Systemic and vastus lateralis muscle blood flow and O₂ extraction during ramp incremental cycle exercise

Abstract

During ramp incremental cycling exercise increases in pulmonary O₂ uptake (V̇o_2p) are matched by a linear increase in systemic cardiac output (Q). However, it has been suggested that blood flow in the active muscle microvasculature does not display similar linearity in blood flow relative to metabolic demand. This study simultaneously examined both systemic and regional (microvascular) blood flow and O₂ extraction during incremental cycling exercise. Ten young men (V̇o_{2 peak} = 4.2 ± 0.5 l/min) and 10 young women (V̇o_{2 peak} = 3.2 ± 0.5 l/min) were recruited to perform two maximal incremental cycling tests on separate days. The acetylene open-circuit technique and mass spectrometry and volume turbine were used to measure Q (every minute) and breath-by-breath V̇o_2p, respectively; systemic arterio-venous O₂ difference (a-vO_2diff) was calculated as V̇o_2p/Q on a minute-by-minute basis. Changes in near-infrared spectroscopy-derived muscle deoxygenation (Δ[HHb]) were used (in combination with V̇o_2p data) to estimate the profiles of peripheral O₂ extraction and blood flow of the active muscle microvasculature. The systemic Q-to-V̇o_2p relationship was linear (∼5.8 l/min increase in Q for a 1 l/min increase in V̇o_2p) with a-vO_2diff displaying a hyperbolic response as exercise intensity increased toward V̇o_{2 peak}. The peripheral blood flow response profile was described by an inverted sigmoid curve, indicating nonlinear responses relative to metabolic demand. The Δ[HHb] profile increased linearly with absolute V̇o_2p until high-intensity exercise, thereafter displaying a “near-plateau”. Results indicate that systemic blood flow and thus O₂ delivery does not reflect the profile of blood flow changes at the level of the microvasculature.

several studies have demonstrated that increments in exercise intensity, and thus oxygen uptake (V̇o₂), are accompanied by a linear increase in systemic blood flow [cardiac output (Q)] to match oxygen (O₂) supply to the O₂ demand within the active tissues (6, 16, 34, 38). According to the Fick equation [V̇o₂ = Q × arterial-venous O₂ difference (a-vO_2diff)] , a linear relationship between whole body Q and V̇o₂ (with a positive intercept), implies a hyperbolic response of the a-vO_2diff and, indeed, such a hyperbolic relationship in systemic O₂ extraction has been demonstrated during incremental maximal cycling exercise (10, 24, 36).

The profiles of systemic Q and a-vO_2diff do not necessarily reflect the dynamic adjustments of blood flow to, and O₂ extraction within, the active tissues. The mechanisms regulating peripheral skeletal muscle hyperemia during incremental exercise have been mainly attributed to the interplay of locally induced vasodilators and sympathetically mediated vasoconstrictors (4, 5, 11, 12, 42, 45, 46). Regional Q and a-vO_2diff are also related to exercise intensity and active muscle mass recruited (11, 45). During incremental exercise, these changes occur to match local O₂ delivery requirements with those of metabolism. To increase O₂ delivery (and thus active muscle blood flow [Q_m]), systemic increases in Q and arterial pressure and systemic reductions in blood flow to nonexercising tissue occur. As oxidative demands of the active tissue increase, Q_m to the vascular beds is regulated by changes in perfusion pressure and vascular tone (14). However, measurements of microvascular blood flow and O₂ extraction dynamics at the level of the active tissue are needed before the mechanisms regulating peripheral skeletal muscle hyperemia during incremental cycling exercise in humans can be further elucidated.

