European Journal of Applied Physiology

, 103:343

Age and microvascular responses to knee extensor exercise in women


  • Beth A. Parker
    • Department of KinesiologyThe Pennsylvania State University
  • Sandra L. Smithmyer
    • Department of KinesiologyThe Pennsylvania State University
  • Samuel J. Ridout
    • Intercollege Graduate Degree Program in PhysiologyThe Pennsylvania State University
  • Chester A. Ray
    • Heart and Vascular InstitutePennsylvania State University College of Medicine
    • Department of KinesiologyThe Pennsylvania State University
    • Intercollege Graduate Degree Program in PhysiologyThe Pennsylvania State University
Original article

DOI: 10.1007/s00421-008-0711-0

Cite this article as:
Parker, B.A., Smithmyer, S.L., Ridout, S.J. et al. Eur J Appl Physiol (2008) 103: 343. doi:10.1007/s00421-008-0711-0


This study examined the relation between femoral artery hyperemic responses to leg exercise and microvascular oxygen extraction in nine young [(mean ± S.E.M) Y: 25 ± 1 years] and 13 older (O: 67 ± 1 years) healthy women. Femoral artery blood flow (FBF; Doppler ultrasound), mean arterial pressure (MAP), and femoral vascular conductance (FVC; FBF/MAP) were assessed at rest and during 9 min of ∼17–18 W single knee extensor exercise. Near-infrared spectroscopy (NIRS) in the vastus lateralis was used to measure changes in deoxyhemoglobin (HHb) and oxyhemoglobin (O2Hb) to assess oxygen extraction (Δ HHb), microvascular blood volume (Δ HHb + O2Hb), and spatial heterogeneity of blood volume (relative dispersion, defined as the coefficient of variation of measurements from all NIR sensors). During exercise, FBF and FVC were lower (p < 0.05) in O compared to Y (Y: 1,430 ± 101 mL min−1 and 16.6 ± 1.2 mL min−1 per mmHg versus O: 1,127 ± 64 mL min−1 and 10.9 ± 0.6 mL min−1per mmHg). While oxygen extraction increased more rapidly in Y, there were no age differences (p > 0.10) beyond the fourth min of exercise. In addition, the relative increase in microvascular blood volume was not significantly different during exercise in Y versus O (age effect: p = 0.87), although Y demonstrated an increase in blood volume heterogeneity in both the active and inactive leg (p = 0.02 and <0.05, respectively) not observed in O (p = 0.71 and 0.28, respectively). NIRS-derived measurements of changes in deoxygenated hemoglobin do not support the hypothesis that older women compensate for the blunted leg hyperemic response to knee extensor exercise through an augmentation of quadriceps oxygen extraction; knee extensor work may be accomplished by other metabolic or microvascular adjustments to exercise in older women.


Oxygen extractionHeterogeneityBlood flow


Older women exhibit lower leg blood flow responses compared to young women during graded two-leg cycling (Proctor et al. 2003) and single-knee extensor exercise (Parker et al. 2008). In the latter study, older women exercised at the same absolute work rates as young women, and neither peak work rate attained nor the duration of the knee extensor exercise protocol (∼18 min) differed with age. We hypothesized that older women may demonstrate compensatory alterations in local oxygen extraction to dynamic knee extensor exercise relative to young women to meet the sustained metabolic demand of the working muscle despite a dramatically lower conduit inflow (Lundgren et al. 1988; Proctor et al. 1998).

Accordingly, the purpose of the present study was to investigate oxygen extraction (estimated indirectly by near infrared spectroscopy, or NIRS) in the vastus lateralis (which has high proportional activation relative to the other quadriceps muscles during kicking exercise in women; (Heinonen et al. 2007)) during an extended bout of strenuous (∼17–18 W) dynamic knee extensor exercise. We hypothesized that (1) there would be a significantly lower femoral artery hyperemic response to knee extensor exercise in older versus young women, and (2) estimated oxygen extraction would thus be augmented in order for older women to continue knee extensor exercise at a similar absolute work rate as young women. To gain additional insight into the mechanisms underlying the vastus lateralis oxygen extraction response in older versus young women, we also assessed microvascular blood volume and the heterogeneity of the microvascular blood volume response to exercise, as blood flow heterogeneity is inversely associated with the efficacy of tissue oxygenation (Walley 1996) and lower heterogeneity may thus improve oxygen extraction in adults exhibiting reduced leg blood flow (Kalliokoski et al. 2001).


