Journal Logo


Health Benefits of an Innovative Exercise Program for Mitochondrial Disorders


Author Information
Medicine & Science in Sports & Exercise: June 2018 - Volume 50 - Issue 6 - p 1142-1151
doi: 10.1249/MSS.0000000000001546


Mitochondrial disorders are the most prevalent types of neuromuscular disease for which no therapy is available. Despite considerable heterogeneity in disease presentation, mitochondrial disorders commonly involve skeletal muscle tissue leading to mitochondrial myopathy and poor exercise capacity (1). The latter is reflected by low levels of peak oxygen uptake (V˙O2peak) or early onset of anaerobic metabolism (i.e., the so-called anaerobic or ventilatory threshold (VT) (2) or by poor muscle oxygen extraction (as assessed with near-infrared spectroscopy) during graded cycle-ergometer/treadmill testing (3,4).

Several studies have demonstrated the benefits of physical training in this patient group, usually in the form of endurance (or “aerobic”) exercise performed for a relatively long period (~2–12 months), on important outcomes, mainly V˙O2peak (5–12). This type of exercise intervention improves muscle mitochondrial biogenesis and oxidative capacity, as reflected by an increase in the activities of citrate synthase (6,7,10) and respiratory chain complexes (6,10). Considerably less work has addressed the effects of a second type of exercise that should also be an integral component of any training program, resistance (“strength”) exercise (13,14).

Muscle-derived molecules, collectively known as “myokines,” might account for some of the multisystemic benefits of exercise, especially anti-inflammatory and healthy metabolic adaptations such as decreased insulin resistance (15). This is an important consideration in view of the high prevalence of physical inactivity coupled to obesity in this patient group (16). No study has evaluated exercise effects on the myokine profile of these patients. Moreover, controversy exists regarding the effects of exercise interventions on quality of life (QoL), usually assessed with the Short Form-36 Item Health Survey (SF-36) questionnaire (5,8–14,16), and scarce data on how exercise training affects patients’ ability to perform activities of daily living (ADL) are available (5,13). Myopathy can also affect ventilatory muscles in these patients, which can further aggravate their exercise intolerance (17), but no previous study has implemented specific inspiratory muscle training (using a threshold-loading device, which can be easily used at home with little time investment). The effects of exercise interventions on other phenotypes related to mitochondrial disorders, and muscle and bone mass remain to be determined.

We determined the effects of a “complete/combined” exercise intervention consisting of innovative aerobic, resistance, and inspiratory muscle exercises, performed by patients with mitochondrial disease on outcomes indicative of aerobic power, muscle strength/power, and maximal inspiratory pressure (PImax; main end points), and ability to perform ADL, body composition, QoL, and blood myokines (both under resting conditions and in response to acute exercise; secondary end points). We hypothesized that our exercise intervention would lead to significant improvements, especially in main end points.


Experimental Design and Ethics

This study was approved by the local ethics committee of the Hospital 12 de Octubre (Madrid, Spain; approval number: 14/359) and was performed in accordance with the Declaration of Helsinki from April 2015 to January 2017. All participants provided written informed consent. Inclusion criteria were as follows: outpatient (male/female) 18–60 yr old; diagnosed with a mitochondrial disorder according to clinical, histomorphologic, respiratory chain, and/or genetic evidence (18); living in the Madrid area; and having no comorbidity or acute condition contraindicating exercise. Study end points were assessed at three time points: baseline (“pretraining”), after an exercise intervention (“posttraining”), and after a subsequent 4-wk training cessation period (“detraining”). Participants were told not to deviate from their habitual dietary habits during the study period. The researchers responsible for assessing participants were blinded to time of evaluation. Assessment/training sessions were performed in the rehabilitation department (for aerobic exercises) or gymnasium (for strength exercises). The patient study design is summarized in Figure 1.

Representation of the study design.

Fourteen patients originally agreed to participate, but two of them withdrew during the first 3 wk (one because of urine infection and the other one because of personal reasons). The main clinical/diagnostic characteristics of the patients who completed the study (n = 12) are shown in Table 1; 46% were overweight (body mass index (BMI) 25–29.9 kg·m−2), 8% obese (BMI >30 kg·m−2), and 8% underweight (BMI <18.5 kg·m−2).

Main clinical and diagnostic characteristics of the patients.

Exercise Training

The intervention lasted 8 wk and included three weekly sessions (Monday, Wednesday, and Friday; total number of planned sessions, 24; session duration, 60–90 min). Each session was supervised by experienced fitness instructors (one per participant) and included three main components: aerobic, muscle resistance (“strength”), and specific inspiratory muscle training. Make-up sessions (on a Tuesday/Thursday) were allowed if a planned session was missed. To determine the adherence to training, we considered a session completed when ≥90% of the prescribed exercises were successfully performed (19).

