Abstract
Fatigability increases while the capacity for mitochondrial energy production tends to decrease significantly with age. Thus, diminished mitochondrial function may contribute to higher levels of fatigability in older adults.
The relationship between fatigability and skeletal muscle mitochondrial function was examined in 30 participants aged 78.5 ± 5.0 years (47% female, 93% white), with a body mass index of 25.9 ± 2.7 kg/m 2 and usual gait-speed of 1.2 ± 0.2 m/s. Fatigability was defined using rating of perceived exertion (6–20 point Borg scale) after a 5-minute treadmill walk at 0.72 m/s. Phosphocreatine recovery in the quadriceps was measured using 31 P magnetic resonance spectroscopy and images of the quadriceps were captured to calculate quadriceps volume. ATPmax (mM ATP/s) and oxidative capacity of the quadriceps (ATPmax·Quadriceps volume) were calculated. Peak aerobic capacity (VO 2 peak) was measured using a modified Balke protocol.
ATPmax·Quadriceps volume was associated with VO 2 peak and was 162.61mM ATP·mL/s lower ( p = .03) in those with high (rating of perceived exertion ≥10) versus low (rating of perceived exertion ≤9) fatigability. Participants with high fatigability required a significantly higher proportion of VO 2 peak to walk at 0.72 m/s compared with those with low fatigability (58.7 ± 19.4% vs 44.9 ± 13.2%, p < .05). After adjustment for age and sex, higher ATPmax was associated with lower odds of having high fatigability (odds ratio: 0.34, 95% CI: 0.11–1.01, p = .05).
Lower capacity for oxidative phosphorylation in the quadriceps, perhaps by contributing to lower VO 2 peak, is associated with higher fatigability in older adults.
Fatigue is common among older adults ( 1 ) and associated with poorer physical function and disability both cross-sectionally ( 2 ) and longitudinally ( 3 ). Fatigue is primarily considered to be an energy disorder and a large proportion cannot be attributed to underlying diseases ( 4 ). It has been hypothesized that age-related decreases in mitochondrial function may contribute to higher levels of age-related fatigue ( 5 , 6 ). The capacity for oxidative phosphorylation in skeletal muscle is lower in older compared with younger adults; however, more recent evidence shows this difference varies across muscle groups and appears largely attributable to decreased physical activity ( 7–12 ). Peak aerobic capacity (VO 2 peak) also decreases significantly with age, independent of muscle loss and physical activity level ( 13 ), and is considered a hallmark manifestation of mitochondrial disorders ( 14 ). Hence, age-related decreased capacity for mitochondrial energy production may contribute to higher levels of fatigability through lower aerobic capacity. Little is known, however, about the etiology of age-related fatigue, particularly as it relates to mitochondrial energetics ( 5 , 6 ).
Fatigability, similar conceptually to exercise tolerance, is fatigue anchored to an activity of a specific intensity and duration ( 6 ). Fatigability, as opposed to global fatigue, provides insight into the degree to which an individual is limited physically due to fatigue ( 6 , 15 ). The primary objective of this research was to determine if skeletal muscle mitochondrial oxidative capacity determined by 31 P magnetic resonance spectroscopy ( 31 P MRS), was related to higher levels of fatigability in older adults. We hypothesized that those with higher compared to lower fatigability would have lower skeletal muscle oxidative capacity. We also hypothesized that physical activity would attenuate this relationship and VO 2 peak would act as a mediator.