Much of what is known regarding peripheral hemodynamics (Q_m) during maximal exercise in humans has been determined from thermodilution blood flow measurements during cycling (10, 12, 24, 28), knee-extension (1, 5, 35, 36), and skiing exercise (11). Assessments of femoral blood flow (FBF) and O₂ extraction during cycling and knee-extension exercise have depicted a linear FBF to leg V̇o₂ relationship (similar to the systemic Q-to-V̇o₂ relationship) with fractional O₂ extraction displaying asymptotic behaviors at ∼80– 85% of maximal O₂ extraction (10, 24, 28, 36). Recently, the use of near-infrared spectroscopy (NIRS) has permitted the assessment of O₂ extraction from the vastus lateralis (VL) muscle, as reflected by changes in the NIRS-derived deoxy-hemoglobin signal ([HHb]), within the exercising muscle microvasculature. Using this technique, two recent studies have estimated the dynamic response of O₂ extraction within the area of NIRS interrogation and have suggested a nonlinear relationship between Q_m and muscle V̇o₂ (V̇o_2m) (9, 17). However, no studies to date have simultaneously described the dynamic adjustment of estimates of systemic Q and a-vO_2diff, and NIRS-derived peripheral (microvascular) blood flow and local muscle O₂ extraction using the same protocol within the same individuals.

Reliable measures of Q during ramp incremental exercise can be obtained noninvasively using the acetylene (C₂H₂) open circuit technique (23) and in combination with measures of pulmonary V̇o₂ (V̇o_2p), estimations of systemic a-vO_2diff can be derived (29, 30). Thus, the goal of this study was to simultaneously compare the dynamic adjustment of systemic and peripheral (vastus lateralis microvascular) blood flow and O₂ extraction (using open-circuit acetylene and NIRS, respectively) during ramp incremental exercise. We hypothesized that the linear relationship between Q and V̇o₂ observed systemically would be distorted in the active tissue.

METHODS

Subjects.

Ten young men [24 ± 5 yr, 82 ± 10 kg, 181 ± 6 cm; body mass index (BMI) = 25.2 ± 2.6 kg/m² (mean ± SD)] and 10 young women (22 ± 2 yr, 65 ± 7 kg, 171 ± 8 cm; BMI = 22.3 ± 1.6 kg/m²) gave written informed consent to participate in this study. All procedures were approved by The University of Western Ontario Research Ethics Board for Health Sciences Research Involving Human Subjects. All subjects were recreationally active, nonobese, and nonsmokers.

Protocol.

Subjects reported to the laboratory on two separate days separated by at least 48 h, but not more than 2 wk, to perform a ramp incremental test to exhaustion (4-min baseline at 20 W with 20 W/min increments) on a cycle ergometer (Lode Corival 400; Lode B.V., Groningen, Holland). Peak V̇o₂ (V̇o_{2 peak}) was determined as the highest 20-s average value from any of the tests. Measures of Q were taken at 2-min intervals during the ramp incremental exercise test, starting at minute 0 for test day 1 and minute 1 for test day 2, respectively. As such, minute-by-minute measures of Q were obtained for each subject after combining data from both days of testing.

Measurements.

Gas-exchange measurements were similar to those previously described (3). Briefly, inspired and expired flow rates were measured using a low dead space (90 ml) bidirectional turbine (Alpha Technologies VMM 110), which was calibrated before each test using a syringe of known volume. Inspired and expired gases were continuously sampled (50 Hz) at the mouth and analyzed for concentrations of O₂, CO₂, and N₂ by mass spectrometry (Perkin Elmer MGA-1100) after calibration with precision-analyzed gas mixtures. Changes in gas concentrations were aligned with gas volumes by measuring the time delay for a square-wave bolus of gas passing the turbine to the resulting changes in fractional gas concentrations as measured by the mass spectrometer. Data were transferred to a computer, which aligned concentrations with volume information to build a profile of each breath. Breath-by-breath alveolar gas exchange was calculated using algorithms of Beaver et al. (7).