Aerobic training

Each session started with aerobic training on a cycle ergometer (Ergoselect 100P; Ergoline, Bitz, Germany). Aerobic exercise was preceded by an initial 10-min warm-up and a final 5-min cool-down period; the central component lasted only 20 min and was performed after a low-volume “high-intensity interval training” (HIT) protocol (20), where 1-min bouts at a target load were interspersed with recovery periods of the same duration. Although one possibility is to use 30-s, “all-out” (i.e., Wingate) repetitions, this might not be safe, tolerable, or appealing for some individuals, such as patients (20). Thus, a more practical model of low-volume HIT that is still time efficient (as it can be performed for 20 min) while also having wider application to different populations including patients consists of 10 × 1-min work bouts at a constant-load, very high intensity (∼90% of maximal heart rate), interspersed with 1 min of recovery (20). Owing to the poor baseline physical condition of our participants and to the fact that HIT has not been previously applied to this patient population, we did not target high intensities until the second half of the program. Thus, the target power output (W) was gradually and individually increased during the program as follows: weeks 1–2, 65% of peak power output (PPO) reached during the baseline cycle-ergometer test until exhaustion (see hereinafter); weeks 3–4, 75% of PPO; weeks 5–6, 85% of PPO; and weeks 7–8, 90%–100% of PPO. The load for recovery periods was kept less than 65% of PPO.

Resistance training

This followed the aerobic training and included a three-time circuit of exercises involving large muscle groups and performed with specific weight training equipment (Gervasport, Pleven, Bulgaria) in the following order: leg and bench press, leg extension, lateral pull-down, and abdominal crunches. Initial loads for each exercise were individually determined, using values of 6–7 in the Borg 1–10 scale of perceived exertion. The rate of load increases was also set individually as follows: when the patient reported a score of <6 for a given exercise in two consecutive sessions, the corresponding load was increased (with the premise that the score for the new load was 6–7). The number of repetitions for each exercise was decreased with corresponding increases in load (kg) as follows: 15 (weeks 1–4) and 7–8 repetitions (weeks 5–8).

Inspiratory muscle training

This included two daily sessions (morning and evening, all at the patient’s home except for Monday–Wednesday–Friday, where the evening session was performed inhospital after the aforementioned resistance training), 6 d·wk−1 (Monday–Saturday). Each of the two daily sessions consisted of 30 inspirations through a specific pressure-load device (Powerbreathe® Classic Medium Resistance; Powerbreathe International Ltd, Southam, UK) against 40% of PImax (session duration ~5 min) (19). Participants’ PImax was reassessed by us at the beginning of each week to adjust the weekly load accordingly.

End Point Assessment

All assessments were consistently performed over a 1-wk period in the following days and order at each of pretraining, posttraining, and detraining, respectively, after familiarization with the equipment and tests (Thursday and Friday): first assessment day (Monday of the next week), venous blood drawing, cycle-ergometer test followed by postexercise blood drawing, and distribution of QoL questionnaires; second day (Wednesday), strength tests followed by functional tests of ADL and PImax test; and third day (Thursday or Friday), body composition analysis. To rule out a potential effect of changes in patients’ levels of habitual physical activity (PA) on study outcomes, during the week before the aforementioned assessments at pretraining, posttraining, and detraining, we determined PA objectively, using a triaxial accelerometer (GT3X monitor device; Actigraph, Pensacola, FL) (21). For each study participant, a minimum of 5 d of monitoring (including 2 weekend days) and a minimum of 10 h·d−1 of complete accelerometry were considered necessary. Data were analyzed using ActiLife5 LITE software (Actigraph). We determined average intensity (counts per minute) by dividing the sum of the total counts per predefined epoch (15 s) for a valid day by the number of minutes of wear time in that day across all valid days. Thereafter, counts were converted to average daily time engaged in moderate-vigorous PA (21).

Main End Points

Aerobic power

Aerobic exercise testing was performed on a cycle ergometer, following a ramp-like protocol (workload increases of 1 W every 6 s (averaging 10 W·min−1) starting from an initial load of 0 W, with a pedal cadence of 60–70 rpm throughout the test). Gas-exchange variables were collected breath-by-breath with an automated metabolic cart (Quark CPET; COSMED, Rome, Italy). The V˙O2peak and the peak pulmonary ventilation rate (V˙Epeak) were computed as the highest value of these two variables obtained for any 20-s period during the tests. Two important physiological indicators of submaximal aerobic power, VT and respiratory compensation point (RCP), were detected using previously defined criteria (22).

