To determine whether conditions for O2 utilization and O2 off-loading from the hemoglobin are different in exercising arms and legs, six cross-country skiers participated in this study. Femoral and subclavian vein blood flow and gases were determined during skiing on a treadmill at ∼76% maximal O2 uptake (V̇o2 max) and at V̇o2 max with different techniques: diagonal stride (combined arm and leg exercise), double poling (predominantly arm exercise), and leg skiing (predominantly leg exercise). The percentage of O2 extraction was always higher for the legs than for the arms. At maximal exercise (diagonal stride), the corresponding mean values were 93 and 85% (n = 3; P < 0.05). During exercise, mean arm O2 extraction correlated with the Po2 value that causes hemoglobin to be 50% saturated (P50: r = 0.93, P < 0.05), but for a given value of P50, O2 extraction was always higher in the legs than in the arms. Mean capillary muscle O2 conductance of the arm during double poling was 14.5 (SD 2.6) ml·min−1·mmHg−1, and mean capillary Po2 was 47.7 (SD 2.6) mmHg. Corresponding values for the legs during maximal exercise were 48.3 (SD 13.0) ml·min−1·mmHg−1 and 33.8 (SD 2.6) mmHg, respectively. Because conditions for O2 off-loading from the hemoglobin are similar in leg and arm muscles, the observed differences in maximal arm and leg O2 extraction should be attributed to other factors, such as a higher heterogeneity in blood flow distribution, shorter mean transit time, smaller diffusing area, and larger diffusing distance, in arms than in legs.
- diffusing capacity
- oxygen extraction
muscular oxygen uptake depends on extrinsic factors such as O2 delivery and the intrinsic factors that regulate both the transfer of O2 from the erythrocytes to the mitochondria and the subsequent utilization of O2 in the mitochondria. However, the diffusive transfer of O2 is not only determined by intrinsic factors, because it also depends on mean capillary O2 tension. It is currently assumed that during exercise with a small muscle mass, intrinsic factors are the main determinants of peak local muscular V̇o2, because the O2 delivery is extraordinary high (3, 44, 61). During exercise with a large muscle mass, the V̇o2 peak of the lower extremities appears to be O2 delivery dependent (6, 7, 16, 33, 35, 57). O2 extraction across the lower extremities may reach maximal values between 90 and 92% of the arterial O2 content (CaO2), and the Po2 in the femoral vein may be close to 10 mmHg in active subjects (6, 7, 16), leaving little room for further extraction. However, in sedentary subjects, the maximal O2 extraction across the legs lies close to 70% of the CaO2 (59), implying that their peak muscular V̇o2 also may be limited by intrinsic factors (20). In physically active but non-arm-trained subjects, a low O2 extracting capacity has been reported for the arms (1, 11, 51, 70). Moreover, arm training resulted in only a marginal improvement in the O2 extraction of the arms (51). Therefore, the intrinsic factors may play an important role in limiting the maximal V̇o2 attainable during maximal arm exercise. One of these roles could be differences in muscle capillarization that may affect both mean transit time (MTT) and diffusion conditions in the muscle. Moreover, primarily on the basis of experiments using the isolated hindlimb preparation in rats, it has been shown that mitochondrial oxidative capacity could determine the rate of O2 utilization and thereby O2 extraction (22, 37, 58). However, the limiting factor may be different in humans as suggested by the fact that prolonged bed rest decreases mitochondrial oxidative capacity and V̇o2 max without reducing O2 extraction in humans (13, 62). Moreover, whole body V̇o2 increases by 6% with hyperoxia (fraction of inspired O2 = 0.5) during arm-cranking exercise (27). This observation is also compatible with a O2 delivery limitation of arm peak V̇o2. However, the effect of hyperoxia also could be explained, but only in part, by the increase in the amount of free O2 in the arterial blood in subjects without exercise-induced hypoxemia. Therefore, we hypothesize that O2 extraction across the arms is low because of a less efficient O2 off-loading from the hemoglobin compared with the legs.
The aim of this study was to determine, first, whether O2 extraction of the arm muscles is lower than that of leg muscles in humans with well-trained arm and leg muscles, at given systemic and regional absolute V̇o2. Second, we assessed whether the conditions for the O2 off-loading from the hemoglobin are different for the arms and legs during submaximal and maximal combined arm and leg exercise in arm- and leg-trained humans. Third, we assessed whether differences in muscle oxidative capacity could account for the differences in O2 extraction between arm and leg muscles. To pursue these aims, we studied a group of well-trained cross-country skiers.
