Journal Logo


A High Activity Level Is Required for Augmented Muscle Capillarization in Older Women

Gliemann, Lasse1; Rytter, Nicolai1; Yujia, Liu1,2; Tamariz-Ellemann, Andrea1; Carter, Howard1; Hellsten, Ylva1

Author Information
Medicine & Science in Sports & Exercise: May 2021 - Volume 53 - Issue 5 - p 894-903
doi: 10.1249/MSS.0000000000002566


Oxygen supply to the working skeletal muscle depends on the capillary density. The capillary network in skeletal muscle is plastic and readily adapts to match the oxygen demand, where exercise training increases capillarization while inactivity leads to capillary rarefaction (1,2). Regressive changes in the capillary network can occur with advancing age (3,4) and become a limiting factor for oxygen supply during exercise. In addition, impairments in microvascular function develop with aging and inactivity (5,6) and may, in parallel with capillary rarefaction, contribute importantly to impaired exercise capacity in older and inactive subjects.

Aging and inactivity may impair matching of oxygen delivery and oxidative metabolism both at the onset of and during steady-state exercise (7,8). In young healthy subjects, a slight delay in the increase in mitochondrial respiration, independent of the delivery of oxygen to the mitochondria, limits the oxygen uptake in the initial phase of low- to moderate-intensity exercise (9). However, in older subjects in whom vascular function is impaired, the delivery of oxygen to the mitochondria may be limiting and account for the slower matching of oxygen delivery and oxygen uptake compared with young subjects (7). One important factor in oxygen supply to mitochondria is the density of the capillary network in the muscle, as this limits the diffusion capacity for oxygen and, thus, muscle oxygen extraction (10–12). Therefore, as a reduction in capillary density is known to occur with inactive aging (4), it may be a contributing factor in the inadequate matching of oxygen delivery to oxygen demand in this population (13).

Women experience a decline in vascular function and impaired blood flow to working skeletal muscle as they age, in particular after the menopausal transition (14–16). In addition, sedentary, older women have been reported to have impaired mitochondrial capacity (17,18) and reduced capillarization compared with young women (17,18), although the latter is not a consistent finding (19,20). It remains unknown to what extent capillary density and matching of oxygen delivery and uptake are influenced by the level of habitual physical activity in older women.

The process of growing new capillaries, angiogenesis, is a balance between angiogenic and angiostatic compounds. One of the most important angiogenic compounds is vascular endothelial growth factor (VEGF), which, in skeletal muscle tissue, primarily is localized in the myocytes and vascular cells (21,22) and, within the blood, bound to platelets (23). An importance of platelets for skeletal muscle angiogenesis has been evidenced by a reduced angiogenic response in rodents experimentally depleted of platelets (24,25). Platelets are therefore likely contributors to the angiogenesis in muscle. The importance of platelets in human skeletal muscle angiogenesis is however unknown, and the influence of lifelong training on the magnitude of VEGF carried by platelets and on their ability to stimulate endothelial cell proliferation has never been examined.

The overall hypothesis of the present study was that older women with lifelong moderate and very high physical activity levels would have a superior capacity for oxygen supply to skeletal muscle compared with sedentary women. A secondary hypothesis was that the angiogenic potential of platelets would be enhanced in trained compared with sedentary women.


Ethical approval

The study was approved by the Ethics Committee of Copenhagen and Frederiksberg communities (H-16042441) and conducted in accordance with the guidelines of the Declaration of Helsinki. Before enrollment in the study, written informed consent was obtained from all subjects.


Forty-one healthy postmenopausal women were included in the study (age, 61 ± 4 yr; 11 ± 5 postmenopausal years; Table 1). Menopausal status was verified by measurements of hypothalamic and reproductive hormones. All subjects had normal resting ECG, were nonsmokers for >20 yr, and had not been diagnosed with hypertension, other cardiovascular diseases, renal dysfunction, insulin resistance, diabetes mellitus, or hypercholesterolemia. None of the subjects were on regular medication and none had been on hormone replacement therapy. Subject characteristics have previously been reported (26).

TABLE 1 - Subject characteristics.
Sedentary (n = 14) Moderately Active (n = 12) Very Active (n = 15)
 Age (yr) 62.4 ± 3.8 61.1 ± 4.2 60.4 ± 4.3
 Time since last menstrual cycle (yr) 12.3 ± 4.4 11.8 ± 5.6 9.6 ± 4.9
 Height (m) 1.68 ± 0.06 1.66 ± 0.05 1.68 ± 0.05
 Weight (kg) 68.8 ± 12.7 68.6 ± 9.1 60.2 ± 4.7 a,b
 Body mass index (kg·m−2) 24.3 ± 3.4 24.7 ± 2.8 21.2 ± 1.5 a,b
 Experimental leg mass (kg) 11.8 ± 2.7 12.2 ± 1.9 11.1 ± 1.3
 Experimental leg lean mass (kg) 7.0 ± 1.3 7.2 ± 0.7 7.3 ± 0.9
 V˙O2peak (mL O2·min−1) 1707.8 ± 402.7 2115.7 ± 307.1 a 2401.2 ± 320.0 a,b
 V˙O2peak kg−1 (mL O2·min−1·kg−1) 25.1 ± 5.2 31.9 ± 5.9 a 39.4 ± 4.4 a,b
 Resting heart rate (bpm) 65.5 ± 7.5 62.1 ± 8.2 56.5 ± 8.1 a,b
 Maximal heart rate (bpm) 164. ±12.3 173.3 ± 12.9 a 169.3 ± 10.7
 20 yr PA at moderate intensity (h·wk−1) 0.75 ± 0.87 2.00 ± 0.85 a 3.60 ± 2.06 a,b
 20 yr PA at high intensity (h·wk−1) 0.17 ± 0.39 0.83 ± 0.83 2.53 ± 1.46 a,b
 20 yr MET per day 213.3 ± 102.1 422.2 ± 160.3 a 785.4 ± 374.4 a,b
 Current PA at moderate intensity (h·wk−1) 0.58 ± 0.67 1.92 ± 1.17 a 3.73 ± 2.09 a,b
 Current PA at high intensity (h·wk−1) 0.00 ± 0.00 0.25 ± 0.62 1.73 ± 0.88 a,b
 Current MET per day 153.0 ± 56.9 336.0 ± 110.5 a 675.0 ± 315.1 a,b
Fitness level is reported as peak oxygen consumption (V˙O2peak) and metabolic equivalent (MET), and history of physical activity (PA) is based on questionnaires and interviews. Data are presented as mean ± SD.
aSignificantly different from the sedentary group.
bSignificantly different from the moderately active group.