Methods
Participants
Community-dwelling ( n = 37) men and women aged 70–89 years from Pittsburgh, Pennsylvania, were enrolled into the Study of Energy and Aging Pilot. Inclusion criteria was body weight ≤285 lbs for men and ≤250 lbs for women, body mass index (BMI) 20–32 kg/m 2 , ability to walk without an assistive device, and free of difficulty performing basic activities of daily living. Exclusion criteria: symptomatic cardiovascular or pulmonary disease, heart attack, angioplasty, or heart surgery within the past 3 months, or a cerebral hemorrhage within the past 6 months, stroke within the past 12 months, or chest pain during walking in the past 30 days ( 16 ). Participants were telephone screened and reassessed at the magnetic resonance imaging center for scanner eligibility, including ability to lie in a supine position for 1 hour, no metal or other implants, joint replacements, or tattoos. Participants had to be willing and able to sign an informed consent. This study was approved by the University of Pittsburgh and California Pacific Medical Center Institutional Review Boards.
Clinical Examination and Measurements
Body height (cm) was measured using a wall-mounted stadiometer and weight (kg) with a certified and calibrated scale and used to calculate BMI (kg/m 2 ). Participants completed demographic, medical, and disease history questionnaires. Depression was assessed using a short form of the Center for Epidemiologic Studies Depression Scale (CES-D) ( 17 ).
Physical function was measured by the Short Physical Performance Battery (SPPB), which includes a 6-meter walk, chair stands, and balance tests ( 18 ). Usual gait-speed was derived from the faster of two 6-meter trials. Time to walk 400 meters (seconds) at a usual pace was also measured. Seven-day free-living physical activity was assessed using the SenseWear (BodyMedia Pittsburgh, PA) Pro armband. The manufacturer’s proprietary algorithm was used to calculate minutes-per-day of moderate (≥3METs) physical activity.
VO 2 peak Test
VO 2 peak was determined using a modified Balke protocol ( 19 ) and participant’s usual 6 meter walking speed was used throughout the test. Treadmill grade was increased 2% every 2 minutes until volitional exhaustion. A resting 12-lead electrocardiogram was conducted prior to and monitored continuously during the test. VO 2 measurements were obtained with a metabolic cart (Moxus, AEI Technologies, Pittsburgh, PA). Gas analyzers and flow, using a 3-L syringe, were calibrated before each test. Studies employing similar protocols in older adults report excellent reproducibility of maximal treadmill walking time (ICC = .87) ( 20 ), which correlates highly with VO 2 peak ( r = .92) ( 21 ).
Determination of ATPmax by 31 P MRS
Maximal mitochondrial ATP production (ATPmax) following an acute bout of knee extensor exercise was determined in vivo using 31 P MRS. Phosphocreatine (PCr) recovery after exercise was used to quantify rates of mitochondrial ATP production. 31 P MRS has been validated by animal and human studies showing that ATPmax varies in direct proportion to oxidative enzyme activity of healthy muscle ( 22 , 23 ) and corresponds to mitochondrial content in human muscle ( 24 ). ATPmax had good reproducibility illustrated by a high correlation ( r = .92) between same day repeat scans of eight participants and Bland Altman analysis ( Supplementary Figure 1 ).
The exercise protocol was performed in a magnetic resonance imaging magnet (3T TIM Trio, Siemens’ Medical System). Participants laid supine with the right knee (unless contraindicated) elevated at ~30°. Straps were placed over the legs and a 2.5″ surface RF coil tuned to 31 P was placed over the quadriceps. Signal was collected by a hemisphere defined by the coil radius (1.25″), which we previously confirmed using an image generated by a similarly sized coil ( 24 ). Participants kicked repeatedly as hard and as fast as they could for two bouts (30 and 36 seconds), each followed by a 6-minute rest. The protocol was designed to deplete PCr stores by 33%–66% to ensure high signal to noise defining PCr recovery without inducing acidosis (pH <6.8), which inhibits oxidative phosphorylation. PCr recovery rate (ATPmax) was measured after exercise until PCr returned to baseline levels.