Q was measured using the acetylene (C₂H₂) open-circuit inert gas wash-in method and analyzed using custom data acquisition software. Although test-retest repeatability was not evaluated in the present study, data from previous studies in our laboratory have demonstrated highly reproducible values at both submaximal and peak intensities, so that measurements were consistent within ∼1 l/min with a mean of 14.6 ± 3.2 l/min on test 1 and 14.5 ± 3.2 l/min on test 2 (P = 0.43) and a reliability of r = 0.97 (29, 30). Further evidence for the high repeatability of this method comes from the fact that the Q measurements performed on two separate days in the present study (measured at odd- and even-minute intervals between test days) fell perfectly in line with each other within each individual (see results). Additionally, this technique was described and validated previously (23). Briefly, a pneumotachograph (Hans Rudolph model 3800; Kansas City, MO; transducer, Validyne MP45–871; Northridge, CA) was attached to a nonrebreathing Y valve (Hans Rudolph 7900), which was connected to a manual valve that allowed switching inspired gases between room air and a bag containing a mixture of C₂H₂ (0.7%), O₂ (21%), He (9%), and balance N₂. Changes in gas concentrations were aligned with gas volumes by measuring the time delay of the gases as described above. During measurements of Q, subjects were asked to avoid coughs, swallows, and partial breaths. At any given measurement, subjects were asked not to alter their breathing pattern when the source of inspired air was switched from room air to the bag containing the gas mixture. The concentration of C₂H₂ and He were displayed on a computer screen on a breath-by-breath basis, and after 10 breaths, the protocol was terminated. Data analysis for the calculation of Q was performed immediately after each maneuver using equations reported previously (23). The a-vO_2diff was calculated from the Fick equation as a-vO_2diff (ml O₂/100 ml blood) = V̇o₂ (l/min)/Q (l/min) × 100.

Local muscle deoxygenation (Δ[HHb]) profiles of the quadriceps vastus lateralis muscle were made with NIRS (Hamamatsu NIRO 300; Hamamatsu Photonics, Hamamatsu, Japan) throughout exercise. Briefly, optodes were placed on the belly of the muscle midway between the lateral epicondyle and greater trochanter of the femur. The system consisted of both an emission probe that carries NIR light from the laser diodes and a detector probe (interoptode spacing = 5 cm); optodes were housed in an optically dense plastic holder and secured on the skin surface with tape and then covered with an optically dense, black vinyl sheet, thus minimizing the intrusion of extraneous light. The thigh was wrapped with an elastic bandage to minimize movement of the optodes. Four laser diodes (λ = 775, 810, 850, and 910 nm) were pulsed in a rapid succession, and the light returning from the tissue was detected by the photodiode for online estimation and display of the concentration changes from the resting baseline of Δ[HHb]. Changes in light intensities were recorded continuously at 2 Hz and transferred to a computer for later analysis. The NIRS-derived signal was zero set with the subject sitting at rest on the cycle ergometer prior to the onset of baseline (i.e., 20 W) exercise. Given the uncertainty of the optical path length in the vastus lateralis at rest and during exercise, Δ[HHb] data are presented as normalized delta (%Δ; see below for normalization procedures) units.

Data analysis.

V̇o_2p data were filtered by removing aberrant data points that lay outside 4 SD of the local mean and then linearly interpolated to 1-s intervals. The second-by-second V̇o_2p data from tests 1 and 2 were time-aligned and ensemble-averaged to yield a single averaged response for each subject for the ramp incremental exercise protocol; subsequently, the min-by-min V̇o_2p values were calculated using 10-s averages from the 5 s prior to and following the rounded minute. The second-by-second Δ[HHb] data were time-aligned and ensemble-averaged in the same manner. The ensemble-averaged Δ[HHb] responses were then normalized (%Δ[HHb]), such that 0% represented the steady-state value observed during 20 W cycling and 100% represented the highest average (i.e., Δ[HHb]_peak) value observed in any continuous 20 s of exercise. As described by Boone et al. (9), %Δ[HHb] data were time-aligned with V̇o_2p by left-shifting the V̇o_2p signal by 20 s to account for the circulatory transit delay between muscle and lung; this was undertaken so that changes in “muscle V̇o₂” (represented by V̇o_2p) were aligned with changes in the %Δ[HHb] signal. Although this 20-s value may not precisely match the circulatory time lag in all individuals, our laboratory has recently described the limitations and challenges associated with its determination (31), and overall, this 20-s value represents a reasonable estimate for the group tested.

