American Journal of Respiratory and Critical Care Medicine

Exercise intolerance is a common complaint, the cause of which often remains elusive after a comprehensive evaluation. In this report, we describe 28 patients with unexplained dyspnea or exertional limitation secondary to biopsy-proven mitochondrial myopathies. Patients were prospectively identified from a multidisciplinary dyspnea clinic at a tertiary referral center. All patients were without underlying pulmonary, cardiac, or other neuromuscular disorders. Patients underwent history, physical examination, complete pulmonary function testing, respiratory muscle testing, cardiopulmonary exercise testing, and muscle biopsy. Results were compared with a group of normal control subjects. The estimated period prevalence was 8.5% (28 of 331). Spirometry, lung volumes, and gas exchange were normal in patients and control subjects. Compared with control subjects, the patient group demonstrated decreased exercise capacity (maximum achieved V˙ o 2 67 versus 104% predicted; p < 0.0001) and respiratory muscle weakness (Pi max 77 versus 115% predicted; p = 0.001). These patients have a characteristic exercise response that was hyperventilatory (peak Ve/V˙ co 2; 55 versus 42) and hypercirculatory (maximum heart rate − baseline heart rate/V˙ o 2max − baseline V˙ o 2max; 91 versus 41) compared to control subjects. Patients stopping exercise due to dyspnea (n = 16) (as compared with muscle fatigue, n = 11) displayed weaker respiratory muscles (Pdimax 61 versus 115 cm H2O; p = 0.01) and were more likely to reach mechanical ventilatory limitation (V˙ emax/ MVV 0.81 versus 0.58; p = 0.02). The sensation of dyspnea was related to indices of respiratory muscle function including respiratory rate and inspiratory flow. We conclude that mitochondrial myopathies are more prevalent than previously reported. The characteristic physiological profile may be useful in the diagnostic evaluation of mitochondrial myopathy.

Keywords: metabolic myopathy; dyspnea; mitochondrial myopathy

Normal exercise requires the adequate delivery of oxygen and nutrients to the muscle and requires that the muscle be capable of utilizing oxygen to metabolize the nutrients and generate energy necessary to perform work. Dysfunction in any of these complex processes can be responsible for exercise intolerance resulting in the perception of abnormal fatigue or dyspnea. Dyspnea during exertion, a common complaint, is usually associated with a cardiac or pulmonary cause. Other etiologies such as deconditioning, obesity, malignancy, anemia, and sleep apnea are described less frequently (1-4), and often the cause remains elusive after an aggressive evaluation (3). Metabolic disorders of muscle have recently been considered a potential diagnostic category in a subset of these patients (5).

Metabolic myopathies are disorders of muscle energy metabolism. These disorders have been grouped into three broad categories: defective carbohydrate utilization, abnormal lipid utilization, and mitochondrial myopathies (6). The term mitochondrial myopathy refers to various syndromes with diverse pathological, histochemical, and biochemical characteristics. These syndromes are often multisystemic with varying signs and symptoms affecting any organ system (7, 8). The final pathogenesis of the different syndromes is a decline in mitochondrial adenosine triphosphate (ATP)-generating capacity leading to a deficit in energy production (9). Exercise intolerance is one manifestation described in some patients (6, 9). As muscle requires oxidative phosphorylation for production of adenosine triphosphate (ATP), mitochondrial dysfunction can also produce muscular symptoms such as myalgia or weakness. The prevalence of mitochondrial myopathies has been estimated at 1 in 50,000 from patients referred to a neurology practice (7), although the prevalence of mitochondrial myopathies presenting as unexplained exercise intolerance, particularly exertional breathlessness, remains unclear (10).

In the present work we describe the prevalence of biopsy-proven mitochondrial myopathy in a group of patients evaluated in a multidisciplinary dyspnea clinic for unexplained exertional dyspnea or exertional limitation. In addition, we provide detailed physiological testing in this cohort of patients and shed insight on the mechanisms of exertional breathlessness in these patients.

Patient Recruitment

Patients were prospectively identified among those referred to the University of Michigan Dyspnea Clinic during a period from November 1, 1994 to November 1, 1998. Only patients with a referral diagnosis of unexplained exertional dyspnea or limitation were included. Patients were referred from a wide variety of physicians and, as such, the evaluation prior to referral varied. The Dyspnea Clinic is a multidisciplinary clinic with pulmonary, cardiology, and neurology physicians. Those patients with exertional limitation, including myalgia, fatigue, and/or dyspnea, without cardiopulmonary or other explanation, were evaluated as described below. A final diagnosis required muscle biopsy confirmation (see below). Patients were excluded if they had airflow obstruction (FEV1/FVC < 70%), parenchymal (interstitial) lung disease, neuromuscular disease, or cardiovascular disease sufficient to account for exertional intolerance. The latter included pulmonary hypertension, valvular disease, ischemic heart disease, or cardiomyopathy associated with impaired right or left ventricular function.

Control Recruitment

Control subjects with normal health were recruited from the staff at the University of Michigan Health System. All subjects volunteered and were not paid for their participation. Potential subjects underwent a history and physical examination to exclude comorbidity. The protocol was approved by the Institutional Review Board of the University of Michigan Health System.

Physiological Assessment

Pulmonary function testing. Pulmonary function studies, including spirometry and lung volumes, were performed on the same day but prior to a cardiopulmonary exercise test (CPET). All spirometric studies were performed on a pulmonary function system with a calibrated pneumotachograph (Medical Graphics Co., St. Paul, MN), and values were expressed as a percent of the predicted values published by Morris and coworkers (11). Lung volumes were measured in a body plethysmograph (Elite; Medical Graphics Co., St. Paul, MN) and the data were expressed as a percent of the predicted values published by Goldman and Becklake (12). Diffusing capacity for carbon monoxide was measured by single breath technique using normal values of Miller and coworkers (13). Maximum voluntary ventilation (MVV) was measured using a 12-s maneuver. The product of FEV1 × 40 was also calculated as a surrogate measure of maximum ventilatory capacity.

