Abstract

Aging is intimately associated with multisystem functional decline and an increased risk of chronic diseases. A pivotal cytological basis underlying this process is the progressive dysregulation of the mitochondrial quality control (MQC) network. Emerging evidence suggests that MQC is not a singular process but rather a multitiered synergistic system encompassing mitochondrial biogenesis, dynamic remodeling, selective autophagy (mitophagy), proteostasis maintenance, and coordinated mitochondrial–organelle communication. This integrated network is critical for preserving cellular energy homeostasis, redox balance, and stress tolerance. During aging, impairments in mitochondrial genomic coordination, network topology, autophagic flux, and protein import and folding collectively contribute to bioenergetic decline, chronic low-grade inflammation, and metabolic imbalance. As a safe and sustainable nonpharmacological intervention, regular exercise systematically remodels MQC structure and function by integrating signaling axes such as AMPK, SIRT1, and p38 MAPK, thereby promoting coordinated mitochondrial renewal and partially reversing aging-associated mitochondrial dysfunction. On the basis of a systematic elucidation of the core mechanisms of MQC and its dysregulation during aging, this review highlights the differential regulatory effects of distinct exercise modalities—specifically endurance training, high-intensity interval training (HIIT), and resistance training—on mitochondrial dynamics, autophagic flux, proteostasis, and mitochondrial turnover. Furthermore, the intrinsic associations among exercise–MQC coupling, inflammatory responses, metabolic imbalances, and emerging peripheral biomarkers are explored. Finally, current research limitations and challenges in clinical translation are analyzed, and future research directions regarding dose–response relationships, multimodal exercise prescriptions, personalized strategies, and systemic integrated regulation are proposed. This review aims to provide a refined theoretical basis for optimizing exercise-based anti-aging interventions.

1 Introduction

Aging is a systemic biological process characterized by the progressive disruption of tissue homeostasis, and its progression is closely associated with an increased risk of chronic conditions such as neurodegenerative diseases, metabolic syndrome, and cardiovascular diseases (López-Otín et al., 2023). With the deepening of research in mitochondrial biology, accumulating evidence indicates that the role of mitochondria in aging extends far beyond the traditional view that regarded them merely as “passive victims” of reactive oxygen species–induced damage. Current perspectives propose that mitochondria actively drive the initiation and progression of aging-related functional decline through a complex regulatory network encompassing biogenesis, dynamics, selective autophagy, and proteostasis, collectively referred to as mitochondrial quality control (MQC). Dysregulation of this network results in a progressive decline in mitochondrial function, leading to disturbances in energy metabolism, activation of inflammation, and reduced cellular stress tolerance, ultimately manifesting as tissue dysfunction and increased disease susceptibility (Szklarczyk et al., 2014; Banarase et al., 2023).

In contrast to earlier research paradigms that emphasized the accumulation of mitochondrial damage, recent studies increasingly focus on understanding the dynamic coordination among the various mechanisms of MQC at a systems level. Mitochondrial biogenesis, dynamics, and mitophagy are regarded as an integrated functional network, and their coordinated imbalance may precede multisystem functional decline (Koklesova et al., 2022; Gottlieb et al., 2021). Under healthy conditions, MQC maintains the stability of mitochondrial population structure and function through complementary mechanisms of renewal, selection, and clearance (Holt, 2010). However, with advancing age, declines in the quality of newly generated mitochondria, network fragmentation, impaired autophagic flux, and the accumulation of misfolded proteins become exacerbated, rendering cells more susceptible to energy deficiency and chronic low-grade inflammation, thereby accelerating the aging process (Feng and Lu, 2025; Skawratananond et al., 2025).

Regular physical exercise represents a well-established and readily implementable non-pharmacological intervention for delaying aging (Grolaux et al., 2024; Lohman et al., 2023). Although early explanations primarily focused on exercise-induced increases in mitochondrial content and biogenesis, emerging evidence suggests that the protective effects of exercise largely depend on its multilayered regulation of the mitochondrial quality control system (Craige et al., 2024; Díaz‐Castro et al., 2024). Exercise enhances mitochondrial renewal efficiency and adaptive capacity through the integration of multiple signaling pathways, thereby partially reversing aging-related mitochondrial dysfunction.