Phosphorus spectra were collected using a standard one pulse experiment to determine levels of PCr, ATP, Pi, PDE (phosphodiesters), and pH throughout exercise and recovery. PCr, Pi, PDE and ATP peak areas in the fully relaxed spectra were measured by integration using Varian VNMR 6.1C software (Varian Medical Systems, Palo Alto, CA). Areas of the PCr and Pi peaks were expressed relative to the ATP peak ( Supplementary Table 2 ). Previous analyses of human vastus lateralis muscle biopsies revealed that ATP content accounted for the range of PCr/ATP levels determined by MRS among participants aged 65–80 years ( 24 ). In contrast, PCr was stable (as was total creatine) and averaged 27mM. Thus, as reported previously ( 16 ), we used 27mM PCr to determine ATPmax. Changes in PCr and Pi peak areas during the tests were analyzed as described earlier ( 25 , 26 ).
Seven participants were missing ATPmax, four due to inadequate PCr breakdown or acidosis and three because of metal deposits or claustrophobia that precluded testing.
Determination of Muscle Size and Oxidative Capacity of the Quadriceps
Magnetic resonance imaging was used to determine quadriceps volume as previously described ( 27 ). Images were collected every 3 cm from hip to thigh (15–25 slices per participant). Standard stereological techniques were used to determine the largest quadriceps muscle cross-sectional area. Oxidative capacity of the quadriceps was calculated by multiplying ATPmax by quadriceps volume.
Determination of Fatigability and Measurement of VO 2 During Steady State Treadmill Test
Fatigability was determined following a 5-minute treadmill walk at 1.6 mph (0.72 m/s) and 0% grade ( 15 ). This speed and duration were selected to simulate the minimal level of activity needed to function relatively independently in the community ( 15 ). Immediately following the walk, participants rated their perceived exertion (ie, rating of perceived exertion [RPE]) using the 6–20 point Borg scale ( 28 ) and were categorized as high fatigability (RPE ≥ 10) or low fatigability (RPE ≤ 9). This cut point was established in the Baltimore Longitudinal Study of Aging, which showed that this threshold is associated with substantially poorer physical functioning ( 15 ). The same test was used for each person to compare fatigability levels across individuals for the same standardized task ( 5 , 6 ). Oxygen consumption was measured during the walk and mean VO 2 was calculated. Immediately following the test, blood lactate levels were assessed by finger stick using a portable lactate analyzer.
Statistical Analyses
The final analytic sample included those with valid ATPmax values ( n = 30). Means and standard deviations or counts and percents were calculated for the entire cohort and separately by high and low fatigability. Univariate statistics for predictors, covariates, and other variables were compared between fatigability groups using t tests, chi-square and nonparametric tests where appropriate. Least-squared age and sex adjusted means for ATPmax and ATPmax·quadriceps volume were compared between fatigability groups. Separate multivariable logistic regression models were generated to determine the odds of having high fatigability associated with increases in the primary predictors of ATPmax and ATPmax·quadriceps volume. Standard deviation increases in ATPmax and ATPmax·quadriceps volume were used to generate odds ratios (ORs) for ease of interpretation, as ORs using original units resulted in upper limits that approached infinity. Analyses were performed using SAS v9.2.
Results
Demographic Characteristics, Medical History, and Physical Function
Participants were aged 78.5 ± 5.0 years, 46.7% female, 93.3% white, BMI of 25.9 ± 2.7 kg/m 2 ( Table 1 ) and were relatively high functioning with SPPB scores of 10.9 ± 1.4 and usual gait-speeds of 1.2 ± 0.2 m/s. There were no differences in demographic characteristics and medical history by fatigability status (all p > .05, Table 1 ). However, those with high fatigability tended to have a higher prevalence of osteoarthritis compared with those with low fatigability ( p = .06). Those with high fatigability were less physically active (36.8 ± 24.5 vs 100.6 ± 83.7 min/d moderate activity, p < .05) and had slower 400 m walk times (383.5 ± 79.3 s vs 319.5 ± 41.4 s, p = .03) compared with the low fatigability group.