Exercise testing. Exercise testing was performed on an electronically braked, calibrated cycle ergometer at a time of clinical stability. Before exercise, patients and control subjects received identical instructions to continue exercise until they were exhausted. No form of encouragement was offered during exercise to avoid variability due to different levels of encouragement offered by different technicians. The initial 2 min consisted of resting data collection followed by 1 min of unloaded cycling. Subsequently, workload was increased by 20 W/min until maximal symptom-limited exercise was achieved. Expired gases and ventilation were measured on a metabolic cart that uses a pneumotachograph positioned at the mouth with O2 and CO2 analyzers (Collins CPXII; Warren E. Collins, Inc., Braintree, MA). This allowed breath-by-breath measures of oxygen consumption (V˙o 2), carbon dioxide production (V˙co 2), minute ventilation (V˙e), respiratory rate (fb), maximal inspiratory flow (Vi), and tidal volume (Vt). The continuous output of the automated system was recorded and displayed on an on-line PC computer where all data were saved for later analysis. The system was calibrated to ensure an appropriate phase response. The predicted values of Hansen and coworkers were used for the exercise measurements (14). In all patients, heart rate, heart rhythm, blood pressure, and oxygen saturation were continuously monitored. In addition, full 12-lead electrocardiograms were monitored during each minute of exercise and recovery. CPET results were interpreted by four experienced observers (F.J.M., J.Z., I.W., and K.F.) to ensure agreement in determination of cardiopulmonary response to exercise. Anaerobic threshold (AT) was estimated using the nadirs of ventilatory equivalents and the V-slope method; both methods were used concurrently looking for consistency (15, 16). If AT was clearly discernible using either of the noninvasive methods, this value was reported. When differences in AT were observed between both techniques, the average value was used. However, in situations in which AT was not discernible using either method, the AT was categorized as indeterminate.

Respiratory muscle testing. In all patients and control subjects, maximal inspiratory mouth pressure (Pi max) was measured using the techniques of Black and Hyatt (17). The maneuvers were performed at rest and within 5 min of ending exercise. In all patients, maximal expiratory mouth pressure (Pe max) was also measured at rest using the techniques of Black and Hyatt (17) and compared with predicted values (18). Pe max was not routinely measured in control subjects.

In 14 patients and seven control subjects, pleural (Ppl) and gastric pressure (Pg) were measured using endoesophageal and gastric balloons, respectively. A separate transducer (Validyne Co., Northridge, CA) measured each pressure, and the calibrated output was continuously recorded on an on-line PC computer. Phase relationship between pressure and flow was within 5% up to 5 Hz and was subsequently corrected for any phase differences. Transdiaphragmatic pressure (Pdi) was calculated as the electronic subtraction of Pg and Ppl (Pdi = Pg − Ppl). Maximal Pdi (Pdimax) and Ppl (Pplmax) were measured at FRC by having the patient perform a maximal inspiratory effort against a partially occluded shutter. The patients were asked to maximally expand the chest and the abdomen and were coached in the performance of this maneuver until three reproducible results were obtained. Transdiaphragmatic measurements were also recorded during sharp, maximal sniffs (Pdisniff) (19). These maximal maneuvers were performed at rest and within 5 min of finishing exercise.


Sensation of breathlessness was measured in patients (n = 28) and control (n = 11) subjects at rest and during each minute of exercise using a 100-mm visual analog scale (VAS) (20). The extremes of this scale were defined as “not at all breathless” and “extreme breathlessness.” In 24 patients, a modified scale was additionally used to assess breathlessness during exercise. This scale consisted of a vertical line labeled 0 to 10, with verbal descriptors at fixed points on the scale. The patients were asked to concentrate on respiratory sensation. These tools have been extensively used to grade breathlessness during exercise (21).

Pathologic Assessment

The skin over the belly of the selected limb muscle was prepped and draped in the standard surgical fashion. After local anesthesia was obtained with 1% lidocaine, a linear incision was made in the skin and taken down sharply through the subcutaneous tissues and muscle fascia. Three specimens were taken of the selected muscle: in a muscle clamp, as a stretched specimen, and as a free piece. Fresh biopsies were flash frozen in isopentane cooled by liquid nitrogen. Blocks of frozen tissue were cut in 6-μm thick serial sections in cryostat and placed on cover slips. Besides hematoxylin and eosin (H&E) and modified Gomory trichrome, the sections were stained with a routine panel of histochemical reactions done for every muscle biopsy in our laboratory. Among these, NADH-tetrazolium reductase (NADH-TR) and succinic dehydrogenase (SDH) were utilized to stain mitochondria. The slides from each case were initially reviewed by a neuropathologist (M.B.) and subsequently by a group of neurologists specializing in neuromuscular disorders who regularly review their patients' muscle biopsies. To be included in this prospective study, all patients demonstrated abnormal mitochondrial staining on muscle biopsy; additional changes varied from subsarcolemmal mitochondrial proliferation to clumping of the mitochondria in the cytosol to classic “ragged red fibers” (22-24). Biochemical analysis confirmed mitochondrial abnormalities in the 26 patients analyzed.

Statistical Analysis

Descriptive/pulmonary function. Statistical comparison of qualitative variables between the patients (n = 28) and control (n = 11) subjects was completed using Chi square analysis and quantitative data were compared using Student's t tests. Variables for comparison were chosen prior to analysis, and no correction for multiple comparisons was utilized. The 28 patients were subsequently categorized by the predominant symptom limiting exercise during CPET. In this way, a group with breathlessness (n = 16) as the limiting symptom was identified and a second group of patients discontinued exercise because of muscle fatigue or pain (n = 11). One individual discontinued exercise due to syncope, and thus, quantification of symptoms at peak exercise was not completed. All quantitative, dyspnea, pulmonary function, respiratory muscle function, and exercise test data in these two patient groups were compared using an unpaired Student's t test. Quantitative data are reported as mean ± standard deviation of the mean.

