Can a Blood Test Really Grade Your Mitochondria? What the Science Says About Mitochondrial Testing

Edited by Samael A. Tejada, Founder & Editor-in-Chief

This article walks through the main approaches being marketed and recommended today — a German blood-cell respiration index, a UK/German ATP and respiration panel menu, and a US gene-expression test built around the "cell danger response" — and grades each one on the same question: has the test itself, not just the biology behind it, been validated?

What would a test of "mitochondrial health" actually have to prove?

Before any test can be called "the best" at measuring something, three things have to exist. First, a definition: what exactly is being measured? Second, a reference standard: an accepted way to know the true answer, so the test's results can be checked against it. Third, published accuracy: sensitivity and specificity in the kind of people the test will be used on, plus evidence the test gives the same result when repeated, and ideally evidence that acting on the result improves outcomes.

"Mitochondrial health" in a person who does not have a diagnosed mitochondrial disease currently fails at step one. There is no consensus definition and no gold-standard score in mainstream mitochondrial medicine. Specialist centres do not grade a person's mitochondria from 0 to 10. They answer narrower questions — is there a genetic defect? Is a specific enzyme complex underperforming in tissue? Is exercise capacity limited in a pattern that fits? — and they combine several methods precisely because no single one is sufficient.

That gap between the marketing language and the clinical reality is the lens for everything below. (Evidence: strong — this reflects the published diagnostic pathway for primary mitochondrial disease.)

How do doctors actually diagnose mitochondrial problems?

The established pathway for suspected primary mitochondrial disease is genetics first. Modern sequencing of mitochondrial and nuclear DNA identifies a definite cause in roughly half to two-thirds of suspected cases. When sequencing finds an uncertain variant — or nothing — specialists turn to functional confirmation: enzyme activity assays and respiration measurements in muscle or in skin fibroblasts, where a major defect (for example, an enzyme complex running below 20% of the normal mean) counts as strong evidence.

The most rigorous recent example of functional testing shows both what good validation looks like and what it honestly delivers. A 2026 study in EMBO Molecular Medicine, led by Children's Hospital Colorado, tested a battery of four functional assays on skin-biopsy fibroblasts from 204 patients with genetically confirmed mitochondrial disease, 51 healthy controls, and 53 patients with look-alike conditions. Combined, the battery reached 76% sensitivity and 93% specificity — good, not perfect, published with its failures included (it performed poorly for one subtype), and presented by its own authors as a complement to genetic sequencing, not a replacement and not a wellness score.

Supportive blood markers exist — lactate, certain metabolites, and stress signals such as GDF15 — but specialists treat them as clues, not verdicts: many patients with genuine mitochondrial disease have normal values, and abnormal values occur in many other conditions. Exercise testing (CPET) can reveal a characteristic pattern of impaired oxygen use. Urine organic acids have a real but narrow role, mainly in children and acute metabolic presentations; specialists note their sensitivity is low when a patient is stable.

Hold that benchmark in mind: named population, genetic reference standard, published accuracy, honest limits. (Evidence: strong for the diagnostic pathway; the fibroblast battery is a single large study.)

So why do thoughtful doctors use these tests anyway?

If the validation is missing, why do serious clinicians — including academic researchers with reputations to protect — keep ordering and developing blood-based mitochondrial tests? The honest answer is that the underlying science is genuinely promising. It is unfinished, which is a different thing from wrong.

Start with the biology. Mitochondria are not only power plants; they are sensors, and immune cells rewire their metabolism as they activate — a field called immunometabolism with a solid literature behind it. Blood cells also circulate through every tissue, so their metabolic state plausibly reports something systemic in a way no single biopsy can.

Then there is the best surrogacy evidence, and it deserves to be stated at full strength. In a 2019 study from the University of Pittsburgh published in JCI Insight, platelet respiration in 32 older adults correlated with the same people's muscle-biopsy respiration (r = 0.595) and with magnetic-resonance measurement of muscle ATP production (r = 0.643) — and platelet proton leak tracked walking speed. In nonhuman primates, blood-cell respiration has tracked skeletal-muscle, cardiac, and even brain mitochondrial respiration. Disease signals at group level point the same way: platelets in diabetic kidney disease show reduced respiration before kidney damage appears in urine tests, and blood cells in ME/CFS and fibromyalgia cohorts show lower respiration and lower BHI than controls.

Set against that is the largest head-to-head study, published in Heliyon in 2024, which found no significant correlation between blood-cell and muscle respiration — in a younger, healthier cohort. The fairest reading of both is that blood may mirror tissue reasonably well in older and sicker people, where mitochondrial decline is systemic, and poorly in the young and healthy. That is a research question with real stakes, not a settled answer in either direction.