Experimental design

In a cross-sectional design, subjects were divided into three groups based on their past physical activity level obtained from a modified IPAQ questionnaire (modified to cover both current activity level and general activity level in the past 20 yrs) combined with interviews: 1) sedentary lifestyle with less than 1 h of low-intensity exercise (e.g., walking) and no moderate-intensity (e.g., bike commuting or running) or high-intensity exercise (e.g., interval-based running or cycling, sports at a competitive level) per week (SED; n = 14); 2) moderately active lifestyle with 2–4 h of low- to moderate-intensity exercise and 1–2 h of high-intensity exercise per week (MOD; n = 12); and 3) very active lifestyle with more than 4 h of moderate and high-intensity exercise per week (VERY; n = 15, Table 1). Great precautions were made to ensure that only subjects with a true arche type of activity lifestyle were included in the study, and thus subjects who had been more or less active in periods during the past 20 yr compared with their current activity level were excluded. After medical screening, subjects visited the laboratory for two experimental days in a randomized order, separated by 1 wk.

Experimental day 1

Body composition was determined by whole-body dual-energy x-ray absorptiometry scanning (Prodigy; GE Healthcare, Chalfont St. Giles, UK). Pulmonary peak oxygen uptake (V˙O2peak) was then determined (Oxycon Pro; Viasys Healhtcare, Hoechberg, Germany) using an incremental exercise test on a mechanically braked cycle ergometer (Monark Ergomedic 839E; Monark, Vansbro, Sweden). After an 8-min warm-up session at 50, 70, or 90 W, SED, MOD, and VERY respectively, the workload was increased by 20 W·min−1 until exhaustion. V˙O2peak was calculated as the average of the three highest consecutive 15-s values. For recognition of true V˙O2peak, three of the following five criteria had to be met: individual perception of exhaustion, respiratory exchange rate >1.15, V˙O2 curve plateau, heart rate reaching age-predicted maximum, and inability to maintain pedaling frequency above 80 rpm. Verbal encouragement was given throughout the test.

Experimental day 2

Subjects refrained from caffeine, alcohol, and exercise for 24 h and had a light standardized breakfast before arriving at the laboratory at 7:30 am. Catheters (20 Ga; Arrow Int., Reading, PA) were inserted into the femoral artery and vein of the experimental leg and in the femoral artery of the nonexperimental leg under local anesthesia (lidocaine, 20 mg·mL−1; Astra Zeneca, Copenhagen, Denmark). By ultrasound guidance, the catheters were inserted 2–3 cm below the inguinal ligament and advanced 10 cm in the proximal direction. Under local anesthesia (lidocaine, 20 mg·mL−1, Astra Zeneca) of the skin and muscle facia, a muscle biopsy was obtained from the vastus lateralis muscle using the percutaneous needle biopsy technique with suction. The biopsies were immediately frozen in liquid nitrogen and stored at 80°C until further analysis.

After 70 min of rest, subjects were seated in a semirecumbent position (their hip angle was fixed at ~120°) in the one-leg knee extensor model and rested 20 min before the first measurement. Blood samples, blood pressure, and blood flow were the measured at rest, after 3 min of passive leg movement (60 rpm), at the onset of exercise (6 W, 15–25 s, and 55–65 s), and after 180 s of exercise (6 W). The initial passive leg movement was done to accelerate the ergometer flywheel and to ensure constant power output from the onset of exercise. Passive leg movement was achieved by instructing the subject to keep the leg fully relaxed while the modified Krogh ergometer was run by the same skilled operator at exactly 60 rpm.

The biopsy collection and the exercise bout were conducted in the same experimental day as previously described after arterial infusion of acetylcholine, sodium nitroprusside, and epoprostenol (26). There was a 90-min break between the last infusion and the first biopsy sampling to ensure minimal carryover effect, and infusion rates, amount of blood collected, and timing and break duration were strictly controlled and identical for all subjects.

Measurements and calculations

Femoral arterial blood flow was measured using ultrasound Doppler (Vivid E9; GE Healthcare, Pittsburgh, PA) equipped with a linear probe (L9) operating at an imaging frequency of 4/8 MHz and a Doppler frequency of 4.2 MHz. The site of measurement in the common femoral artery was distal to the inguinal ligament but above the bifurcation into the superficial and profound femoral branch to avoid turbulence from the bifurcation. All recordings were obtained at the lowest possible insonation angle and always below 60°. Sample volume was maximized by choosing the widest section of the vessel, and recordings were made without interference of the vessel wall. A low-velocity filter (<1.8 m·s−1) rejected noise caused by turbulence at vascular wall. Doppler traces and B-mode images were recorded continuously and averaged over 30 s at rest and 180-s exercise and over 10 s at the onset of exercise. Arterial diameter was assessed during systole from B-mode images for each Doppler recording with the probe positioned perpendicular to the vessel. Intra-arterial and intravenous blood pressure and heart rate were monitored by use of transducers (Pressure Monitoring Set; Edwards Lifesciences, Irvine, CA) positioned at the level of the heart. Leg vascular conductance (LVC) was calculated as femoral arterial blood flow (LBF)/(mean femoral arterial pressure − mean femoral venous pressure), and leg O2 uptake was calculated as arteriovenous O2 difference × LBF. Experimental leg mass was calculated from data obtained during the dual-energy x-ray absorptiometry scan and previously reported (26). Blood lactate, hemoglobin, and blood gases were measured using an ABL800 FLEX analyzer (Radiometer, Bronshoj, Denmark). Blood sampling time was adjusted to account for transit times from the collection site at the femoral artery and vein at the onset of exercise. Mean transit time from femoral artery to femoral vein has previously been reported to be 5–7 s during moderate-intensity knee extensor exercise wherefore arterial samples was collected 6 s before the venous samples (27).