Breathlessness. We analyzed the effect of absolute minute ventilation and V˙e/maximal voluntary ventilation (MVV) on breathlessness (as measured by VAS) using linear regression as described by our laboratory (25) and others (26). Single and multiple variable regression models to predict breathlessness (measured by VAS) were created. The final variables utilized were parameters of respiratory muscle function, including respiratory rate (fb), respiratory timing (inspiratory time, Ti, divided by total respiratory cycle duration, Ttot), tidal volume/forced vital capacity (Vt/FVC), maximum inspiratory flow (Vi), and pleural pressure/maximal pleural pressure (Ppl/Pplmax). These parameters have been previously shown to contribute to the intensity of breathlessness using a similar analysis (27). Each control subject and patient had VAS recordings at each minute of exercise. Regression models used all available data. Due to the fact that observations over time of exercise were correlated, mixed models permitting assessment of these correlations were used. These analyses were performed first for the patients only (n = 27) as categorized by type of discontinuance of exercise: predominantly because of dyspnea (n = 16) and for those with fatigue (n = 11). The analyses were then completed for all subjects as categorized into three groups: control subjects (n = 11), dyspneic patients (n = 16), and fatigued patients (n = 11).

Patient Population

Twenty-eight patients were identified who met inclusion/exclusion criteria. A summary of the demographic data for the patient group and the control subjects is presented in Table 1. The groups had a similar mean age, height, and weight. The duration of symptoms for the patients varied widely ranging from 2 mo to 60 yr. Thirteen muscle biopsies were performed for suspected mitochondrial biopsy during the study period and were negative. The diagnoses in these patients included myositis (n = 6), nonspecific Type II atrophy (n = 4), and normal muscle (n = 3).


CharacteristicPatientsControl Subjectsp Value
Age, yr36 ± 933 ± 40.36
Sex, m/f6/226/50.04
Weight, kg70 ± 1871 ± 90.85
Height, cm168 ± 9159 ± 530.36
Duration of symptoms,
 mean; yr6.8 ± 7.9NA

Definition of abbreviation: NA = not applicable.


During the period from November 1, 1994 to November 1, 1998, a total of 1,446 patients were referred to the Dyspnea Clinic at the University of Michigan Health System. In 331 (23%) of these patients, the referral diagnosis was unexplained exertional dyspnea or limitation. All patients in the current cohort were included in these 331 patients. From our experience, we estimate the period prevalence of mitochondrial myopathies as 28 of 331 or 8.5% of patients referred to a pulmonary subspecialty clinic with unexplained exertional dyspnea or limitation. The final diagnoses for the remaining patients included airway disease (n = 103, 31%); cardiovascular disorders (n = 59, 18%); interstitial lung disease (n = 48, 15%); nonmitochondrial neurological disease (n = 17, 5%); multifactorial (n = 35, 11%); and miscellaneous conditions (n = 34, 10%). Seven patients had a history and CPET that was typical of a mitochondrial myopathy, but no biopsy was performed.

Physiological Assessment

Pulmonary function/respiratory muscle testing. Pulmonary function data are summarized in Table 2. The patient group demonstrated a significantly lower absolute MVV, despite both patients and control subjects having normal spirometry and lung volumes. The mean diffusing capacity of the lung for CO (Dl CO) was normal in the patient and control groups although there was a trend for a lower Dl CO in the patients. Six patients had an abnormal Dl CO (60% to 78% predicted), with two being current or former smokers. High-resolution computed tomograms of the chest and ventilation/perfusion lung scanning failed to reveal a pulmonary parenchymal or vascular cause for this abnormality in all six of these patients.


VariablesPatientsControl Subjectsp Value
Pulmonary function
 FVC, % pred100 ± 14108 ± 150.13
 FEV1, % pred105 ± 16110 ± 100.33
 MVV, L/min111 ± 39186 ± 34< 0.0001
 FEV1 × 40, L132 ± 25165 ± 300.001
 RV, % pred116 ± 31 99 ± 220.15
 TLC, % pred106 ± 16108 ± 150.83
 Dl CO, % pred 90 ± 15103 ± 30.06
Respiratory muscle function
 Pi max, % pred 77 ± 34115 ± 380.01
 Pe max, % pred 50 ± 23NA
 Pdisniff, cm H2O 84 ± 22 81 ± 350.82
 Pdimax, cm H2O 80 ± 36144 ± 290.0004

Definition of abbreviations: Dl CO = diffusion capacity for carbon monoxide; FEV1 = forced expiratory volume in 1 s; FVC = forced vital capacity; MVV = maximum voluntary ventilation; NA = not available; Pdimax = maximal transdiaphragmatic pressure; Pdisniff = transdiaphragmatic sniff pressure; Pe max = maximal expiratory mouth pressure; Pi max = maximal inspiratory mouth pressure; RV = residual volume; TLC = total lung capacity.

Respiratory muscle function is also presented in Table 2. Compared with controls, the Pi max was significantly lower in the patient group. Pe max was not routinely measured in control subjects. Fourteen patients consented for and underwent additional measurement of transdiaphragmatic pressures. Compared with control subjects the Pdimax was lower in the patient group, although no difference was seen in Pdisniff. However, in nine of these 14 patients and in two control subjects, the Pdisniff was below the lower limit of normal for our laboratory (122 ± 40 cm H2O) (28).

Exercise testing. Table 3 summarizes selected cardiopulmonary exercise test parameters for the group of patients and control subjects. The mean maximal achieved V˙o 2 was decreased for the patient group although the response was variable with five patients demonstrating a maximal achieved V˙o 2 greater than 90% predicted.