Despite continuous advances in research, several key questions remain unresolved, including how different MQC components achieve effective coordination under exercise-induced stress; whether specific exercise modalities exert preferential regulatory effects on particular MQC pathways; and how mechanistic insights at this level can be translated into clinically generalizable exercise prescription strategies. In this context, the present review systematically integrates current evidence regarding the role of exercise-mediated MQC remodeling in aging. We first delineate the structural and functional logic of MQC, identify critical nodes of dysregulation during aging, and summarize the regulatory characteristics of different exercise modalities. On this basis, we further analyze the association between tissue-level mechanisms and systemic adaptations, and explore how this association relates to aging-associated inflammation and metabolic imbalance. It should be clarified that the majority of existing literature focuses on skeletal muscle at the mechanistic level, as it is metabolically active, highly responsive to exercise stimuli, and serves as a classical tissue model for investigating exercise-induced MQC remodeling (Choi et al., 2025). However, exercise-triggered MQC adaptations are not confined to local muscle but are accompanied by systemic signal reprogramming. In recent years, peripheral samples—including blood, plasma/serum, peripheral blood mononuclear cells (PBMCs), platelets, extracellular vesicles, and circulating cell-free mitochondrial DNA (ccf-mtDNA)—have increasingly been utilized to assess exercise-related mitochondrial adaptations, and the corresponding indicators to some extent reflect alterations in whole-body mitochondrial homeostasis and inflammatory status. In addition to skeletal muscle tissue, this review also incorporates systemic peripheral biomarkers into the discussion. Finally, we address current limitations in the field and challenges in clinical translation, and propose directions for future research. We anticipate that this review will facilitate an integrated understanding of the relationships among exercise, mitochondrial quality control, and aging, and provide a reference framework for optimizing anti-aging intervention strategies.

2 Methodology2.1 Literature search strategy

A systematic search was conducted in the PubMed and Elsevier ScienceDirect databases. The search period was defined from 2015 to 2025. The search strategy combined controlled vocabulary terms and free-text keywords, which were integrated using Boolean logical operators. Reference lists of relevant review articles and key research studies were manually screened, and citation tracking was performed to supplement potentially overlooked important publications, thereby ensuring comprehensive literature coverage. The search string was as follows: (“exercise” OR “physical training” OR “endurance training” OR “high-intensity interval training” OR “resistance training”) AND (“mitochondrial quality control” OR “mitochondrial biogenesis” OR “mitochondrial dynamics” OR “mitophagy” OR “autophagy” OR “mitochondrial unfolded protein response”) AND (“aging” OR “ageing” OR “sarcopenia” OR “frailty”)

2.2 Inclusion and exclusion criteria

The inclusion criteria were as follows:

Explicit inclusion of exercise interventions or structured physical training protocols;

Assessment of at least one MQC-related indicator, including but not limited to: markers of mitochondrial biogenesis; proteins associated with mitochondrial dynamics; indices related to mitophagy; respiratory chain complex activity or mitochondrial respiratory function parameters;

A research context involving aging.

The exclusion criteria were as follows:

1) Studies limited to in vitro cellular experiments without an organismal aging context; 2) Studies that did not report quantifiable MQC-related outcome measures; 3) Non-English publications; 4) Conference abstracts lacking complete supporting data.

2.3 Literature screening process

Two investigators independently conducted title and abstract screening. Full texts of preliminarily eligible studies were subsequently retrieved for further evaluation. Discrepancies arising during the screening process were resolved through discussion, and when necessary, adjudicated by the corresponding author.

2.4 Quality assessment

To evaluate the quality of the included studies, we applied a modified version of the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach. This method was used to assess risk of bias, methodological rigor, and the strength of evidence for each study. High-quality studies (e.g., well-controlled randomized trials and studies with large sample sizes) were prioritized in the evidence synthesis (Figure 1).

Structural and functional basis of mitochondrial quality control (MQC).

3 Structural and functional foundations for maintaining mitochondrial quality control homeostasis3.1 Cross-genomic coordination safeguards the intrinsic quality of mitochondrial biogenesis