The whole body Q-to-V̇o₂ relationship was determined using linear regression analyses. Data from baseline (20 W) to peak exercise were included to derive regressions for each individual. The whole body a-vO_2diff-to-V̇o₂ relationship was described by the following hyperbolic function: y = a·x / (B + x), where a represents the asymptotic value and B is the x value corresponding to 50% of the total amplitude. As with whole body Q-to-V̇o₂, data spanned from baseline to peak exercise.

The profile of peripheral (vastus lateralis) blood flow (Q_VL) was determined, as described by Ferreira et al. (17). Briefly, peripheral a-vO_2diff was estimated from the profile of %Δ[HHb] using published values for muscle a-vO_2diff (24, 26, 37); baseline and peak exercise a-vO_2diff were assumed to equal 10 ml O₂/100 ml blood (26) and 18 ml O₂/100 ml blood (24), respectively. To convert the normalized change in %Δ[HHb] to a-vO_2diff (in ml O₂/100 ml blood), we used the following formula: a-vO_2diff = 10 + [%Δ[HHb]/100]·8, where 10 ml O₂/100 ml blood corresponds to assumed a-vO_2diff during baseline cycling and 8 ml O₂/100 ml blood is the assumed change in a-vO_2diff from baseline to peak exercise. Thus, second-by-second Q_VL was calculated (Fick equation) for each individual as the quotient of V̇o₂ (measured) and peripheral a-vO_2diff (as described immediately above).

We have previously reported that the profile of %Δ[HHb] (plotted as a function of V̇o_2p) is better described by a piecewise equation that includes two linear segments [ hereafter referred to as “double-linear” as described by Spencer et al. (40)] compared with a symmetrical sigmoid function previously proposed. This has been confirmed in the present data set, and therefore, no data from the sigmoid functions are presented. The “double-linear” model characterizes the predominant increase in %Δ[HHb] observed throughout the middle portion of the exercise protocol (beginning at the point where the %Δ[HHb] signal began a systematic increase above baseline as determined by visual inspection) and the “plateau,” which follows (40): y = segment1(x) = [y1·(BP-x) + y2·(x-x1)] /(BP-x1); segment2(x) = [y2·(x2-x) + y3·(x-BP)] /(x2-BP), f = if [x ≤ BP, segment1(x); else, segment2(x)], where x1 and x2 represent the minimum and maximum x values, respectively; y1 and y3 represent the predicted %Δ[HHb] at x1 and x2, respectively; BP represents the x value at the break point between the two segments; and y2 represents the predicted %Δ[HHb] at BP (i.e., y2 = %Δ[HHb], where segments intersect). Thus, this “double-linear” analysis yields: y = m₁·x + b₁ for x < BP and y = m₂·x + b₂ for x > BP, where m represents the slope and b is the y-intercept value. The model parameters were estimated for each participant by least-squares linear and nonlinear (Q-to-V̇o₂ and a-vO_2diff-to-V̇o₂, respectively; Origin, OriginLab, Northampton, MA); the “double-linear” function was fit using SigmaPlot 12.0 (Systat Software, Point Richmond, CA). For all regressions, the best fit was defined by minimization of the residual sum of squares and minimal variation of residuals around the y axis (y = 0) (40). All statistical analyses were performed using SPSS version 19.0 (SPSS, Chicago, IL). Descriptive data are presented as mean ± SD. Independent pairwise t-tests were used to detect differences between men and women. Statistical significance was accepted at a P value <0.05.