Demographic Characteristic by High and Low Fatigability
| High Fatigability (RPE ≥ 10, n = 12) | Low Fatigability (RPE ≤ 9, n = 18) | Entire Cohort ( n = 30) | |
|---|---|---|---|
| Age, years | 79.0 (4.6) | 78.2 (5.3) | 78.5 (5.0) |
| Sex, % female | 58.3 (7) | 38.9 (7) | 46.7 (14) |
| Race, % white | 100 (12) | 88.9 (16) | 93.3 (28) |
| Body weight, kg | 72.3 (15.6) | 71.6 (11.2) | 71.9 (12.9) |
| Body mass index, kg/m 2 | 26.2 (3.0) | 25.7 (2.6) | 25.9 (2.7) |
| Moderate physical activity, min/d | 36.8 (24.5) | 100.6 (83.7)* | 71.2 (66.5) |
| Smoker, current/former | 16.7 (2) | 44.4 (8) | 33.3 (10) |
| Alcohol intake, 6+ drinks/wk | 8.3 (1) | 0.0 (0) | 3.3 (1) |
| Diabetes, yes | 0 (0.0) | 4.8 (1) | 3.3 (1) |
| History of myocardial infarction, yes | 16.7 (2) | 5.46 (1) | 10.0 (3) |
| History of COPD, yes | 8.3 (1) | 0.0 (0) | 3.3 (1) |
| History of osteoarthritis, yes | 50.0 (6) | 16.7 (3) | 30.0 (9) |
| History of cancer, yes | 41.7 (5) | 61.1 (11) | 53.3 (16) |
| CES-D score | 7.7 (3.3) | 6.8 (2.8) | 7.2 (3.0) |
| Usual gait speed, m/s | 1.1 (0.2) | 1.3 (0.2) | 1.2 (0.2) |
| Time to walk 400 m, s | 383.5 (79.3) | 319.5 (41.4)* | 343.8 (65.5) |
| SPPB score, 0–12 | 10.8 (1.5) | 10.9 (1.3) | 10.9 (1.4) |
| High Fatigability (RPE ≥ 10, n = 12) | Low Fatigability (RPE ≤ 9, n = 18) | Entire Cohort ( n = 30) | |
|---|---|---|---|
| Age, years | 79.0 (4.6) | 78.2 (5.3) | 78.5 (5.0) |
| Sex, % female | 58.3 (7) | 38.9 (7) | 46.7 (14) |
| Race, % white | 100 (12) | 88.9 (16) | 93.3 (28) |
| Body weight, kg | 72.3 (15.6) | 71.6 (11.2) | 71.9 (12.9) |
| Body mass index, kg/m 2 | 26.2 (3.0) | 25.7 (2.6) | 25.9 (2.7) |
| Moderate physical activity, min/d | 36.8 (24.5) | 100.6 (83.7)* | 71.2 (66.5) |
| Smoker, current/former | 16.7 (2) | 44.4 (8) | 33.3 (10) |
| Alcohol intake, 6+ drinks/wk | 8.3 (1) | 0.0 (0) | 3.3 (1) |
| Diabetes, yes | 0 (0.0) | 4.8 (1) | 3.3 (1) |
| History of myocardial infarction, yes | 16.7 (2) | 5.46 (1) | 10.0 (3) |
| History of COPD, yes | 8.3 (1) | 0.0 (0) | 3.3 (1) |
| History of osteoarthritis, yes | 50.0 (6) | 16.7 (3) | 30.0 (9) |
| History of cancer, yes | 41.7 (5) | 61.1 (11) | 53.3 (16) |
| CES-D score | 7.7 (3.3) | 6.8 (2.8) | 7.2 (3.0) |
| Usual gait speed, m/s | 1.1 (0.2) | 1.3 (0.2) | 1.2 (0.2) |
| Time to walk 400 m, s | 383.5 (79.3) | 319.5 (41.4)* | 343.8 (65.5) |
| SPPB score, 0–12 | 10.8 (1.5) | 10.9 (1.3) | 10.9 (1.4) |
Notes: Values are mean ± ( SD ) or % ( n ). COPD = chronic obstructive pulmonary disease; CES-D = Center for Epidemiologic Studies Depression Scale; RPE = rating of perceived exertion; SPPB = short physical performance battery.
