Getting results back is the point at which private testing most often goes wrong. A number outside a range triggers alarm; a number inside it triggers dismissal. Both reactions misunderstand what the ranges are.
What a reference range actually is
A reference range is usually the central 95% of results from a reference population of apparently healthy people. It is a description of what is common, not a definition of what is healthy.
Trend beats snapshot
A single result is a point. Three results over two years is information. A ferritin of 30 means little in isolation; a ferritin that has gone 95 → 60 → 30 means something, even though every value is technically in range.
This is the strongest argument for testing periodically rather than once — not to catch more abnormalities, but to establish what normal looks like for you.
Biological variation is larger than people expect
| Marker | Typical variation | Practical implication |
|---|---|---|
| Testosterone | Up to 30% day to day | Never diagnose on one sample |
| Ferritin | Rises sharply with any inflammation | A cold can normalise a deficient result |
| TSH | Varies with time of day, highest overnight | Sample at a consistent time when tracking |
| Cortisol | Falls roughly 80% from morning to evening | Timing is the whole result |
| HbA1c | Very stable — reflects ~3 months | A genuine change means something |
Things that quietly distort results
- High-dose biotin — common in hair and nail supplements, and a well-documented cause of falsely abnormal thyroid results. Stop two days before testing.
- Recent hard training — raises creatine kinase, liver enzymes and inflammatory markers for days.
- Acute illness — suppresses testosterone and raises ferritin and CRP. Wait a fortnight after recovery.
- Time of day — decisive for testosterone and cortisol.
- Dehydration — concentrates almost everything measured.
- Iron supplements taken that morning — make iron studies unreadable.
When to act
- Markedly out of range — contact your GP, and do so promptly rather than researching first.
- Slightly out of range with symptoms that fit — worth a GP appointment with the result in hand.
- Slightly out of range with no symptoms — usually worth repeating before doing anything else.
- In range but at an extreme, with a clear trend — worth mentioning, particularly with previous results to show.
- In range, no symptoms, no trend — no action. This is the majority of flagged results.
Sources
Association for Clinical Biochemistry and Laboratory Medicine
Reference interval derivation and the statistics of multiple testing.
MHRA drug safety update: biotin interference in immunoassays
Documented interference with thyroid and hormone assays at high supplemental doses.
