Impact of aging and exercise on skeletal muscle mitochondrial capacity, energy metabolism, and physical function

The study addresses how aging affects skeletal muscle mitochondrial function and how this relates to declines in muscle mass, physical function, and metabolic health, collectively contributing to sarcopenia and loss of independence. Although mitochondrial respiratory activity appears to decrease with age and is hypothesized to underlie functional decline, the relative contributions of chronological aging versus reduced physical activity (PA) are unclear. The authors aimed to: (1) determine whether mitochondrial function is reduced in older versus younger adults matched for habitual PA and how it relates to muscle function; and (2) assess whether regular exercise training in older adults preserves mitochondrial function and muscle health, also comparing to physically impaired older adults. They further examined correlations among mitochondrial capacity, physical function, and muscle health.

Prior work shows age-related decreases in skeletal muscle mitochondrial respiration in humans, and PGC-1α levels correlating with walking speed in older adults. Animal models (e.g., SOD1 knockout mice, aged rats) implicate mitochondrial dysfunction and apoptosis in accelerated sarcopenia, fiber splitting, and atrophy. In humans, evidence linking mitochondrial alterations to muscle function is limited, with some studies focusing separately on either function or mitochondria. Declines in PA with age can reduce mitochondrial capacity, while exercise training stimulates biogenesis via PGC-1α, suggesting PA partly explains observed declines. However, studies report mixed findings: some show age-related MAPR declines even in trained individuals, while others (e.g., Distefano et al.) report no chronological age effect on ex vivo respiration when accounting for PA. This background motivates disentangling effects of age versus PA using well-phenotyped cohorts.

Design: Cross-sectional study conducted at Maastricht University Medical Center+ (Sept 2017–Mar 2020), approved by ethics committee; informed consent obtained; registered NCT03666013. Participants: n=59; young normally active (Y, 20–30 y, n=17), older normally active (O, 65–80 y, n=17), trained older (TO, 65–80 y, n=19), physically impaired older (IO, 65–80 y, n=6). Normally active: ≤1 structured exercise session/week; trained: ≥3 sessions/week ≥1 h for ≥1 year; IO: SPPB ≤9. Exclusions: MRI contraindications, uncontrolled hypertension, medications interfering with outcomes, CVD, type 2 diabetes, other safety concerns; lab assessments ensured liver/kidney function normal. Experimental schedule: Five visits over ~5 weeks; participants maintained habitual diet/PA; refrained from strenuous activity for 3 days before tests. Assessments: - Habitual physical activity: ActivPAL3 accelerometer worn for 5 consecutive days (incl. weekend); steps/day, stepping time as % waking time; PA classified as high-intensity (>110 steps/min) and low-intensity (≤110 steps/min). - Body composition: Air displacement plethysmography (BodPod) to quantify fat mass and fat-free mass (FFM). - Maximal aerobic capacity: Graded cycling test with ECG to volitional exhaustion; VO2max via indirect calorimetry. - Submaximal exercise and energy efficiency: 1-h cycling at 50% Wmax (from maximal test), cadence 60–70 rpm; indirect calorimetry at 15 and 45 min; calculated energy expenditure (Weir equation), substrate oxidation (Péronnet & Massicotte); gross exercise efficiency (GEE = work/EEE × 100) and net exercise efficiency (NEE = work/(EEE−REE) × 100); REE measured by ventilated hood for 45 min and adjusted for FFM (residuals). - Insulin sensitivity and substrate selection: Hyperinsulinemic-euglycemic clamp with [6,6-2H2]-glucose tracer; insulin infusion 40 mU m−2 min−1; M-value (GIR) for whole-body insulin sensitivity; glucose kinetics via Steele equations; insulin-stimulated glucose disposal (Rd) corrected for plasma insulin and glucose (S), expressed per kg BW and per kg FFM; endogenous glucose production (EGP) suppression; indirect calorimetry during basal and steady-state insulin for RER and substrate oxidation. IO largely excluded from clamp comparisons due to safety (n=3 remained). - Muscle strength: Isokinetic knee extension/flexion (Biodex System 3 Pro), 30 consecutive movements at 120°/s; peak torque normalized to FFM. - Muscle volume and in vivo mitochondrial function: 3T MRI for T1-weighted upper leg imaging to quantify muscle volume; 31P-MRS of vastus lateralis during rest, knee-extension exercise, and recovery; phosphocreatine (PCr) recovery rate constant k (s−1) as in vivo oxidative capacity (pH maintained >6.9; mono-exponential fit). - Muscle biopsy and ex vivo mitochondrial respiration: Vastus lateralis biopsy (Bergström method); permeabilized fiber bundles assessed in OROBOROS oxygraph under hyper-oxygenated conditions (~400 µM O2). Substrate-uncoupler-inhibitor titrations for: state 2 (substrate only: malate+octanoyl-carnitine (MO) or malate+glutamate (MG)); state 3 (ADP-stimulated) with MG; lipid-supported (MO); convergent Cx I+II (MOG, MOGS, MGS); maximal uncoupled (state 3u, FCCP); leak (state 4o, oligomycin). Cytochrome c test to ensure membrane integrity (<15% increase criterion). Quadruplicate measurements. - Mitochondrial content: Western blot OXPHOS complexes I–V in muscle lysates (Abcam ab110411), quantified via infrared imaging. - Gait performance and stability: Self-paced 6-min walk test (6MWT) on CAREN system; fixed-speed treadmill trials (0.4–1.8 m/s) for step time/length/width/double support time means and CVs; perturbation protocol with unilateral belt accelerations (10 perturbations; analyze Pert1, Pert2, Pert9). Margin of stability (MoS) computed (Base, Pre, Post1–8 steps); number of recovery steps to baseline stability. IO excluded from perturbation analyses (only 3 completed). Statistics: Sex distribution by χ2. Y vs O by two-sided independent t-tests. O vs TO vs IO by one-way ANOVA with Tukey’s post-hoc or Kruskal–Wallis with Bonferroni correction as appropriate. Gait tasks: mixed-effects models across speeds and groups; two-way repeated measures ANOVAs for perturbations; nonparametric tests for recovery steps. Correlations: Pearson/Spearman and partial correlations (adjusted for sex, age, BMI). Significance at p<0.05.