*Significant difference between fatigability groups, p < .05.
Aerobic Capacity, Muscle Size, and Mitochondrial Function
Aerobic capacity, muscle size, and mitochondrial function, both stratified by fatigability status and overall, are shown in Table 2 . Mean VO 2 peak was 22.3 ± 5.9 mL/kg/min (range: 7.8–33.4), ATPmax: 0.52 ± 0.13 mM ATP/s (range: 0.30–0.83) and ATPmax·quadriceps volume: 591.25 ± 203.93 mM ATP·mL/s (range: 222.50–965.04).
Perceived Exertion, Aerobic Capacity, Lactate Levels, Mitochondrial Function and Muscle Size by High and Low Fatigability
| High Fatigability (RPE ≥ 10, n = 12) | Low Fatigability (RPE ≤ 9, n = 18) | Entire Cohort ( n = 30) | |
|---|---|---|---|
| VO 2 peak, mL/kg/min | 18.9 (4.4) | 24.4 (5.8)* | 22.3 (5.9) |
| VO 2 peak, mL/min | 1326.5 (324.7) | 1730.0 (432.5)* | 1576.9 (436.9) |
| VO 2 during steady state walk, mL/kg/min | 10.4 (1.0) | 10.4 (1.8) | 10.4 (1.5) |
| VO 2 during steady state walk, mL/min | 744.3 (150.4) | 738.8 (156.9) | 741.0 (151.8) |
| % VO 2 peak reached during steady state walk | 58.7 (19.4) | 44.9 (13.2)* | 50.2 (16.9) |
| Blood lactate end of steady state walk | 1.81 (1.09) | 1.63 (1.02) | 1.71 (1.03) |
| ATPmax, mM ATP/s | 0.47 (0.12) | 0.55 (0.14) | 0.52 (0.13) |
| Quadriceps volume, mL | 1081.2 (427.1) | 1204.1 (307.9) | 1154.9 (358.5) |
| High Fatigability (RPE ≥ 10, n = 12) | Low Fatigability (RPE ≤ 9, n = 18) | Entire Cohort ( n = 30) | |
|---|---|---|---|
| VO 2 peak, mL/kg/min | 18.9 (4.4) | 24.4 (5.8)* | 22.3 (5.9) |
| VO 2 peak, mL/min | 1326.5 (324.7) | 1730.0 (432.5)* | 1576.9 (436.9) |
| VO 2 during steady state walk, mL/kg/min | 10.4 (1.0) | 10.4 (1.8) | 10.4 (1.5) |
| VO 2 during steady state walk, mL/min | 744.3 (150.4) | 738.8 (156.9) | 741.0 (151.8) |
| % VO 2 peak reached during steady state walk | 58.7 (19.4) | 44.9 (13.2)* | 50.2 (16.9) |
| Blood lactate end of steady state walk | 1.81 (1.09) | 1.63 (1.02) | 1.71 (1.03) |
| ATPmax, mM ATP/s | 0.47 (0.12) | 0.55 (0.14) | 0.52 (0.13) |
| Quadriceps volume, mL | 1081.2 (427.1) | 1204.1 (307.9) | 1154.9 (358.5) |
Notes: Values are mean ± ( SD ). ATP = adenosine triphosphate; RPE = rating of perceived exertion.
*Significant difference between fatigability groups, p < .05.