Muscle strength/power

We assessed upper/lower-body muscle power with bench/leg press tests using the same equipment as for training, connected to a linear encoder (Power Encoder; Smartcoach Europe, Stockholm, Sweden). This methodology has proved valid to determine muscle force (N) and power (W) (23) and has been recently applied by us for assessing patients with different types of myopathy (19,24). Patients performed one set of three repetitions for each exercise at the maximum possible speed; each set was followed by a recovery period of 60–90 s. The load (or “resistance”) was increased by 5 or 10 kg (leg press) and by 2 kg (bench press) in each successive set. Because power is the product of force and velocity, it initially increases with resistance and then decreases when the resistance causes a substantial decrease in velocity. Thus, the test is stopped when the decrease in velocity is so pronounced that it also causes a decrease in average muscle concentric power. An advantage of this test is that it ends with a submaximal load, thereby avoiding major/haemodynamic problems. We recorded the highest value of average power (W) and load (kg) in the concentric–propulsive phase, which typically coincides with the start of a decline in this variable together with the occurrence of the highest value of average force (N) (19,24).

Maximal inspiratory capacity

We measured participants’ PImax at the residual volume using a mouth pressure metre (Micro Medical Inc, Chatham, Kent, UK) (19).

Secondary End Points

Functional capacity during ADL

Walking capacity was evaluated with the 6-min walking distance (6MWD) test in accordance with established standards (19). Mobility was assessed with the timed up and go test (TUG) and 15-step stair tests (25).

Body composition

Total and regional body composition was assessed by dual-energy x-ray absorptiometry (DXA; Hologic Serie Discovery QDR, Physician’s Viewer, APEX System Software Version 3.1.2. Bedford, MA) as described (26). An additional examination was conducted to estimate bone mass and femoral T-score (number of SD above or below the young adult mean bone mineral density) at the proximal region of the femur. Femoral T-scores were calculated using the manufacturer’s reference values based on the revised National Health and Nutrition Examination Survey III reference data (26). Patients were also compared with a sex- and age-matched healthy control group (21 men and 12 women; age, 45.7 ± 2.0 yr) previously assessed by us with the same equipment.


We used the Spanish version of the SF-36 questionnaire (version 2) to assess patients’ health-related QoL (27).

Blood variables

Twenty myokines/cytokines were assessed in plasma samples before (“basal”) and after aerobic power tests using the MILLIPLEX® MAP Human High Sensitivity T Cell Magnetic Bead Panel and the MILLIPLEX® MAP Human Myokine Magnetic Bead Panel with Luminex® Technology on a MAGPIXTM instrument (EMD Millipore Corporation, Billerica, MA). Basal plasma levels of creatine kinase (CK; laboratory reference values, <190 U·L−1 (men) and <170 U·L−1 (women)), creatinine (reference, <1.2 mg·dL−1 (men) and <0.9 mg·dL−1 (women)) and lactate (reference, 0.5–2.2 mmol·L−1) were also measured using a Cobas C spectrophotometer (Cobas® 8000 modular analyser series; Roche Diagnostics, S.L., Barcelona, Spain).

Statistical Approach

We compared end point values over time using the nonparametric Friedman test. We also compared acute changes (preexertion vs postexertion) in myokines at each of the three study time points using the Wilcoxon test. Finally, DXA variables were compared between patients and controls using the Mann–Whitney U test. To minimize the risk of statistical error type I (given the large number of end points), analyses were corrected for multiple comparisons using the stringent Bonferroni method in which the threshold P value is obtained by dividing 0.05 by the number of comparisons; furthermore, post hoc paired comparisons were performed only for those end points showing a significant time effect.


Valid accelerometry data in each of the three time points were obtained from seven patients. Mean (±SEM) levels of moderate-vigorous PA did not differ (P = 0.8426) between pretraining (56 ± 20 min·wk−1), posttraining (56 ± 21 min·wk−1), and detraining (45 ± 29 min·wk−1). Vigorous PA was virtually absent in all individuals in the three points of assessment.

Adherence and safety

Mean adherence to the program was 94% ± 5% (range, 83%–100%). Main reasons for missing 1+ sessions were family (n = 5 sessions) or work (n = 4) obligations, medical visits (n = 1), or minor health issues temporarily disabling exercise (n = 7; e.g., upper respiratory tract infection and migraine). No health issues other than the expected muscle discomfort were noted. In two subjects, the 15 repetitions that were originally planned for strength exercises during weeks 1–4 were divided into two sets of 7–8 repetitions to attenuate muscle pain.