Six elite cross-country skiers, age 24 (SD 4) yr, height 180 (SD 6) cm, and weight 74 (SD 6) kg, volunteered to participate in the study. One week before the experiment, their maximal O2 uptake (V̇o2 max) was 5.1 (SD 0.3) l/min or 72 (SD 4) ml·kg−1·min−1, assessed during an incremental intensity test to exhaustion. The incremental exercise test was carried out on skiers using the diagonal stride technique while skiing uphill with roller skis on a modified treadmill (Refox, Falun, Sweden). This V̇o2 max value is referred to as V̇o2 max DS. All subjects were informed about the possible risks and discomfort involved in the study before they gave their written consent to participate. This study was carried out according to the Declaration of Helsinki and was approved by the Ethical Committee of the Karolinska Institute, Stockholm, Sweden. The reason for choosing cross-country skiers was not only that they have well-trained arm and leg muscles but also that they are able to skillfully perform exercise with the upper or lower extremities, as well as exercise combining the upper and lower extremities in the upright position.
All subjects were familiar with the use of roller skies. On the experimental day, the subjects reported to the laboratory at 8:00 AM, and catheters were placed under local anesthesia (2% lidocaine) and advanced to the final position under fluoroscopic guidance, as previously described (9, 69). An 18-gauge catheter (Hydrocath; Ohmeda, Swindon, UK) was inserted percutaneously, using the Seldinger technique, into the left or right femoral artery 2–5 cm below the inguinal ligament and was advanced 5–10 cm in the proximal direction. This catheter was connected to a blood pressure transducer positioned at the height of the fourth intercostal space (T100209A; Baxter, Unterschleissheim, Germany) and was also used to sample arterial blood. A 20-gauge catheter was inserted in the left femoral vein 2 cm below the inguinal ligament and was advanced 5–7 cm in the distal direction for femoral venous blood sampling. In the right femoral vein, a venous catheter with side holes (Radiopack TFE; Cook, Bjaerverskov, Denmark) was inserted and advanced ∼5 cm proximal to the inguinal ligament for the injection of iced physiological saline solution. A thin polyethylene-coated thermistor (model 94-030-2.5F T.D. Edslab probe; Baxter, Irvine, CA) was inserted through the venous catheter for blood flow measurements by using the constant infusion thermodilution technique (3). An additional 18-gauge catheter also was inserted into the left femoral vein 2–3 cm below the inguinal ligament and was advanced under fluoroscopic guidance until the tip was positioned in the center of the right atrium, to sample blood from the right atrium. The last catheter, a Swan-Ganz triple-lumen catheter (model 132F5 Edslab) was inserted into an antecubital vein and, under fluoroscopic guidance, was advanced into the subclavian vein until the tip was positioned at 5 cm before the merger with the jugular vein. One lumen was used for blood sampling and another for infusion of iced saline solution for blood flow measurements. Infusate temperatures were measured with a thermistor set in a flow-through chamber (model 93-505 Edslab) connected to the venous catheters. All sampling catheters were connected to a three-way stopcock and, along with the thermistor, were sutured to the skin to minimize the risk of movement during exercise.
Classic skiing involves different techniques. The diagonal stride technique involves both the arms and the legs and is used uphill (Fig. 1). The double poling technique mainly involves the upper body (arms) and is used on flat terrain and slightly uphill. Leg skiing is diagonal skiing without poles, which means that, in contrast to diagonal skiing, all the propulsive forces are generated only by the legs. The protocol consisted of 40 min of continuous diagonal stride (referred to as continuous arm + leg) at 77% of V̇o2 max DS, followed without breaks by 10 min of double poling (arm) at 74% of V̇o2 max DS, 10 min of diagonal stride (arm + leg) at 77% of V̇o2 max DS, and 10 min of leg skiing (leg) at 78% of V̇o2 max DS. These conditions elicited a whole body V̇o2 close to 4 l/min. The speed of the treadmill was then reduced, and after 3 min of active recovery, while subjects skied with the diagonal technique at 30–40% of V̇o2 max DS, the speed and slope of the treadmill was increased every minute until subjects reached exhaustion, which occurred within 6–8 min. Blood samples were taken after 21, 24, and 36 min of continuous arm + leg skiing and then ∼5–7 min after the start of arm, arm + leg, and leg skiing, respectively.
At the end of the exercise, the position of the subclavian vein and right atrium catheters was checked with fluoroscopy. No catheter was found displaced during the experiment.
Whole body V̇o2, CO2 production, and ventilation were measured continuously using an ergo spirometry system (AMIS 2001; Innovision, Odense, Denmark), which was calibrated with high-precision gases (16.00 ± 0.04% O2 and 4.00 ± 0.1% CO2; Air Liquide, Kungsängen, Sweden). During submaximal exercise, the V̇o2 values obtained during the last 4 min were averaged. The V̇o2 max was calculated as the average of the three highest 10-s consecutive measurements of O2 uptake.