Platelets and platelet-free plasma

Venous blood was drawn in citrate-anticoagulated tubes and kept at room temperature until centrifuged (10 min at 200g and 20°C) to obtain platelet-rich plasma. The platelet-rich plasma was layered on top of a gradient density medium (iodixanol solution; OptiPrep, Sigma-Aldrich, St. Louis, MO) and centrifuged (15 min at 450g and 20°C) after which a fraction of isolated platelets and platelet-poor plasma was collected. To obtain plasma completely free of platelets, the platelet-poor plasma was centrifuged (15 min at 450g and 20°C) once more on the density gradient medium, and a fraction of the plasma was collected as platelet-free plasma. In all collected samples, platelets were counted using a hematology analyzer (XP-300; Sysmex, Kobe, Japan). Isolated platelets were lysed in a fresh batch buffer (MSD Tris Lysis Buffer; Meso Scale Diagnostics, Rockville, MD) on ice before being centrifuged (10 min at 10,000g and 4°C), and the supernatant was kept as platelet lysate. All samples were stored at −80°C until later analysis.

Platelet releasate

Platelet releasate was obtained by adding a platelet agonist to samples of isolated platelets. In brief, thrombin receptor activating peptide 6 (TRAP-6) was added to the platelets before the sample was incubated on an orbital shaker (5 min at 2000 rpm and 37°C). Afterward, the sample was centrifuged (10 min at 2000g and 4°C) and the supernatant was collected as the platelet releasate. A similar approach was used as control with regard to platelet-free plasma collected for endothelial cell proliferation assay.


Primary skeletal muscle microvascular endothelial cells were isolated from human muscle samples by use of antibody coated magnetic beads (Dynabeads; ThermoFischer Scientific, Waltham, MA) as previously described (28). The cells were cultured in medium 200 with low serum growth supplement containing fetal bovine serum, fibroblast growth factor, heparin, and epidermal growth factor (Cascade Biologics Inc., Portland, OR). The cells were grown in 96-well plates for 24 h before the medium was replaced with medium 200 supplemented with 10% platelet releasate or 2% platelet-free plasma. Medium 200 supplemented with low serum growth supplement and gradient density medium was added as controls, and all samples were made in duplicates. After 24 h of incubation, bromodeoxyuridine was added, and the plates were incubated for an additional 12 h. Proliferation of the microvascular endothelial cells was measured by incorporation of bromodeoxyuridine into the DNA and determined by an immunoassay according to the manufacture’s recommendations (Roche, Mannheim, Germany) and as previously described (28).

Immunohistochemical determination of skeletal muscle capillarization and fiber type

The embedded muscle samples were cut using a cryostat, and transverse sections 8 μm in thickness were mounted on glass slides. To verify the cross-sectional orientation of the individual muscle fiber, multiple samples were cut and examined under light microscopy until a cross section of desirable size, orientation, and uniform polygonal appearance was visible. For immunohistochemical staining, the cross sections were fixed for 2 min in phosphate-buffered saline (pH 7.2; Gibco 70013-016, Life Technologies Denmark, Nærum, Denmark) containing 2% formaldehyde followed by a washing sequence in a 1:10 wash buffer (Dako S3006; DakoCytomation, Glostrup, Denmark) and blocking for 10 min in phosphate-buffered saline containing 1% BSA. Capillaries were visualized using biotinylated Ulex europaeus agglutinin I lectin (1:100; VECTB-1065, VWR) with secondary antibody (1:1000; Streptavidin, Alexa Fluor® 568 conjugate, Invitrogen S11226, Life Technologies Denmark). Myofiber borders were visualized using an antibody against laminin (1:500; Dako Z0097) with secondary antibody (1:1000, AF 405 Goat antirabbit). MHC-I was stained with MB421(1:1000; Sigma-Aldrich, Merck, Darmstadt, Germany) with secondary antibody (1:1000, Alexa Fluor® 568, Invitrogen, Life Technologies Denmark), and MHC-IIa was stained with sc-71 (1:200; SC-71 was deposited to the DSHB by S. Schiaffino; DSHB Hybridoma Product SC-71, University of Iowa, Iowa City, IA), with secondary antibody (1:1000, Alexa Fluor® 488, Invitrogen, Life Technologies Denmark).

Specificity of the staining was assessed by staining without the primary antibody. Visualization was performed on a computer screen using a fluorescence microscope (Carl Zeiss, Oberkochen, Germany), and all morphometric analysis were performed using a digital analysis program (ImageJ, NIH ImageJ). Two or more separate sections of a cross section were used for analysis, and only sections without artifacts or tendency to longitudinal cuts were analyzed. The type and number of muscle fibers and the capillaries within each section were counted, and capillary supply was subsequently expressed as capillaries per fiber (C:F ratio) and capillary density (mm−2). All analyses were carried out by the same blinded investigator.