Main end points

Peak values of heart rate (161 ± 6, 164 ± 6, and 161 ± 7 bpm at pretraining, posttraining, and detraining, respectively; P = 0.242 for time effect) and respiratory exchange ratio (1.18 ± 0.04, 1.18 ± 0.02, and 1.23 ± 0.03; P = 0.368) reached by the participants in the aerobic power tests did not differ over time, which, besides indicating a comparable level of effort, suggests that patients performed a maximal effort in the three assessments. A significant time effect was found for all end points indicative of aerobic power, muscle strength, and inspiratory muscle power (all P ≤ 0.004) except for V˙O2 at the RCP relative to muscle mass (P = 0.006), indicating a consistent training improvement (Table 2). Statistical significance was also reached for numerous end points in paired post hoc comparisons of pretraining vs posttraining values. Importantly, some training improvements were not completely lost after detraining, with several values at detraining remaining significantly higher than those at pretraining (P ≤ 0.004). All main end points showed the same individual response (i.e., training-induced improvement) for virtually all patients (see Figure, Supplemental Digital Content 1, Individual data for patients’ main end points,

Results of patients’ main end points by training status.

Secondary end points

A significant time effect (P ≤ 0.003) was found for all outcomes indicative of functional capacity in ADL, total/trunk and leg muscle mass, total fat mass, femoral fracture risk (T-score), and general health perception in the SF-36 questionnaire. Similar to the main outcomes, this indicated a training-induced improvement (which for fat mass and T-score was reflected in a decrease in mean values) followed by a subsequent loss, at least partially, of such improvements after detraining (Table 3). In post hoc analyses, we found a significant improvement at posttraining compared with pretraining for all the functional tests as well as for total lean and leg muscle mass (all P ≤ 0.003). Furthermore, detraining values of 6MWD and 15-step stair tests, and total lean mass remained significantly higher than pretraining values (P ≤ 0.003). For ADL tests and DXA variables, virtually all patients showed the same individual response (i.e., training-induced improvement; see Figure, Supplemental Digital Content 2, Individual data for patients’ secondary end points, Furthermore, lean mass was significantly lower in patients than in controls (see Table, Supplemental Digital Content 3, Results of dual-energy x-ray absorptiometry in mitochondrial disease patients and age- and gender-matched healthy controls,

Results of patients’ secondary end points by training status.

No significant time effect was found for the basal levels of blood variables (all P > 0.002; Table 4). A significant acute exertional increase was found only for interleukin 8 (IL-8) at posttraining and detraining (both P = 0.002) and for fatty acid binding protein 3 (FABP3) at detraining (P = 0.002; Table 4).

Results of blood variables by patients’ training status.


The main finding of our study was that, despite its relatively short duration (8 wk), a “complete” and innovative exercise intervention including aerobic, resistance, and specific inspiratory muscle exercises performed by patients with mitochondrial disorders produced significant benefits in numerous indicators of their physical capacity, including aerobic power, muscle strength, and inspiratory muscle power. It also improved patients’ ability to manage ADL and their general health perception. The training program also induced a shift toward a healthier body composition phenotype, with increased and decreased muscle and fat mass, respectively. Importantly, despite the heterogeneity in clinical profiles, these improvements were corroborated in virtually all patients individually. Furthermore, many of the training improvements were partly retained after detraining. Our intervention was also safe and well tolerated by the patients: no major health issues, no significant increases in muscle damage (CK levels), and no alterations in renal function (creatinine) were found. To the best of our knowledge, this is the first study to use such an “integrative” training approach and to show benefits in essentially all of the functional end points we tested, including also improvements in important novel outcomes, muscle mass, and bone fracture risk.

Peak aerobic power levels are usually poor in mitochondrial disorders (1,4–6). The V˙O2peak levels varied widely among our patient cohort, which is likely attributable to the variability in the magnitude of OXPHOS impairment in this patient group (1). Nevertheless, before starting the training program, the majority (75%) of patients had a V˙O2peak level of <25 mL·kg−1·min−1 (or 7 METs, where 1 MET reflects resting metabolic rate or 3.5 mL·kg−1·min−1 for most humans). The finding that mean V˙O2peak approached 8 MET is clinically relevant because this is the minimum threshold for optimal health, above which the risk for all-cause mortality is significantly reduced (28).