The practical reasons are just as real. A muscle biopsy is painful and cannot be repeated every quarter; a blood draw can. Clinicians treating disabling fatigue — a symptom mainstream medicine often has no objective measure for at all — can trend a patient against their own baseline, which is a weaker claim than diagnosis but not a useless one. And the tests are physically low-risk; the risk is in over-reading them.

None of this rescues a commercial score that has not published its accuracy. Preliminary plausibility is why the research continues; it is not a substitute for validation, and a careful clinician who uses these tools will tell you exactly that. But it does mean the doctors reaching for them are not chasing nothing. They are working ahead of the evidence — and the sections below grade how far ahead each commercial test has run. (Evidence: moderate for group-level associations and tissue correlations in older cohorts; not established for individual diagnosis.)

What is the Bioenergetic Health Index — and does it hold up?

The Bioenergetic Health Index (BHI) began as an academic idea. In 2014, researchers at the University of Alabama at Birmingham proposed combining measurements of how blood cells breathe — their oxygen consumption under a series of chemical stresses on an instrument called a Seahorse analyser — into a single index of cellular bioenergetic health. It was published as a concept for research, not as a validated clinical test with cut-offs.

Two commercial services now sell versions of it:

BioVis Diagnostik (Germany) offers "BHI Plus." White blood cells (PBMCs) are isolated from a blood draw and their oxygen consumption is measured while the respiratory chain is blocked step by step. The results — basal respiration, inferred ATP production, proton leak, maximum and reserve capacity — are combined by a proprietary, re-weighted formula into one number on a 0.3–3.3 scale. Add-on tests cover mitochondrial DNA copy number and related markers. The report goes further than a number: it sorts patients into bands that point to specific therapies — interval hypoxia training at higher scores, ozone treatment or CO₂ baths at lower ones — with treatment durations from one month to over a year. No clinical trial behind those therapy bands is cited in the company's own document. The validation offered for the re-weighted BHI Plus formula is an in-house study of 183 samples stressed in a dish — unpublished, not peer-reviewed, with no patient-versus-control accuracy figures. BioVis's price for the test is not public. The laboratory itself is accredited to the ISO 15189 medical-laboratory standard, but whether this specific test sits inside that accredited scope could not be confirmed from public sources — and lab accreditation certifies a quality system, not the truth of a novel score.

The Academy of Nutritional Medicine (AONM, UK) routes blood to MMD, a laboratory in Magdeburg, Germany, for a menu built on the same concept: an ATP Profile (£125) measuring ATP capacity split between mitochondria, glycolysis, and reserve; a Mitochondrial Health Index (£195) using the same Seahorse approach; and add-ons such as mtDNA copy number, lactate/pyruvate ratio, and a fuel-flexibility assay (the full combination is £485, plus courier). AONM's own order form states that it "cannot provide a medical diagnosis" and "makes no claims whatsoever to be able to diagnose or treat medical conditions."

Both services face the same scientific problem, and it is a serious one: blood cells are not muscle, brain, or heart. In the largest study that measured both in the same people (Heliyon, 2024), respiration of blood cells did not correlate significantly with respiration of skeletal muscle — and athletes' superior muscle respiration was not mirrored in their blood cells. The authors concluded blood cells cannot accurately detect smaller changes in muscle respiration. A blood-cell score may be measuring something real about blood cells. Reading it as your whole body's mitochondrial health is an unproven leap. (Evidence: the underlying measurement method is legitimate research technology; the commercial scores and their clinical claims are vendor-only — grade D.)

Can gene expression measure mitochondrial function?

A different approach comes from ProgeneDx in the United States. Its test, GENIE, does not measure respiration, ATP, or enzyme activity at all. It measures messenger RNA — how strongly 188 genes are being expressed in blood cells — using a NanoString counting platform. One group of those genes is labelled the mitochondrial domain: genes for electron-transport chain parts, membrane transport proteins, ATP synthase components, and mitochondrial ribosomes. When those transcripts run low, the report infers a state the company calls "hypometabolism," displayed alongside domains for immune signalling, coagulation, apoptosis, and others, and staged on a curve tied to a specific treatment protocol.

The clinical price is $720, plus roughly $150 for the dry-ice return shipping the RNA samples require; only licensed providers can order it. The science it rests on is two small studies from the test's own developers (2015 and 2016, the second with 14 patients). No published study of the GENIE panel itself — its accuracy, its repeatability, or an independent group's replication — could be found.