Mitochondrial biogenesis constitutes the initiating step of mitochondrial quality control (MQC). It integrates multiple signals, including cellular energy status, redox homeostasis, and nutrient sensing, thereby determining the structural integrity, metabolic capacity, and long-term adaptive potential of newly formed mitochondria (Scarpulla, 2012; Kaur and Naqvi, 2025). Peroxisome proliferator-activated receptor gamma coactivator-1 alpha (PGC-1α) serves as the central transcriptional coactivator within this regulatory network (Dumesic et al., 2025; Gidlund et al., 1985; Moissidis et al., 2025). Upon sensing cellular energy stress and redox fluctuations, PGC-1α is activated through post-translational modifications, such as AMP-activated protein kinase (AMPK)-mediated phosphorylation and sirtuin 1 (SIRT1)-mediated deacetylation. Activated PGC-1α subsequently translocates to the nucleus, where it cooperates with nuclear respiratory factors to initiate the transcription of genes encoding respiratory chain subunits, fatty acid oxidation enzymes, and factors involved in mitochondrial DNA (mtDNA) replication and transcription, thereby systemically activating the mitochondrial biogenesis program (Clark and Parikh, 2021; Ke et al., 2021; Uittenbogaard and Chiaramello, 2020).

Concurrently, PGC-1α drives the coordinated expression of nuclear-encoded respiratory chain subunits and mitochondrial transcription factor A (TFAM) by activating transcription factors such as NRF1/2 and ERRα. TFAM represents a critical node in coordinating transcription between the nuclear and mitochondrial genomes. By binding to the mtDNA control region and inducing conformational changes, TFAM plays an essential role in transcription initiation, replication, and structural packaging. The co-expression of oxidative phosphorylation (OXPHOS) subunits encoded by both nuclear and mitochondrial genomes maintains the proper stoichiometry of respiratory chain complexes. The assembly of these complexes depends on the coordinated integration of subunits encoded by the dual genomes; dysregulation of either genome can disrupt this homeostatic balance, leading to defective respiratory chain assembly, abnormal mitochondrial architecture, and functional impairment (Hees and Harbauer, 2022; Cuppari et al., 2019; Papier et al., 2022).

During aging, although certain tissues may exhibit transient compensatory upregulation of PGC-1α signaling, the persistent accumulation of mtDNA mutations, progressive decline in mitochondrial translational machinery, and structural impairment of supercomplex assembly mechanisms collectively disrupt mitochondrial proteostasis and compromise the functional integrity of the OXPHOS system. These defects may reduce the capacity of newly generated or remodeled mitochondria to establish membrane potential, diminish oxidative phosphorylation efficiency, and weaken antioxidant stress resistance, thereby exacerbating aging-associated declines in energy metabolism (De Smalen et al., 2023; Safdar et al., 2016; Parmar et al., 2024).

3.2 Spatial organization and network topology shape the platform for mitochondrial quality control

Beyond molecular composition, the subcellular spatial distribution and network topology of mitochondria profoundly influence their functional performance and MQC efficiency (Wai and Langer, 2016; Pickles et al., 2018; Steffan et al., 2025). Such structural heterogeneity establishes the basis for spatial recognition and functional compartmentalization in subsequent selective clearance mechanisms. Distinct mitochondrial subpopulations exhibit significant differences in anatomical localization and functional responsibilities. For example, in skeletal muscle, mitochondria demonstrate clear subcellular functional specialization: subsarcolemmal mitochondria (SSMs) are primarily distributed near capillaries and perinuclear regions, facilitating nutrient and oxygen sensing and predominantly participating in substrate uptake and metabolic regulation; in contrast, intermyofibrillar mitochondria (IFMs), characterized by a high surface area-to-volume ratio and tight embedding within the sarcomeric structure, are responsible for rapid ATP generation and localized energy supply to sustain continuous contractile function (Willingham et al., 2021).

In most eukaryotic cells, mitochondria form dynamically interconnected networks through fusion and fission. This connectivity facilitates the uniform distribution of electrochemical gradients, metabolites, and signaling molecules across the mitochondrial population, thereby enhancing systemic functional robustness. However, in mature skeletal muscle cells, mitochondria are spatially constrained by myofibrils and the cytoskeleton, aligning along sarcomeres to form a distribution pattern tightly coupled to contractile units. Their degree of network connectivity varies according to muscle fiber type and is particularly heterogeneous in oxidative fibers (Holt et al., 2024; Katti et al., 2022; Milner et al., 2000).

Aging systemically disrupts tissue structural order, resulting in the loss of subcellular localization specificity and fragmentation of the mitochondrial network. These alterations are associated with decreased activity of fusion proteins, increased activity of fission proteins, weakened cytoskeletal support, alterations in mitochondria-associated membrane structures, and lipid peroxidation (Smith et al., 2023; Ma et al., 2024; Zorov et al., 2019). Disintegration of the network structure not only impairs the efficiency of energy and repair signal transmission across the mitochondrial population but also renders mitochondria more susceptible to diverse stressors.