Femoral and subclavian venous blood flow were measured using constant-infusion thermodilution, as described in detail elsewhere (3). Briefly, iced saline was infused (Harvard pump; Harvard Apparatus, Millis, MA) through the femoral and subclavian veins simultaneously at flow rates sufficient to decrease blood temperature at the thermistor by 0.5–1°C. At rest, saline infusions were continued for at least 60 s, whereas during exercise, infusions 15–20 s long were used until femoral vein temperature had stabilized at its new lower value. Blood flow was calculated on the basis of thermal balance principles, as detailed by Andersen and Saltin (3). Resting blood flow and arterial blood pressure were measured six times over 60 min and averaged. During submaximal exercise, blood flow measurements were performed in duplicate. The reported submaximal blood flow values represent the average of at least two measurements. At peak effort, the measurements were made within 1 min of exhaustion.
Leg and Arm V̇o2.
Leg and arm V̇o2 values were computed separately using the Fick method, i.e., leg V̇o2 = leg blood flow × (CaO2 − Cfv), and arm V̇o2 = subclavian vein blood flow × (CaO2 − Csv), where Cfv represents the O2 content in the femoral vein and Csv represents the O2 content in the subclavian vein.
Histochemical and enzymatic analysis.
Serial transverse sections were stained for myofibrillar ATPase as described by Brooke and Kaiser (5). Muscle capillary density was analyzed, visualized, and quantified as described by Qu et al. (50). Muscle biopsies were analyzed for citrate synthase (CS) and 3-hydroxyacyl-CoA dehydrogenase (HAD) activity (69).
Blood samples and analytical procedures.
Blood was sampled anaerobically in heparinized syringes and immediately analyzed for hemoglobin (Hb), O2 saturation (OSM3 hemoxymeter; Radiometer, Copenhagen, Denmark), blood pH, CO2, and O2 tension (ABL5; Radiometer). Blood gases were corrected for measured femoral vein blood temperature (femoral venous and arterial blood gases) and subclavian vein blood temperature (subclavian venous blood gases). Blood O2 content was computed from the saturation (So2) and Hb concentration ([Hb]), i.e., (1.34 × [Hb] × So2) + (0.003 × Po2). Another blood sample was taken, and the blood was collected in ice-cold tubes that contained 10 μl of 0.33 M EDTA per milliliter of blood and was immediately centrifuged at 4°C for 10 min and stored at −50°C until analysis. Plasma was analyzed enzymatically for lactate (Roche Unikit; Hoffman-LaRoche, Basel, Germany) on an automatic analyzer (Cobas Fara; Roche Diagnostics, Basel, Switzerland). Before biochemical analysis, muscle biopsy samples were freeze-dried and dissected free of connective tissue, visible fat, and blood with the use of a stereomicroscope. The standard P50, defined as the value of Po2 that causes Hb to be saturated by 50% when the O2Hb equilibration curve is determined at 37°C, pH = 7.40, Pco2 = 40 mmHg, was calculated from the whole set of arterial and venous gases obtained in each experiment. The in vivo P50 was calculated using Kelman's equation (32). The in vivo P50 is the Po2 value that causes Hb to be saturated at 50% at the temperature, Pco2, and pH of the blood in the femoral and subclavian veins during exercise. Arm and leg muscle O2 conductance and mean capillary Po2 values were determined as previously described by Wagner (71, 73).
Capillary muscle O2 conductance and mean capillary Po2.
To calculate the diffusing capacity for O2 (D), we used an iterative numerical integration procedure to find the value of O2 conductance (i.e., in ml·min−1·mmHg−1) that yields the measured femoral muscle venous Po2. The calculation of D assumes 1) that the intracellular Po2 is negligibly small at V̇o2 max (15, 54), 2) that the O2 remaining in the femoral and subclavian venous blood is wholly accountable for by diffusion limitation of O2 from the microcirculation to the mitochondria, and 3) that perfusion/V̇o2 heterogeneity and perfusional or diffusional shunt are negligible. Mean capillary Po2 is the numerical average of all computed Po2 values, equally spaced in time, along the capillary from the arterial to venous end (49).
Descriptive statistics were performed on each variable to confirm the assumptions of normality and homoscedasticity. The effect of the skiing technique during submaximal exercise on the dependent variables was assessed using a one-way repeated-measures analysis of variance (ANOVA). Mauchly's test of sphericity was run before the ANOVA, and in case of violation of the sphericity assumption, the degrees of freedom were adjusted according to the Huynh and Feldt test. Pairwise comparisons were carried out with Tukey's test. The relationship between muscular O2 conductance and O2 extraction was determined using linear regression. Maximal exercise values were obtained in only three subjects. Thus we decided not to perform comparisons between submaximal and maximal exercise. However, to test for differences between arm and leg variables during maximal exercise in these three subjects, we used a paired Student's t-test. The significance level was set at P < 0.05. Data are expressed as means ± SD, unless otherwise stated.
O2 delivery and consumption.
