Tiredness that sleep does not fix is one of the most common reasons anyone starts reading about mitochondria at all. At first it looks like a scheduling problem. Then a holiday helps for about a week. Somewhere around that point the word "mitochondria" shows up in a search box, and with it a stack of pages that retell a school textbook and end in a supplement ad.

The real picture sits between those two extremes. Mitochondrial dysfunction does belong on the canonical list of aging hallmarks, but whether it is a cause or a consequence remains unsettled. And the best-documented lever for changing it turned out to be nothing that comes in a capsule.

What actually changes with age

Mitochondria are the intracellular stations that turn oxygen and nutrients into ATP, the cell's universal energy currency. They carry their own DNA, their own cycle of fission and fusion, and their own disposal system for worn-out copies. The review by Nunnari and Suomalainen describes them not as passive "batteries" but as a signalling hub that affects inflammation, apoptosis and metabolism broadly (Cell, 2012).

When researchers assembled the list of aging hallmarks, mitochondrial dysfunction went in as its own entry (Cell, 2013), and in the updated version ten years later it stayed there (Cell, 2023). That detail matters: a decade of close criticism did not remove it.

What exactly breaks is laid out in the review by Sun, Youle and Finkel (Molecular Cell, 2016). The changes are several and differ in kind: respiratory chain efficiency falls, mutations accumulate in mitochondrial DNA, the balance between organelle fusion and fission drifts, and mitophagy weakens, so damaged copies are cleared less well. That last one is routinely underrated, because the problem is not only that working mitochondria become fewer, but that broken ones stop disappearing on time.

This was measured in people, and the numbers exist

Discussions here usually lean on mice. Direct human measurement exists too. Short and colleagues took skeletal muscle biopsies from 146 volunteers aged 18 to 89 and saw a progressive decline in mitochondrial enzyme activity and ATP synthesis rate with age (PNAS, 2005). That is tissue biochemistry, not a questionnaire about how people feel.

Cause or consequence, still an open question

Worth pausing here, because this is exactly where most articles make a leap. The fact that mitochondria work worse in an older person does not establish that they cause aging.

The strongest argument for causation came from mice with a defective mitochondrial DNA polymerase. Mutations in their mitochondrial genome accumulate at an accelerated rate, and the animals do age prematurely (Nature, 2004). That is a model with a deliberately broken enzyme, not proof that the same route runs in a healthy human. It does show the link is not purely correlational.

What actually moves the needle, per the data

The best-documented way to change mitochondrial function in a human turned out not to be a molecule. Robinson and colleagues trained young and older volunteers on three different programmes and took muscle biopsies before and after. High-intensity intervals produced the largest gain in mitochondrial respiration, and the gain in the older group was larger than in the younger one (Cell Metabolism, 2017). The review by Cartee and colleagues pulls this line of work together (Cell Metabolism, 2016).

For anyone selling compounds this is an inconvenient fact. We include it anyway, because a section on mitochondria that omits training is incomplete no matter who wrote it.

Where research compounds fit

A separate and much younger line runs through molecules aimed at the mitochondrion directly. The directions that come up most often:

  • MOTS-c is unusual from the start, being encoded by mitochondrial rather than nuclear DNA. In work by Lee and colleagues it affected metabolic homeostasis and insulin sensitivity in mice (Cell Metabolism, 2015). Humanin belongs to the same class. We covered the compound in more depth in a separate review.
  • SS-31 binds cardiolipin in the inner membrane, aiming at the organelle's structure rather than at one enzyme. Details in the elamipretide review.
  • Nicotinamide riboside comes at it from another direction, at the coenzyme level. Human data are most abundant here, and more contradictory than commonly assumed: the evidence base, reviewed.
  • Mitophagy itself is part of the wider process of autophagy, which we wrote about here.

What the whole list shares: the evidence base is mostly preclinical. These are research-use compounds and the catalogue labels them that way. A promised outcome from a compound at this stage means the seller either has not read the primary sources or is counting on you not having read them.

What you can check before paying

Since nobody can honestly guarantee a functional result here, one question does still have an answer: what is actually in the vial. Batches come with a certificate of analysis from an independent laboratory, and the document opens without registration and without an order. The lot number from the label can be checked in the same place. The seller can be verified before payment rather than after it.

Open the batch certificates and read what they say before ordering anything.