Importantly, V˙O2peak also increased significantly after training when normalized to leg muscle mass, which was most involved during the exercise mode of leg pedaling used here both for aerobic power assessment and aerobic training. This suggests an improved ability of the trained muscle tissue to extract and use O2 during exercise, which is considered a major limiting factor of the V˙O2peak that can be reached by patients (5,11,29). Our hypothesis of increased patient muscle aerobic power after training is consistent with previous studies reporting improved endothelial function, mitochondrial biogenesis, and oxidative capacity (6–11,30), as well as increased peak oxygen extraction (as assessed by near-infrared spectroscopy) in the muscle tissue of trained patients (12). Unfortunately, we were not authorized by our ethics committee to collect muscle biopsies from the patients. Our observations of significant improvements in V˙O2peak through training are in agreement with previous intervention studies of similar or longer duration that have focused on aerobic exercise (5–12). The success of our training program despite its relatively short duration might be attributable to the low-volume HIT protocol we used, at least for the second half of the intervention, where participants performed ~10 repetitions of 1-min intervals per session, at ~80% (weeks 5–6) and 90%–100% of peak watts achieved during the baseline cycle-ergometer test (weeks 7–8). Indeed, growing evidence suggests that this type of training stimulates physiological adaptations (e.g., V˙O2peak increases) comparable with moderate-intensity continuous training despite a lower time commitment and exercise volume (20,31). Our findings are also important in practical terms given that “lack of time” remains the most commonly cited barrier to regular exercise participation among Westerners (20). By contrast, previous studies have used interventions of continuous exercise during longer bouts but at lower intensities, such as 30 min at 70% of peak watts (13), 30–45 min at 65%–80% of maximal heart rate (10–12), or 30 min at 70%–80% of V˙O2peak (5–7). Of note, here we chose to use the term “V˙O2peak” instead of maximal oxygen uptake (“V˙O2max”). A second, constant work rate test performed at ~110% of the work rate achieved on the initial ramp test would have served as a validation tool to assess actual V˙O2max levels (32). This validation method might be incorporated for the assessment of mitochondrial disease patients in future training studies.

Patients with mitochondrial disorders often show a deteriorated maximal inspiratory capacity, as manifested by a low PImax (17). In this regard, the mean baseline values of PImax in our patients (79 ± 11 cm H2O) were only slightly (+3.8%) higher than those we recently reported using the same method in patients of a very similar mean age (~46 vs 47 yr here) suffering from a devastating disease, pulmonary arterial hypertension, which typically provokes marked exercise intolerance together with frequent dyspnea and also myopathy affecting ventilatory muscles (19). However, not only PImax (+29%, P < 0.001) but also another indicator of muscle ventilatory capacity, V˙Epeak, showed remarkable increases with training (+25%, P < 0.004). Two patients (nos. 2 and 12; Table 1) required noninvasive nocturnal ventilatory support (with bilevel positive airway pressure, or biPAP). Although after the training program, they both reported feeling less ventilation limited during daily living, the biPAP was still kept at night for safety purposes. Furthermore, any improvement in respiratory muscle capacity as the one shown here is of potential medical relevance because in the long term, respiratory failure is a main cause of death among adults with mitochondrial disorders (33).

Patients with mitochondrial disorders have high blood lactate levels at rest and especially during exercise, even of submaximal intensity, as a consequence of OXPHOS dysfunction (1). In this regard, we also found a significant training-induced improvement in two important indicators of submaximal exercise capacity, the VT (also referred to as the “anaerobic threshold”) and the RCP (also known as “second VT”). Previous studies have also found an increase in the VT with training in this patient group (5,13), but ours is the first study to assess training effects in the RCP. Although the VT represents the first increase in pulmonary ventilation that is proportional to the increase in CO2 output generated by the HCO3 buffering of lactic acidosis, the RCP represents a higher work intensity at which blood lactate accumulation increases considerably (i.e., production exceeds clearance) and is accompanied by an additional hyperventilation in an attempt to buffer increasing blood acidosis (22). Therefore, the increments in VT and RCP provoked here were likely mediated by (i) a higher ability to buffer lactic acidosis and (ii) a higher ability of working muscles to rely on aerobic metabolism, as reflected by the finding that V˙O2peak normalized to leg muscle mass increased with training.

We used a novel, practical approach to assess the effects of strength training on the ability of muscles to produce power by studying the force–velocity curve, allowing the patients to avoid performing maximal efforts and the Valsalva maneuver and therefore limiting the risk of adverse events (as opposed to “classical one-repetition maximum” weight-lifting exercises). Moreover, compared with conventional one-repetition maximum tests, muscle power improvements have more direct transference into patients’ ability to perform ADL (e.g., standing from a chair and stair climbing) (19). This was indeed reflected in the training-induced improvement we found for all outcomes indicative of functional capacity during ADL (6MWD, TUG, and 15-step stair tests) in our patients, demonstrating that ADL was better tolerated after training.

The increments in strength/muscle power levels were accompanied by improvements in muscle mass measured by DXA. No study has assessed exercise training effects on the muscle mass in this patient group; yet, it is a potentially important issue because our patients showed significantly lower total lean mass (~18%) than did their healthy peers. Thus, the training improvement in total lean mass is of special relevance, particularly the higher values at the lower extremities. This training effect remained after the detraining period. The phenotype of patients’ muscle tissue was likely changed with resistance training toward hypertrophy. A previous study has reported that strength training increases the number of satellite cells (as a result of increased muscle damage) and of intermediate muscle fibers, while decreasing the number of cytochrome c oxidase–deficient fibers (14). Although the mechanisms involved in this process remain to be clearly elucidated, Murphy et al. (14) proposed that the activation of quiescent mitotic cells by strength training results in the shifting of normal mitochondrial templates to mature muscle and restoration of a normal mitochondrial genotype and phenotype. However, a similar proportion of mitochondrial DNA (mtDNA) mutations (single large-scale deletions) has been reported in satellite cells and mature muscle of patients with mitochondrial disorders (34). Unfortunately, the fact that we did not perform muscle biopsies precluded assessing a potential training effect on mtDNA mutational load at the muscle tissue level.