3.3 Cooperative quality control mechanisms of mitochondrial dynamics and autophagy

Mitochondrial dynamics, through the dynamic balance between fusion and fission, enable real-time surveillance and quality selection within the mitochondrial network. Mitochondrial fusion is mediated by outer membrane proteins MFN1 and MFN2 and the inner membrane protein OPA1, promoting the exchange of mitochondrial contents and facilitating functional complementation, damage repair, and maintenance of metabolic homeostasis. In contrast, mitochondrial fission is executed by dynamin-related protein 1 (DRP1), which is recruited to the mitochondrial surface by membrane receptors such as MFF and MiD49/51. A critical function of fission is to actively segregate severely dysfunctional mitochondria from the network, generating independent depolarized fragments, thereby preventing damage propagation and creating conditions for subsequent clearance (Kleele et al., 2021; Alavi and Fuhrmann, 2013; Song et al., 2024). Aging frequently leads to an imbalance in mitochondrial dynamics, characterized by reduced expression or activity of OPA1 and MFN1/2, accompanied by enhanced DRP1 signaling, resulting in network fragmentation and loss of continuity. These structural alterations are often associated with impaired respiratory function, decreased membrane potential stability, and increased production of reactive oxygen species (ROS), particularly in highly metabolic tissues such as the heart and skeletal muscle (Sukhorukov et al., 2024; Tokuyama et al., 2022; Leija et al., 2022).

Regulation of mitochondrial dynamics lies between biogenesis and autophagic degradation, determining whether damaged mitochondria undergo repair or are targeted for removal. Dysfunctional mitochondria identified through dynamic selection are primarily processed via PINK1–Parkin pathway–mediated selective autophagy (mitophagy). Upon loss of mitochondrial membrane potential, PINK1 kinase stabilizes on the outer mitochondrial membrane, subsequently recruiting and activating the E3 ubiquitin ligase Parkin, which induces extensive ubiquitination of outer membrane proteins. These ubiquitin chains serve as “eat-me” signals recognized by autophagy adaptor proteins, ultimately targeting damaged mitochondria for encapsulation and delivery to lysosomes for degradation (Xu et al., 2025; Jiménez-Loygorri et al., 2024). However, aging also impairs this terminal clearance pathway, resulting in a generalized decline in autophagic flux. Even when mitochondrial damage is successfully marked by the PINK1–Parkin system, the efficiency of subsequent steps—including autophagosome formation, trafficking, and fusion with lysosomes—is markedly reduced (Ordureau et al., 2020; Shan et al., 2025; Chen et al., 2023).

Reduced autophagic clearance capacity leads to the accumulation of dysfunctional mitochondria and their contents within the cytoplasm. Notably, mitochondrial DNA (mtDNA) leaked into the cytosol can be recognized by the pattern recognition receptor cyclic GMP-AMP synthase (cGAS), thereby activating the cGAS–STING signaling pathway and inducing innate immune responses, including type I interferon production. Meanwhile, mtDNA and excessive ROS may function as danger-associated molecular patterns that promote the assembly and activation of the NLRP3 inflammasome, driving the maturation and secretion of pro-inflammatory cytokines such as interleukin-1β (IL-1β) (Vringer and Tait, 2023; Xian et al., 2022; Xu et al., 2022). Animal model studies further confirm that under conditions of Parkin deficiency–induced mitophagy impairment, inflammasome activation is more pronounced and closely associated with aggravated tissue injury (Li et al., 2019; Wang et al., 2024). These findings indicate that diminished autophagic efficiency is not only a driving factor for the accumulation of mitochondrial dysfunction but also a critical link connecting organelle dysfunction to systemic chronic inflammatory states (Figure 2).

Synergistic mechanisms of exercise in regulating the dysregulation of mitochondrial quality control during aging.