Those with high fatigability had significantly lower VO 2 peak (18.9 ± 4.4 vs 24.4 ± 5.8 mL/kg/min, p < .05) and ATPmax·quadriceps volume (493.69 ± 203.95 vs 656.30 ± 181.19 mM ATP·mL/s, p < .05, Figure 1 ) compared with those with low fatigability. ATPmax was lower in those with high fatigability compared with those with low fatigability (0.47 ± 0.12 vs 0.55 ± 0.14 mM ATP/s, p = .09, Table 2 ); after adjustment for age and sex, the difference became significant (0.46 ± 0.03 vs 0.56 ± 0.03 mM ATP/s, p = .04). Differences in ATPmax·quadriceps volume were slightly attenuated after adjusting for age and sex (519.61 ± 48.55 vs 639.02 ± 39.46 mM ATP·mL/s, p = .07). Mean oxygen consumption during the 5 minute, 0.72 m/s treadmill bout was nearly identical between those with high and low fatigability (10.4 ± 1.8 vs 10.4 ± 1.0 mL/kg/min, p > .99); however, those with high fatigability required a significantly higher proportion of VO 2 peak to walk at the same speed (58.7 ± 19.4 vs 44.9 ± 13.2%, p < .05, Table 2 ). No difference in muscle volume was observed between fatigability groups ( p = .37). Pearson correlations between age, MRS, VO 2 peak, and physical activity measures are in Supplementary Table 1 . The PCr, ATP, Pi, and PDE levels did not differ significantly between groups (all p > .05, Supplementary Table 2 ).
Oxidative capacity of the quadriceps by high and low fatigability. Maximal mitochondrial adenosine triphosphate production defined as phosphocreatine recovery in the quadriceps, following an acute bout of exercise, measured by 31 P magnetic resonance spectroscopy, multiplied by quadriceps volume. Fatigability was defined using rating of perceived exertion (RPE, 6–20 point Borg scale) after a 5-minute treadmill walk at 0.72 m/s. High fatigability = RPE ≥10 and low fatigability = RPE ≤9.
Oxidative capacity of the quadriceps by high and low fatigability. Maximal mitochondrial adenosine triphosphate production defined as phosphocreatine recovery in the quadriceps, following an acute bout of exercise, measured by 31 P magnetic resonance spectroscopy, multiplied by quadriceps volume. Fatigability was defined using rating of perceived exertion (RPE, 6–20 point Borg scale) after a 5-minute treadmill walk at 0.72 m/s. High fatigability = RPE ≥10 and low fatigability = RPE ≤9.
Relationship Between Fatigability and Mitochondrial Function
One standard deviation increase in ATPmax was associated with 2.94 times lower odds of having high fatigability (OR: 0.34, 95% CI: 0.11–1.01, p = .05, Table 3 ) after adjustment for age and sex. One standard deviation increase in ATPmax·quadriceps volume was associated with 2.56 times lower odds of having high fatigability (OR: 0.39, 95% CI: 0.16–0.96, p = .04). Adjusting for age and sex attenuated this relationship to borderline significance (OR: 0.37, 95% CI: 0.13–1.10, p = .07, Table 3 ). The relationships between ATPmax and ATPmax·quadriceps volume and fatigability were attenuated after adjustment for physical activity or VO 2 peak ( Table 3 ). Similarly quadriceps volume was not significantly associated with fatigability univariately ( p = .35) or after age and sex adjustment ( p = .98). Finally, adjustment for osteoarthritis had no effect on the relationship between either ATPmax or ATPmax·quadriceps volume with fatigability.