RESULTS

Absolute (men, 4.2 ± 0.5 l/min; women, 3.2 ± 0.5 l/min; P < 0.05) but not relative (men, 52.0 ± 7.5 ml·kg⁻¹·min⁻¹; women, 49.0 ± 7.9 ml·kg⁻¹·min⁻¹; P > 0.05) V̇o_{2 peak} was greater in men compared with women. Figure 1 displays the dynamic adjustment of systemic blood flow and O₂ extraction (a-vO_2diff) and the profile of the adjustment of peripheral blood flow response (Q_VL) and exercising muscle O₂ extraction (Δ%[HHb]) as a function of V̇o₂ during the ramp incremental test. Table 1 provides the parameter estimates of the model fits to the data. Systemic Q displayed the characteristic linear relationship with V̇o₂ that was similar in men (Q = 5.8 × V̇o₂ + 4.2) and women (Q = 5.7 × V̇o₂ + 2.9). Similarly, the hyperbolic response for a-vO_2diff was the same in men and women. The derived Q_VL was depicted as an inverted sigmoid response, whereas the %Δ[HHb] was best described with two linear segments. The slope and y-intercept of the %Δ[HHb] signal before BP (hereafter referred to as Δ[HHb] − BP) (i.e., m₁ and b₁) were different between men and women, showing a greater rate of increase in women.

Fig. 1.

Profiles of systemic and peripheral blood flow (Q_sys and Q_VL, respectively) and O₂ extraction (a-vO_2diff and %Δ[HHb], respectively) as a function of V̇o₂. Left: Qsys, a-V̇o_2diff, Q_VL, and %Δ profiles for men (open circles). Right: Qsys, a-V̇o_2diff, Q_VL, and %Δ profiles for women (open triangles). Data points are the mean across all subjects at the given V̇o_2p; model fits are the fit parameters from the mean of the fits to data in each individual. Dashed line indicates location of Δ[HHb] -BP. Numbers displayed above individual data points indicate the number of subjects from which the specific point was derived.

Table 1.

Parameter estimates for linear (Q), hyperbolic (a-vO_2diff), and double-linear (%Δ[HHb]) models as a function of V̇o_2p displayed for men and women

	Q		a-vO2diff			%Δ[HHb]
	m	b	a	B	m1	b1	m2	b2
Men	5.8 ± 1.1	4.2 ± 2.4	18.3 ± 3.5	0.9 ± 0.5	42.4 ± 6.4	−55.7 ± 20.6	2.0 ± 9.7	90.7 ± 38.8
Women	5.7 ± 1.0	2.9 ± 1.6	17.2 ± 3.4	0.5 ± 0.3	76.8 ± 23.4*	−89.3 ± 39.0*	9.8 ± 10.1	66.6 ± 31.2

Values are expressed as means ± SD, derived from each individual. m, slope of the linear regression; b, y-intercept of the linear regression; a, asymptotic value; B, x value corresponding to 50% of the total amplitude; m₁ and m₂ slope of linear regression before and after Δ[HHb] -BP, respectively; b₁ and b₂ y-intercept of linear regression before and after Δ[HHb] -BP, respectively.

^*Significant difference from men, P < 0.05; n = 10 for men and women.

DISCUSSION

This study examined the responses of systemic and peripheral (vastus lateralis muscle) blood flow and O₂ extraction during ramp incremental exercise. The novelty of the study design resided in obtaining simultaneous measures and estimates of central and regional (microvascular) responses in the same subjects during cycling exercise throughout the full range of exercise intensities. The main finding was that as V̇o₂ progressively increased, the dynamic adjustments of systemic blood flow and O₂ extraction were different compared with that in the periphery in that there was a nonlinear increase in Q_VL concomitant with a rapid increase of muscle deoxygenation throughout “moderate”-intensity exercise followed by a plateauing of local deoxygenation in heavy-intensity to maximal exercise.

