Introduction
Muscle mitochondrial mass and function are positively affected by regular endurance exercise (Tonkonogi and Sahlin
2002). Due to the pivotal role of mitochondria in determining endurance capacity, there is a need for robust and non-invasive measurements of muscle mitochondrial function (Lanza and Nair
2009). Mitochondrial function in skeletal muscle is classically analysed ex vivo by measuring oxygen consumption in muscle biopsies. Less-invasive techniques have emerged over the last quarter century, allowing the measurement of mitochondrial function in vivo. These techniques are both based on the recovery of muscle homeostasis after exercise (Meyer
1988), assessed by measuring either the regeneration of phosphocreatine (PCr) using magnetic resonance spectroscopy (
31P-MRS) or by the return of muscle oxygen consumption (m
\(\dot{V}\)O
2) to basal levels using near-infrared spectroscopy (NIRS). Mitochondrial function analysed by both techniques have been shown to be in good agreement with each other (Ryan et al.
2013), but NIRS offers advantages over
31P-MRS due to its higher portability and relatively low costs, making it more suitable for on-site and routine measurements.
NIRS uses the difference in light absorption of oxygenated and deoxygenated haemoglobin and myoglobin in the near-infrared region (Grassi and Quaresima
2016). By emitting light at different wavelengths, it is possible to differentiate between the oxygenated and deoxygenated states. When used on muscle and combined with arterial occlusions, it allows for measurement of m
\(\dot{V}\)O
2, as the change from oxygenated to deoxygenated haemoglobin and myoglobin reflects the use of oxygen in the tissue underneath the NIRS probe when blood flow is occluded (Van Beekvelt et al.
2001). Multiple, transient arterial occlusions after a short bout of exercise allows for the measurement of post-exercise recovery of m
\(\dot{V}\)O
2, a procedure used to assess mitochondrial capacity (Motobe et al.
2004). The underlying assumption is that post-exercise regeneration of readily available energy carriers (i.e. ATP and PCr) is directly linked to aerobic metabolism and, therefore, a higher mitochondrial capacity will be associated with a faster return of m
\(\dot{V}\)O
2 to the pre-exercise state (McMahon and Jenkins
2002). Indeed, the NIRS procedure to assess recovery kinetics of m
\(\dot{V}\)O
2 in vivo showed a strong correlation with maximal ADP-stimulated respiration of permeabilised muscle fibres in situ (Ryan et al.
2014).
On a whole body level, the regeneration of readily available energy carriers is assumed to contribute to the transient elevation of whole body oxygen consumption (m
\(\dot{V}\)O
2) above resting values in the immediate post-exercise period, also known as excess post-exercise oxygen consumption (EPOC). EPOC can be divided into a rapid and a prolonged phase, in which the mechanisms that contribute to the elevated m
\(\dot{V}\)O
2 are different (Gaesser and Brooks
1984). In particular, the rapid phase is defined to reflect the myofibrillar consumption of the readily available energy substrates in the beginning of exercise, such as PCr and ATP, as well as the replenishment of tissue and haemoprotein oxygen stores and lactate removal (Chance et al.
1992; Børsheim and Bahr
2003). In accordance with an important role for PCr regeneration in EPOC, Rossiter et al. showed that whole body
\(\dot{V}\)O
2 is related to muscle PCr kinetics in the recovery phase (Rossiter et al.
2002). As the latter may be related to skeletal muscle mitochondrial capacity, an inverse relationship between EPOC and NIRS assessment of m
\(\dot{V}\)O
2, reflecting skeletal muscle mitochondrial function, can be hypothesised (Kemp et al.
2015). However, it should be noted that despite clear effects on mitochondrial capacity (Lanza and Nair
2009), the effect of endurance training status on EPOC is controversial, most likely as a result of methodological difficulties (Børsheim and Bahr
2003). When comparing low- with high-endurance capacity subjects, no research design can control for relative exercise intensity, total work and exercise duration at the same time. Still, by controlling for relative intensity and exercise duration, different
\(\dot{V}\)O
2 recovery dynamics between trained and untrained subjects have been observed (Short and Sedlock
1997). Thus, when using such an approach, post-exercise whole body VO
2 recovery dynamics may be a reflection of aerobic fitness, and be related skeletal muscle mitochondrial capacity.
NIRS has been used as a non-invasive measure for muscle mitochondrial function in various clinical populations such as COPD and cystic fibrosis patients. In general, NIRS studies indicate skeletal muscle mitochondrial dysfunction under these pathological conditions (Adami et al.
2017; Willingham and McCully
2017). On the other hand, endurance athletes, characterised by a high whole body peak oxygen uptake (
\(\dot{V}\)O
2peak), showed a faster post-exercise recovery of m
\(\dot{V}\)O
2 than fully sedentary subjects (Brizendine et al.
2013). Still, the difference in
\(\dot{V}\)O
2peak between the two groups was considerable (74 vs 34 ml.kg-1.min-1, respectively). It would be of interest to study whether this technique is also sensitive enough to detect differences within a more normally active, healthy population, as it is as yet unclear to what extent NIRS assessment of skeletal muscle mitochondrial capacity is related to other established measures of oxidative metabolism related to exercise, such as
\(\dot{V}\)O
2peak and EPOC in a normally active, healthy population. This information would further support the applicability and physiological relevance of NIRS assessment of mitochondrial capacity.