Training-induced gains in muscle mass were also accompanied by a reduction in fat mass and consequently a healthier body composition phenotype. While keeping in mind the limitation that we did not assess potential dietary effects on the body composition changes observed during the study, this is also an important finding given that many of these patients are at high risk for metabolic conditions such as obesity and type 2 diabetes (16,35–37), or impaired glucose tolerance and diminished insulin-stimulated glucose influx to skeletal muscle and adipose tissue (36). An additional novelty of our study was the finding that exercise training decreased the risk for femoral fractures (T-score), which is also clinically relevant in light of the fact that mitochondrial disorders are often associated with endocrine conditions that can adversely affect bone mass (diabetes mellitus, insulin resistance/glucose intolerance, hypoparathyroidism, renal tubular acidosis, chronic renal insufficiency, and gastrointestinal disease) (38). Interestingly, one of our patients was reclassified from osteoporotic to osteopenic after the training period (T-score of −2.5 and −2.3, respectively). The finding that seven patients were classified as having osteopenia or osteoporosis at the beginning of the study and the proposed link between mitochondrial dysfunction and a bone phenotype (38) further support the need to implement exercise interventions such as those used here in this patient group.

Skeletal muscle can act as an endocrine organ, especially during contractions, by releasing molecules that are collectively known as “myokines” (mainly proteins and peptides) into the bloodstream (15). Myokines account for some of the beneficial effects of exercise, such as anti-inflammatory effects, or healthy metabolic adaptations (e.g., decreased insulin resistance). In this regard, patients with mitochondrial disorders can present high basal concentrations of several proinflammatory cytokines (such as tumor necrosis factor α) (36). Accordingly, we evaluated the effects of exercise training on basal myokine levels and found no statistically significant differences. By contrast, a significant acute increase in IL-8 was found at posttraining and detraining, and FABP3 at detraining. Previous research on healthy adults has reported an increase in circulating levels of IL-8 after an exercise bout (39). Similarly, IL-8 receptor levels are elevated in the working skeletal muscles of healthy individuals (15) and in patients with mitochondrial disorders after an acute exercise bout (36), with IL-8 receptor–mediated signaling potentially stimulating local angiogenesis (15). A stimulating effect of IL-8 on angiogenesis after acute exercise might be mediated by CXC receptor 2 signaling, which in the longer term may increase muscle capillarity (15). In a previous study, capillary density increased after 12-month aerobic training in four patients with different mtDNA mutations and high mtDNA mutational load (7), although the underlying mechanism remains unknown. However, the same authors found a training-induced decrease in capillary density after 12-wk aerobic training (6). Clearly, more research is needed to examine the effects of exercise on muscle capillarization in patients with mitochondrial disorders and to determine whether IL-8–related pathways are involved. Our data on FABP3 are in agreement with a previous study in healthy individuals where a single bout of endurance cycle-ergometer exercise provoked an increase in FABP3 expression in vastus lateralis muscle (40). FABP3 facilitates the cytoplasmic transport of fatty acids through intracellular membranes and provides the primary means for energy production in working skeletal muscle together with carbohydrates (40). Future studies might determine if longer exercise training interventions are needed to elicit significant changes in the myokine blood profile.

Besides unavailability of patients’ muscle biopsies due to ethical constraints, a main limitation of our study was that we did not perform a randomized controlled trial on patients. This is partly compensated by the fact that we assessed patients at three time points (including detraining), thereby preventing subject bias as the same subjects were used as their own controls. In addition, the exercise intervention did not have a beneficial effect on patients’ moderate-vigorous PA levels after it had ended (i.e., detraining period). Nevertheless, our study has several strengths. It is the first to use an “integrative” training approach in this patient group. Other strengths were as follows: the myriad benefits obtained, and the novelty of several end points (such as bone and muscle mass) and of our training methods, including low-volume HIT, “specific” resistance training, and inspiratory muscle exercises. Concerning the latter method, this was easy to perform and little time consuming, and it induced a significant improvement in inspiratory muscle capacity. Finally, we minimized type I statistical error by adjustment for multiple comparisons.

In conclusion, an 8-wk “complete” exercise intervention produced benefits in numerous indicators of physical capacity in patients with mitochondrial disorders, as well as a shift toward a healthier body composition phenotype, with increased and decreased muscle and fat mass, respectively.