Inclusion/exclusion criteria and initial screening

Nine young women (ages 20–30) and 13 older women (ages 60–79) completed the study. All subjects were non-obese (BMI ≤ 30), nonsmokers, had clinically normal blood chemistry (i.e. hemoglobin concentrations ranged from 111–162 g L−1, total cholesterol ≤6.22 mmol L−1, LDL cholesterol ≤3.89 mmol L−1), and resting supine ankle-brachial index ratings (ABI between 0.90 and 1.30; VP2000, Colin Medical). All subjects were normotensive (resting blood pressure ≤140/90 mmHg) and free of overt chronic diseases as evaluated by medical history questionnaire, a physical examination and resting ECG. Additionally, no subjects were taking medications having significant hemodynamic effects (including oral contraceptives and hormone therapy) for at least the last 12 months. Young females were studied in days 1–7 of their menstrual cycle to standardize the influence of female hormones (Williams et al. 2001). On study day, subjects were asked to refrain from alcohol, exercise, caffeine, aspirin, ibuprofen, or herbal supplements for at least 12 h prior to testing. All subjects gave their written, informed consent to participate. This study was approved by the Office for Research Protections and the Institutional Review Board at The Pennsylvania State University in agreement with the guidelines set forth by the Declaration of Helsinki.

Fitness and physical activity status

Subjects were neither extremely sedentary nor extremely trained or fit (as assessed by treadmill maximal oxygen uptake, or \( \ifmmode\expandafter\dot\else\expandafter\.\fi{V}O_{{2}\rm{max}} \), values referenced to age-predicted norms (ACSM 2006) and scores on either the Yale Physical Activity Questionnaire (Dipietro et al. 1993) for older subjects or the Baecke Questionnaire of Habitual Physical Activity (Baecke et al. 1982) for young subjects). None of the subjects participated in moderate to high intensity aerobic exercise >3 day week−1 or regular lower body resistance training >2 day week−1during the past 12 months. A continuous incremental treadmill test (SensorMedics, Yorba Linda, CA) to maximal exertion to was used to determine \( \ifmmode\expandafter\dot\else\expandafter\.\fi{V}O_{{2}\rm{max}}.\)

Total and regional body composition was estimated using dual-energy X-ray absorptiometery (DXA; model QDR 4,500 W, Hologic, Waltham, MA) with subjects in the supine position as described previously (Proctor et al. 2005). In addition, thigh volume was estimated as described previously (Parker et al. 2007).

Study procedures

Exercise modality

Single leg knee extensor exercise, designed to isolate the quadriceps muscle group, was performed as described previously (Parker et al. 2007). Briefly, subjects were reclined in a seat in the supine position with knees flexed at an angle of 90°. The subject’s torso and both legs were fixed by straps attached to the chair, and the left leg was strapped into a boot attached by lever arm to the pedal of a cycle ergometer placed behind the subject. The right leg was allowed to hang free. One extension of the quadriceps muscle moved the subject’s lower leg 90°–170° and the ensuing flexion was a passive return pulled by the flywheel of the ergometer. Subjects kicked at a constant cadence of 40 kicks min−1 (0.67 Hz). Resistance was increased by increasing the weight attached to a belt surrounding the flywheel such that friction on the flywheel increased proportionately. Subjects participated in three familiarization visits prior to the study visit.

Graded exercise protocol

To determine maximal knee extensor work rate, subjects completed a graded protocol as described previously (Parker et al. 2007). Following 3 min of quiet rest, and 3 min of unloaded passive exercise, the subject was instructed to begin kicking against no resistance (0W) for 3 min, after which resistance increased incrementally (by 4.8 W) every 3 min until the subject could no longer maintain cadence.

Study protocol

On a separate day, again following 3 min of quiet rest and 3 min of unloaded passive exercise, subjects performed 0.67 Hz knee extensor exercise at an absolute work rate approximating 70–85% of the maximal work rate attained in the graded protocol. Subjects maintained a constant cadence at this work rate until they reached exhaustion. Subjects who did not reach exhaustion by 24 min were stopped to avoid quadriceps overuse and due to the technical difficulty of collecting continuous high-quality Doppler ultrasound measurements during sustained exercise.