Quantification of protein expression

Biopsies were freeze dried and dissected free from fat, blood, and connective tissue. Five milligrams dry weight of muscle tissue was homogenized in a fresh batch buffer (10% glycerol, 20 mM sodium pyrophosphate, 150 mM NaCl, 50 mM HEPES, 1% Nonidet P-40, 20 mM β-glycerophosphate, and proteolytic inhibitors) two times for 30 s (Qiagen Tissuelyser II; Retsch, Haan, Germany). After rotation end over end for 1 h, the samples were centrifuged for 30 min at 17,500g at 4°C, and the lysates were collected as the supernatant. Protein concentrations were determined in the lysates using BSA standards (Pierce Reagents, Rockford, IL). The lysates were diluted to appropriate protein concentrations in a concentrated sample buffer (0.5 M Tris base, dithiothreitol, sodium dodecyl sulfate, glycerol, and bromphenol blue), and equal amounts of total protein were loaded for each sample in different wells on precasted Tris–HCL gels (Bio-Rad, Hercules, CA). Samples from the same subject were always loaded on the same gel. After gel electrophoresis, the proteins were transferred (semidry) to a polyvinylidene difluoride membrane (Immobilon Transfer Membrane; Millipore, Billerica, MA), which was incubated with primary antibody overnight and then washed three times for 5 min in Tris-buffered saline–Tween before incubation with secondary antibody for 1 h. The membranes were incubated with the total OXPHOS human antibody cocktail primary antibody (ab110411; Abcam, Cambridge, MA). Secondary antibodies used were goat antirabbit or rabbit antigoat HRP-conjugated antibodies (dilution 1:5000, P-0448 and P-0449, DakoCytomation). After detection and quantification (ChemiDoc MP system; Bio-Rad), the protein content was expressed in arbitrary units. The protein content of each sample was set relative to a pool of all samples loaded on the respective gel. Standard curves with a range of protein content were run. Representative blots are presented in the supplemental material [see Figure, Supplemental Digital Content 1, Representative Western blot of total OXPHOS cocktail in duplicates from one sedentary (SED), moderately (MOD) and very trained (VERY) subject,].

Quantification of VEGF protein by electrochemiluminescence assay

Muscle lysate, platelets, and platelet-free plasma, including releasates, were used for VEGF protein determination by electrochemiluminescence multiplex assay for VEGF165 (Meso Scale Diagnostics) according to the manufacturer’s guidelines. The muscle lysate samples were normalized to the total protein content in each sample, and the platelet contents was normalized to the platelet counts.

Data analysis

The rate of changes in LBF and the leg oxygen uptake were fitted to a nonlinear regression model using a monoexponential one-phase association using the following equation:


where Y(t) is the dependent variable (LBF or leg oxygen uptake) at any time (t), Y(baseline) is the resting value immediately before the onset of exercise, Amp is the amplitude of the response, TD is the time delay, and τ represents the time to achieve 63% of the steady-state amplitude (GraphPad Prism v8.; GraphPad Software, San Diego, CA).

To test the hypothesis that the level of physical activity influences leg vascular kinetics at the onset of one-legged knee extensor exercise and to identify a possible mediating effect of capillarization, a multiple linear regression model was applied. First, the direct effect of fitness level on the leg vascular kinetics was modeled. Next, a model was applied to assess the effect of fitness level on capillary density as well as the effect of capillary density on leg vascular kinetics. To explore the interrelationships between all three variables, linear regression modeling of the combined effect of fitness level and capillary density on leg vascular kinetics was performed.

Statistical analysis

The number of subjects was based on a priori power calculations of the expected differences in LVC during 6 W exercise between untrained and well-trained subjects. Significance level for all tests was set at an α level of P < 0.05 at a power level of 0.8. Data are reported as the mean ± SD, unless otherwise stated. Statistical analyses and data modeling were performed with R (version 3.4.1; R Foundation for Statistical Computing, Vienna, Austria) using the interface RStudio (version 1.1.463; RStudio Team, Boston, MA) and with the extension packages lme4 (29) and multcomp (30). A linear mixed-model approach was used to investigate the effect of different activity levels in the three groups. For all measures, a linear mixed-model approach was used to detect differences with the intervention. Subjects were specified as a repeated factor and identifier of random variation. Residual and Q-Q plots confirmed homogeneity of covariance and normal distribution of the data set. Post hoc procedure was used to detect all pairwise differences and performed with multiple comparison. Reported P values are adjusted by single step within each model to avoid type 1 error (30).


Subject characteristics

There were no differences between the three groups in height, age, or time since menopause (Table 1). The very active group had lower body weight and BMI compared with the sedentary (P = 0.016 and P = 0.005) and moderately active group (P = 0.028 and P = 0.001; Table 1). Systolic, diastolic, and mean arterial blood pressure were lower in the very active (P < 0.001) and moderately active groups (P < 0.002) compared with the sedentary groups (Table 1).

Peak oxygen consumption and self-reported activity level

Peak oxygen consumption (V˙O2peak) was higher in the very active group compared with the sedentary (P < 0.001) and moderately active group (P = 0.040; Table 1). The moderately active group had higher V˙O2peak compared with the sedentary group (P = 0.043). Self-reported weekly duration of moderate- and high-intensity physical activity was higher in the very active group compared with the two other groups (P < 0.001), and higher in the moderately active group compared with the sedentary group (P = 0.011; Table 1)