We acknowledge POWERbreath International Ltd., through BIOCORP EUROPA S.L./POWERbreath Spain, for the kind support with equipment for inspiratory muscle training, as well as the blood bank of Hospital 12 de Octubre for facilitating patients’ blood drawing.

This study was funded by Fondo de Investigaciones Sanitarias (grant numbers PI14/01085, PI17/00093, PI15/00431, and PI15/00558), and FEDER funds from the European Union. M. M. is supported by Miguel Servet contracts (CPII16/00023) from the Institute of Health Carlos III (ISCIII) and FEDER funds. C. F.-L. is supported by Sara Borrell contract (CD14/00005) from the Institute of Health Carlos III.

The authors report no conflict of interest and affirm that the results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation. The results of the present study do not constitute endorsement by the American College of Sports Medicine.


1. Taivassalo T, Jensen TD, Kennaway N, DiMauro S, Vissing J, Haller RG. The spectrum of exercise tolerance in mitochondrial myopathies: a study of 40 patients. Brain. 2003;126(Pt 2):413–23.
2. Fernández J, Montemayor T, Bautista J, et al. The use of cardiopulmonary exercise test in patients with mitochondrial myopathies. Med Clin (Barc). 2000;114(4):121–7.
3. Grassi B, Marzorati M, Lanfranconi F, et al. Impaired oxygen extraction in metabolic myopathies: detection and quantification by near-infrared spectroscopy. Muscle Nerve. 2007;35(4):510–20.
4. Porcelli S, Marzorati M, Belletti M, Bellistri G, Morandi L, Grassi B. The “second wind” in McArdle’s disease patients during a second bout of constant work rate submaximal exercise. J Appl Physiol (1985). 2014;116(9):1230–7.
5. Bates MG, Newman JH, Jakovljevic DG, et al. Defining cardiac adaptations and safety of endurance training in patients with m.3243A > G-related mitochondrial disease. Int J Cardiol. 2013;168(4):3599–608.
6. Jeppesen TD, Schwartz M, Olsen DB, et al. Aerobic training is safe and improves exercise capacity in patients with mitochondrial myopathy. Brain. 2006;129(Pt 12):3402–12.
7. Jeppesen TD, Dunø M, Schwartz M, et al. Short- and long-term effects of endurance training in patients with mitochondrial myopathy. Eur J Neurol. 2009;16(12):1336–9.
8. Taivassalo T, De Stefano N, Argov Z, et al. Effects of aerobic training in patients with mitochondrial myopathies. Neurology. 1998;50(4):1055–60.
9. Taivassalo T, De Stefano N, Chen J, Karpati G, Arnold DL, Argov Z. Short-term aerobic training response in chronic myopathies. Muscle Nerve. 1999;22(9):1239–43.
10. Taivassalo T, Shoubridge EA, Chen J, et al. Aerobic conditioning in patients with mitochondrial myopathies: physiological, biochemical, and genetic effects. Ann Neurol. 2001;50(2):133–41.
11. Taivassalo T, Gardner JL, Taylor RW, et al. Endurance training and detraining in mitochondrial myopathies due to single large-scale mtDNA deletions. Brain. 2006;129(Pt 12):3391–401.
12. Porcelli S, Marzorati M, Morandi L, Grassi B. Home-based aerobic exercise training improves skeletal muscle oxidative metabolism in patients with metabolic myopathies. J Appl Physiol (1985). 2016;121(3):699–708.
13. Cejudo P, Bautista J, Montemayor T, et al. Exercise training in mitochondrial myopathy: a randomized controlled trial. Muscle Nerve. 2005;32(3):342–50.
14. Murphy JL, Blakely EL, Schaefer AM, et al. Resistance training in patients with single, large-scale deletions of mitochondrial DNA. Brain. 2008;131(Pt 11):2832–40.
15. Fiuza-Luces C, Garatachea N, Berger NA, Lucia A. Exercise is the real polypill. Physiology (Bethesda). 2013;28(5):330–58.
16. Apabhai S, Gorman GS, Sutton L, et al. Habitual physical activity in mitochondrial disease. PLoS One. 2011;6(7):e22294.
17. Flaherty KR, Wald J, Weisman IM, et al. Unexplained exertional limitation: characterization of patients with a mitochondrial myopathy. Am J Respir Crit Care Med. 2001;164(3):425–32.
18. Parikh S, Goldstein A, Koenig MK, et al. Diagnosis and management of mitochondrial disease: a consensus statement from the Mitochondrial Medicine Society. Genet Med. 2015;17(9):689–701.
19. González-Saiz L, Fiuza-Luces C, Sanchis-Gomar F, et al. Benefits of skeletal-muscle exercise training in pulmonary arterial hypertension: The WHOLEi + 12 trial. Int J Cardiol. 2017;231:277–83.