4 Coordinated mechanisms by which exercise modulates mitochondrial quality control dysregulation during aging4.1 Reactivation of upstream signaling networks attenuated in aging

With advancing age, the PGC-1α–centered mitochondrial regulatory network in skeletal muscle exhibits functional decline, manifested by reduced levels of key proteins such as PGC-1α and TFAM, as well as diminished activity of the upstream energy sensor AMPK. Exercise—particularly intermittent exercise accompanied by pronounced metabolic fluctuations—induces periodic perturbations in intracellular metabolite levels and calcium signaling, thereby activating signaling pathways including AMPK, p38 MAPK, and CaMKII. Within this regulatory network, AMPK acts as a primary energy sensor and is activated early; it cooperates with p38 MAPK to activate PGC-1α via phosphorylation and simultaneously elevates intracellular NAD+ levels to stimulate the deacetylase activity of SIRT1. SIRT1 further enhances the transcriptional coactivator function of PGC-1α through deacetylation (Chen WK. et al., 2018; Kang et al., 2013; Combes et al., 2015). This coordinated modification restores the mitochondrial gene expression program in aged muscle, promoting mitochondrial protein synthesis and functional improvement.

Activated PGC-1α broadly coordinates downstream gene expression, not only facilitating coordinated transcription between nuclear and mitochondrial genomes to drive mitochondrial biogenesis, but also upregulating genes related to fatty acid oxidation and antioxidant defense. Following long-term exercise training, the mitochondrial network in skeletal muscle undergoes structural and functional remodeling, characterized by increased mitochondrial density, enhanced respiratory chain complex activity, and improved overall oxidative phosphorylation capacity. In this process, reactive oxygen species (ROS) exert dual regulatory roles: on the one hand, exercise-induced ROS function as key signaling molecules that positively regulate exercise adaptation through activation of pathways such as MAPK; on the other hand, the antioxidant defense system enhanced in parallel with PGC-1α activation mitigates oxidative damage while preserving the signaling function of ROS, thereby re-establishing cellular homeostasis at an elevated metabolic level (Iijima et al., 2025; Halling et al., 2017).

4.2 Remodeling mitochondrial dynamics and optimizing damage segregation

In aged skeletal muscle and other tissues, the overall efficiency of the mitochondrial quality control system declines. Studies have shown that under stress conditions, the mitochondrial fusion protein OPA1 is prone to proteolytic cleavage and inactivation, thereby inhibiting the fusion of damaged mitochondria with the healthy network; meanwhile, DRP1-mediated fission signaling becomes relatively dominant, promoting the segregation of dysfunctional mitochondrial fragments from the network. This imbalance between fusion and fission is considered a major mechanism leading to mitochondrial network fragmentation (Marzetti et al., 2025a).

Regular exercise, through activation of the AMPK/SIRT1–PGC-1α axis, increases the expression of fusion proteins such as OPA1 and MFN1/2, while suppressing phosphorylation of Drp1 at Ser616 and reducing the recruitment of its adaptor proteins FIS1/Mff. These effects restrain excessive fission, enabling fragmented mitochondria to be reutilized and re-integrated into a highly interconnected network structure, thereby facilitating the sharing of membrane potential and matrix contents across the mitochondrial population, restoring oxidative phosphorylation efficiency, and contributing to the delay of aging phenotypes such as sarcopenia (Campos et al., 2023; Moore et al., 2019; Ruegsegger et al., 1985; Fealy et al., 1985; Long et al., 2022).

Restructuring of mitochondrial dynamics may provide the necessary structural foundation for subsequent selective clearance. Evidence indicates that exercise-induced selective mitophagy is frequently accompanied by mitochondrial fission. Exercise primarily activates the AMPK–ULK1 signaling axis and may also involve the PINK1/Parkin pathway, thereby promoting the labeling of damaged or dysfunctional mitochondria for entry into the autophagic process. Concurrently, exercise enhances the nuclear translocation and transcriptional activity of transcription factor EB (TFEB), promotes lysosomal biogenesis, and increases cellular degradative capacity. These processes help ameliorate the impaired autophagic flux commonly observed in aged tissues, reduce intracellular retention of damaged mitochondria, and consequently lower the risk of mtDNA leakage and mtDNA-mediated inflammatory responses. Under exercise intervention, mitochondrial dynamic balance and autophagic clearance mechanisms exhibit coordinated interactions, collectively maintaining mitochondrial population quality and functional homeostasis (Laker et al., 2017; Zhou et al., 2025; Huang et al., 2019).