Data acquisition and systemic measurements

All systemic variables were collected on-line at a sampling frequency of 400 Hz and stored using a Powerlab system (AD Instruments, Castle Hill, Australia). Heart rate and beat-to-beat systolic and diastolic blood pressure (radial tonometry of the right hand; Colin, Medical Instruments Corporation) were measured continuously throughout the study. Mean arterial pressure (MAP, in mmHg) was calculated as (1/3 systolic pressure) + (2/3 diastolic pressure). Knee kick cadence was captured using a Cateye Astrale 8 (Cateye, Boulder, CO) cycle computer attached to the flywheel.

Femoral artery hemodynamic measurements

A Doppler ultrasound machine (HDI 5000, Philips, Bothell, Washington) equipped with a high resolution 7–4 MHz linear-array transducer was used to measure mean blood velocity and vessel diameter of the left common femoral artery, as described previously (Parker et al. 2007). For velocity measurements, the artery was insonated at a constant angle of 60° with the sample volume adjusted to cover the width of the artery, while diameter measurements were obtained with the artery insonated perpendicularly. Velocity measurements were taken continuously during the protocol, with the exception of high-resolution diameter measurements (taken in 2D mode to optimize imaging) taken for 20 s every 3 min starting at minute two of rest. A custom interface unit processed the high-resolution angle-corrected, intensity-weighted Doppler audio information (i.e., mean blood velocity) from the ATL system into a lower frequency velocity signal (frequency range 0–20 Hz) that could be sampled in real time by Powerlab.

Diameter measurements were stored on VHS tape and digitized at 4 frames s−1 using Brachial Imager software (Medical Imaging Applications; Iowa City, IA). Post-test analysis of diameters was performed using edge-detection software (Brachial Analyzer Software, Medical Imaging Applications) as described previously (Parker et al. 2007). Briefly, the technician (always the same and blind to any subject information) selected a region of interest along the arterial wall and the edge of the wall was detected by pixel density and represented by a line of best fit. Diameter measurements were calculated from the intima-lumen interface. Femoral artery blood flow (FBF), averaged at rest, during passive exercise, and during every 3 min of loaded kicking, was calculated according to the formula:
$$ {\text{FBF}}:{\text{Blood velocity}}\, \times \,\pi \, \times \,({\text{femoral diameter}}/2)^{2} \, \times \,60 $$
where the FBF is in ml min−1, the blood velocity is in cm s−1, the femoral diameter (averaged across the cardiac cycle) is in cm, and 60 is used to convert from ml s−1 to ml min−1. Femoral vascular conductance (FVC) was calculated as FBF/MAP.

Quadriceps microvascular oxygenation measurements

Changes in oxygen levels in the vastus lateralis of both the active and inactive kicking leg were estimated noninvasively by NIRS (Near Infrared Spectroscopy; LEDI; NIM, Inc., Philadelphia). The NIRS unit uses a dual light source, continuous wavelength spectrophotometer with six detectors per probe and light-source separation of 30 mm to define the relationship between absorption of electromagnetic radiation and the relative concentration of deoxyhemoglobin (HHb) and oxyhemoglobin (O2Hb) with respect to the Beer-Lambert Law. Light emitted into the larger arteries and veins (>1 mm) is almost completely absorbed by the large concentration of hemoglobin such that detectable changes in absorption at 730 nm (HHb) and 850 nm (O2Hb) are attributable to changes in tissue oxygenation in the microcirculation (Kalliokoski et al. 2006). Differences in oxygenation between hemoglobin and myoglobin cannot be distinguished; however, the NIRS signal is generally assumed to be coming primarily from HHb and O2Hb (Mancini et al. 1994; Wilson et al. 1989). In addition, the assumption of a constant path length of photons (i.e., constant scattering) in this unit dictates that measurements be expressed as a relative measure.