During submaximal exercise, whole body V̇o2 was similar during the different exercise conditions at 4 l/min, which represented ∼76% of V̇o2 max measured during diagonal skiing (V̇o2 max DS; Table 1). At peak exercise, a V̇o2 close to 95% of V̇o2 max DS was reached. The contribution of the legs to whole body V̇o2 varied between 46% (during double poling) and 67% (during leg skiing). Conversely, for the arms, the corresponding values were 37 and 9%, respectively (Table 1). V̇o2 peak during double poling is a little lower (∼86% of the V̇o2 max DS) (Holmberg et al., unpublished observations). This means that compared with the specific double poling V̇o2 peak, the skiers were working at 86% of the double poling V̇o2 peak.
Activities of muscle oxidative enzymes.
The effect of the exercise on the blood gases, pH, and lactate is depicted in Table 2. The systemic and leg O2 content arterial-venous differences (a-vDif) were lower during arm exercise compared with the other exercise modes, whereas the arm O2 a-vDif was lowest during leg skiing (Fig. 2A). Regardless of the exercise mode, the percentage of O2 extraction was always higher for the legs than for the arms, even during double poling (Fig. 2B). During diagonal skiing, the femoral vein O2 extraction values were slightly greater when the exercise was preceded by the double poling bout than during the 40 min of continuous diagonal skiing (Fig. 2B).
The exercise Po2 and O2 content values were always lower in the femoral vein than in the subclavian vein, whereas the O2 tension and content values in the blood of the right atrium were in between (Fig. 3, A and B). Even when the subjects used the double poling technique, which elicited a similar V̇o2 in the arms and legs (Table 1), the Po2 and O2 content values were lower in the femoral vein than in the subclavian vein (Fig. 3, A and B). Leg O2 extraction was closely related to arm O2 extraction during submaximal exercise with the diagonal stride technique (r = 0.89, P < 0.05). As shown in Fig. 4A, leg mean O2 extraction correlated closely with mean leg V̇o2 across conditions (r = 0.89, P < 0.05; n = 6), whereas this relationship was not significant at the arm level (r = 0.61, P = 0.28; n = 6). During arm exercise, there also was a close relationship between the mean O2 extraction and the corresponding P50 value (r = 0.93, P < 0.05) (Fig. 4B).
O2 off-loading and arm diffusing capacity.
As shown in Table 3, during the 40 min of continuous diagonal skiing, the in vivo P50 value was lower than during the second bout of diagonal exercise, which was performed just after the double poling (33.7 (SD 1.2) and 35.2 (SD 1.5) mmHg, respectively. During combined leg and arm exercise, the blood pH was similar in the femoral and subclavian veins (Fig. 3C). However, during double poling, the blood pH was lower in the subclavian than in the femoral vein, whereas the opposite was true during leg skiing. The lower extraction capacity of the arm was not related to differences in the variables that regulate the unloading of O2 from the Hb, because the degree of acidification of blood, blood temperature (not shown), and Pco2 (Fig. 3D) were very similar in the effluent blood from the legs and arms. However, compared with the subclavian vein, a small, but significantly higher blood temperature (+0.6 degrees) and Pco2 (6–10 mmHg) were observed in the femoral vein during leg skiing. Thus the calculated in vivo P50 value was almost the same in the effluent blood from the arm muscles during double poling and the effluent blood from the leg muscles during leg skiing (35.9 (SD 1.7) and 35.8 (SD 1.5) mmHg, respectively, P = 0.66) (Table 3). Despite this high P50 value in the arms, which was significantly higher than the 34.6 (SD 1.5) mmHg observed in the legs, the O2 extraction was markedly higher in the legs than in the arms during double poling (Fig. 2B).
At maximal exercise with the diagonal technique, the arm and leg in vivo P50 values were 38.8 (SD 4.3) and 37.5 (SD 3.1) mmHg (mean of 3 subjects). In the latter condition, one skier was able to extract 97 and 89% of the O2 supplied to legs and the arms, respectively. This superb O2 extraction was achieved with in vivo P50 values of 41.0 and 43.7 mmHg, respectively.
The calculated mean capillary muscle O2 conductance of the arm muscles during double poling was 14.5 (SD 2.6) ml·min−1·mmHg−1, and the mean capillary Po2 was 47.7 (SD 2.6) mmHg. The corresponding values for the leg muscles during maximal exercise with the diagonal technique were 48.3 (SD 13.0) ml·min−1·mmHg−1 and 33.8 (SD 2.6) mmHg. Arm muscle capillary O2 conductance tended to correlate with maximal arm O2 extraction percentage (r = 0.76, P = 0.08).
This study shows that in arms, but not in legs, mean O2 extraction is closely related to the mean in vivo P50 value. Moreover, for a given P50 value, the upper extremities extract less O2 than the lower extremities in humans with highly trained arm and leg muscles during exercise. This lower O2 extraction is associated with a lower O2 conductance in the upper compared with the lower extremities. Hence, for a given O2 demand, a greater O2 delivery is needed for exercising arm than leg muscles, which is the cause of the relatively high blood flow to the arms.
