle 95% of results.
Two consequences follow directly, and neither is controversial.
The range describes that population, not an ideal. If the people measured include a substantial number who are unwell in ways the marker reflects, the range widens to accommodate them.
By construction, one in twenty healthy people falls outside it. A result marginally outside a reference range is common and frequently means nothing at all — which is the opposite error, and worth saying on a page like this.
Reference ranges also differ between laboratories, because each builds them from its own population and methods. A result "in range" at one lab can be marginally out of range at another.
For a large share of what gets measured, the reference range is exactly the right target and there is no interesting gap to discuss.
Tightly regulated markers. Sodium, potassium, calcium, and pH are held within narrow limits by active physiological control. There is no "optimal within the normal range" for sodium. Being in range is the goal, and being out of it matters.
Organ-safety markers. Creatinine, liver enzymes and similar tests are screening for damage. In range means no evidence of damage, which is the entire question being asked.
Most of a comprehensive panel is like this. That is worth saying, because the impression left by a great deal of health content is that reference ranges are broadly untrustworthy. They are not.
Three specific situations produce a real gap. Each has a mechanism rather than a claim.
The reference population is not a healthy population. Ranges are built from people who present for testing. For markers reflecting metabolic health, that population includes a great many people with early metabolic disease — so the range absorbs them and widens. Fasting insulin is the clearest case: the upper end of a typical range reflects what is common in a population where insulin resistance is common, not what is desirable.
Ranges are often age-adjusted, and ageing is not the same as health. Where a range shifts with age, it describes what is typical for that age, which quietly builds decline into the definition of normal. Testosterone is the example most often raised.
Outcome evidence sometimes points somewhere narrower.** For a handful of markers there is direct evidence linking levels to events, and it does not always align with the reference interval. Cardiovascular risk markers are the clearest example: cardiology has for years worked to targets below the laboratory's "normal" for people at elevated risk, and that is mainstream practice rather than an alternative view.
Where one of these applies, the distinction is real. Where none does, it is not.
The useful version of this argument is narrow and it is real: for a few markers, the reference range answers a different question than the one you are asking.
The unhelpful version is a published list of optimal numbers applied to everyone, which is how the phrase is usually used and is not something we do. We read your results against your symptoms, your history and each other.
Where the evidence supports aiming somewhere specific, we say so and explain why. Where it does not, treating a number instead of a person creates problems rather than solving them.
— Dr. David Oliak, board-certified physician and surgeon
Being straightforward about this is part of the point of the page.
Published "optimal ranges" are usually not evidence. A narrower interval presented as optimal generally reflects a practitioner's preference rather than outcome data showing that people inside it do better. Sometimes those preferences are reasonable. They are still not the same as evidence, and they should not be presented as if they were.
Some markers are contested. Whether there is an optimal thyroid level narrower than the reference range is an active disagreement among endocrinologists, not a settled matter that mainstream medicine has overlooked. Our thyroid testing page treats it as the open question it is.
Treating a number has costs. Medication carries risk, over-treatment produces its own problems, and a person who feels well and is treated toward a target can end up worse rather than better. The direction of the error matters less than the fact that it is an error.
And a symptom without an abnormal marker still needs explaining. Often the explanation is that the cause has not been found yet — not that the range is wrong.
It would be straightforward to put a table of optimal values on this page. Most practices in this field do. We do not, for three reasons.
A number means different things in different people. A ferritin at the low end means one thing in a man and another in a woman with heavy periods. A borderline TSH means one thing with antibodies present and another without.
Published targets get self-applied. A number on a web page becomes something to chase, without the history and symptoms that make it interpretable.
And where a target is well supported by evidence, it is specific to a situation — a risk level, an age, a diagnosis — not a universal value that belongs in a table.
This is the more useful way to think about the whole question. A blood panel is one input into an evaluation. It is not the evaluation.
What makes a result mean something is everything around it: your symptoms, your history, your family history, what you take, what has changed since the last time you were measured, and what an examination finds. The same value in two people can call for two different responses, and the difference is not in the value.
So "what is an optimal level" has no general answer. It has a specific one, for a specific person, once the rest of the picture is in front of you.
That is also the practical difference between ordering a panel and having an evaluation.
Anyone can buy a list of markers. What those markers mean for you takes someone who has taken your history, examined you, and is sitting with both the results and you. And someone who will still be there in six months when it is time to see what moved.
In range is usually reassuring, and should be taken as such. For most of a panel, that is the end of the question.
A few markers deserve a closer look even inside the range — fasting insulin, vitamin D, ferritin, and the cardiovascular markers, for the reasons above.
Slightly out of range is often nothing. One in twenty healthy people falls outside any given range by construction. A single marginal result is a reason to repeat the test, not to start treatment.
The trend matters more than the snapshot. A value that has moved substantially within the range over two years is more informative than one reading at either end of it. This is the strongest argument for establishing a baseline while everything still looks unremarkable.
And the panel is read alongside you, not instead of you. Every one of these judgements depends on knowing who the result belongs to.
Let's begin
Your evaluation includes an unhurried hour going through what your results actually mean for you — what is reassuring, what is worth watching, and what is worth acting on. In Brea. To speak with us first, call 714-671-2005.
What a comprehensive blood panel measures
Fasting insulin: the marker most panels skip
ApoB and Lp(a): better measures of cardiovascular risk
Thyroid testing: what a full panel shows
That your result falls within the middle 95% of a reference population the laboratory measured. It describes what is common in that group. It is not by itself a statement about what is ideal, and it is not identical between laboratories.
Mostly no. For tightly regulated markers and organ-safety tests, in range is exactly the right target. The distinction between normal and optimal is real for a specific and fairly short list of markers, for identifiable reasons — not as a general rule.
Usually that the cause has not been found yet, rather than that the ranges are wrong. It is worth checking whether the right markers were measured. A standard panel does not include fasting insulin, ApoB, Lp(a), or a full thyroid panel.
And it is worth looking at the non-laboratory explanations: sleep, the perimenopausal transition, iron status, medication effects, mood.
We do not publish one, and the reason is on this page. The reference range for insulin is wide, and it reflects a population in which insulin resistance is common — so a result at the upper end deserves attention.
What "good" looks like for you depends on your glucose, your lipids, your weight history and your family history.
Often not. By construction, one in twenty healthy people falls outside any given reference range. A single marginal result is usually a reason to repeat the test in context rather than to act on it.
We do not work from a published alternative range. Where evidence supports aiming at something narrower than the laboratory interval, we use it and explain why. Where it does not, we read the result against your symptoms and the rest of your panel rather than against a different table.