Despite certain reports suggesting otherwise (28, 41), there is compelling information indicating that Q adjustments to incremental exercise occur in a linear fashion and with a positive intercept when evaluated as a function of metabolic rate (V̇o₂) (1, 6, 16, 34, 36). This implies that the adjustment of a-vO_2diff during incremental exercise displays a hyperbolic response and, indeed, this response has been demonstrated previously (10, 24, 36). Data from the present investigation support that contention and are in line with past reports indicating ∼5–6 l/min increase in Q for a 1 l/min increase in V̇o₂ (2, 22, 27, 34, 39). Importantly, the regression lines obtained in this study were derived from many data points during the ramp incremental test (i.e., minute by minute), which increases confidence in the quality of the fit and Q-to-V̇o₂ linearity.

Recently, several studies have characterized the dynamic response of the NIRS-derived [HHb] signal during ramp incremental testing (9, 17, 19) to characterize the profile of O₂ extraction within some of the active tissues (vastus lateralis muscle) and, as a consequence, make inferences with regard to the blood flow response within the microcirculation. Similar to the results from the present investigation, the implication from those studies was that there is a nonlinear muscle blood flow-to-V̇o₂ relationship in the area of NIRS interrogation. Several factors may play a role for the loss of this linearity within the measured muscle. For instance, a rapid increase in blood flow during the exercise on-transient has been proposed to be connected to the mechanical effects of the muscle pump and rapid vasodilation of unknown origin (43, 44). Following this rapid increase, there is a period during exercise (until ∼60–70% of V̇o_{2 peak}) in which peripheral (i.e., vastus lateralis) blood flow increases modestly, despite a comparatively larger increase in systemic blood flow. This larger reliance on peripheral O₂ extraction for a given increase in V̇o₂ might reflect an initial sympathetically mediated vasoconstrictor effect (controlling blood pressure). Thereafter, peripheral O₂ extraction reaches a “plateau-like” response and further increases in blood flow in the active tissues are necessary to support a further increase in V̇o₂ by ∼1 l/min. The idea that adequate redistribution of blood flow is required to support an increase in V̇o₂ at higher intensities of exercise has previously been proposed (33). Such a redistribution of blood flow to the active tissues might be supported by increased sympathetic vasoconstriction of peripheral nonactive tissues and the splanchnic and renal regions (38). However, this increase of blood flow may also be supported by an intensity-dependent functional sympatholysis (45), such that sympathetic vasoconstriction in active muscles is greatly attenuated (by metabolic events in contracting skeletal muscle) to optimize muscle perfusion (21, 42). In support of this hypothesis, Tschakovsky et al. (45) reported a diminished reduction of forearm vascular conductance (while evoking high endogenous norepinephrine release) during rhythmic handgrip exercise, as the intensity of exercise increased from rest to heavy exercise. Additionally, VanTeeffelen and Segal (46) demonstrated that functional hyperemia increased in response to intensity of contraction in the presence of prevailing vasoconstriction in the feed and proximal arterioles and vasodilation in distal arterioles. These “arteriole level”-dependent vasomotor response differences implicate the necessity for blood flow and O₂ extraction measurement at the level of the microvasculature when examining the relationship of blood flow/O₂ delivery to muscle O₂ utilization. Recently, the Δ[HHb] -BP has been shown to be coincident to the intensity associated with the maximal lactate steady state (8). It is possible that beyond this marker of exercise intensity, metabolite accumulation within the interstitial fluid (e.g., K⁺, lactate) may contribute to vasomotor relaxation and thus improved exercise hyperemia.

Although it was expected that the overall profiles of the systemic and peripheral responses would not be affected by sex (18), we have displayed these profiles in men and women separately in Fig. 1. V̇o_{2 peak} was smaller in women compared with men (3.2 l/min vs. 4.2 l/min); therefore, by including both sexes when displaying systemic and peripheral responses, the profiles would be distorted by the reduction in subject number, especially at higher V̇o₂ values. For example, at 3.0 l/min, 17 subjects would contribute to the profile; however, the number would be reduced to 13 at 3.5 l/min, seven at 4.0 l/min, and two at 4.5 l/min. However, it should be noted that the lines of best fit (presented in all panels in Fig. 1) are derived from individual response profiles and not the mean response profile. Nevertheless, the general responses were similar in men and women; although women displayed a steeper slope (m₁) for %Δ[HHb], these differences were to be expected as the absolute V̇o_{2 peak} was smaller in women [O₂ extraction is associated with relative (rather than absolute) intensity, such that peak O₂ extraction will be reached when approaching maximal intensities, independent of absolute V̇o_{2 peak}].

