Research prototype, not medical advice.

07 — Mitochondrial dysfunction

"Antagonistic" hallmark (López-Otín et al., Cell 2023, doi:10.1016/j.cell.2022.11.001). In plain terms: mitochondria are the "power plants" of the cell. With age, they produce less energy, leak more toxic waste, and the cell is worse at recycling worn-out plants.

1. Definition

Mitochondria make ATP (the cell's "energy currency") through the respiratory chain. They also regulate calcium, cell death and part of cell signalling. With age, their efficiency falls, their own DNA (mtDNA) accumulates mutations, quality control (mitophagy, formation of new mitochondria) weakens and NAD+ declines. The hallmark is called "antagonistic" because at low doses, mitochondrial by-products (reactive oxygen species, ROS) are useful signals; it is their chronic excess that becomes harmful (Sun 2016, doi:10.1016/j.molcel.2016.01.028).

2. Mechanisms and sources

Mitochondrial DNA mutations. mtDNA (37 genes) is copied often and poorly protected. "Mutator" mice, whose mitochondrial DNA polymerase proofreads its copies poorly (POLG), accumulate mutations and age prematurely: hair loss, osteoporosis, anaemia, shortened lifespan (Trifunovic 2004, doi:10.1038/nature02517). This aging works mainly through cell death (apoptosis), not through excess oxidative stress (Kujoth 2005, doi:10.1126/science.1112125). In older humans, mutations are distributed as a "mosaic": some muscle fibres or neurons lose their respiratory chain and others do not (Kauppila 2017, doi:10.1016/j.cmet.2016.09.017).

The free radical theory scaled back. The old idea ("ROS damage everything, so antioxidants extend life") is largely refuted:

NAD+ decline. NAD+ is essential to the respiratory chain and to sirtuins (hallmark 06). It falls with age in several tissues, partly because the enzyme CD38, carried by inflammatory immune cells, consumes more of it (Camacho-Pereira 2016, doi:10.1016/j.cmet.2016.05.006; review: Covarrubias 2021, doi:10.1038/s41580-020-00313-x).

Failing quality control. The cell normally removes worn-out mitochondria (mitophagy, the PINK1/Parkin pathway, whose mutations cause familial forms of Parkinson's disease: Pickrell 2015, doi:10.1016/j.neuron.2014.12.007) and makes new ones (biogenesis, via PGC-1α). Both processes decline with age.

Leaks and inflammation. A damaged mitochondrion can release its DNA into the cell. The immune system then mistakes it for a virus and triggers an interferon response via cGAS-STING (West 2015, doi:10.1038/nature14156). This is a direct bridge to chronic inflammation.

Sources: sedentary lifestyle (the leading modifiable factor), overeating, drugs toxic to mitochondria, inherited mtDNA mutations.

3. Symptoms and clinical manifestations

4. Interactions with the other hallmarks

5. Existing or investigational remedies

InterventionTargetLevel of evidenceKey resultSource
Endurance and interval exerciseBiogenesis (PGC-1α), mitophagyHuman trials (RCT); recommendationHigh-intensity interval training increases the mitochondrial capacity of old muscle and protein synthesisRobinson 2017, doi:10.1016/j.cmet.2017.02.009
Urolithin AMitophagyPreclinical, then human phase 2 trialsWorm: extends life; mouse: improves muscle (Ryu 2016). Human: safe, improved mitochondrial signature (Andreux 2019); muscle endurance and biomarkers improved at 4 months in 65-90-year-olds (Liu 2022) and in overweight 40-64-year-olds (Singh 2022)Ryu 2016, doi:10.1038/nm.4132; Andreux 2019, doi:10.1038/s42255-019-0073-4; Liu 2022, doi:10.1001/jamanetworkopen.2021.44279; Singh 2022, doi:10.1016/j.xcrm.2022.100633
Nicotinamide riboside (NR)NAD+Human phase 1-2 trialsWell tolerated, blood NAD+ increased (Martens 2018); in old muscle, the NAD+ metabolome rises and inflammatory signals fall, with no bioenergetic gain (Elhassan 2019)Martens 2018, doi:10.1038/s41467-018-03421-7; Elhassan 2019, doi:10.1016/j.celrep.2019.07.043
Nicotinamide mononucleotide (NMN)NAD+Preclinical; human phase 2 trial (n=25)Mouse: mitigates physiological decline (Mills 2016). Prediabetic women: improved muscle insulin sensitivity (Yoshino 2021)Mills 2016, doi:10.1016/j.cmet.2016.09.013; Yoshino 2021, doi:10.1126/science.abe9985
NAD+ precursor in old miceStem cellsPreclinicalStem cell function restored, life slightly extendedZhang 2016, doi:10.1126/science.aaf2693
MitoQ (mitochondria-targeted antioxidant)Mitochondrial ROSHuman trial (crossover RCT, n=20)Improved vascular function in 60-79-year-oldsRossman 2018, doi:10.1161/HYPERTENSIONAHA.117.10787
Elamipretide (SS-31, Forzinity)Cardiolipin (inner membrane)Approved (FDA accelerated approval, 19 Sept. 2025) for Barth syndrome, on muscle strength (intermediate endpoint); confirmatory trial requiredFirst approved drug that targets mitochondria; no aging-related indicationFDA, https://www.fda.gov/news-events/press-announcements/fda-grants-accelerated-approval-first-treatment-barth-syndrome
Classic antioxidants (vitamins A, E, beta-carotene)Global ROSMeta-analysis of RCTsNo benefit; slightly increased mortality; hinder adaptation to exerciseBjelakovic 2012, doi:10.1002/14651858.CD007176.pub2; Ristow 2009, doi:10.1073/pnas.0903485106
CD38 inhibitionNAD+ consumptionPreclinicalMaintains NAD+ and mitochondrial function in old miceCamacho-Pereira 2016, doi:10.1016/j.cmet.2016.05.006

6. Open questions

  1. Is NAD+ really limiting in humans? Precursors do raise NAD+ in the blood, but functional gains are slim or absent. It is not yet known which tissues benefit.
  2. Cause or consequence? In humans, mitochondrial decline largely reflects the decline in physical activity. It remains hard to separate "intrinsic" aging from a sedentary lifestyle.
  3. mtDNA mosaicism: must a threshold of failing cells be reached for a tissue to suffer? How can it be measured in a living person?
  4. Which ROS should be reduced, where and when? Targeted antioxidants (MitoQ, SkQ1) remain to be validated on solid clinical endpoints.
  5. Mitophagy: does urolithin A really act through mitophagy in humans? Direct measurements in tissue are missing.
  6. Biomarkers: there is still no standard, accessible measure of mitochondrial health (phosphorus spectroscopy, respirometry on biopsy, mtDNA in blood).

Note: a frequently cited study on PINK1/Parkin and STING inflammation (Sliter 2018, Nature) was retracted; it is not used here.

7. References