Logistic Regression Models for the Association between High Fatigability and ATPmax or ATPmax·Quadriceps Volume
| ATPmax | ATPmax·Quadriceps Volume | |||
|---|---|---|---|---|
| Model | Odds Ratio* (95% CI) | Wald χ2p Value** | Odds Ratio* (95% CI) | Wald χ2p Value** |
| Model 1, unadjusted | 0.45 (0.20–1.17) | .11 | 0.39 (0.16–0.96) | .04 |
| Model 2 † | 0.34 (0.11–1.01) | .05 | 0.37 (0.13–1.10) | .07 |
| Model 3 ‡ | 0.57 (0.18–1.84) | .35 | 0.46 (0.13–1.59) | .22 |
| Model 4 § | 0.41 (0.09–1.83) | .24 | 0.62 (0.14–2.71) | .53 |
| ATPmax | ATPmax·Quadriceps Volume | |||
|---|---|---|---|---|
| Model | Odds Ratio* (95% CI) | Wald χ2p Value** | Odds Ratio* (95% CI) | Wald χ2p Value** |
| Model 1, unadjusted | 0.45 (0.20–1.17) | .11 | 0.39 (0.16–0.96) | .04 |
| Model 2 † | 0.34 (0.11–1.01) | .05 | 0.37 (0.13–1.10) | .07 |
| Model 3 ‡ | 0.57 (0.18–1.84) | .35 | 0.46 (0.13–1.59) | .22 |
| Model 4 § | 0.41 (0.09–1.83) | .24 | 0.62 (0.14–2.71) | .53 |
Notes: *Per standard deviation increase.
† Adjusted for age and sex.
‡ Adjusted for age, sex, and VO 2 peak.
§ Adjusted for age, sex, and physical activity.
** p value corresponds to the relationship between ATPmax or ATPmax·Quadriceps volume and high fatigability, respectively.
Discussion
The capacity for oxidative ATP synthesis, assessed with 31 P MRS (ATPmax and ATPmax·muscle volume), was lower in older adults with higher levels of fatigability. To our knowledge, this is the first study to have examined the relationship between mitochondrial energetics and fatigability in older adults. The inverse association between ATPmax and fatigability is consistent with research showing patients with mitochondrial disorders possess lower levels of exercise tolerance than controls ( 29 ). This is also consistent with data from mitochondrial gene ANT1-knockout mice, a model for chronic ATP deficiency, displaying lower exercise tolerance and higher fatigability compared with wild type ( 30 ). Additionally, ATPmax and ATPmax·quadriceps volume are highly related to aerobic capacity in older adults ( 31 ), as skeletal muscle ATP production is one of the two major components comprising VO 2 peak, and both decrease with age ( 13 , 24 ). Reduced cardiac output, the other major component of VO 2 peak, may also contribute to higher fatigability independent of ATPmax ( 32 ). Nonetheless, impaired mitochondrial energy production may be a specific factor that contributes to higher levels of fatigability via lower aerobic capacity.
Another key finding was that participants with high fatigability had to utilize a higher proportion of VO 2 peak to walk at the same speed compared with those with lower fatigability. There are several mechanisms related to mitochondrial function to consider. First, it is possible that the source of energy, ie, mitochondrial and glycolytic, to maintain walking speed was related to fatigability; however, this was not reflected by blood lactate levels during steady-state walking. Second, impaired ATPmax may contribute to a reduced ability to replenish high energy phosphates (eg, PCr) from inorganic phosphate (Pi) ( 33 ), leading to a buildup of Pi, which is associated with muscle fatigue ( 34 ). Additionally, it is possible that when mitochondrial energy production approaches maximum capacity or is depleted, a sensation of fatigue is elicited as a response ( 5 ). For example, Fiser and his coworkers ( 35 ) showed that those with slower gait-speed reached a significantly higher proportion of VO 2 peak and had higher RPE (ie, fatigability) during a submaximal walking test. Our data extend these findings by suggesting that mitochondrial function may be in the etiologic pathway of fatigability and ultimately slowed gait ( 16 ). This is also supported by our finding that time to walk 400 m was significantly slower in the high fatigability group. These data also advance the “Energetic Pathway for Mobility Loss” theory postulated by Schrack and colleagues ( 36 ), which states that the maximum capacity for energy (VO 2 peak) decreases with age ( 13 ), restricting the availability of energy for everyday activities eventually leading to mobility loss in older age. Our data suggest that impaired mitochondrial function may contribute to lower availability of energy and VO 2 peak in older adults, which induces fatigability and may subsequently exacerbate decreased physical activity levels and mobility. The potential role of impaired mitochondrial function in age-related mobility loss is supported by the finding of Coen and his coworkers ( 16 ) in this cohort that lower ATPmax is significantly associated with slower usual-paced 400 m walking time. This is bolstered by two other studies showing that mitochondrial function is impaired in sedentary as compared to active older adults ( 11 ) and that lower functioning older adults possess greater mitochondrial dysfunction compared with higher functioning ( 37 ). Furthermore, fatigue is independently associated with lower SPPB scores and slower walking speeds ( 2 ) as well as a significant predictor of future functional decline ( 3 ). Thus, interventions aimed at improving mitochondrial energy production may be effective at lowering fatigability (exercise tolerance), resulting in sustainable increased physical activity levels and ultimately mobility in older adults. To establish directionality and causality, it is important to study the longitudinal relationships between age, physical activity, VO 2 peak, fatigability, mitochondrial function, and mobility.