ParameterPatient GroupControl Subjectsp Value
Metabolic response
 Maximum achieved watts114 ± 43240 ± 66< 0.0001
 Maximum achieved V˙ o 2, % pred67 ± 22104 ± 26< 0.0001
 AT, % V˙ o 2max pred46 ± 17 65 ± 150.03
 Resting lactate, mmol/L (n = 16) 0.9 ± 0.55NA
 Peak lactate, mmol/L (n = 16) 5.4 ± 3.65NA
Cardiovascular response
 HRmax, % pred84 ± 10 85 ± 110.42
 HRR91 ± 62 41 ± 210.01
 O2 pulse, % pred80 ± 23124 ± 40< 0.0001
Ventilatory response
 V˙ e max, L/min71 ± 31104 ± 230.0008
 MVV, L/min111 ± 39186 ± 34< 0.0001
 V˙ e max/MVV0.70 ± 0.280.59 ± 0.120.22
 fb, breaths/min47 ± 15 40 ± 70.15
 Vtmax/FVC0.40 ± 0.110.50 ± 0.070.01
 V˙ e/V˙ o 2, peak59 ± 22 41 ± 60.02
 V˙ e/V˙ co 2, peak55 ± 1942 ± 170.002
 V˙ e/V˙ o 2, ≈ 50% V˙ o 2 peak49 ± 27 25 ± 40.0001
 V˙ e/V˙ co 2, ≈ 50% V˙ o 2 peak57 ± 23 29 ± 2< 0.00001
Gas exchange (n = 14)
 P(a–a)o 2, mm Hg, rest5 ± 7.5NA
 P(a–a)o 2, mm Hg, max8 ± 7.6NA
 Vd/Vt rest0.40 ± 0.13NA
 Vd/Vt max0.32 ± 0.14NA
 PaCO2 , mm Hg, rest37 ± 4NA
 PaCO2 , mm Hg, max32 ± 6NA
 SpO2 , % (n = 28)95 ± 1.5NA

Definition of abbreviations: AT = anaerobic threshold; fb = respiratory rate; FEV1 = forced expiratory volume in 1 s; FVC = forced vital capacity; HRmax = maximal heart rate; HRR = heart rate response; MVV = maximum voluntary ventilation; NA = not available; V˙ co 2 = carbon dioxide output; Vd = dead space volume; V˙ emax = ventilation at peak exercise; V˙ o 2 = oxygen uptake; Vt max = tidal volume at peak exercise.

The cardiovascular response was evaluated by analyzing the maximum heart rate, heart rate response ([HRmax − HRinitial]/[V˙o 2max − V˙o 2initial]) (29), and oxygen pulse (V˙o 2max/maximum heart rate). The heart rate response was uniformly abnormal, being increased (normal < 50) in all patients (Figure 1). The oxygen pulse was also significantly lower in patients despite the exclusion of any cardiac disease prior to entry into the study. Although not presented, the electrocardiographic response to exercise was normal in all patients.

Anaerobic threshold was determined using a combined approach of the ventilatory equivalents method and the modified V-slope in 17 patients but was undetermined in 11 patients. We did not routinely monitor lactate levels with enough frequency during exercise to allow a more accurate measure of lactate threshold. The mean AT was lower in patients compared with control subjects with four patients having AT below normal (40% of predicted V˙o 2max).

The ventilatory response to exercise differed between patients and control subjects with the V˙emax being significantly lower in patients. However, the ventilatory reserve described by V˙emax/MVV was variable ranging from 0.35 to 1.44 and without significant difference noted between patients and control subjects. Ventilatory efficiency was estimated by measurement of ventilatory equivalents (V˙e/V˙o 2, V˙e/co 2). The maximal V˙e/V˙o 2 and V˙e/V˙co 2 were impressively elevated in comparison with normal subjects. Fourteen patients had resting and peak blood gas analysis. In these patients, oxygenation was normal and a decrease in the dead space/tidal volume ratio (Vd/Vt) was noted during exercise. The remainder of the patients had normal oxygenation as measured by pulse oximetry. The exercise response for a typical patient is shown in Figure 2.

Table 4 demonstrates that the patient-reported symptom-limiting exercise was dyspnea in 16 patients and peripheral muscle fatigue or myalgia in 11 patients. One patient ended exercise with syncope and neither the presence of dyspnea nor fatigue could be evaluated. As expected, the patients ending exercise due to complaints of muscle fatigue or myalgia had significantly lower peak dyspnea scores as measured by either Borg score or VAS compared with patients ending exercise due to dyspnea. When the patients with dyspnea and fatigue were compared, no difference was noted in demographic data. MVV and FVC were significantly lower in the dyspnea-limited patients. The Pdimax was also lower in patients with dyspnea, and there was a trend for a lower Pi max and Pdisniff. No difference was noted in aerobic capacity or measures of gas exchange in those patients ending exercise because of dyspnea, although a significantly higher V˙emax/MVV was noted in this group. Importantly, there was no difference in V˙emax between the two groups or in ventilatory equivalents.


ParameterLimiting Symptomp Value
Dyspnea (n = 16)Fatigue (n = 11)
 Age, yr37 ± 11 38 ± 60.97
 Sex, m/f13/39/20.97
 Weight, kg 68 ± 17 74 ± 200.42
 Height, cm167 ± 9170 ± 100.44
Pulmonary function
 FVC, % pred 95 ± 14107 ± 130.02
 FEV1, % pred100 ± 17111 ± 150.11
 MVV, L/min 93 ± 33128 ± 280.01
 FEV1 × 40, L122 ± 23143 ± 240.03
 TLC, % pred105 ± 15111 ± 190.32
 RV, % pred122 ± 26109 ± 370.35
 Dl CO, % pred 85 ± 16 98 ± 120.08
Respiratory muscle function
 Pi max, % pred 64 ± 29 95 ± 360.06
 Pe max, % pred 45 ± 31 57 ± 320.22
 Pdisniff, cm H2O 70 ± 22100 ± 430.10
 Pdimax, cm H2O 61 ± 23115 ± 310.01
 Peak VAS, mm 82 ± 17 61 ± 270.05
 Peak Borg8.0 ± 2.7 4.6 ± 2.60.01
 V˙ o 2max, % pred65 ± 19 75 ± 240.26
 AT, % V˙ o 2max pred44 ± 20 55 ± 30.27
 O2 pulse, % pred78 ± 17 88 ± 260.20
 HRR87 ± 45 75 ± 360.52
 V˙ e max, L/min68 ± 21 74 ± 260.56
 V˙ e max/MVV0.81 ± 0.3080.58 ± 0.1530.02
 fb, breaths/min49 ± 16 44 ± 150.36
 Vtmax/FVC0.40 ± 0.080.41 ± 0.150.75
 V˙ e/V˙ o 2, peak60 ± 17 53 ± 180.31
 V˙ e/V˙ co 2, peak58 ± 20 50 ± 180.32
 SpO2 , %95 ± 4 95 ± 40.91