Probes were wrapped in saran wrap, placed longitudinally on the belly of the vastus lateralis muscle of the active and inactive leg, and secured by wrapping elastic bandage around the leg and probe. After calibration and balancing, data were sampled at 3 Hz continuously throughout the exercise protocol and stored offline in a text file. Total Hb was calculated as the sum of O2Hb and HHb and used as an index of change in regional microvascular blood volume (Belardinelli et al. 1995; Van Beekvelt et al. 2001). Changes in the HHb signal were used to represent oxygen extraction (De Blasi et al. 1994; Ferrari et al. 1997). Measurements of total Hb and HHb were averaged in 30 s increments and normalized to reflect changes from rest (arbitrarily defined as 0 μM). Heterogeneity of the total Hb signal at each data point was calculated as relative dispersion, or the coefficient of variation of values from all eight detectors, and averaged for rest and exercise (Kalliokoski et al. 2006).

Statistical analysis

Statistical analyses were performed using SAS (SAS 9.1, Cary, North Carolina). All data are reported as mean ± S.E.M. with significance set at p < 0.05. A one-way ANOVA (Proc GLM) and Tukey post-hoc analysis were used to compare baseline differences between groups. For comparisons of responses to knee extensor exercise, a repeated measures ANOVA (Proc Mixed) model with time as the within-individual factor and age as the between-individual factor and an auto-regressive variance–covariance structure was used to determine differences between young and older subjects in outcome variables. A Bonferroni post-hoc adjustment was performed when significant age × time differences were detected.


Subject characteristics

Subject characteristics are presented in Table 1
Table 1

Subject Characteristics




Sample size



Age (years)

25 ± 1

67 ± 1*

Resting SBP (mmHg)

107 ± 5

123 ± 3*

Resting DBP (mmHg)

56 ± 1

59 ± 2

BMI (kg m−2)

22.0 ± 1.0

24.1 ± 0.6

LDL Cholesterol (mmol L−1)

2.5 ± 0.2

3.1 ± 0.1*

Triglycerides (mmol L−1)

0.7 ± 0.1

0.9 ± 0.1

Quad muscle (kg)

2.0 ± 0.1

1.7 ± 0.1*

\( \ifmmode\expandafter\dot\else\expandafter\.\fi{V}O_{{2}\rm{max}} \)(mL kg−1 min−1)

38.2 ± 1.6

24.6 ± 0.7*

a\( \ifmmode\expandafter\dot\else\expandafter\.\fi{V}O_{{2}\rm{max}} \) (‰)

55 ± 9

40 ± 3

Resting FBF (mL min−1)

199 ± 37

114 ± 12*

Resting FVC (mL min−1 mm Hg−1)

2.7 ± 0.4

1.4 ± 0.2*

Femoral diameter (mm)

7.2 ± 0.2

7.1 ± 0.2

Data are expressed as group means ± S.E.M for young and older women

a According to ACSM 2006

* Significant (p < 0.05) difference between young and older subjects

Exercise protocol characteristics

During the graded knee extensor exercise test to exhaustion, there was no age difference in the peak exercise work rate attained in young versus older subjects (Y: 25 ± 2 W vs. O: 22 ± 1 W; p = 0.21); accordingly, there was also no age difference in the absolute (Y: 18 ± 1 W vs. O: 17 ± 1 W; p = 0.47) or relative (Y: 74 ± 3% of max vs. O: 78 ± 2% of max; p = 0.28) work rates at which the subjects exercised. In addition, there was no age difference in the duration of the exercise protocol maintained (Y: 693 ± 75 s vs. O: 644 ± 58 s; p = 0.67) among subjects who stopped due to exhaustion (n = 6Y and 9O). For the purpose of the ensuing leg hemodynamic comparisons, the first 9 min of measurements were analyzed to avoid the confounding influence of subject dropouts on data interpretation (all but one young and one older women remained in the sample until 9 min).

Femoral hemodynamic responses to exercise (Fig. 1)

Mean arterial pressure (MAP), femoral blood flow (FBF), and femoral vascular conductance (FVC) at rest, during passive exercise, and during loaded kicking exercise are shown in Fig. 1. Older women exhibited higher MAP, lower FBF, and lower FVC at rest and during 9 min of active kicking exercise.
Fig. 1

Mean arterial pressure (MAP; top), femoral blood flow (FBF; middle), and femoral vascular conductance (FVC; bottom) expressed as group means ± S.E.M. at rest, passive exercise, and loaded kicking exercise in 9 young and 13 older women. Asterisk indicates significant (p < 0.05) difference between young and older women. From 6 to 9 min of exercise, one young and one older woman dropped out of the sample due to fatigue. Arrows indicate the onset of active kicking