The relationship between ATPmax and fatigability was attenuated after adjustment for physical activity. This was not surprising as ATPmax is closely associated with physical activity ( 8 ), and we did not have a large enough sample to detect an independent relationship. Mitochondrial dysfunction may be a contributor to, as well as a consequence of, age-related declines in physical activity. For example, mitochondrial dysfunction activates apoptotic pathways in skeletal muscle and activation of these pathways may contribute to sarcopenia ( 38 ), which may result in decreased physical activity levels. However, more recent work shows that aged, compared with young, skeletal muscle fibers are more susceptible to mitochondrial mediated apoptosis independent of lower oxidative capacity ( 9 ). Although decreases in physical activity likely initiate the process, lower ATPmax may exacerbate age-related decreases in physical activity by contributing to the age-related loss of VO 2 peak and exercise tolerance (ie, higher fatigability). Furthermore, increasing ATPmax may be a mechanism by which physical activity can lower fatigability. These relationships warrant further longitudinal study.
This study had several strengths. Mitochondrial function was measured in vivo, which reflects actual mitochondrial energy production in the living skeletal muscle as opposed to energy production measured in isolated mitochondria from muscle biopsy ( 39 ). Fatigability was measured following a standardized performance test, eliminating any contextual and recall biases associated with self-reported fatigability. This study also had limitations. The small sample size limited our ability to detect independent relationships. Thus, other benefits of physical activity that may lower fatigability could not be separated from higher ATPmax. Although fatigability was assessed following a standardized performance test, a certain degree of subjectivity remained as participants rated their perceived exertion, as opposed to an observed deterioration in performance. This cohort of older adults was also relatively healthy and high functioning; thus, it remains unclear whether mitochondrial function contributes to fatigability in lower functioning, more frail populations. However, it was encouraging that in a relatively homogenous cohort in regard to physical function, we saw relationships between ATPmax and fatigability.
In conclusion, we provide novel evidence showing that impaired mitochondrial function may be implicated in the etiologic pathway of age-related fatigability. Understanding the etiology of fatigability is vital to prevent and treat declines in physical function. Impaired mitochondrial function may lead to higher levels of fatigability by contributing to lower maximal aerobic or reserve capacity. Physical inactivity may initiate the process; however, improving mitochondrial function may improve fatigability, which may then increase physical activity levels due to improved exercise tolerance. The causal role of impaired mitochondrial function and lower aerobic capacity in age-related fatigability needs to be studied longitudinally and in a larger population of older adults across a wider range of physical function.
Funding
This work was supported by the National Institute on Aging at the National Institutes of Health (1RC2AG036594-01 and 1RC2AG036606)
Acknowledgments
Participants were recruited from the Pittsburgh Claude D. Pepper Older Americans Independence Center Research Registry (P30 AG024827). A preliminary version of this work was presented in poster form at the 2011 Annual Meeting of the Gerontological Society of America in Boston, Massachusetts.