Definition of abbreviations: AT = anaerobic threshold; Dl CO = diffusion capacity for carbon monoxide; fb = respiratory rate; FEV1 = forced expiratory volume in 1 s; FVC = forced vital capacity; HRR = heart rate response; MVV = maximum voluntary ventilation; NA = not available; Pdimax = maximal transdiaphragmatic pressure; Pdisniff = transdiaphragmatic sniff pressure; Pe max = maximal expiratory mouth pressure; Pi max = maximal inspiratory mouth pressure; RV = residual volume; TLC = total lung capacity; VAS = visual analog scale; V˙ co 2 = carbon dioxide output; V˙ emax = ventilation at peak exercise; V˙ o 2 = oxygen uptake; V˙ o 2max = oxygen uptake at peak exercise; Vtmax = tidal volume at peak exercise.


There was a strong correlation between V˙e and VAS (r = 0.68; p < 0.05) and V˙e/MVV and VAS (r = 0.68; p < 0.05). Mixed models utilizing all available minute-by-minute measurements of breathlessness (VAS) over the period of exercise were developed for selected parameters (see Methods) of respiratory muscle function. Each of these models also included subject or patient group effects and group-by-corresponding parameter interactions. A summary of the results of these models is presented in Table 5. Part A shows the findings for the patients only, the two group analyses, in which the group comparisons were between patients with dyspnea and fatigue. It is noted that there were significant regressions of VAS with respiratory rate (fb; VAS increases by 0.54 for each unit increase in fb) and Vi (VAS decreases by 11.14 for each unit increase of Vi, i.e., less inspiratory flow). There were no interactions between patients with dyspnea and fatigue-limited patients or with the parameters studied. Table 5, Part B, indicates the results for models that included control subjects and patients, as categorized into patients with dyspnea and fatigue. In these analyses, control subjects were used as the referent level. The findings are similar to those for patients only; that is, increasing respiratory rate (1.59) or Vt/FVC (1.43) significantly increases VAS, whereas increasing Vi significantly decreases (15.48) VAS. There were effects of the patient or control group on VAS for Vt/FVC and an interaction of these groups with respiratory rate.


ParameterRegression Coefficientp Value
A. Two-group comparison for patients only: dyspnea (n = 16) versus fatigue (n = 11)
fb (n = 25)  0.540.009
Ti/Ttot (n = 12) 11.850.39
Vt/FVC (n = 25)  5.970.28
Vi (n = 12)−11.140.001
Ppl/Pplmax (n = 12)−0.490.96
Two groups (dysp-  nea, fatigue)No significant group effect or group-by-  parameter interaction was found in any of  the models
B. Three-group comparison, all subjects: patients (categorized as dyspnea or fatigue) and control subjects
fb (n = 31)  1.590.0001
Ti/Ttot (n = 18) −5.360.32
Vt/FVC (n = 31)  1.430.0001
Vi (n = 18)−15.480.0001
Ppl/Pplmax (n = 18) 12.160.21
Three groups (pa-  tients with dys-  pnea, fatigued  patients, control  subjects)For Vt/FVC, each of the two patient groups  (dyspnea, p = 0.05 and fatigue, p = 0.04)  had significant increases in VAS as com-  pared with controls, that is, a group effect;  for fb, there was a significant (p = 0.01) di-  minution in VAS as an interaction of fb for  the fatigued patients as compared to con-  trol subjects

Definition of abbreviations: Ppl/Pplmax = pleural pressure/maximum pleural pressure; Ti/Ttot = inspiratory time/total time of respiratory cycle; Vi = maximum inspiratory flow rate; Vt/FVC = tidal volume/forced vital capacity.

*Each model also included group and group-by-corresponding parameter interaction. Sample size, n, depended on available data.

The present work describes a large cohort of patients with biopsy-proven mitochondrial myopathies that was initially referred for evaluation of unexplained exertional dyspnea or intolerance. Close examination of this cohort reveals several important observations: mitochondrial myopathy was the diagnosis in 8.5% of patients referred to our Dyspnea Clinic with unexplained exertional dyspnea or intolerance over a 4-yr period; aerobic exercise capacity was reduced and was characterized by a hypercirculatory and hyperventilatory response with normal pulmonary gas exchange (see Figure 1); and those patients stopping exercise due to breathlessness approached their ventilatory limit and demonstrated a greater likelihood of respiratory muscle dysfunction.

Our data confirm that mitochondrial myopathies should be considered in the differential diagnosis of unexplained exertional breathlessness or fatigue (5, 30). From our experience, we estimate the prevalence of these disorders as 8.5% in patients presenting to an internal medicine subspecialty clinic with these complaints. Over the same time period, seven additional patients had typical histories, physical examinations, and CPETs, but muscles biopsies were not performed. If these patients were included as patients with mitochondrial myopathies the period prevalence would increase to 10.6%. These prevalences are much higher than a previous point prevalence estimate of 1 in 50,000 taken from patients referred to a specialty neurology center in the Northern United Kingdom (7). To our knowledge, this is the first estimated prevalence in patients presenting with unexplained exertional dyspnea or intolerance to an internal medicine subspecialty clinic. The prevalence in other patient populations requires further investigation. A wide spectrum of clinical presentations has been reported (7, 10) in patients with mitochondrial myopathies. However, those presenting with breathlessness have been described less frequently with several uncontrolled reports of small numbers of patients (6, 10, 23, 31-33). We extend these findings to a large number of biopsy-proven patients who are well characterized and compared with a group of age-matched control subjects.

