Labs and Testing Explained

A normal result is a statement about a population, not a verdict on you, and one flagged value on a big panel is often just arithmetic.

17 answers·Written and medically reviewed by Zahraa Sater, M.D., M.P.H.

How to read results4

Hormone and cortisol tests3

Nutrient and heart markers6

Tests worth skipping2

Scans and imaging2

Not sure which one you need?

The question worth asking about any test is not whether it is interesting but whether a plausible result changes the plan. Here is how several commonly requested tests score.

TestWhat it actually answersWhen it changes managementWhen it does not
TSH with free T4Is the thyroid making enough hormone nowSymptoms of thyroid disease, a dose change, pregnancy or planning itRepeated at short intervals in a stable person on a stable dose
Lp(a)Inherited lipoprotein risk you cannot see on a standard panelOnce in adulthood, and in relatives of anyone with a high levelRepeat testing, since the level is largely fixed for life
ApoBHow many atherogenic particles are present, not how much cholesterol they carryTriglycerides are high or LDL looks reassuring but risk does notAs a substitute for treating an already clear elevated risk
Ferritin with iron studiesWhether iron stores explain fatigue or hair sheddingFatigue, heavy periods, hair loss, restless legsDrawn during an acute illness, when inflammation raises ferritin regardless of stores
25(OH)DVitamin D statusMalabsorption, bone disease, chronic kidney disease, certain medicationsRoutine screening in a healthy adult
Fasting insulin and HOMA-IRA research-derived index of insulin resistanceNormal glucose with a strong metabolic picture, as one input among severalUsed as a diagnosis, since no diagnostic threshold is agreed across assays
Reverse T3Nothing decision-useful in a healthy outpatientEffectively never in this settingOrdered to explain symptoms with normal thyroid function

The answers below take each of these tests in turn and explain what a specific result should and should not lead to.

Background: how to think about labs and testing explained A longer read from Dr. Sater, for context rather than a specific question

A reference range is a statement about a population, not a statement about you. It is built by measuring a reference group and drawing the lower limit at the 2.5th percentile and the upper limit at the 97.5th, as Yesim Ozarda set out in Biochemia Medica in 2016. One in twenty perfectly healthy people therefore falls outside the range on any given test, by construction. Order twenty tests on a healthy person and the chance that at least one comes back flagged is roughly 64 percent. That is arithmetic, not illness. The reverse error costs more: a result sitting inside the range is not evidence that the system is working, because the range was designed to include almost everyone.

What this area actually covers

Every test does one of three jobs, and the same test is interpreted differently depending on which job it is doing. Screening looks for disease in someone with no symptoms. Diagnosis confirms or excludes a specific suspicion in someone who has symptoms. Monitoring tracks something already known. A TSH drawn as a screen in a well person, a TSH drawn to explain fatigue, and a TSH drawn six weeks after a levothyroxine dose change are three different tests wearing the same name, and a result of 4.0 mIU/L means something different in each.

This section also covers the mechanics that quietly wreck results: what you ate, what time you attended, where you were in your cycle, and which supplements you took. Biotin is the clearest example, since it interferes with many immunoassays and can make thyroid results look like disease that is not there.

Where testing goes wrong in both directions

Over-testing and under-testing produce different kinds of harm. Large direct-to-consumer panels generate flagged values that are statistically expected, and each one then costs a repeat test, an imaging study or a specialist visit to unwind. Under-testing is quieter: a workup that stops at TSH and a complete blood count and never reaches ferritin, B12, glucose with insulin, or an assessment for sleep apnea.

A third pattern sits between them. A single value gets treated as a trend. Most hormones and many metabolic markers vary day to day and hour to hour, so the difference between two results is often the assay and the timing rather than a change in the person.

How to decide whether a test is worth ordering

One rule handles most of it: a test earns its place only if a plausible result would change what happens next. Lipoprotein(a) passes that test. The National Lipid Association's 2024 focused update recommends measuring Lp(a) at least once in all adults, treating levels at or above 125 nmol/L as high risk, and a high value changes how aggressively other risk factors are treated and prompts testing in relatives. It is inherited and largely stable, so once is usually enough.

Routine vitamin D screening fails the same test. The Endocrine Society's 2024 guideline states that in healthy adults it suggests against routine screening for 25(OH)D levels, and the US Preventive Services Task Force concluded in 2021 that the evidence is insufficient to assess the balance of benefits and harms of screening asymptomatic adults. Reverse T3 fails it too: the American Thyroid Association states that in healthy, non-hospitalized people it "is not clinically useful."

The genuinely contested territory is the idea of an optimal range, narrower than the laboratory reference range, marketed heavily by direct-to-consumer and functional testing services. Narrowing a target for a specific person with a specific condition is ordinary clinical practice. Publishing a universal optimal range is not, because those numbers are usually not derived from outcome data at all, and no specialty society has endorsed them.

What changed recently

The direction of travel is fewer routine screens and more targeted one-time tests. The Endocrine Society moved against routine vitamin D screening in healthy adults in 2024. The National Lipid Association added a once-in-a-lifetime Lp(a) measurement in 2024. The USPSTF updated its osteoporosis screening recommendation in January 2025, recommending screening for women 65 and older and for postmenopausal women under 65 at increased fracture risk, both grade B, while concluding that evidence in men remains insufficient.

The answers below cover preparation and timing first, then individual tests and what each one is actually for, then the tests worth skipping and how to read your own report. The table above is the fastest way to find the test you are holding.

Other topics

Every answer here is written and reviewed by a board-certified endocrinologist. To have that judgment applied to your own history, book a consultation with Dr. Sater.

Book a consultation