Hemodynamics at rest and at the onset of exercise

At rest, there was no difference in any of the measured hemodynamic parameters (Fig. 1A–I). LBF was not different between the three groups at 20- or 60-s exercise but was higher in the sedentary group after 3 min exercise compared with the moderately active and the very active group (P = 0.003 and P = 0.004; Fig. 1A). Mean arterial pressure was higher in the sedentary group at all exercise time points compared with the moderately active and the very active groups (P = 0.032 and P = 0.007; Fig. 1B). LVC was higher in the very active groups compared with the sedentary (P = 0.006) and moderately active groups (P = 0.004) after 20 s exercise but not at 60 s or 3 min exercise (Fig. 1C). Heart rate was higher in the sedentary group at all exercise time points compared with the very active group (P = 0.004; Fig. 1D). Leg oxygen uptake was higher in the very active group after 20 s exercise compared with the sedentary group (P = 0.016). There was no difference in leg oxygen uptake at 60 s exercise, but the level was higher in the sedentary group at 3 min of exercise compared with the moderately active group (P = 0.009; Fig. 1E). Leg oxygen delivery was higher in the very active group compared with the sedentary and moderately active group at 20 s exercise (P = 0.016 and P = 0.018; Fig. 1F). There was no difference in oxygen delivery between the groups at 60 s or 3 min of exercise. Leg arterio-venous oxygen difference was not different between the three groups at any time point (Fig. 1G). Leg arterio-venous lactate difference was higher in the sedentary group compared with the very active group at 20 s exercise (P = 0.029), but not at 60 s or 3 min of exercise (Fig. 1H). Leg lactate release was not different between the three groups at any time point (Fig. 1I).

Leg hemodynamics in the sedentary, moderately active and very active group of postmenopausal women during 3 min of one-legged knee extensor exercise. LBF (A), mean arterial pressure (B), LVC (C), heart rate (D), leg oxygen uptake (V˙O2, E), leg oxygen delivery (F), leg arteriovenous oxygen difference (G), leg arterio-venous lactate difference (H), leg lactate release (I) at rest and during the first 20, 60, and 180 s of one-leg knee extensor exercise at 6 W. Blood flow (J) and leg oxygen uptake kinetics (K) are presented as time to reach 63% amplitude. *Very active significantly different from sedentary. #Moderately active significantly different from sedentary. §Very active significantly different from moderately active. Data are presented as mean ± SD (A–I) or 5–95 percentile box and min/max whiskers (J–K).

Estimation of blood flow and oxygen uptake kinetics

Time to reach 63% amplitude in LBF from rest to exercise was significantly lower in the very active group compared with the sedentary and moderately active groups (P = 0.008 and P = 0.04; Fig. 1J). Time to reach 63% amplitude in leg oxygen uptake from rest to exercise was significantly lower in the very active group compared with the sedentary and moderately active groups (P = 0.021 and P = 0.040; Fig. 1K).

VEGF levels in platelets, plasma, and skeletal muscle

Platelet levels of VEGF (Fig. 2A) and platelet releasate levels of VEGF (Fig. 2B) were not different between the three groups. Platelet-free plasma levels of VEGF were not different between the three groups (Fig. 2D). Platelet-free plasma stimulated with TRAP-6 VEGF levels were not different between the three groups (Fig. 2E). Skeletal muscle VEGF levels were higher in the very active group compared with the sedentary group (P = 0.014; Fig. 2G) but not different than the moderate group.

Protein content of VEGF and angiogenic potential of platelets in the sedentary (SED), moderately active (MOD), and very active group (VERY) of postmenopausal women. VEGF content in platelets (A), in platelet releasate stimulated by TRAP-6 (B), and endothelial cell (EC) proliferation induced by platelet releasate (C). VEGF content in platelet-free plasma (D), platelet-free plasma stimulated with TRAP-6 (E), and EC proliferation induced by platelet-free plasma (F). VEGF (G) and OXPHOS complex I to complex V protein expression in skeletal muscle (H). Data are presented as mean, box is 5–95 percentile, and whiskers are min/max. *Very active significantly different from sedentary. #All groups significantly different from control and M200.

Proliferative potential of platelet-free plasma and platelet releasate

Platelet releasate and platelet-free plasma induced an approximately nine-fold increase (P < 0.001) in endothelial cell proliferation, but there was no significant difference in magnitude between any of the groups (Fig. 2C and F).

Mitochondrial OXPHOS protein levels in skeletal muscle

Complexes I, II, and V were higher in the very active group compared with the sedentary group (P < 0.01; Fig. 2H) with no difference compared with the moderate group. Complexes III and IV were not different between the groups.

Capillarization and fiber area

Capillary-to-fiber ratio was higher in the very active group compared with the sedentary and the moderately active group (P = 0.020 and P = 0.025; Table 2).

TABLE 2 - Capillarization and fiber type distribution.
Sedentary (n = 14) Moderately Active (n = 12) Very Active (n = 15)
Capillary–fiber ratio 1.30 ± 0.30 1.30 ± 0.19 1.65 ± 0.48 a,b
Capillary density (mm−2) 356.2 ± 66.3 365.0 ± 40.2 409.3 ± 57.5 a
Mean fiber area (μm2) 3669.1 ± 698.7 3622.2 ± 791.8 4035.2 ± 1090.1
Muscle fiber type distribution
MHC-I % 50.4 ± 16.5 40.1 ± 12.3 44.1 ± 12.5 b
MHC-IIa + x % 49.6 ± 16.5 57.9 ± 12.4 55.8 ± 12.5 b
Immunohistochemical analysis of muscle capillarization and fiber type distribution. Percent distribution of skeletal muscle fiber type, determined by myosin heavy chain (MHC) I and pooled IIa and IIx. Data are presented as mean ± SD.
aSignificantly different from the sedentary group.
bSignificantly different from the moderately active group.

Capillary density was higher in the very active group compared with the sedentary group (P = 0.039 and P = 0.033; Table 2). Representative immunohistochemical images are presented in Figure 3.

Histochemical staining. Representative histochemical staining of skeletal muscle fibers and capillaries from one sedentary, one moderately active, and one very active subject. Myofiber borders were visualized by using an antibody against laminin (blue), and capillaries were visualized using biotinylated Ulex (red). Muscle fiber type was determined by staining for MHC-I (green) and MHC-II (black).

Mean muscle fiber area was not different between any of the groups (Table 2).