We describe normal spirometry in all patients, although a mild decrement in Dl CO was noted in a minority. The latter finding is of unclear clinical relevance given the absence of pulmonary parenchymal or vascular disease on imaging studies and the presence of normal arterial oxygenation during exercise in all patients. Previous investigators have noted normal pulmonary function in patients with mitochondrial disorders (9, 10). Interestingly, none of 14 patients previously described had respiratory muscle weakness (9). We describe a significantly lower Pi max and Pdimax in patients compared with control subjects. Additionally, nine patients demonstrated diaphragmatic strength two standard deviations below the limit of normal for our laboratory. The finding of muscle weakness in a significant percentage of our patients was unexpected as previous reports of respiratory muscle weakness associated with this disorder have been limited to case reports of respiratory failure (34, 35). Our data suggest that milder degrees of respiratory muscle weakness are common in patients with mitochondrial myopathies and may contribute, at least in part, to their symptoms.

As a group, our patients demonstrated limited aerobic capacity, although five patients achieved a maximal V˙o 2 greater than 90% of predicted. It would appear that a normal V˙o 2max does not a priori preclude the diagnosis of mitochondrial myopathy and may reflect an individual's initial high level of fitness, which, despite significant reduction, still remains within normal limits. The pattern of exercise response in patients was hypercirculatory and hyperventilatory. All patients had an elevated (> 50) heart rate response ([HRmax − HRinitial]/ [V˙o 2max − V˙o 2initial]). Noninvasive estimates identified an anaerobic threshold in 17 patients, being below 40% maximal predicted V˙o 2 in four of these patients. Indeterminate AT in 11 patients most probably reflected hyperventilation. Importantly, the AT was lower in the patient group compared with the control subjects. The lack of regular monitoring of lactate levels in our patients limited our ability to determine the anaerobic threshold as the marked hyperventilation present in some patients precluded its noninvasive identification of AT by four independent observers. Our findings support those of previous investigators who have noted decreases in AT and abnormal heart rate responses (6, 9, 33). The low work rate, aerobic capacity, AT, and elevation in blood lactate are likely a reflection of abnormalities in O2 utilization, as a normal cardiac output, O2 content, and O2 delivery have been reported in a small group of patients (10). A single case report has confirmed an increased O2 delivery but abnormal O2 extraction (markedly reduced arteriovenous concentration difference [C(a–v)O2]) consistent with abnormal skeletal muscle oxidative metabolism (32). An additional study, using noninvasive estimates of cardiac output during maximal and submaximal exercise, confirmed exercise intolerance related to impaired peripheral oxygen extraction (33). A hyperdynamic circulatory response appears to be one of the more consistent findings in patients with metabolic myopathies (6, 32). It has been postulated that it may be due to the regulatory role that abnormalities in muscle oxidative metabolism may play on the cardiovascular response as a reflection of the normal coupling between O2 utilization and delivery (31, 33). Unfortunately, this hypercirculatory response is not specific for mitochondrial myopathies, as some investigators have described impaired muscle oxidative metabolism in patients with exercise intolerance of unexplained origin (36). As such, our data clearly support a conclusion that myopathic disorders be considered in patients with an abnormal circulatory response during exercise testing (5, 30).

An abnormal ventilatory response was seen in our patient population with an excessive ventilation for the metabolic demand as measured by increased V˙e/V˙o 2 and V˙e/V˙co 2. The ventilatory equivalents were markedly elevated in patients compared with the control subjects. The etiology of the hyperventilatory response in patients with mitochondrial myopathy is unknown. It has been postulated that it occurs in response to the excess of V˙co 2 produced by the buffering of lactate (10). This seems less plausible given the abnormally elevated V˙e/V˙co 2 in the present study. Another hypothesis relates the hyperventilation to an increase in respiratory drive originating in metabolically sensitive chemoreceptors localized in peripheral skeletal muscles, similar to the mechanism described to explain the hypercirculatory response (31). Another potential explanation is the stimulation of mechanoreceptors from respiratory muscle weakness and the inability of patients to generate tidal volumes appropriate for the levels of work. Indeed, our patients had decreased inspiratory muscle strength and a decreased Vtmax/FVC ratio at the end of exercise compared with control subjects. In our study, the V˙emax/MVV was also higher in the patient group compared with the control subjects, although the difference did not reach significance. As the V˙emax/MVV ratio exceeds the normal level (0.70–0.75), there is an increased likelihood for mechanical ventilatory limitation. The majority of patients terminated exercise because of breathlessness (n = 16); a minority stopped because of peripheral muscle fatigue or myalgia (n = 11) and one because of exercise-induced syncope. Importantly, those patients experiencing predominantly breathlessness demonstrated a lower MVV, Pdimax, and a higher V˙emax/MVV. Furthermore, the lower MVV in many instances was disproportionately reduced compared with FEV1 × 40, suggesting respiratory muscle weakness. These patients' Pi max and Pdisniff values were also lower than the fatigue-limited subjects and approached significance. These data would support an increased role of respiratory muscle weakness in these patients. Finally, the slightly abnormal Vd/Vt at peak exercise should also be considered a possible contributing factor of the hyperventilatory response observed in these patients. Without additional arterial blood gases during exercise, it is difficult to discern the mechanisms responsible for the hyperventilation observed in these patients.

Although Vd/Vt decreased with exercise in our patients, the value achieved at peak exercise was slightly elevated (normal < 0.28). This could be due to the lower intensity of exercise achieved by the patients as compared with controls, the smaller tidal volumes resulting from altered breathing patterns as demonstrated by the lower Vt/FVC in patients as compared with controls, and the realization that a 2 mm Hg change in PaCO2 (well within the limits of reproducibility) could account for differences of 0.03 to 0.04 in Vd/Vt. Based on the comprehensive evaluation including HRCT of the chest and ventilation-perfusion scintigraphy, there was no clinical evidence to suggest that the slightly abnormal Vd/Vt observed at peak exercise was caused by pulmonary vascular disease. The low PaCO2 observed at peak exercise as compared with rest may suggest a relative alveolar hyperventilation relative to metabolic demands.