The aim of this study is to measure mitochondrial function using NIRS in a recreationally active, healthy population divided into relatively low- and relatively high-aerobic fitness groups and relate it to parameters of aerobic fitness. The recovery of m
\(\dot{V}\)O
2 in both the frequently activated gastrocnemius muscle and in the often undertrained forearm will be measured (Hamner et al.
2010). We hypothesised that the recovery of m
\(\dot{V}\)O
2 and whole body
\(\dot{V}\)O
2 recovery, i.e. EPOC, is faster in the relatively high-fitness group. Furthermore, we expect muscle and whole body
\(\dot{V}\)O
2 recoveries to correlate, since post-exercise replenishment of energy stores in the muscle encompasses an important component in the rapid phase of EPOC.
Materials and methods
Subjects
Healthy males between the age of 18 and 28 years were recruited from the local university and community population. None of the subjects had a history of cardiovascular, respiratory or metabolic disease. None of the subjects identified as regular smoker ( > 5 cigarettes per week) used recreational drugs during the study or reported recent use of performance enhancing drugs or supplements. Subjects were non-anaemic (haemoglobin concentration > 13 g/dL), verified using HemoCue Hb 201 microcuvette (HemoCue AB, Sweden). Main exercise modalities in high-fitness group were cycling (3x), lacrosse (2x), triathlon (1x), rowing (1x) and running (1x). Main exercise modalities in low-fitness group were sailing (1x), running (1x), weight lifting (1x), volleyball (1x) or no regular exercise at all. Only males were selected in this study due to the limited penetration depth of the NIRS device used and sex differences in subcutaneous adipose tissue thickness.
Pre-experimental screening protocol
Subjects were selected based on whole body peak oxygen uptake (\(\dot{V}\)O2peak) measured using an incremental exercise test on electrically braked bicycle ergometer (Corival CPET, Lode, The Netherlands). Subjects were asked to refrain from vigorous exercise for 48 h and to have consumed their last meal 2 h before this test. Oxygen consumption, carbon dioxide production and air flow were measured using MAX-II metabolic cart (AEI technologies, USA). Exhaled air was continuously sampled from a mixing chamber and heart rate was measured with a strap-on chest heart rate monitor (Polar Electro, Finland). After 3 min of warming-up, the protocol started at a workload of 75 W, or 125 W for subjects who exercised > 3 times a week, and was increased every minute in increments of 25 W. Subjects were instructed to maintain a self-selected pedal rate between 70 and 80 revolutions per minute (RPM). Inability to pedal at a rate above 60 RPM for 15 s was considered point of exhaustion and the end of the test. For the test to be valid, two out of three of the following criteria should have been met: (1) A maximal heart rate within 10 beats of the predicted maximum (220—age), (2) attainment of a plateau in \(\dot{V}\)O2, i.e. \(\dot{V}\)O2 failing to increase with 150 mL/min, despite an increase in work load, (3) achievement of an RER ≥ 1.1. \(\dot{V}\)O2peak was determined by binning data in 15-s intervals.
Eight relatively high-aerobic fitness (\(\dot{V}\)O2peak ≥ 57 mL/kg/min) and eight low-aerobic fitness subjects (\(\dot{V}\)O2peak ≤ 47 mL/kg/min) were selected to take part in the study, based on chosen cut-offs. A total of 24 subjects were screened to end up with the desired sample size.
Experimental protocol
All measurements were done fasted, i.e. subjects were not allowed to eat after 08:00 PM the night before. The subjects refrained from heavy physical exercise 48 h prior to testing and from any exercise and consumption of alcohol 24 h prior to testing. Maximal Voluntary Contraction (MVC) hand grip strength of the dominant hand was measured using a Jamar Hydraulic Hand Dynamometer (Performance Health, IL, USA). Highest value out of three 5-s isometric contractions was set as MVC. Body fat percentage was measured according to the four-site method by Durnin–Womersley using the skinfold measurements of the triceps, biceps, subscapula and suprailiac, measured using a skinfold caliper (Harpenden, UK). Furthermore, skinfold between NIRS receiver and transmitter was measured on the calf and the forearm.
NIRS measurements
Deoxyhaemoglobin (HHb) and oxyhaemoglobin (O
2Hb) were continuously measured using the continuous wave PortaMon wireless, dual-wavelength NIRS system (760 and 850 nm; PortaMon, Artinis Medical Systems, Netherlands). The 40-mm channel was used for analysis. Data were collected via bluetooth at 10 Hz using Oxysoft software (Artinis Medical Systems). The NIRS probe was placed longitudinally on the lateral gastrocnemius 4 cm distal to the knee joint and on the flexor digitorum superficialis (FDS). To secure the probe and protect it from environmental light, the probe was tightly taped to the skin. To measure oxygen consumption, a blood pressure cuff (Hokanson SC5 and SC12; D.E. Hokanson Inc., Bellevue, WA) was placed proximally of the probe above the knee joint and on the upper arm. The cuff was powered and controlled by a rapid cuff inflator system (Hokanson E20 and AG101 Air source; D.E. Hokanson Inc.) set to a pressure of 230–250 mm Hg. Post-exercise muscle oxygen consumption recovery was assessed similar to previously published protocols (Ryan et al.