4.3 Restoration of mitochondrial protein import and folding homeostasis

Aging is accompanied not only by abnormalities in mitochondrial morphology but also by marked disruption of mitochondrial proteostasis. Studies have demonstrated widespread downregulation of the mitochondrial translation program in aged muscle, resulting in reduced abundance of respiratory chain subunits (De Smalen and Handschin, 2025). As a potent stimulus, exercise promotes mitochondrial biogenesis while simultaneously regulating the synthesis, transport, and assembly of related proteins through pathways such as the PGC-1α/ERRα axis, thereby supporting renewal of the mitochondrial proteome. At the molecular level, exercise activates the highly conserved mitochondrial unfolded protein response (UPRmt), which cooperates with the mitochondrial quality control system to maintain organelle homeostasis. Experimental evidence indicates that both acute and chronic exercise induce the expression of key transcription factors, including ATF5, CHOP, C/EBP-β, and ATF4. These factors may translocate to the nucleus and coordinately upregulate downstream molecular chaperones (e.g., mtHsp70 and Hsp60) and proteases (e.g., ClpP and LONP1), thereby promoting proper protein folding and facilitating degradation of damaged proteins. In addition, the JNK pathway plays an important regulatory role in exercise-triggered UPRmt signaling (De Smalen et al., 2023; Gaspar et al., 2023). Furthermore, mitochondrial protein quality control is disrupted to varying degrees in aging, skeletal muscle disuse, and Parkinson’s disease (PD) models. In PD models, mitochondrial protein translocation mechanisms are markedly impaired, characterized by reduced expression of outer membrane translocases (TOM20/TOM40) and the inner membrane channel TIM23. This impairment is associated with abnormal accumulation of α-synuclein, which may bind to TOM20 and inhibit the import of nuclear-encoded proteins. In aged skeletal muscle, the import rate of matrix proteins may not necessarily decline; however, the stability of cytosolic precursor proteins is reduced, suggesting impairment at pre-import stages, and aging attenuates exercise-induced adaptive enhancement of protein import. Exercise intervention may ameliorate these processes through multiple mechanisms. In PD models, long-term treadmill exercise reduces α-synuclein accumulation and restores TOM/TIM expression. In other neuroprotective models, exercise-induced upregulation of Caveolin-1 contributes to maintenance of TOM20 levels and preservation of protein import function. Collectively, these changes promote efficient transport of nuclear-encoded proteins into mitochondrial subcompartments, alleviate cytosolic proteotoxicity, and restore mitochondrial homeostasis (Koo et al., 2017; Pan et al., 2021; Zhang et al., 2020).

In summary, exercise synergistically enhances multiple processes—including protein import, folding maintenance, and terminal clearance—thereby reconstructing and preserving mitochondrial proteostasis at the molecular level (see Table 1 for details). This mechanism complements exercise-mediated regulation of mitochondrial biogenesis, dynamics, and autophagy pathways, collectively forming a multilayered and highly adaptive quality maintenance network. These findings provide a systematic mechanistic framework for understanding how exercise delays aging-associated mitochondrial dysfunction.