Interestingly, previous investigators prospectively exploring the etiology of unexplained breathlessness described a significant number of patients with deconditioning or unexplained dyspnea (1, 3). Similarly, in a retrospective study of 32 patients with unexplained dyspnea, 14 exhibited a hyperventilatory response (4). It is plausible that at least a portion of these patients represented undiagnosed mitochondrial myopathies. A response to exercise training was used by several groups to define deconditioning as the mechanism for unexplained dyspnea (1-3, 5). However, recently it has been shown that deconditioning plays an important role in the exercise response of patients with mitochondrial disorders (9, 37, 38). A 30% increase in aerobic capacity, reduction in blood lactic acid, and improvement in ADP recovery have been reported after 8 wk of training in patients with mitochondrial myopathy (38). The same group has demonstrated that patients with mitochondrial myopathy improve their aerobic capacity (30%) to a greater extent than patients with nonmetabolic myopathy (16%) and normal control subjects (10%) after 8 wk of exercise training (37). Importantly, the aerobic capacity of these patients after training was still reduced as compared with the sedentary normal control group before training. As such, a positive training response cannot be used to exclude mitochondrial disorders.

An additional novel finding of our work results from an analysis of the pathophysiology of breathlessness in patients with mitochondrial myopathy using the principles described by previous investigators (27, 39, 40). We describe strong linear correlations between breathlessness (measured by either VAS or Borg) and V˙e. This is similar to the description in normal subjects and patients with chronic obstructive pulmonary disease (COPD) (39, 40). In addition, breathlessness is felt not only to relate to increased ventilatory demand, but also to increased impedance to ventilatory muscle function and functionally weakened respiratory muscles (41). As such, the lower respiratory muscle function in our patients with dyspnea with mitochondrial myopathies likely plays an important role. Multivariable analysis confirms that indices of respiratory muscle function are highly correlated with exertional breathlessness in our patients. In our patients, the variables most significantly correlated with dyspnea during exercise as measured by VAS were respiratory rate and Vi. Interestingly, there was little difference between the response in patients with dyspnea and fatigue, although differences were noted when patients were compared with control subjects for Vt/FVC and respiratory rate.

Our data are limited by the known difficulty in diagnosing mitochondrial disorders (22, 24), which is particularly true in patients presenting with an atypical clinical picture (24). As such, we included only patients with histological abnormalities (22, 42), biochemical abnormalities (22), and/or documented mutations of mitochondrial DNA (22). In addition, our data are limited by the lack of routine measurement of lactate levels. It has been suggested that patients with mitochondrial myopathies have elevated resting (8) and exercise (37) lactate levels. However, our patients and others recently described (9, 43) have failed to show elevated resting lactate levels. Similarly, markedly elevated lactate levels during exercise were not demonstrated in our patients or in those reported by Dandurand and coworkers (9). The difference in results may reflect the heterogeneity of mitochondrial myopathies with diverse pathological, histochemical, and biochemical characteristics. In addition, the different methods utilized and mediums in which lactate is analyzed (arterial, venous, whole blood, plasma, etc.) likely also contributed to the reported differences. In view of these conflicting results, a reasonable approach appears to be that if lactates are elevated the possibility of a mitochondrial disorder is strengthened, but the finding of normal lactate levels does not exclude a mitochondrial disorder.

Our study confirms that mitochondrial myopathies are more prevalent than previously appreciated among patients presenting to a multidisciplinary subspecialty clinic with exertional intolerance. Consequently, mitochondrial myopathy should be considered in the differential diagnosis of unexplained exertional intolerance or breathlessness. These patients generally demonstrate normal spirometry but have a significant incidence of respiratory muscle weakness. Cardiopulmonary exercise testing reveals hypercirculatory and hyperventilatory responses with normal pulmonary gas exchange. This characteristic physiological profile can be useful in the diagnostic evaluation of patients with unexplained exercise intolerance.

Supported in part by National Institutes of Health NHLBI Grant P50HL46487, NIH/NCRR 3 MO1 RR00042-33S3, and NIH/NIA P60 AG08808-06.