Young (Y) vs older (O) with similar habitual PA (~10,000 steps/day; no differences in HPA/LPA): - Physical function and fitness: Older adults walked ~9% less distance on 6MWT (p=0.032); VO2max ~26% lower (p<0.001); maximal power output lower (p<0.001); isokinetic strength lower (p<0.001). Upper leg muscle volume similar (p=0.151). REE (adjusted for FFM) similar (p=0.275). RER at rest and during submaximal exercise similar (p=0.852 and p=0.624, respectively). Gross and net exercise efficiency both lower in O (both p<0.001). - Insulin sensitivity: M-value ~22% lower in O (p=0.050). Peripheral insulin sensitivity S per kg BW lower in O (0.050 (0.028) vs 0.073 (0.024) µmol kgBW−1 min−1, p=0.014); trend when normalized to FFM (p=0.069). EGP suppression similar (p=0.850). Basal and insulin-stimulated RER, and metabolic flexibility, similar. - Gait: Older had greater step length variability during unperturbed walking (group effect F(1,30)=7.077, p=0.012 at 1.2–1.6 m/s). In perturbations, MoS worse in O only at first perturbation (F(1,28)=7.7, p=0.010), with adaptation preserved by subsequent perturbations. - Mitochondrial capacity: Ex vivo state 2 respiration similar except MO lower in O (p=0.044). State 3 respiration lower in O for MG (p=0.032); MO not different. Parallel Cx I+II state 3 lower in O: MOG (p=0.014), MOGS (p=0.006), MGS trend (p=0.077). Maximal uncoupled respiration ~17% lower in O (p=0.008). Leak (state 4o) similar (p=0.379). In vivo PCr recovery rate constant ~16% lower in O (p=0.003). OXPHOS protein expression (Complexes I–V) similar between Y and O. Older groups (O vs trained older TO vs impaired older IO): - PA and body composition: TO had lower BMI and fat mass than O (BMI p=0.046; fat % p=0.050) and IO (BMI p=0.024; fat % p=0.021). Steps/day higher in TO vs O (p=0.050) and IO (p=0.004); HPA time higher in TO vs O (p=0.049) and IO (p=0.005). LPA time higher in O than IO (p=0.038). - Function and fitness: IO covered ~22% less 6MWT distance vs O (p=0.013) and ~29% less vs TO (p=0.005). VO2max highest in TO (TO vs O ~1.2-fold, p=0.003; TO vs IO ~1.4-fold, p=0.004). Maximal power higher in TO vs O and IO (both p<0.001). Knee extension torque ~24% and ~27% lower in IO vs O (p=0.026) and TO (p=0.006); knee flexion torque highest in TO vs O (~1.1-fold, p=0.046) and vs IO (~1.5-fold, p<0.001). Muscle volume similar across groups. REE and RER similar across older groups. Gross exercise efficiency higher in TO vs O (p=0.001) and IO (p<0.001); net efficiency higher in TO vs O (p=0.006) and IO (p=0.002). - Insulin sensitivity: M-value ~1.3-fold higher in TO vs O (p=0.023). S per kg BW tended to be higher in TO (p=0.072); S per kg FFM not different (p=0.165). EGP suppression and RER (basal/insulin) similar; metabolic flexibility similar. - Gait: No significant differences in gait variability or MoS between O and TO across perturbations. - Mitochondrial capacity: In permeabilized fibers, TO higher than O in state 2 (all substrates, p<0.05), state 3 MG (~1.2-fold, p=0.049) and MO (~1.4-fold, p<0.001), and combined Cx I+II: MOG (p=0.001), MOGS (p=0.002), MGS (p=0.018). Maximal uncoupled respiration ~1.4-fold higher in TO (p=0.001). Leak (state 4o) higher in TO vs O (~1.3-fold, p=0.006) and vs IO (~1.3-fold, p=0.026). In vivo PCr k ~1.2–1.3-fold higher in TO vs O and IO, but not significant (TO vs O p=0.169; TO vs IO p=0.135; O vs IO p=0.815). OXPHOS content: Complexes I and III higher in TO vs O (p=0.010; p=0.042); Complex II and IV higher in TO vs both O and IO (II p=0.003 and 0.025; IV p<0.001 and 0.003, respectively). Correlations (combined cohort): Max coupled respiration (MOGS3) correlated with VO2max (r=0.436, p=0.002) and daily steps (r=0.326, p=0.014). Max uncoupled respiration correlated with M-value (r=0.407, p=0.003) and gross exercise efficiency (r=0.402, p=0.003); negatively with double support time variability (r=−0.324, p=0.019). PCr k negatively correlated with double support time variability (r=−0.313, p=0.025). Gross exercise efficiency strongly correlated with 6MWT distance (r=0.679, p<0.001) and inversely with chair-stand time (r=−0.485, p<0.001).