Muscle fiber distribution was different between the very active group and the moderately active group with a higher percent of MHC1 and a lower percent of MHC2 in the very active group compared with the moderately active group (P = 0.008; Table 2) with no difference compared with the sedentary group.

Peak oxygen uptake–capillarization–O2 extraction relationships

Linear regression models of the effect of peak oxygen uptake on capillary density showed a positive relationship (P = 0.012; Fig. 4A), a tendency for a relationship between peak oxygen uptake and on leg vascular kinetics (P = 0.070; Fig. 4B), and a significant effect of capillary density on the leg vascular kinetics (P = 0.037; Fig. 4C). When modeling a linear regression with both peak oxygen uptake and capillary density as independent variables, the combined effect on leg vascular kinetics was significant (P = 0.017). A regression model of the effect of peak oxygen uptake on the arterial venous O2 difference at 3 min of one-legged knee extensor exercise showed a significant linear relationship (P = 0.034).

Correlations between muscle capillarization, maximal oxygen uptake (V˙O2max), and blood flow kinetics. Correlations between capillary density and V˙O2max (A), time to 63% amplitude blood flow and V˙O2max (B), and time to 63% amplitude blood flow uptake and capillary density (C).


The main findings in the present study were that the very active postmenopausal women presented a faster increase in femoral arterial blood flow and oxygen uptake at the onset of exercise compared with the moderately active and sedentary women. In addition, the very active women had a higher capillary density and capillary-to-fiber ratio as well as a greater content of skeletal muscle mitochondrial proteins and VEGF, whereas the level of VEGF in platelets and platelet-free plasma was similar between the groups. Platelet releasate and platelet-free plasma added to endothelial cells induced a marked increase in proliferation, but this effect was not different between the groups.

The current study included women with sedentary, moderately active, or very active lifelong physical activity levels. The women were divided into these three groups based on questionnaires and interviews regarding the volume and intensity of weekly training during the last 20 yr (Table 1). The division into groups was determined by set limits of regular physical activity, and there was no overlap in terms of either volume or intensity between the individuals in each group. This aspect is important to bear in mind for the below discussion.

Onset of exercise

At a functional level, the main findings in this study were that the very active postmenopausal women both had a faster increase in blood flow and thus oxygen delivery upon initiation of exercise compared with the two other groups of women. The functional implication of this observation would be an improved matching of oxygen delivery and uptake at the onset of exercise in lifelong trained women. To our knowledge, this is the first study to evaluate matching of leg oxygen delivery and uptake in postmenopausal women. In men of similar age, matching of oxygen delivery and uptake is impaired compared with young subjects (31), and in line with our current observations, exercise training has been shown to improve matching of leg oxygen delivery and uptake in older men (32).

Only a few studies have investigated blood flow and oxygen kinetics by invasive techniques (9,27), but there is evidence of improved oxygen uptake kinetics after a period of exercise training in young subjects (33,34). Moreover, in postmenopausal women, pulmonary oxygen uptake kinetics has been shown to be faster in endurance-trained compared with untrained women (35) and to be improved by 6 wk of exercise training in sedentary women (36). On this note, it is somewhat surprising that we did not find a clear difference in leg oxygen uptake kinetics between the sedentary and the moderately active group. The reason for this is unclear, but it is probable that the low-intensity exercise and thereby oxygen extraction demand, limited the sensitivity of the measurement and thereby the possibility to attain clear group differences. The low exercise intensity was necessary to allow the untrained subjects to complete the exercise.

Steady-state exercise

The steady-state level of blood flow and oxygen uptake at 3 min was lower in the very active than in the sedentary women. The lower oxygen uptake suggests greater mechanical efficiency, as also previously reported in trained men (37), but the mechanism underlying this difference is unclear. One possible factor is the distribution of Type I and Type II muscle fibers that express different mechanical efficiency. However, the present study found that the highest numerical fraction of the more efficient Type I fibers was present in the sedentary group (~50%), lowest in the moderately active group (~40%), whereas the very active group was in the middle (~44%). The lower blood flow in the very active women could be due to the lower oxygen uptake but also to a greater oxygen extraction. Endurance training has been shown to increase oxygen extraction (12), and although there was no detectable difference in oxygen extraction between the groups, there was a linear relationship between fitness level and oxygen extraction during exercise. Moreover, we identified capillarization as a significant mediator of the relationship between fitness level and oxygen extraction, which suggest that the higher capillarization in the very active group is an underlying mechanism of lower LBF during exercise.

The intensity of exercise was low but selected to allow all subjects to complete the bout; for some of the sedentary subjects, the intensity was experienced as hard, but it was very easy for the very active subjects. Accordingly, a higher exercise-induced lactate release was noted in the sedentary group compared with the very active group. It is plausible that exercise at a relative intensity for the three groups would have resulted in a different outcome.


The finding that older women with a high level of physical activity through life have a greater capillarization than women with a less active lifestyle was an expected finding (38), and the women included here express capillarization comparable with that of female master cyclist (39). Previous studies have shown that endurance-trained men may have up to a twofold higher capillary per fiber ratio and capillary density compared with untrained men of similar age (40). Nevertheless, although the present data agree well with these findings, the ~0.3 difference in capillary per fiber ratio between the sedentary and the very active older women was somewhat less than the difference of ~0.6 in capillary per fiber ratio previously observed in young sedentary versus endurance-trained women (18,41) and the difference of ~0.7 in capillary per fiber ratio in young endurance-trained versus sedentary men (42). This smaller difference in magnitude between the groups of older women is unclear but may be related to a lower training volume in older compared with young women and compared with men. Another possibility is that skeletal muscle microvascular endothelial cells of older women appear to have a markedly reduced capacity for proliferation (28), which could indicate that it is more difficult to maintain capillarization in women with aging. The absolute capillary per fiber ratio and capillary density in the group of sedentary women was also low compared with our own findings on men of similar age (43), as also previously reported (17,44).