Estimated changes in blood volume and oxygen extraction in the active and inactive leg (Fig. 2a–d)

One older subject’s data was excluded from analysis due to questionable probe placement. The changes, relative to rest, in total hemoglobin (O2Hb + HHb; estimated microvascular blood volume) and HHb (estimated oxygen extraction) in the active leg are shown in Fig. 2a, b. For changes in blood volume (2a), there was no overall effect of age (p = 0.87) nor was there an age × time interaction (p = 0.27). For changes in oxygen extraction (2b), there was no significant overall effect of age (p = 0.13); however, the age × time interaction was significant (p < 0.01) as young women demonstrated a more rapid increase in oxygen extraction in the first through 4 min of loaded kicking exercise. To test whether the above-described age differences in the active kicking leg were unique to the heightened metabolic demands of exercise (i.e., not attributable to age-related measurement artifact), changes in blood volume and oxygen extraction were also examined in the inactive leg (Fig. 2c, d). There was no effect of age (p = 0.76 and p = 0.94, respectively) or age × time interaction (p = 0.23 and p = 0.58, respectively) for either changes in blood volume (2c) or oxygen extraction (2d).
Fig. 2

Microvascular blood volume (Δ total Hb, or O2Hb + HHb) and oxygen extraction (Δ HHb) in the active (kicking; a and b) and inactive (nonkicking; c and d) vastus lateralis expressed as group means ± S.E.M. averaged by thirty second intervals at rest, during passive exercise, and loaded kicking exercise in nine young and 13 older women. Asterisk indicates significant (p < 0.05) difference between young and older women. From 6 to 9 min of exercise, one young and one older woman dropped out of the sample due to fatigue. Arrows indicate the onset of active kicking

Heterogeneity of microvascular blood volume (Fig. 3)

There was no age difference in resting or exercise blood volume heterogeneity (p = 0.45 and 0.14, respectively); however, blood volume heterogeneity significantly increased during exercise in young women (p = 0.02) but not older women (p = 0.71). Microvascular blood volume heterogeneity in the inactive leg was also not different between young and older women at rest or exercise (p = 0.21 and 0.63, respectively) and again increased significantly in young (p < 0.05) but not older (p = 0.28) women during exercise.
Fig. 3

Microvascular blood volume heterogeneity expressed as group means ± S.E.M. at rest and during exercise in nine young and 13 older women in the active (left) and inactive (right) leg. dagger indicates significant (p < 0.05) difference between rest and exercise within an age group. There were no age-group differences for comparisons at rest or during exercise


The purpose of the study was to test the hypothesis that local oxygen extraction, as estimated indirectly by NIRS, is higher in older women who exhibit blunted femoral hyperemic responses to knee extensor exercise relative to young women. In contrast to our hypothesis, however, we did not find evidence of augmented oxygen extraction in older women, despite the observed age-related reduction in femoral artery blood flow. Several explanations could underlie this finding, including methodological aspects of NIRS-based estimates of muscle oxygen extraction as well as physiological mechanisms associated with aging, such as muscle metabolism, the matching of flow to metabolic demand, and microvascular blood volume.

Relation between femoral blood flow and local oxygen extraction

We previously observed that the attenuated leg blood flow and vascular conductance responses exhibited by older women during knee extensor exercise did not influence their peak work rate attained or duration of the graded exercise protocol (Parker et al. 2008). In fact, doubling the length of the graded knee extensor protocol in that study by reducing the work rate increments also did not influence the age group comparisons of peak work rate/duration of the exercise protocol. Similarly, in the current study, there was no age difference in the duration for which young and older women maintained ∼17–18 W knee extensor exercise to exhaustion, although not all of the women fatigued before measurements were stopped. Regardless, older women demonstrated reduced leg blood flow and leg vascular conductance relative to young women while exercising at a similar absolute work rate (Fig. 1). While hemoglobin was lower in young women (Y: 127 ± 2 g L−1 vs. O: 133 ± 2 g L−1; p = 0.03), estimated oxygen delivery ((multiplying the estimated arterial oxygen content [1.34 × Hb × SaO2 (assuming SaO2 of 97%) mL O2/dL blood) by femoral blood flow] was still significantly higher in young women during exercise (data not shown). Collectively, these findings suggest that older women accomplish the knee extensor work with reduced oxygen delivery to the working muscle.