Because O2 extraction increases with exercise intensity, and the relative intensity during double poling was ∼86% of double poling V̇o2 peak, O2 extraction should have been higher for the arms than for the legs. However, the experimental findings showed higher extraction in the legs than in the arms. This aspect confers additional robustness to our findings.
Arm O2 extraction in trained and untrained muscles.
Rasmussen et al. (51) reported mean axillary vein O2 saturation values of 38% (extraction: 60%) during arm cranking at an intensity eliciting a heart rate of 170 beats/min. In the current investigation, the subclavian vein saturation during submaximal exercise achieved mean values between 24 and 31% (extraction: 68–75%). The high arm O2 extraction capacity in the current investigation is likely the result of several years of regular training, given that Rasmussen et al. (51) reported that after 5 wk of intensive arm training (1 h × 5 days/wk), the extraction capacity of the arm muscles was only slightly improved. During maximal exercise, the arm muscles of our skiers extracted 85% of the O2 supplied, which is higher than reported during maximal cycle ergometry in the leg muscles of untrained people before (59) and even after 9 wk of a training program resulting in a 35% higher V̇o2 max (59). These skiers attained in their arms the same O2 extracting values as we have observed in the legs of physically active subjects (V̇o2 max between 47 and 63 ml·kg−1·min−1) during upright maximal leg cycle ergometry (7, 68). Roca et al. (59) reported maximal leg extraction values of 72% in sedentary subjects that increased to 82% after 9 wk of training. Compared with these values, our skiers also reached remarkably high maximal O2 extraction values in their legs (93%). However, untrained subjects may reach rather high leg O2 extractions after bed rest (62) or during ischemic exercise (46).
Several factors may account for the observed differences in muscular O2 extraction, which depends on the interaction of the following: 1) kinetics of O2 off-loading from the Hb; 2) capillary muscle O2 conductance (72); 3) blood flow, mean transit time (47), and degree of mismatch between the metabolic demand and blood flow distribution and/or degree of shunt (48); and 4) muscle maximal oxidative capacity (22) and exercise intensity.
Kinetics of O2 off-loading from Hb.
The off-loading of O2 from Hb during exercise is facilitated by acidification of the blood and the increase of temperature and Pco2 (67). Experiments in dogs have shown that O2 extraction at V̇o2 max decreases when the P50 of the Hb is reduced from its normal value of 32 to 23 mmHg (24), whereas it increases when the P50 is raised to 53.2 mmHg (56). In agreement, we observed a close correlation between O2 extraction and P50. Moreover, during diagonal skiing, slightly higher femoral vein O2 extraction values were observed after than before the double poling bout, which elicited a high degree of blood acidification, despite similar V̇o2 and leg blood flows in both phases of diagonal skiing. The latter implies that for a given V̇o2, facilitating the off-loading of O2 from the Hb is associated with increased arm and leg O2 extraction values during exercise. Thus our experiment shows that the conditions for the off-loading of O2 from the Hb may influence O2 extraction in the arm and leg muscles. However, despite small differences in blood temperature and the greater release of lactate and protons during the arm exercise (29, 69), the conditions at which the O2 dissociation curve of Hb operates are similar in the arm and leg muscles (see Fig. 3). Thus the lower arm O2 extraction capacity cannot be explained by a slower off-loading of O2 from the Hb in the arms, because the in vivo affinity of Hb for O2 was similar in the arms and legs during both submaximal and maximal exercise.
Capillary muscle Po2 and O2 conductance.
In this investigation we calculated the capillary muscle O2 conductance by assuming that mitochondrial Po2 is very close to 0–2 mmHg during double poling and during maximal exercise with the diagonal technique. This assumption is likely to be true in muscles working near their maximum (53, 64). Muscle mass is a major determinant of the muscle capillary O2 conductance. Although it is impossible to know exactly what was the actual muscle mass recruited during the double poling, some estimations can be made. The mean capillary O2 conductance of the whole arm of our skiers was 14.5 ml·min−1·mmHg−1, i.e., a bit lower than the value of 16.6 ml·min−1·mmHg−1 reported for the leg muscles in 24 sedentary subjects with a mean body weight of 70 kg (including 7 women) (10). Assuming that the mean muscle mass of the skiers arms was 4–5 kg and that the sedentary subjects had a leg muscle mass of 8 kg, it can be estimated that the skiers had ∼40–75% higher muscle capillary O2 conductance in their arm muscles than the sedentary subjects studied by Cardus et al. (10) had in their legs. In physically active subjects we have obtained values of 25 ml·min−1·mmHg−1 during knee extensor exercise (57), 42.2 ml·min−1·mmHg−1 during maximal exercise on the cycle ergometer in physically active subjects (8), and 55 ml·min−1·mmHg−1 in amateur cyclists (16). Together, these data indicate that cross-country skiers have a rather high capillary muscle O2 conductance in their arms, similar to that observed in the legs of physically active subjects but lower than reported in the leg muscles of well-trained subjects having almost the same maximal cardiac output as our skiers (16). Using the data obtained at maximal exercise, we calculated that the muscle mass-normalized leg capillary muscle O2 conductance is at least 50% higher in the leg muscles than in the arm muscles, assuming an arm muscle mass of 4–5 kg and a leg muscle mass of 10–11 kg. These data, combined with the positive relationship observed between the percentage of arm O2 extraction and capillary muscle O2 conductance, suggest that the lower arm-to-leg capillary muscle O2 conductance may explain part of the lower extraction capacity of the arm muscles than the leg muscles in the elite cross-country skiers. However, the question remains, why is capillary muscle O2 conductance lower in the arm than in the leg muscles?
