How mitochondria adapt with age: mechanisms, resilience, and therapies

Mitochondria have long been cast as the batteries of the cell, and their decline has been framed for decades as the molecular equivalent of a battery running flat: less ATP, more reactive oxygen species, and faster ageing. A new review published in the journal Biogerontology argues that this familiar picture is fundamentally incomplete. Piotr Paweł Chmielewski of Wroclaw Medical University proposes a framework he calls mitochondrial homeodynamics, which treats the organelle’s contribution to ageing not as a single pathway gone wrong but as the progressive erosion of three linked capacities: maintenance, adaptation, and recovery. In doing so, the review reframes mitochondrial dysfunction as something that can be measured dynamically, rather than merely described, and it issues a stern warning about what popular anti-ageing supplements can and cannot actually deliver.

The core of the argument is that mitochondria are not static energy factories. They integrate oxidative metabolism with redox signalling, calcium handling, biosynthesis, apoptosis, innate immunity, and constant communication with other organelles. What determines whether an aged tissue retains function, the review contends, is not the raw abundance of mitochondria or their resting respiration, but the network’s ability to sustain itself under stress. Maintenance safeguards the mitochondrial genome, proteome, and membrane integrity. Adaptation adjusts metabolism and remodels network and cristae architecture to match changing demand. Recovery restores function and reserve after a challenge such as exercise, infection, or injury. Ageing, in this view, contracts the range over which these transitions can occur, even when basal measurements look deceptively normal.

Take the mitochondrial genome. The review synthesises evidence showing that the accumulation and clonal expansion of mtDNA mutations during ageing arise predominantly from replication error rather than oxidative damage, overturning a long-standing assumption. Heteroplasmic variants can remain functionally silent until their proportion crosses mutation- and tissue-specific biochemical thresholds, and mutation load alone does not determine mitochondrial failure. Recent work adds a provocative twist: ribonucleotide incorporation into mitochondrial DNA can drive inflammatory signalling, meaning that defective genome maintenance has consequences well beyond impaired oxidative phosphorylation. When damaged mitochondrial DNA or RNA leaks into the cytosol of senescent cells, it activates innate immune pathways including cGAS-STING and the NLRP3 inflammasome, reinforcing the senescence-associated secretory phenotype and feeding the chronic low-grade inflammation known as inflammageing.

Proteostasis is equally central. The human mitochondrial proteome exceeds one thousand proteins, most encoded in the nucleus, imported into the organelle, and assembled into complexes under the surveillance of chaperones, ATP-dependent proteases, ribosome quality control, the mitochondrial unfolded protein response, and mitophagy. This network matters most in long-lived postmitotic cells such as neurons and cardiomyocytes, where cumulative proteotoxic stress cannot be diluted by cell division. Strikingly, cryoelectron tomography studies have recently linked an age-related decline in mitoribosome abundance and organisation in primary human T cells to impaired mitochondrial biogenesis and reduced cellular function, illustrating how organelle translation can become a cell-type-specific bottleneck in human ageing.

The review is emphatic that neither fragmentation nor elongation of mitochondrial networks is intrinsically good or bad. Experimental manipulation of either fission or fusion can extend lifespan in model organisms under specific conditions, which undermines the habit of treating one morphology as a universal marker of mitochondrial health. The more informative property is network plasticity, the capacity to remodel appropriately. Human studies back this up: three-dimensional imaging of aged skeletal muscle shows that age-related changes in mitochondrial architecture correlate with muscle characteristics, and that mitochondrial fragmentation predicts age-associated decline in physical capacity. Exercise appears to preserve aspects of mitochondria–endoplasmic reticulum contact-site organisation, while recent mechanistic work shows that mitochondrial calcium uptake declines with ageing and directly constrains muscle performance, a finding with implications for how energetic reserve, rather than basal respiration, should be assessed.

Mitophagy, the selective removal of damaged mitochondria, emerges as another domain where quantity is the wrong metric. Suppression of basal mitophagy can drive cellular ageing phenotypes in primary cells, and coordination between mitophagy and biogenesis, regulated by PGC-1 coactivators and TFAM, is necessary for organismal adaptation during ageing. The translational issue is turnover quality, not maximal turnover rate: excessive elimination without adequate replacement reduces energetic capacity, while biogenesis without effective quality control simply expands a dysfunctional pool. Crucially, enhanced mitophagy can restrain the cGAS-STING inflammatory axis in ageing models, positioning mitochondrial housekeeping as an anti-inflammatory intervention in its own right.