2013). In summary, the protocol consists of three 30-s rest measurements of basal oxygen consumption. To calibrate the signal between individuals, the minimal oxygenation of the tissue underneath the probe was then determined by 30-s maximal hand grip exercise for FDS or by plantar flexion exercise using a rubber resistance band for gastrocnemius, followed by an arterial occlusion until baseline or with a maximum of 4 min total occlusion time. The hyperaemic response after the cuff was released was considered maximal oxygenation. Recovery oxygen consumption after exercise was measured after 30 s of intermittent handgrip exercise at 50% of MVC for the FDS or plantar flexion exercise using a rubber resistance band until 50% of maximal oxygenation for gastrocnemius. Right after exercise, a series of transient occlusions (5 × 5 s on/5 s off, 5 × 7 s on/7 s off, 10 × 10 s on/10 s off) was used to measure the recovery of muscle oxygen consumption after exercise. Recovery measurements were performed in duplicates with 2-min rest between tests.
Analysis of muscle oxygen consumption data
NIRS data were analysed using Matlab-based (The Mathworks, MA, USA) analysis software (NIRS_UGA, GA, USA). Data were analysed as 100% of maximal oxygenation. m
\(\dot{V}\)O
2 was calculated during every arterial occlusion using the slope of the change in HHb and O
2Hb (Hb difference) for 3 s for the 5-s occlusions, for 5 s for the 7-s occlusions, 7 s for the 10-s occlusions and 15 s for the basal measurements. A blood volume correction factor was used for each data point (Ryan et al.
2012) to correct for redistribution of blood distally from the cuff. In short, changes in HHb and O
2Hb should be proportional during arterial occlusions. A blood volume correction factor (
β) was calculated to account for possible changes and was used to correct each data point. m
\(\dot{V}\)O
2 recovery measurements post-exercise were fitted to a mono-exponential curve:
$$y~\left( t \right) = {\text{End}} - ~\Delta *e^{{ - k \cdot t}}$$
with Y representing the m\(\dot{V}\)O2 during the arterial occlusions, End being the m\(\dot{V}\)O2 immediately after the cessation of exercise, delta (∆) being the difference between m\(\dot{V}\)O2 after exercise and m\(\dot{V}\)O2 during rest, K being the rate constant expressed in time units, and t being time. Rate constants of duplicates were averaged. Rate constants calculated from curve fitting with R2 < 0.95 were excluded from analysis as a measure of poor data quality.
EPOC measurements
Basal oxygen consumption (method see below) was measured in supine position after an overnight fast. Subject was rested 30 min before the facemask was attached. After 20 min of basal measurement, the subject cycled for 20 min at a work rate adjusted to 55% of
\(\dot{V}\)O
2peak (Maresh et al.
1992). This protocol resulted in equal relative intensity for each subject. Upon cessation of exercise, subjects were placed in supine position for 20 min.
Analysis of whole body oxygen consumption data
Exhaled air was continuously sampled using a strap-on face mask, and binned in 15-s intervals. Due to the individualization of the exercise protocol, absolute oxygen consumption was different between subjects. The recovery of
\(\dot{V}\)O
2 expressed as a percentage of EPOC where the last 10 min of exercise was averaged and expressed as 100% EPOC and the values during the last 5 min of basal measurements were averaged and expressed as 0% EPOC (Short and Sedlock
1997). The recovery data were plotted using a two-phase exponential decay according to the formula:
$$Y = {\rm{Plateau}} + ({\rm{YOFast}}*\exp \left( { - K{\rm{Fast}}*X} \right) + ({\rm{YOslow}}*\exp \left( { - {\rm{KSlow}}*X} \right),$$
where
$${{\rm Y0}}{\text{Fast}} = \left( {{\rm Y0} - {\text{Plateau}}} \right)*{\text{PercentFast}}*0.01\;{\text{and}}$$
$${\rm YO}{\text{Slow}} = \left( {{\rm Y0} - {\text{Plateau}}} \right)*\left( {100 - {\text{PercentFast}}} \right)*0.01.$$
In the formula, Plateau represents the basal oxygen consumption, or 0% EPOC. Y0 represents the oxygen consumption during exercise, or 100% EPOC. KFast and KSlow represent the rate constants of recovery as an inverse unit of time. PercentFast is the fraction of the Y that is represented by the fast phase, as percent.
Statistical analyses
Data are presented as mean ± SD. Statistical analyses were performed using GraphPad Prism v.5 (GraphPad Software, CA, USA). Means between the two groups were compared using a Students unpaired t test. Correlations between variables were compared using regression analysis. Significance was accepted at p < 0.05. Normality was tested using Shapiro–Wilk normality test. Not normal data were compared using Mann–Whitney tests.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.