MTT and heterogeneity in blood flow distribution.
A larger MTT may facilitate the transfer of O2 from the erythrocyte to the mitochondria (47). The MTT of the erythrocytes crossing the muscle capillaries during maximal exercise is estimated as MTT = CBV/MBF, where CBV is the capillary blood volume and MBF is the muscle blood flow. In turn, the capillary volume may be calculated from the capillary density (31). We did not measure the capillary density in the current investigation, but from previous studies in elite cross-country skiers of the same level of performance as the present subjects, it appears that the capillary density is similar in arm and leg muscles (Table 4) (38, 39, 42). With a mean capillary density of 500 capillaries in arm and leg muscles of elite cross-country skiers, a similar MTT through the capillaries is expected in the muscles of the upper and lower extremities. For example, if at maximal exercise the active muscle mass is 4 and 11 kg in the arm and legs, respectively, the calculated MTT values will be 672 and 674 ms, respectively (31). However, the corresponding values will be 1,195 and 1,198 ms if an inner mean capillary diameter of 6.0 μm rather than 4.5 μm is assumed. What these numbers suggest is that MTT is rather high in both upper and lower extremities at maximal exercise. At first sight, MTT does not appear to explain the observed differences in O2 extraction between the arms and legs. However, heterogeneity in macrovascular MTT has been related to O2 extraction during low exercise intensity in humans (30). Nevertheless, macrovascular MTT heterogeneity decreases with exercise intensity, but the degree of skeletal muscle capillary MTT heterogeneity existing at maximal exercise remains unknown. In theory, MTT values lying between 0.3 and a little more than 1 s are possible (63).
In our experimental conditions, where exercise is performed at maximal intensity, there is no reason to suspect a mismatch, shunt, or differences in capillary length between the perfused and the active fibers (72) in arm and leg muscles (14, 26). However, we cannot rule out a greater degree of heterogeneity in the distribution of blood flow between muscles in the arms than in the legs. In fact, electromyographic activity data show large interindividual differences in the degree of recruitment of upper arm muscles during double poling (25). Lower O2 extraction may be expected in the muscles that are less activated, hence also contributing to a low mean O2 extraction of the whole extremity. Moreover, upper compared with lower extremities differ in degree of freedom of movement. Leg flexion and extension is performed in a rigid back and forward pattern with only minor variation in the activation of the different muscles and their functional portions. Arm and shoulder muscles, on the other hand, perform their movements in different positions, resulting in larger variation in the degree of muscle recruitment. Hence the mass of active muscles in the arm and shoulder region may be overestimated in cross-country skiing. In turn, the actual MTT across the active arm muscles may be shorter than the calculated value of 672 ms if flow is passing through a smaller capillary bed. In case of a lower active muscle mass in the upper extremity than assumed in this study, the real skeletal muscle O2 conductance of the arms would have been underestimated. However, as explained below, this fact cannot explain alone the observed differences in muscle O2 conductance between arms and legs.
The transfer of O2 from the Hb to the mitochondria should overcome the resistance imposed by the elements interposed between the Hb molecule and the mitochondrial matrix. It seems that most of the resistance to the diffusion of O2 from the Hb to the mitochondria relies on the capillary wall (23, 24, 56). Increasing the number of capillaries around each fiber (i.e., the amount of capillary-to-muscle fiber exchange surface) is likely the most efficient way to improve the muscle conductance for O2 (72). In fact, endurance training induces an increase in the number of capillaries (2). Capillary proliferation appears more accentuated in elite endurance athletes (60, 63, 65) and the longer the time of sport participation (60). Conditions in which the conductance of O2 has been lowered by reducing the mitochondrial oxidative capacity (36) or by eliminating the presence of myoglobin in the sarcoplasm (18) are associated with capillary proliferation as a compensation mechanism. Although it seems that capillary density is similar in the leg and arms muscles of cross-country skiers (Table 4), the capillary-to-fiber perimeter index (mean number of capillaries around each fiber/mean muscle fiber perimeter) is slightly higher in the leg than in the arm muscles because of the higher mean fiber cross-sectional area of the arm muscles. Using the data in Table 4, we have calculated a mean capillary index of 0.011 and 0.010 capillaries per micrometer of capillary perimeter in the leg and arm muscles, respectively. Thus the available morphometric data suggest that the diffusional area is lower and the diffusional distance longer in the arm than in the leg muscles.