This is where the framework becomes provocative for the booming longevity industry. The review systematically evaluates candidate interventions and finds the human evidence strikingly lopsided. Exercise provides the strongest human evidence for coordinated mitochondrial and functional adaptation, engaging all three homeodynamic domains at once as a repeated multidimensional challenge followed by recovery. Energy restriction improved cardiometabolic risk factors in the CALERIE trial but never tested human lifespan. NAD+ precursors such as nicotinamide riboside and nicotinamide mononucleotide reliably raise NAD+-related metabolites, yet their effects on metabolic, vascular, inflammatory, and functional outcomes remain inconsistent; biochemical target engagement, the review insists, establishes biological activity rather than geroprotection. Urolithin A produced molecular evidence of mitophagy engagement and improved selected muscle-endurance outcomes in a randomized trial of older adults, while spermidine supplementation failed to improve its primary cognitive endpoint. MitoQ improved endothelial function in a small study, and elamipretide acutely increased muscle ATP production but flopped in a longer heart-failure trial. Notably, metformin attenuated some exercise-induced mitochondrial adaptations in older adults, a caution that a drug can engage a plausible target while blunting the response to a stimulus of established benefit.

The review’s most quotable conclusion is unambiguous: no mitochondrial intervention has been shown to slow ageing or extend lifespan in healthy humans, and the movement of a biomarker towards a younger reference value does not establish rejuvenation. Circulating markers such as GDF15 and FGF21, which rise with age and report activation of the mitochondrial integrated stress response, are confounded by inflammation, renal function, malignancy, and common medications including metformin. They report that a stress pathway is active, not that the underlying capacities are preserved.

To guard against circularity, Chmielewski frames the concept as three falsifiable predictions. First, among individuals matched for age, sex, and resting mitochondrial capacity, the amplitude and rate of recovery after a standardised physiological challenge should predict subsequent functional decline, whereas resting measures alone should predict it less well or not at all. Second, an intervention that improves resting biomarkers without improving adaptation or recovery should not deliver the functional outcomes attributed to its mitochondrial action. Third, because the limiting domain differs across tissues and individuals, the response to a domain-specific intervention should be predictable from which domain is limiting at baseline. If measures of adaptation and recovery add no predictive information beyond resting phenotypes, the framework should, by its own logic, be abandoned.

The implications for precision geroscience are considerable. Because skeletal muscle, heart, neurons, immune cells, and haematopoietic stem cells face entirely different constraints, the review argues against a single mitochondrial ageing score unless its components are explicitly tissue-aware. Even in stem cells, mitochondrial abundance identifies states with distinct self-renewal properties rather than functioning as a simple damage marker, so the relevant phenotype is the ability to transition between metabolic states, not maximal output in one of them. Sex modifies mitochondrial phenotypes too: human blood-cell respiration differs by sex across age and can even increase with age, complicating any assumption of uniform decline.

The practical obstacle is measurement. Standardised challenge-based phenotyping, quantifying respiratory reserve, substrate switching, mitophagic flux, redox recovery, and functional performance after exercise, metabolic stress, or vaccination, remains to be validated for reproducibility, tissue specificity, and prediction of clinical outcomes. Muscle biopsy permits detailed phenotyping but is invasive; blood-based assays may not represent brain, heart, or muscle. The review sets out five priorities for the field, including identifying which measurements reproducibly capture the three domains in humans, and determining whether they predict functional decline independently of chronological age and established risk factors.

The conclusion is a measured reframing rather than a breakthrough claim. Mitochondrial ageing, the review argues, is best described as erosion of coordinated maintenance, adaptation, and recovery, not uniform loss of ATP production or simple oxidative stress. Mitochondrial changes may be initiating, amplifying, compensatory, or consequential depending on tissue and context, and more mitochondrial activity is not invariably beneficial; the objective is regulated flexibility rather than maximisation. For a field saturated with promises of youthful NAD+ levels and mitochondrial rejuvenation, the message is sobering but constructive: preserve adaptive capacity and function without disrupting compensatory biology, and judge interventions by recovery kinetics and clinically meaningful outcomes, not by biomarkers drifting towards a younger reference range.