The most telling detail is on the company's own website, in its own words: the testing is provided "on a research use only basis," and the company "cannot make any specific claims as to test results unless and until receiving FDA approval." A test that is research-use-only is, by its own disclaimer, not a validated clinical diagnostic — however confident the marketing around it sounds. GENIE was also built for one specific framework, chronic inflammatory response syndrome (CIRS), whose diagnostic criteria have themselves not been validated in independent systematic reviews; major bodies including the CDC do not recognise the broader syndrome. Using the panel as a general test of mitochondrial health in ordinary fatigue goes beyond even its intended use. (Evidence: vendor-only, research-use-only — grade D.)

What happened when an earlier mitochondrial score was independently tested?

This field has been here before. An earlier generation of testing — an ATP-profiles test scored as a "Mitochondrial Energy Score" and widely used in chronic fatigue care — published promising results from its own developers in 2009. When independent researchers at Newcastle University replicated the protocol in Scientific Reports in 2019, they found no difference between patients and healthy controls on any component of the score — and found that simply delaying sample processing by 24 hours, well within the window the original method allowed, significantly changed the results. Their conclusion: the test did not have the reliability required of a diagnostic test and should not be offered as one.

Fairness matters here: that failed test is not identical to today's ATP Profile and respiration index, which use different methods. But the newer tests have published no validation of their own to put in its place. In a field where the one score that got independently checked failed, "trust us" is not evidence. (Evidence: strong — peer-reviewed independent replication.)

What is the cell danger response — and can a test measure it?

The cell danger response (CDR) is often cited as the theory tying these tests together. It comes from Dr. Robert Naviaux of UC San Diego, a respected mitochondrial and metabolomics researcher. His 2020 paper in the journal Mitochondrion — "Cell danger response biology: The new science that connects environmental health with mitochondria and the rising tide of chronic illness" — argues that mitochondria act as environmental sentinels, that cells under threat shift into defensive metabolic states, and that healing requires those states to complete in sequence.

It is an influential hypothesis. It is also, in that paper, exactly that: a Perspective — an essay. It reports no new experiment, no patient cohort, no biomarker, and no test. The CDR stages have no agreed laboratory criteria, which means "stuck in the cell danger response" cannot currently be confirmed or ruled out by any lab result. When a commercial report claims to show your CDR status — through gene transcripts or a respiration score — it is adding a claim the framework's own author did not make in that paper. The idea may guide future research. Today, it cannot validate a test. (Evidence: hypothesis-level framework — grade C as a theory, not applicable as test validation.)

What should you ask before paying for a mitochondrial test?

If a practitioner recommends a test of your mitochondrial health, these questions separate evidence from marketing:

1. Has this exact test — not the technology, not the theory — published its accuracy? Ask for the study showing sensitivity and specificity against a reference standard, in people like you.
2. What is its regulatory status? "Research use only" means the maker itself cannot claim it diagnoses anything. An accredited laboratory is not the same as a validated test.
3. Does the result change treatment — and is that treatment itself proven? A score that routes you to months of therapy should come with trial evidence for both the routing and the therapy.
4. What tissue is actually being measured, and what is the evidence it reflects the tissue you care about? For blood-based scores, that bridge is currently unproven.
5. What would we do differently with a normal result versus an abnormal one? If the answer is the same programme either way, the test is decoration.

If the answers are vague, the honest options are the established ones: a proper clinical work-up for your symptoms, and — where mitochondrial disease is genuinely suspected — referral for genetic and functional testing at a specialist centre.

The bottom line

The desire behind these tests is reasonable: fatigue and slow recovery are real, and mitochondria sit at the centre of how cells make energy. But as of today, no commercial test has earned the title of "best test for mitochondrial health" — or "validated test" at all. The German respiration index and the UK/German panel menu rest on a legitimate research method wrapped in unpublished, proprietary scoring, with therapy recommendations that outrun the evidence. The US gene-expression test measures transcripts, not mitochondrial function, and says so itself in its research-use-only disclaimer. The cell danger response remains a hypothesis without a measurable biomarker. And the one earlier score in this field that faced independent replication failed it.

That is not a reason to dismiss mitochondrial science — the specialist diagnostic pathway shows how powerful it is when it is done rigorously. It is a reason to demand of any mitochondrial test what we would demand of any other: publish the accuracy, then make the claim.

Where the evidence stands

• Established: Mitochondrial disease is diagnosed genetics-first, with functional testing of muscle or fibroblasts as confirmation — a 2026 fibroblast battery study published 76% sensitivity and 93% specificity against genetic diagnoses.
• Vendor-only (grade D): The commercial blood scores reviewed here — the BioVis BHI Plus index, the AONM/MMD panel menu, and the ProgeneDx GENIE panel — have no published accuracy data and no independent replication; GENIE is sold on a research-use-only basis by its own maker.
• Hypothesis (grade C): The cell danger response is an influential framework with no agreed laboratory criteria; no test can currently confirm or rule out a CDR state.