Limitations.

A limitation of this study was the use of the NIRS-derived Δ[HHb] signal as a measure of O₂ extraction as a replacement for a-vO_2diff in the Fick equation to derive an estimate of Q_VL. Our recent study has criticized this approach to calculate the profile of peripheral blood flow during exercise transitions from baseline to moderate intensity (32). The major problem presented in that study was related to the uncertainties with the use of micromolar units as a valid replacement for a-vO_2diff in the equation. The normalization of the Δ[HHb] used in the present study overcomes this limitation. This normalization requires assumed “usual” values of the muscle a-vO_2diff. However, the range of values for resting and peak exercise a-vO_2diff would certainly vary among subjects and differ from those assumed in this paper. Accepting this limitation, the real profile of Q_VL should remain similar to those presented here, unless large deviations occurred between the actual a-vO_2diff and those previously reported (24, 26, 37), which were used to normalize the Δ[HHb] data and derive Q_VL.

Another factor to consider is that the number of subjects composing the overall response was progressively reduced as exercise intensity increased. However, two facts should be noted: 1) despite the reduced number of subjects at higher-exercise intensities, the values obtained for Q in the “reduced” data set (with Q used as an example due to its well-established relationship with metabolic demand) fell directly within what would be predicted from its linear relationship with V̇o₂; and 2) the fits displayed in Fig. 1 were modeled from the mean values derived from the individual responses from each subject. As such, the fits presented in this study were not affected by the loss of subjects at higher-exercise intensities.

Finally, it is acknowledged that choosing to evaluate the vastus lateralis muscle only during ramp incremental exercise does not provide a complete picture of the skeletal muscle blood flow response to exercise. It has previously been reported that the profile of rectus femoris deoxygenation during ramp incremental exercise differed compared with either the vastus lateralis and vastus medialis muscle groups (13), and it has been shown that muscle group activation during cycling changes nonuniformly with increasing relative intensity, and there is a considerable heterogeneity with respect to individual muscle group activation (e.g., gluteus maximus has been reported to increase ∼40-fold at maximal exercise) (20). Although these factors may have contributed to some of the differences observed between central and peripheral components, it is important to note that different studies have shown no differences (15) or minimal differences in spatial heterogeneities in muscle deoxygenation within the quadriceps (25).

Perspectives and Significance

This study demonstrated that the linear relationship that exists between blood flow and metabolic demand during ramp incremental cycling exercise does not occur at the level of the active muscle vasculature. The inverse sigmoidal-like response of Q_VL indicates that blood flow and, thus, O₂ delivery are regulated by different mechanisms in the vascular beds compared with those observed systemically as exercise intensity increases. Active muscle microvascular perfusion and O₂ extraction differ considerably from the responses observed systemically during maximal incremental exercise and, therefore, caution should be used when applying systemic circulatory responses to interpret those of the periphery.

GRANTS

This study was supported by Natural Sciences and Engineering Research Council of Canada research and equipment grants.

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

Author contributions: J.M.M., M.D.S., and D.H.P. conception and design of research; J.M.M. and M.D.S. performed experiments; J.M.M., M.D.S., and D.A.K. analyzed data; J.M.M., M.D.S., D.A.K., and D.H.P. interpreted results of experiments; J.M.M. and M.D.S. prepared figures; J.M.M. and D.A.K. drafted manuscript; J.M.M., M.D.S., D.A.K., and D.H.P. edited and revised manuscript; J.M.M., M.D.S., D.A.K., and D.H.P. approved final version of manuscript.