20. Gibala MJ, Little JP, Macdonald MJ, Hawley JA. Physiological adaptations to low-volume, high-intensity interval training in health and disease. J Physiol. 2012;590(5):1077–84.
21. Santos-Lozano A, Santín-Medeiros F, Cardon G, et al. Actigraph GT3X: validation and determination of physical activity intensity cut points. Int J Sports Med. 2013;34(11):975–82.
22. Lucía A, Hoyos J, Pérez M, Chicharro JL. Heart rate and performance parameters in elite cyclists: a longitudinal study. Med Sci Sports Exerc. 2000;32(10):1777–82.
23. González-Badillo JJ, Sánchez-Medina L. Movement velocity as a measure of loading intensity in resistance training. Int J Sports Med. 2010;31(5):347–52.
24. Santalla A, Munguía-Izquierdo D, Brea-Alejo L, et al. Feasibility of resistance training in adult McArdle patients: clinical outcomes and muscle strength and mass benefits. Front Aging Neurosci. 2014;6:334.
25. Bennell K, Dobson F, Hinman R. Measures of physical performance assessments: Self-Paced Walk Test (SPWT), Stair Climb Test (SCT), Six-Minute Walk Test (6MWT), Chair Stand Test (CST), Timed Up & Go (TUG), Sock Test, Lift and Carry Test (LCT), and Car Task. Arthritis Care Res (Hoboken). 2011;63(1 Suppl):S350–70.
26. Binkley N, Kiebzak GM, Lewiecki EM, et al. Recalculation of the NHANES database SD improves T-score agreement and reduces osteoporosis prevalence. J Bone Miner Res. 2005;20(2):195–201.
27. Vilagut G, Valderas JM, Ferrer M, Garin O, López-García E, Alonso J. [Interpretation of SF-36 and SF-12 questionnaires in Spain: physical and mental components]. Med Clin (Barc). 2008;130(19):726–35.
28. Kodama S, Saito K, Tanaka S, et al. Cardiorespiratory fitness as a quantitative predictor of all-cause mortality and cardiovascular events in healthy men and women: a meta-analysis. JAMA. 2009;301(19):2024–35.
29. Taivassalo T, Abbott A, Wyrick P, Haller RG. Venous oxygen levels during aerobic forearm exercise: an index of impaired oxidative metabolism in mitochondrial myopathy. Ann Neurol. 2002;51(1):38–44.
30. Siciliano G, Manca ML, Renna M, Prontera C, Mercuri A, Murri L. Effects of aerobic training on lactate and catecholaminergic exercise responses in mitochondrial myopathies. Neuromuscul Disord. 2000;10(1):40–5.
31. Gibala MJ, McGee SL. Metabolic adaptations to short-term high-intensity interval training: a little pain for a lot of gain? Exerc Sport Sci Rev. 2008;36(2):58–63.
32. Poole DC, Jones AM. Measurement of the maximum oxygen uptake V˙o(2max): V˙o(2peak) is no longer acceptable. J Appl Physiol (1985). 2017;122(4):997–1002.
33. Barends M, Verschuren L, Morava E, Nesbitt V, Turnbull D, McFarland R. Causes of Death in adults with mitochondrial disease. JIMD Rep. 2016;26:103–13.
34. Spendiff S, Reza M, Murphy JL, et al. Mitochondrial DNA deletions in muscle satellite cells: implications for therapies. Hum Mol Genet. 2013;22(23):4739–47.
35. Schaefer AM, Walker M, Turnbull DM, Taylor RW. Endocrine disorders in mitochondrial disease. Mol Cell Endocrinol. 2013;379(1–2):2–11.
36. Rue N, Vissing J, Galbo H. Insulin resistance and increased muscle cytokine levels in patients with mitochondrial myopathy. J Clin Endocrinol Metab. 2014;99(10):3757–65.
37. Gandhi SS, Muraresku C, McCormick EM, Falk MJ, McCormack SE. Risk factors for poor bone health in primary mitochondrial disease. J Inherit Metab Dis. 2017;40(5):673–83.
38. Langdahl JH, Frederiksen AL, Hansen SJ, et al. Mitochondrial point mutation m.3243A > G associates with lower bone mineral density, thinner cortices and reduced bone strength. A case–control study. J Bone Miner Res. 2017;32(10):2041–48.
39. Chan MH, Carey AL, Watt MJ, Febbraio MA. Cytokine gene expression in human skeletal muscle during concentric contraction: evidence that IL-8, like IL-6, is influenced by glycogen availability. Am J Physiol Regul Integr Comp Physiol. 2004;287(2):R322–7.
40. Schmutz S, Däpp C, Wittwer M, Vogt M, Hoppeler H, Flück M. Endurance training modulates the muscular transcriptome response to acute exercise. Pflugers Arch. 2006;451(5):678–87.


Supplemental Digital Content

Copyright © 2018 by the American College of Sports Medicine