The absolute levels of capillaries per fiber are very similar to the recent observations in lifelong trained versus sedentary postmenopausal women (18), but a surprising finding in the present study was that the moderately trained group had the same capillarization as the sedentary group. This could suggest that more intense exercise training is needed to maintain a better capillarization in postmenopausal women. However, there is one longitudinal study which indicates that exercise training, at a level comparable with that of the moderately active group, improves capillary-to-fiber ratio after 9–12 months in women of similar age (45), supporting that exercise-induced angiogenesis is achievable in ageing women. The accumulated evidence on capillarization in women is still very sparse, and future studies are needed in this area.

In parallel with a greater capillarization, the very active women also had a higher content of VEGF in the muscle. This finding is in line with the observation that a period of regular exercise training increases the muscle VEGF level in middle-age and in older men and women (2,43). Compared with previous findings in younger individuals (46), the muscle content of VEGF in the less active groups of women was lower, whereas it was similar to that in the very active women; thus, a very physically active lifestyle appears to maintain VEGF levels in the muscle with age.

A striking finding in the current study was that there was no difference in the level of capillarization between the habitually active and the sedentary women. This finding was paralleled by a similar content of mitochondrial OXPHOS proteins, a similar rate of increase in oxygen delivery in these two groups, and a similar change in vascular conductance in response to arterial acetylcholine infusion, as published previously on the same group of women (26). Combined, our data would suggest that, in women, a substantial amount of regular physical activity is required through life to counteract age-induced impairments in microvascular function and growth as well as in mitochondrial capacity. At a functional level, this would have clear implications for the oxygen delivery and extraction capacity. However, it should be noted that arterial blood pressure at rest and during exercise was markedly higher in the sedentary women compared with the moderately active and very active groups, which had similar blood pressures (Fig. 1B). Thus, other cardiovascular health factors such as arterial compliance may be more influenced by moderate levels of physical activity in women (47).

Platelets and plasma VEGF

Although skeletal muscle is an important source of VEGF for angiogenesis, VEGF carried in plasma or by platelets may also influence muscle angiogenesis (23,25). In this study, we included an assessment of platelet VEGF levels as well as their functional effect on proliferation of endothelial cells as biomarkers for angiogenic potential. Platelets and platelet-free plasma were isolated from blood samples with a novel method developed in our laboratory. Both the platelet fraction and the platelet-free plasma were found to contain VEGF protein, and both fractions induced a nine-fold increase in endothelial cell proliferation emphasizing their potential to stimulate capillary growth. Unlike the VEGF concentrations in muscle, the platelet and the plasma concentrations of VEGF and also their effect on endothelial cell proliferation were similar. This finding may imply that the angiogenic potential of platelets and platelet-free plasma is not influenced by habitual training. Nevertheless, it is important to point out that this finding does not exclude a role for platelets in exercise-induced muscle angiogenesis as it may be other aspects such as activation and binding of the platelets in response to each exercise bout that is more important than the absolute angiogenic potential. Animal studies have supported a role for platelets in angiogenesis in both skeletal muscle (25) and other tissues (24), and further studies evaluating their potential role in muscle angiogenesis humans are clearly warranted.


We compared skeletal muscle hemodynamics, capillarization, and the angiogenic potential of platelets in three groups of older postmenopausal women with varying lifelong habitual physical activity levels. Our findings show that very active women present more optimal skeletal muscle hemodynamics, both in terms of a faster increase at the onset of exercise and in terms of lower absolute submaximal blood flow, than moderately active and sedentary women. These aspects were not different between moderately physically active and sedentary women. Combined, these findings suggest that a high degree of regular physical activity through life is required for improvements in skeletal muscle hemodynamics and oxygen extraction capacity, although other cardiovascular variables such as blood pressure may require a lesser degree of physical activity. Moreover, the marked effect of platelet releasate on endothelial cell proliferation provides further support on the role of platelets in skeletal muscle angiogenesis in humans.

The study was supported by the Independent Research Fund Denmark–Medical Sciences, the Danish Ministry of Culture Fund for Sports Research, and the Memorial Foundation of Eva and Henry Frænkel.

The authors declare no conflicts of interest. The results of the present study do not constitute endorsement by the American College of Sports Medicine. The results of the study are presented clearly, honestly, and without fabrication, falsification, or inappropriate data manipulation.