1. DePaso WJ, Winterbauer RH, Lusk JA, Springmeyer SCChronic dyspnea unexplained by history, physical examination, chest roentgenogram, and spirometry: analysis of a seven year experience. Chest100199112931299
2. Pratter MR, Curley FJ, Dubois J, Irwin RSCause and evaluation of chronic dyspnea in a pulmonary disease clinic. Arch Intern Med149198922772282
3. Martinez FJ, Stanopoulos I, Acero R, Becker FS, Pickering R, Beamis JFGraded comprehensive cardiopulmonary exercise testing in the evaluation of dyspnea unexplained by routine evaluation. Chest1051994168174
4. Sridhar MK, Carter R, Banham SW, Moran FAn evaluation of integrated cardiopulmonary exercise testing in a pulmonary function laboratory. Scot Med J401995113116
5. Weisman IM, Zeballos RJClinical evaluation of unexplained dyspnea. Cardiologia411996621634
6. Elliot DL, Neil RM, Buist M, Goldberg L, Kennaway NG, Phil D, Powell BR, Kuehl KSMetabolic myopathies: evaluation by graded exercise testing. Medicine681989163172
7. Chinnery PF, Turnbull DMMitochondrial medicine. QJM901997657667
8. Andreu AL, Hanna MG, Reichmann H, Bruno C, Penn AS, Tanji K, Pallotti F, Iwata S, Bonilla E, Lach B, et al.. Exercise intolerance due to mutations in the cytochrome b gene of mitochondrial DNA. N Engl J Med341199910371044
9. Dandurand RJ, Matthews PM, Arnold DL, Eidelman DHMitochondrial disease: pulmonary function, exercise performance, and blood lactate levels. Chest1081995182189
10. Hooper RG, Thomas AR, Kearl RAMitochondrial enzyme deficiency causing exercise limitation in normal-appearing adults. Chest1071995317322
11. Morris J, Koski A, Johnson LSpirometric standards for healthy non-smoking adults. Am Rev Respir Dis10319715767
12. Goldman HI, Becklake MRRespiratory function tests: normal values at median altitudes and the prediction of normal results. Am Rev Tuberc791959457467
13. Miller WF, Scacci R, Gast LR. Laboratory evaluation of pulmonary function. Hagerstown, MD: Harper & Row; 1982.
14. Hansen JE, Sue DY, Wasserman KPredicted values for clinical exercise testing. Am Rev Respir Dis1291984S49S55
15. Zeballos RJ, Weisman IMBehind the scenes of cardiopulmonary exercise testing. Clin Chest Med151994193213
16. Wasserman K, Hansen JE, Sue DY, Whipp BJ. Principles of exercise testing and interpretation. Philadelphia: Lea & Febiger; 1987.
17. Black LE, Hyatt REMaximal respiratory pressures: normal values and relationship to age and sex. Am Rev Respir Dis991969696702
18. Rochester DF, Arora NSRespiratory muscle failure. Med Clin North Am671983573597
19. Mier-Jedrzejowicz A, Brophy C, Moxham J, Green MAssessment of diaphragm weakness. Am Rev Respir Dis1371988877883
20. Muza SR, Silverman MT, Gilmore GC, Hellerstein HK, Kelsen SGComparison of scales used to quantitate the sense of effort to breathe in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis1411990909913
21. Mahler DA. The measurement of dyspnea during exercise in patients with lung disease. Chest 1992;101(5 Suppl):242S–247S.
22. Walker UA, Collins S, Byrne ERespiratory chain encephalomyopathies: a diagnostic classification. Eur Neurol361996260267
23. Jackson MJ, Schaefer JA, Johnson MA, Morris AAM, Turnbull DM, Bindoff LAPresentation and clinical investigation of mitochondrial respiratory chain disease. A study of 51 patients. Brain1181995339357
24. Argov ZFunctional evaluation techniques in mitochondrial disorders. Eur Neurol3919986571
25. Martinez FJ, Orens JB, Whyte RI, Graf L, Becker FS, Lynch JPLung mechanics and dyspnea after lung transplantation for chonic airflow obstruction. Am J Respir Crit Care Med153199615361543
26. Mador MJ, Kufel TJReproducibility of visual analog scale measurements of dyspnea in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis14619928287
27. Leblanc P, Bowie DM, Summers E, Jones NL, Killian KJBreathlessness and exercise in patients with cardiorespiratory disease. Am Rev Respir Dis13319862125
28. Martinez FJ, Strawderman RL, Flaherty KR, Cowan M, Orens JB, Wald JRespiratory response during arm elevation in isolated diaphragm weakness. Am J Respir Crit Care Med1601999480486
29. Eschenbacher WL, Mannina AAn algorithm for the interpretation of cardiopulmonary exercise tests. Chest971990263267
30. Weisman IM, Zeballos RJAn integrated approach to the interpretation of cardiopulmonary exercise testing. Clin Chest Med151994421445
31. Haller RG, Lewis SF, Estabrook RW, DiMauro S, Servidei S, Foster DWExercise intolerance, lactic acidosis, and abnormal cardiopulmonary regulation in exercise associated with adult skeletal muscle cytochrome c oxidase deficiency. J Clin Invest841989155161
32. Haller RG, Henriksson KG, Jorfeldt L, Hultman E, Wibom R, Sahlin K, Areskog N, Gunder M, Ayyad K, Blomqvist CG, et al.. Deficiency of skeletal muscle succinate dehydrogenase and aconitase. J Clin Invest88199111971206
33. Vissing J, Galbo H, Haller RGExercise fuel mobilization in mitochondrial myopathy: a metabolic dilemma. Ann Neurol401996655662
34. Cros D, Palliyath S, DiMauro S, Ramirez C, Shamsnia M, Wizer BRespiratory failure revealing mitochondrial myopathy in adults. Chest1011992824828
35. O'Brien A, Blaivas M, Albers J, Wald J, Watts CA case of respiratory muscle weakness due to cytochrome c oxidase enzyme deficiency. Eur Respir J121998742744
36. Argov Z, DeStefano N, Taivassalo T, Chen J, Karpati G, Arnold DLAbnormal oxidative metabolism in exercise intolerance of undetermined origin. Neuromusc Disord7199799104
37. Taivassalo T, DeStefano N, Chen J, Karpati G, Arnold DL, Argov ZShort-term aerobic training response in chronic myopathies. Muscle Nerve22199912391243
38. Taivassalo T, DeStefano N, Argov Z, Matthews PM, Chen J, Genge A, Karpati G, Arnold DLEffects of aerobic training in patients with mitochondrial myopathies. Neurology50199810551060
39. Mador MJ, Kufel TJReproducibility of visual analog scale measurements of dyspnea in patients with chronic obstructive pulmonary disease. Am Rev Respir Dis14619928287
40. Noseda A, Carpiaux JP, Schmerber J, Valente F, Yernault JCDyspnoea and flow-volume curve during exercise in COPD patients. Eur Respir J71994279285
41. O'Donnell DE. Exertional breathlessness in chronic respiratory disease. In: Mahler D, editor. Dyspnea, lung biology in health disease. New York: Marcel Dekker, Inc.; 1998. p. 97–147.
42. Stadhouders AM, Sengers RCA. Morphological observations in skeletal muscle from patients with a mitochondrial myopathy. J Inherit Metab Dis 1987;10(Suppl 1):62–80.
43. Siciliano G, Manca ML, Renna M, Prontera C, Mercuri A, Murri LEffects of aerobic training on lactate and catecholaminergic exercise responses in mitochondrial myopathies. Neuromusc Disord1020004045
Correspondence and requests for reprints should be addressed to Fernando J. Martinez, M.D., TC 3916, 1500 E. Medical Center Dr., Ann Arbor, MI 48109-0360. E-mail:


No related items
American Journal of Respiratory and Critical Care Medicine

Click to see any corrections or updates and to confirm this is the authentic version of record