Despite similar habitual physical activity, older adults exhibit lower skeletal muscle mitochondrial capacity (ex vivo respiration and in vivo PCr recovery), reduced aerobic capacity, exercise efficiency, strength, and whole-body insulin sensitivity, and greater gait variability compared to younger adults. These findings support an age-associated decline in mitochondrial function independent of habitual PA levels and not explained by reduced mitochondrial content. In contrast, exercise-trained older adults show preserved or enhanced mitochondrial respiration and improved exercise efficiency, aerobic capacity, strength, and insulin sensitivity compared to normally active peers. The enhancements in trained older adults are largely attributable to increased mitochondrial content (higher OXPHOS protein levels), suggesting training-induced mitochondrial biogenesis mitigates age-related functional declines. Significant correlations between mitochondrial capacity and exercise efficiency, aerobic capacity, insulin sensitivity, walking performance, and gait stability indicate mitochondrial energetics are closely linked to muscle function and metabolic health. Collectively, the results suggest that while normal daily activity may be insufficient to prevent age-related mitochondrial and functional declines, sustained exercise training can counteract these effects, likely via increased mitochondrial content and improved oxidative capacity.

This study demonstrates that aging, even with adequate habitual physical activity, is associated with reduced skeletal muscle mitochondrial capacity, exercise efficiency, aerobic capacity, muscle strength, gait stability, and insulin sensitivity. Regular exercise training in older adults largely preserves mitochondrial function and muscle health, with improvements linked to increased mitochondrial content. The strong associations between mitochondrial capacity, exercise efficiency, and insulin sensitivity underscore mitochondria as a promising therapeutic target to mitigate age-related deterioration of skeletal muscle and maintain physical function. Future research should explore mechanisms underlying age-related mitochondrial declines independent of activity, the role of exercise intensity and modality in optimizing mitochondrial adaptations, sex-specific effects, and longitudinal interventions to establish causality.

- Cross-sectional design limits causal inference regarding aging versus activity effects. - Small sample size in the physically impaired older group (IO, n=6) reduced power, especially for mitochondrial and gait perturbation analyses; many IO excluded from clamp. - Differences in adiposity and body composition between groups may confound insulin sensitivity and potentially mitochondrial outcomes. - In vivo 31P-MRS PCr recovery can be influenced by perfusion and oxygen delivery, potentially obscuring between-group differences. - Plasma insulin levels during clamp differed between young and older groups, requiring correction and potentially affecting comparability. - Cohorts were relatively active (~10,000 steps/day), which may limit generalizability to more sedentary older populations. - Potential sex differences were not the primary focus; mixed-sex groups may mask sex-specific effects.