A count of mitochondrial DNA in blood predicts a person's death and disease
Mitochondria, the cell's power plants, carry a little DNA of their own, and the number of copies per cell can be counted in a blood sample. In 2014 a research group reported, in over sixteen thousand people followed for years, that those with the lowest counts were more likely to die, and in 2017 the same group reported that low counts went with more heart disease and stroke in over twenty-one thousand. The count can be read from stored blood. But a review by a mitochondria research group warns that it mostly reflects which blood cells and how many platelets a sample holds, not how well the power plants work, and no trial in this project's records uses it to judge a treatment. It is a research measure, not a test anyone can order.
What would count
The number of copies of mitochondrial DNA per cell, counted in a blood sample, predicts time to death or to heart or blood-vessel disease after allowing for age and sex, in two or more independent groups of at least a thousand people followed for years, in peer-reviewed reports. For the clinic rung: a regulator clears the count as a medical test, or a named health system uses it in routine care and publishes that practice. For wide use: offered in routine care in several countries.
Who settles it: Peer-reviewed cohort reports; the regulator's clearance record, or a named health system's published practice, for the clinic rung.
Where it stands
First human. Reached 4 December 2014, on the record of Ashar and 11 others, 4 December 2014 (doi.org). Next rung within 5 to 20 years.
When it reaches the clinic: the project's estimate
This is the one thing on this page that is a judgment rather than a record. The project's estimate of the year it reaches the clinic is 2045, no earlier than 2031 and no later than 2100; the model draws from that range every run, with the likely year the most probable draw. Set 18 October 2026, a first pass awaiting an outside review.
| Rung | No earlier than | Most likely | No later than | |
|---|---|---|---|---|
| In clinical use | 2031 | 2045 | 2100 | used for the year at the top |
Why. The reports are large and agree on direction, but no regulator clearance and no health-system use of the count turned up: a search of the United States device clearance database (openFDA, 2026-10-18) lists no device named for mitochondrial DNA, and a web search the same day found no guideline or clinic using it (laboratory-run tests need no clearance, and other countries were not searched). The one measurement problem the group raises, platelets and the mix of white cells, is not settled in the large reports and would have to be before a laboratory could sell the count as a test; and no trial in the records uses the count to judge a treatment, which a measure for judging repairs needs. The earliest year assumes a health system publishes use of the count inside a risk score within five years; the latest, 2100, means perhaps not this century, because the count may stay a research measure.
Rests on: Ashar and 11 others, 4 December 2014; Ashar and 13 others, 1 November 2017; Rausser and 11 others, 26 October 2021; Picard, 19 June 2021.
Changes to this record
None since it was written on 18 October 2026. Every change to the stage or the estimate lands here with its date, the old value, the new one and the reason, and a change to the estimate never happens because a number looked wrong: it needs a dated reason and a source.
The last readings
No weekly reading has run yet. Once a week, from the site's launch, every treatment is read against the news, against what would count above, and the last three readings of this one are shown here with the date and the verdict. About the weekly reading.
What the story so far says about it
On the record
- 4 December 2014Association of mitochondrial DNA levels with frailty and all-cause mortality
- 1 November 2017Association of mitochondrial DNA copy number with cardiovascular disease
- 26 October 2021Mitochondrial phenotypes in purified human immune cell subtypes and cell mixtures
- 19 June 2021Blood mitochondrial DNA copy number: What are we counting?
Notes
Added on 2026-10-18, after a reading of the work of Martin Picard's group at Columbia University on mitochondria and ageing. The count is of mitochondrial DNA copies per cell in a blood sample. The group's own measures (a health index built from enzyme activities and this count, panels on sorted blood cells, blood markers) are left off until a peer-reviewed report follows people to an outcome with them. Checked on 2026-10-18 against the papers' open copies at PubMed Central. Ashar and colleagues (Journal of Molecular Medicine, online 2014-12-04, https://doi.org/10.1007/s00109-014-1233-3) used two groups, 4,892 and 11,509 people: the fifth with the lowest counts were 47 per cent more likely to die than the fifth with the highest (hazard ratio 1.47, 95% range 1.33 to 1.62), after allowing for age and sex. Ashar and colleagues (JAMA Cardiology, 2017-11-01, https://doi.org/10.1001/jamacardio.2017.3683) followed 21,870 people in three studies for 13.5 years on average: each standard deviation below the average count went with 23 per cent more heart disease or stroke (hazard ratio 1.23, 95% range 1.19 to 1.26), unchanged after allowing for the usual risk factors. Searching both papers for platelets and white-cell counts found no mention of allowing for them. Rausser and colleagues (eLife, 2021-10-26) found in 21 healthy adults that platelets stuck to samples made mitochondrial measures read 9 to 22 per cent high and that platelet count fell about 6 per cent per decade, which could account for part of the fall in blood mitochondrial DNA with age; Picard (Mitochondrion, 2021-06-19) reviews why the count is set mostly by the mix of cells and platelets and is not a measure of how well mitochondria work. A review of 22 studies (Fu and colleagues, QJM, 2025-01-01, https://doi.org/10.1093/qjmed/hcae208) agrees on direction; only its abstract was read, so it is not listed as a source here. Open question: whether the prediction survives allowing for platelets and the mix of white cells. What would count does not ask for it, and the stage stays on the 2014 report.