Data regarding compensatory changes in oxygen extraction in aged individuals who exhibit reduced leg blood flow during dynamic exercise are equivocal, with reports of both greater leg oxygen extraction during cycling exercise in middle-aged (Wahren et al. 1974), older sedentary (Poole et al. 2003) and older trained (Proctor et al. 1998) men as well as similar oxygen extraction in older sedentary men (Beere et al. 1999) and older normally active women (Proctor et al. 2003). Limited data regarding knee extensor exercise are similarly ambiguous (Donato et al. 2006; Ferri et al. 2007; Lawrenson et al. 2003); discrepancies may be attributable to the age, fitness or sex of the subject populations studied as well as the methodology used to investigate oxygen extraction (i.e., direct vs. indirect measurements). Using NIRS based estimates of microvascular oxygenation, we did not find evidence of augmented oxygen extraction in older women (Fig. 2a–b). In fact, estimated quadriceps oxygen extraction was lower in the initial minutes of exercise in older versus young women, suggestive of delayed kinetics and/or onset of metabolic demand in older women (DeLorey et al. 2007).

Given the inverse relation between blood flow heterogeneity and oxygen extraction (Kalliokoski et al. 2001; Walley (1996), we hypothesized that older women might also potentiate oxygen extraction through lower microvascular blood flow heterogeneity, as estimated through measurements of relative dispersion of microvascular blood volume (Kalliokoski et al. 2006). Interestingly, while there were no age differences in heterogeneity of the total Hb signal at rest, microvascular blood volume heterogeneity increased significantly in the young but not older women during exercise in both the active and inactive leg (Fig. 3). However, this age effect was observed in both the active and inactive leg, and did not appear to result in augmented oxygen extraction in older women.

Oxygen extraction: methodological considerations

There are several methodological issues associated with NIRS-based estimates of oxygen extraction that must be considered in the present investigation. Given that arterial deoxygenated hemoglobin (HHb) concentrations are negligible at normal PO2 (De Blasi et al. 1994), it has been concluded that changes in HHb estimated by NIRS reflect changes in oxygen extraction independent of increased perfusion (Ferrari et al. 1997). NIRS estimates of oxygenation cannot be considered a substitute for directly measured \( {\text{a}} - \overline{{\text{v}}} {\text{O}}_{2} \) difference, especially given that the venous oxygen content draining from a whole limb (as sampled by a catheter) is quantitatively different from NIRS measures of microvascular oxygen content in the regional muscle vascular space, of which ∼85% may be capillaries (DeLorey et al. 2003; MacDonald et al. 1999; Poole et al. 1995). In addition, age- and exercise-associated effects on the microvasculature, with respect to microvessel rarefaction (Behnke et al. 2006), arteriolar tortuosity (Bearden et al. 2004), and the efficacy of the muscle pump, could also influence the HHb signal. However, the validity of NIRS is supported by studies showing that the magnitude and nature of the change in HHb is in agreement with expected changes in oxygen extraction during exercise in young and older adults (DeLorey et al. 2007) as well as diabetics (Bauer et al. 2007).

Another consideration of using NIRS technology is that the propagation of light through tissue is influenced by both the muscle and subcutaneous fat through which it travels. A thicker adipose tissue layer, such as that observed in older subjects (caliper-derived measurements of quadriceps adipose thickness were 11.3 ± 1.6 mm in young and 16.0 ± 1.0 mm in older women), may confound NIRS measurements in muscle by preventing light from passing through the muscle tissue and biasing oxygenation measurements since adipose tissue metabolism differs from muscle metabolism (van Beekvelt et al. 2001). However, Homma et al. (1996) found that a source-detector distance of 30 mm (the same as used in the current study) was sufficient to detect changes in leg tissue deoxygenation during exercise in subjects with varying thicknesses of adipose tissue. In addition, since these same authors hypothesized that the NIR light penetrates muscle tissue at minimum deep enough to reach half the distance between source and detector, we also compared oxygenation characteristics of young versus older women in the current study whose adipose tissue thickness was 15 mm or less. In this subset analysis, there was still the same delayed onset of oxygen extraction in the initial minutes of exercise in older women yet no age differences in the following minutes. Finally, there were no age-differences in oxygenation parameters in the inactive leg (Fig. 2c, d), which would be expected should adipose tissue thickness be significantly influencing NIR measurements.