Muscle oxidative capacity.
Despite the great amount of training performed by these subjects with the arm muscles, deltoid muscle oxidative capacity was lower than vastus lateralis oxidative capacity, as reflected by the 25 and 11% lower CS and HAD activities in the deltoid compared with the vastus lateralis muscles of our skiers (69). However, no relationship was observed between capillary muscle O2 conductance and the activities of CS and HAD. Previous studies also have reported lower CS and HAD activities in the triceps brachii than in the vastus lateralis in cross-country skiers (39) (Table 4). It is difficult to establish to what extent the role of this reduced oxidative capacity may be in the lower O2 extraction capacity of the arms, but some animal in vitro experiments support this possibility (22, 37, 58) while others do not (36). In humans, however, prolonged bed rest results in reduced mitochondrial volume and muscle oxidative enzyme activities while O2 extraction during maximal exercise is elevated (13, 62). In addition, the fact that V̇o2 max increases with hyperoxia in the trained quadriceps of elite athletes (52) and patients with restricted physical activity due to respiratory insufficiency (55) suggests that humans, regardless of trained status (21), do not use their maximal mitochondrial oxidative capacity even when exercising at V̇o2 max with a small muscle mass (27). In fact, it has been estimated that only 20% of the muscle mass has to be recruited to tax the maximal oxidative capacity in humans (4). This corresponds well with the observation made in rats that maximal muscle O2 extraction remains unchanged despite a 45 and 30% reduction in cytochrome oxidase and of CS activity, respectively (36). Despite the fact that the arm oxidative capacity of the elite cross-country skiers is lower than their leg muscle oxidative capacity (39), it is likely far in excess of their respective maximal O2 demand (4). In support, we can argue that rowing V̇o2 max increases with hyperoxia in elite rowers, a discipline in which there is a substantial contribution of the arms to the exercise V̇o2 max (41, 45). In addition, Clausen et al. (11) showed that after 5 wk of leg training, the whole body V̇o2 during maximal arm cranking increased by 10%, whereas the arm arterial-venous difference in O2 content remain unchanged. Thus our data and those summarized in Table 4 suggest that the small difference in oxidative capacity between arm and leg muscles has a minor role, if any, in the lower maximal arm O2 extraction capacity.
Increased arm vascular reactivity?
Under resting conditions, the forearms have higher vasodilatory responsiveness to acetylcholine, substance P, nitroprusside (40), and isoproterenol (28) and lower vasoconstricting responsiveness to phenylephrine (α1 agonist) (43). These functional differences between arms and legs suggest increased ability to maintain vasodilation in the arms than in the legs under some level of ongoing sympathetic activation. Thus our data could be interpreted to indicate that the lower O2 extraction in the arms only reflects a compensation for the relatively greater perfusion of the arms compared with the legs. However, if the reason for the lower O2 extraction of the arms was that the arms vasodilated in excess, then O2 extraction should have been similar in arms and legs at peak exercise with the diagonal style, because during this condition, perfusion of the arm and legs is limited by the pumping capacity of the heart (9). Nevertheless, peak arm blood flow reached a much lower value during maximal combined arm and leg exercise than during double poling, while O2 extraction remained at a lower level in the arms than in the legs. Although the reported higher vascular reactivity of the forearm at rest may facilitate the vasodilatory response in the arms during combined arm and leg exercise, it remains unknown whether vascular reactivity remains higher in the arm than in the leg muscles during exercise. It may be that it does not, given that vascular reactivity during exercise is modulated by the compound action of several vasodilating and vasoconstricting agents (12). For example, despite a similar sympathetic activation in arms and legs in response to the cold pressor test, vascular resistance increases in the arms but not in the legs (28). Thus it is possible that during near-maximal combined arm and leg exercise, the exercise-induced elevation of muscle sympathetic nerve activity is likely more efficiently counteracted in the legs than in the arms, i.e., reducing the vasodilatory response of the arms more than that of the legs (66). This mechanism will give the perfusing priority to the main locomotory muscles. However, in case of competition for blood flow between the leg muscles and the respiratory muscles, the perfusion priority is likely given to the respiratory muscles (19).