SectionCore MQC moduleAging-associated dysregulationExercise-induced molecular changesFunctional outcomesStudy subjectsLevel of evidenceReferences3.1 Reactivation of Blunted Upstream Signaling Networks in AgingPGC-1α–mediated mitochondrial biogenesis axisDecreased PGC-1α mRNA and nuclear protein levels; reduced p-CREB; decreased TFAM, cytochrome c, and mtDNA/nDNA ratioIncreased nuclear PGC-1α protein; enhanced p-AMPK and p-p38 activity; restoration of TFAM expression, mtDNA content, and cytochrome c levelsRestoration of mitochondrial biogenesis and protein synthesis; attenuation of functional decline; reduced BAX/Bcl-2 ratio and Caspase-3 expressionRatsLevel IIIKang et al. (2013)AMPK–SIRT1 signaling synergy and anti-inflammatory networkMyocardial hypertrophy and elevated inflammation; downregulation of SIRT1/PGC-1α; chronic low-grade inflammation and reduced metabolic flexibilityUpregulation of SIRT1, PGC-1α, and AMPKα1; increased FOXO3a phosphorylation; improved insulin signaling; suppression of inflammatory responsesImproved myocardial architecture; enhanced antioxidant defense; restored insulin sensitivity and metabolic homeostasisRats; Humans (MetaMEx database analysis)Level II–IIIChen et al. (2018a)Lipid metabolic reprogramming and intramyocellular lipid regulationMyofiber degeneration and intramyocellular lipid accumulationModulation of lipid metabolism targets (PPARG, ADIPOQ, FABP4); enhanced fatty acid oxidation; inhibition of lipogenesisSuppression of muscular fat infiltration; improved metabolic microenvironmentHumans (MetaMEx database analysis)Level IIIijima et al. (2025)3.2 Remodeling of Mitochondrial Dynamics and Optimization of Damage SegregationPGC-1α–dependent inhibition of mitochondrial fission and network remodelingExcessive mitochondrial fragmentation; increased FIS1/DRP1 expression; Drp1 activation associated with metabolic impairmentDownregulation of FIS1 and DRP1; reduced Drp1 Ser616 phosphorylation; increased OPA1 expression; exercise effects partially dependent on PGC-1αRestoration of mitochondrial reticular network; improved oxidative metabolism; amelioration of insulin resistanceMice; HumansLevel II–IIIHalling et al. (2017),Fealy et al. (1985)AMPK-mediated remodeling of mitochondrial dynamics and maintenance of adaptive capacityImbalance in mitochondrial fission–fusion cycling; aging-associated decline in exercise capacityAMPK activation of dynamic cycling; tissue-specific regulation of Drp1 expressionDelayed age-related mitochondrial fragmentation; preserved exercise capacity and metabolic homeostasisCaenorhabditis elegans; Mice; HumansLevel II–IIICampos et al. (2023),Moore et al. (2019)Activation of the mitophagy–lysosome axisAccumulation of dysfunctional mitochondria; reduced autophagic flux; impaired lysosomal functionAMPK-dependent phosphorylation of Ulk1 (Ser555); enhanced targeting of damaged mitochondria to lysosomes; AMPK–SIRT1–mediated TFEB nuclear translocation; mechanotransduction and myokine-mediated autophagy regulationClearance of damaged mitochondria; enhanced autophagic flux and lysosomal activity; maintenance of systemic homeostasisMice; HumansLevel II–IIILaker et al. (2017),Zhou et al. (2025),Huang et al. (2019)3.3 Restoration of Mitochondrial Protein Import and Proteostatic HomeostasisPGC-1α/ERRα-mediated recovery of mitochondrial translationReduced mitochondrial translational capacity; decreased PGC-1α and ERRα expressionInduction of the PGC-1α/ERRα axis; upregulation of mitochondrial ribosomal and translational proteinsCorrection of mitochondrial protein synthesis defects; amelioration of sarcopenia-associated functional declineMiceLevel IIIde Smalen et al. (2023)JNK-driven mitochondrial unfolded protein response (UPRmt)Downregulation of UPRmt-related gene expressionJNK activation; upregulation of Hspd1, LONP1, Yme1L1, ClpP, and ATF5Enhanced skeletal muscle oxidative capacity; improved aerobic exercise performanceMiceLevel IIIGaspar et al. (2023)Restoration of mitochondrial protein import machinery (TOM/TIM complex)Decreased TOM-40, TOM-20, and TIM-23 expression; α-synuclein accumulationIncreased TOM-40, TOM-20, and TIM-23 expression; upregulation of COX-I/IV; reduced α-synuclein accumulationAttenuated dopaminergic neuronal loss; improved motor coordination in Parkinsonian modelsMiceLevel IIIKoo et al. (2017)

Exercise-mediated modulation of mitochondrial quality control (MQC) in aging.

Level I —— Randomized controlled trials in human populations.

Level II —— non-randomized human studies or longitudinal intervention studies.

Level III ——mechanistic studies limited to animal or cell experiments.

4.4 The value of peripheral biomarkers in exercise and MQC research

Traditional assessment of mitochondrial quality control (MQC) has primarily relied on tissue biomarkers obtained through invasive procedures such as skeletal muscle biopsy, including measurements of mitochondrial respiratory enzyme activity or protein expression levels. However, the invasive nature of these approaches limits their applicability for repeated sampling in clinical populations and large-scale longitudinal studies. In recent years, peripherally derived biomarkers have attracted increasing attention as complementary tools for evaluating systemic mitochondrial homeostasis and stress status, thereby providing a novel perspective for exploring the relationship between exercise adaptation and mitochondrial function.

Specifically, tissue biomarkers mainly reflect organelle homeostasis within local muscle cells, whereas circulating biomarkers are more indicative of systemic mitochondrial stress levels and inter-organ communication. For example, circulating cell-free mitochondrial DNA (ccf-mtDNA) exists in multiple forms in the bloodstream, including naked mtDNA fragments, complexes bound to TFAM, fragments encapsulated within extracellular vesicles (EVs) or mitochondria-derived vesicles (MDVs), components of neutrophil extracellular traps (NETs), and a small proportion of intact mitochondria (Zhao et al., 2025; Nidadavolu et al., 2023). This structural heterogeneity suggests that ccf-mtDNA may originate from distinct biological processes, arising either from passive release during apoptosis or necrosis, or from regulated extrusion of mitochondrial contents, the latter being considered a component of MQC (Caicedo et al., 2024; Zhang et al., 2021; Fu et al., 2025; Trumpff et al., 2021).