Oxygen extraction: physiological explanations

Leg exercise with similar oxygen extraction and lower leg blood flow in older women relative to young women, in agreement with our previous findings during cycling exercise (Proctor et al. 2003), suggests a lower exercising leg \( \ifmmode\expandafter\dot\else\expandafter\.\fi{V}O_{2} \) in older women, which could reflect age-related changes in exercise efficiency and/or muscle metabolism. For example, greater reliance on oxidative metabolism has been observed in older versus young adults (Lanza et al. 2007). By contrast, however, some studies have shown that oxidative capacity declines by as much as 50% in the leg in older humans (Conley et al. 2000). Changes in fiber type (Coggan et al. 1992; Short et al. (2005), motor unit recruitment (Ferri et al. 2007) and reductions in flow directed to oxidative fibers (Musch et al. 2004) with age may also influence muscle metabolism and alter the relation between leg \( \ifmmode\expandafter\dot\else\expandafter\.\fi{V}O_{2} \)and absolute work.

Alternatively, the NIRS-derived measure of microvascular blood volume—the estimated change in the total hemoglobin (HHb + O2Hb) signal relative to rest—was not significantly different between young and older women during exercise in the current study (Fig. 2). This finding does not likely reflect an augmentation of capillary recruitment with age in women (Coggan et al. 1992; Fuglevand and Segal. 1997). Rather, the preserved microvascular blood volume observed in older women during the current study may be unique to knee extensor exercise, when the increase in cardiac output approximately doubles the increase in leg blood flow (Magnusson et al. 1997), femoral venous oxygen content is still quite high even during intense exercise (Roach et al. 1999) and there is an extremely high flow-to-quadriceps muscle mass ratio (Richardson et al. 1993). Therefore, the increase in leg blood flow during higher-intensity single-knee extensor exercise in young women (greater even than that observed in young men; (Parker et al. 2007)) may exceed the metabolic demand of the working muscle, reflecting a “hyperperfused” condition (Rowell et al. 1986) in which moderate reductions in leg blood in older women do not significantly influence capillary blood volume and consequently oxygen exchange. However, we would caution over-interpretation of microvascular blood volume given that the total hemoglobin signal derived by NIRS is influenced by numerous factors, including local muscle blood flow, the muscle pump, vasodilation, hemoconcentration, and capillary recruitment (DeLorey et al. 2003).

One final possible explanation for the observed oxygen extraction responses in older women is that their reduced femoral blood flow was solely attributable to their ∼15% reduction in quadriceps muscle mass, although we did not find this to be the case previously (Parker et al. 2008). We did not have the statistical power to fully examine the influence of quadriceps muscle mass in older women on leg blood flow and oxygen extraction given the smaller sample size utilized (i.e., age differences in leg blood flow normalized to quadriceps muscle were all non-significant due to low statistical power affected by increased variance associated with normalization). However, matching five young to five older women for quadriceps muscle (∼1.84 kg) did not alter the observed age differences in leg blood flow and vascular conductance or microvascular oxygenation. Thus, the age difference in muscle mass does not seem a likely explanation for the current findings regarding estimated local oxygen extraction.


Use of near infrared spectroscopy to non-invasively assess muscle microvascular responses in the vastus lateralis during higher intensity knee extensor exercise in young versus older women did not provide evidence that older women augment oxygen extraction to compensate for the reduced leg blood flow and vascular conductance measured in the femoral artery with Doppler ultrasound. While there are methodological limitations associated with NIRS-based estimates of oxygen extraction, the current study suggests that there may be metabolic and/or microvascular adaptations that result in the relatively well-preserved capacity for knee extensor exercise despite blunted conduit dilatory responses in older women.


We thank the GCRC clinicians and staff as well as Michael D. Herr, Doug Johnson, and Denny Ripka for their assistance with data collection. This research was supported by R01 AG18246 (D.N. Proctor), NIA Interdisciplinary Training in Gerontology Grant T32 AG00048 (B.A. Parker) and M01 RR10732 (GCRC).

Copyright information

© Springer-Verlag 2008