Although limited by the fact that O2 extraction at maximal exercise intensity was not determined during isolated exercise with either the upper and the lower extremities, our results are clear: O2 extraction across the upper extremities is lower than oxygen extraction across the lower extremities at a given absolute and relative exercise intensity. Given the fact that venous effluent blood from either extremity drains blood principally coming from the active muscles, the skin, and the bone marrow, the calculated extraction values represent the mean extraction capacity of the whole extremity. In theory, the skeletal muscles of both extremities could have a similar O2 extraction capacity if the degree of “venous admixture,” i.e., the amount of blood coming from the bone marrow and the cutaneous circulation at the sampling point, is higher in the arms than in the legs. However, the magnitude of the differences in fractional O2 extraction between legs and arms is too high to be accounted for by different degrees of venous admixture. Of all potential contributors to venous admixture, skin blood flow is quantitatively the most important. Skin blood flow increases with body temperature. In these experiments we took measurements of venous Po2 and O2 extraction after 12 min of diagonal skiing, and they were repeated after 24 and 36 min of diagonal skiing (data not shown). During this period, exercise intensity was constant, but body temperature increased significantly from 38.1 to 38.6°C. Despite this increase in body temperature, and likely in skin blood flow (17), no increase in venous Po2 or reduction in O2 extraction was observed in arms or legs. This suggests that the degree of admixture is small and is not accentuated by a condition that increases skin blood flow. Moreover, we have calculated that during double poling, venous admixture in the upper extremities should have been at least 1.1 l/min greater than in the lower extremities in case the skeletal muscle O2 extraction values in arm muscles during double poling were similar to the O2 extraction in the leg muscles during leg skiing. This means that total venous admixture should have been 2.2 l/min higher in the upper than in the lower extremities, which does not seem reasonable because it would imply an impossible regional difference in cutaneous vasodilation between the upper and the lower extremities (34). Moreover, forearm exercise experiments were performed in the 1960s with venous catheters placed in deep forearm veins (under X-ray control) to avoid skin contamination and with a cuff inflated around the wrist at >200 mmHg to exclude the circulation of the hand (74). In these experiments, maximal exercise extraction values were lower than reported in the current investigation for the whole upper extremity (74).
Even with our state-of-the-art technology it is impossible to know the accurate amount of muscle mass activated during arm or leg exercise. Even worse, it is not possible to know to what extent the motor units are activated. The latter limitation also applies to human models of localized exercise such as the leg extension model. This problem holds for both extremities. The only way to circumvent this limitation is by using animal preparations, which may be stimulated electrically to activate maximally all motor units. However, these preparations show lower D values than reported in vivo simply because with these preparations V̇o2 peak values in most models are ∼50% of the real in vivo V̇o2 peak values. Our lower D values agree well with the observation of lower O2 extraction capacity for the arms. For an active muscle mass of 5 and 11 kg in arms and legs, a local muscle V̇o2 of 138 and 147 ml/kg muscle mass can be calculated in the arms during double poling and in the legs during maximal diagonal stride, respectively. The corresponding “muscle-mass-normalized D” is 2.9 and 4.4 ml·min−1·mmHg−1·kg muscle mass−1. Arm muscle mass-normalized DO2 would have matched leg muscle-mass-normalized D for an active muscle mass of 3.3 kg in the arms and 11 kg in the legs. This proportion 3.3/11 (arm/leg active muscle mass) allows for a match in muscle mass-normalized D between arm and leg muscles but is simply impossible, because the V̇o2 peak per kilogram of active muscle mass would have been 209 and 147 ml/kg, i.e., 42% higher in the arm than in leg muscles. Thus only a lower value of muscle mass-normalized D in the arm muscles fits both our results and the current scientific knowledge.
In summary, the present investigation shows that O2 extraction is closely related to the in vivo P50 during arm exercise in humans with highly endurance trained arm and leg muscles. However, for a given P50 value, the arms have lower O2 extraction and lower capillary muscle O2 conductance values than the leg muscles. Because the conditions for the O2 off-loading from the Hb are similar in leg and arm muscles, the observed differences in maximal arm and leg O2 extraction should be attributed to other factors, such as a higher diffusing distance and higher heterogeneity in the distribution of blood flow between muscles or functional portions of the same muscle, or lower mean transit time and lower diffusing area in the arms than in the legs. These findings have implications to explain the differences in the cardiorespiratory response to arm vs. leg exercise, because the lower the muscular extraction capacity, the greater should be the blood flow, and hence the cardiovascular effort, to maintain a given metabolic rate.
This study was supported by grants from the Swedish Olympic Committee, Team Danmark, and the Sport Research Council of the Ministry of Culture. The Copenhagen Muscle Research Center is funded by Danish National Research Foundation Grant 504-14.
We highly appreciate help from the staff and the use of facilities at the Department of Cardiology and Clinical Physiology, Karolinska Hospital, Sweden. We also acknowledge the excellent technical assistance provided by Birgitte Jessen and Karen Juel.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society