Notably, the previously described mechanisms whereby mtDNA leakage activates the cGAS–STING and NLRP3 inflammatory pathways may be systemically reflected by alterations in circulating ccf-mtDNA levels. In other words, ccf-mtDNA may partially represent the process by which intracellular mitochondrial homeostatic imbalance is translated into systemic inflammatory signaling. In studies of aging and cardiometabolic diseases, elevated ccf-mtDNA levels have been associated with systemic inflammation, frailty status, and reduced exercise tolerance, with some evidence suggesting predictive value independent of traditional inflammatory markers (Mengozzi et al., 2025; Byappanahalli et al., 2023). The effects of exercise on ccf-mtDNA appear to be context-dependent. Acute high-intensity exercise may induce a transient increase in circulating levels, possibly related to acute mitochondrial stress and tissue remodeling responses during exercise. In contrast, long-term regular training is often accompanied by reduced resting ccf-mtDNA levels, suggesting improvement in chronic inflammatory status and enhancement of mitochondrial homeostasis. Therefore, when investigating the impact of exercise on MQC adaptation, differences in exercise modality and physiological status must be considered.

Beyond circulating DNA, peripheral blood mononuclear cells (PBMCs) and platelets provide additional information for assessing mitochondrial function. These cells are relatively easy to obtain and suitable for repeated sampling in research settings. Studies have shown that PBMC mitochondrial respiratory parameters—such as basal respiration, maximal respiration, and spare respiratory capacity—are associated with cognitive function, hippocampal volume, and white matter integrity (Qin et al., 2024; Dang et al., 2025; Mahapatra et al., 2023). Although these indices cannot fully substitute for direct assessment of skeletal muscle tissue, they offer reference information regarding systemic redox status and dynamic changes in energy metabolism.

Exercise-induced muscle adaptations may also be communicated systemically via extracellular vesicles (EVs). Myokines released during muscle contraction—including peptides, microRNAs (miRNAs), mtDNA, and proteins—can be transported through EVs such as exosomes, enter the circulation, reach distal organs (e.g., the brain), and be internalized by target cells through endocytosis (Memme et al., 2021). Meanwhile, MDVs and circulating EVs containing mitochondrial components may serve as peripheral manifestations of early MQC events, participating in inter-tissue signaling by carrying respiratory chain subunits or mtDNA (König and McBride, 2024; Chen et al., 2024). In addition, fibroblast growth factor 21 (FGF21) and growth differentiation factor 15 (GDF15), which are associated with mitochondrial stress responses, are upregulated in various tissue-specific models of mitochondrial dysfunction. Their secretion is linked to the mitochondrial unfolded protein response and the integrated stress response. Upon entering the circulation, these factors exert endocrine regulatory effects and participate in the modulation of energy metabolism. Physiological stimuli such as exercise training and cold exposure can likewise increase their circulating levels. Existing evidence indicates that plasma concentrations of FGF21 and GDF15 are associated with metabolic status and the degree of mitochondrial stress, suggesting their potential utility as auxiliary indicators for evaluating mitochondrial functional state and related metabolic adaptations (Jena et al., 2023).

In addition to regulating mitophagy and mitochondrial dynamics, exercise induces the secretion of specific myokines that participate in local and systemic signaling. Studies have demonstrated that the peptide hormone apelin, produced in response to muscle contraction, declines with aging, and its circulating concentration is positively correlated with muscle function in older adults. Systemic or muscle-specific deletion of apelin or its receptor APLNR exacerbates age-related muscle atrophy and strength decline, whereas exogenous apelin supplementation partially improves muscle fiber cross-sectional area and muscle function in aged mice by activating the AMPK pathway, promoting mitochondrial biogenesis and autophagic flux, and suppressing inflammatory responses. Moreover, apelin may act on muscle stem cells (MuSCs), enhancing their proliferative and differentiative capacities, thereby facilitating regenerative repair in aged muscle (Vinel et al., 2018) (Figure 3).