1. Andersen P, Henriksson J. Capillary supply of the quadriceps femoris muscle of man: adaptive response to exercise. J Physiol. 1977;270(3):677–90.
2. Hansen AH, Nielsen JJ, Saltin B, Hellsten Y. Exercise training normalizes skeletal muscle vascular endothelial growth factor levels in patients with essential hypertension. J Hypertens. 2010;28(6):1176–85.
3. Harris BA. The influence of endurance and resistance exercise on muscle capillarization in the elderly: a review. Acta Physiol Scand. 2005;185(2):89–97.
4. Ryan NA, Zwetsloot KA, Westerkamp LM, Hickner RC, Pofahl WE, Gavin TP. Lower skeletal muscle capillarization and VEGF expression in aged vs. young men. J Appl Physiol. 2006;100(1):178–85.
5. Schrage WG, Eisenach JH, Joyner MJ. Ageing reduces nitric-oxide- and prostaglandin-mediated vasodilatation in exercising humans. J Physiol. 2007;579(1):227–36.
6. Nyberg M, Blackwell JR, Damsgaard R, Jones AM, Hellsten Y, Mortensen SP. Lifelong physical activity prevents an age-related reduction in arterial and skeletal muscle nitric oxide bioavailability in humans. J Physiol. 2012;590(21):5361–70.
7. Poole DC, Musch TI. Mechanistic insights into how advanced age moves the site of V˙O2 kinetics limitation upstream. J Appl Physiol. 2010;108(1):5–6.
8. DeLorey DS, Paterson DH, Kowalchuk JM. Effects of ageing on muscle O2 utilization and muscle oxygenation during the transition to moderate-intensity exercise. Appl Physiol Nutr Metab. 2007;32(6):1251–62.
9. Jones AM, Krustrup P, Wilkerson DP, Berger NJ, Calbet JA, Bangsbo J. Influence of exercise intensity on skeletal muscle blood flow, O2 extraction and O2 uptake on-kinetics. J Physiol. 2012;590(17):4363–76.
10. Egginton S. Invited review: activity-induced angiogenesis. Pflugers Arch. 2009;457(5):963–77.
11. Mortensen SP, Egginton S, Madsen M, et al. Alpha adrenergic receptor blockade increases capillarization and fractional O2 extraction and lowers blood flow in contracting human skeletal muscle. Acta Physiol (Oxf). 2017;221(1):32–43.
12. Krustrup P, Hellsten Y, Bangsbo J. Intense interval training enhances human skeletal muscle oxygen uptake in the initial phase of dynamic exercise at high but not at low intensities. J Physiol. 2004;559(1):335–45.
13. Chilibeck PD, Paterson DH, Cunningham DA, Taylor AW, Noble EG. Muscle capillarization O2 diffusion distance, and VO2 kinetics in old and young individuals. J Appl Physiol. 1997;82(1):63–9.
14. Proctor DN, Newcomer SC, Koch DW, Le KU, MacLean DA, Leuenberger UA. Leg blood flow during submaximal cycle ergometry is not reduced in healthy older normally active men. J Appl Physiol (1985). 2003;94(5):1859–69.
15. Nyberg M, Egelund J, Mandrup CM, et al. Early postmenopausal phase is associated with reduced prostacyclin-induced vasodilation that is reversed by exercise training: the Copenhagen Women Study. Hypertension. 2016;68(4):1011–20.
16. Parker BA, Smithmyer SL, Pelberg JA, Mishkin AD, Proctor DN. Sex-specific influence of aging on exercising leg blood flow. J Appl Physiol. 2008;104(3):655–64.
17. Coggan AR, Spina RJ, King DS, et al. Histochemical and enzymatic comparison of the gastrocnemius muscle of young and elderly men and women. J Gerontol. 1992;47(3):B71–6.
18. Gries KJ, Raue U, Perkins RK, et al. Cardiovascular and skeletal muscle health with lifelong exercise. J Appl Physiol. 2018;125(5):1636–45.
19. Croley AN, Zwetsloot KA, Westerkamp LM, et al. Lower capillarization, VEGF protein, and VEGF mRNA response to acute exercise in the vastus lateralis muscle of aged vs. young women. J Appl Physiol. 2005;99(5):1872–9.
20. Gavin TP, Kraus RM, Carrithers JA, Garry JP, Hickner RC. Aging and the skeletal muscle angiogenic response to exercise in women. J Gerontol A Biol Sci Med Sci. 2015;70(10):1189–97.
21. Olfert IM, Howlett RA, Wagner PD, Breen EC. Myocyte vascular endothelial growth factor is required for exercise-induced skeletal muscle angiogenesis. Am J Physiol Regul Integr Comp Physiol. 2010;299(4):R1059–67.
22. Wagner PD. The critical role of VEGF in skeletal muscle angiogenesis and blood flow. Biochem Soc Trans. 2011;39(6):1556–9.
23. Banks RE, Forbes MA, Kinsey SE, et al. Release of the angiogenic cytokine vascular endothelial growth factor (VEGF) from platelets: significance for VEGF measurements and cancer biology. Br J Cancer. 1998;77(6):956–64.
24. Karina, Wahyuningsih KA, Sobariah S, et al. Evaluation of platelet-rich plasma from diabetic donors shows increased platelet vascular endothelial growth factor release. Stem Cell Investig. 2019;6:43.
25. Packham IM, Watson SP, Bicknell R, Egginton S. In vivo evidence for platelet-induced physiological angiogenesis by a COX driven mechanism. PLoS One. 2014;9(9):e107503.
26. Gliemann L, Rytter N, Tamariz-Ellemann A, et al. Lifelong physical activity determines vascular function in late postmenopausal women. Med Sci Sports Exerc. 2020;52(3):627–36.
27. Krustrup P, Jones AM, Wilkerson DP, Calbet JAL, Bangsbo J. Muscular and pulmonary O2 uptake kinetics during moderate- and high-intensity sub-maximal knee-extensor exercise in humans. J Physiol. 2009;587(Pt 8):1843–56.
28. Olsen LN, Hoier B, Hansen CV, et al. Angiogenic potential is reduced in skeletal muscle of aged women. J Physiol. 2020;598:5149–64. doi:10.1113/JP280189.
29. Bates D, Mächler M, Bolker BM, Walker SC. Fitting linear mixed-effects models using lme4. J Stat Softw. 2015;67(1). Available from: doi:10.18637/jss.v067.i01.
30. Hothorn T, Bretz F, Westfall P. Simultaneous inference in general parametric models. Biom J. 2008;50(3):346–63.
31. Casey DP, Ranadive SM, Joyner MJ. Aging is associated with altered vasodilator kinetics in dynamically contracting muscle: role of nitric oxide. J Appl Physiol. 2015;119(3):232–41.
32. Piil P, Jørgensen TS, Egelund J, et al. Effects of aging and exercise training on leg hemodynamics and oxidative metabolism in the transition from rest to steady-state exercise: role of cGMP signaling. Am J Physiol Regul Integr Comp Physiol. 2018;315(2):R274–83.
33. Gunnarsson TP, Christensen PM, Thomassen M, Nielsen LR, Bangsbo J. Effect of intensified training on muscle ion kinetics, fatigue development, and repeated short-term performance in endurance-trained cyclists. Am J Physiol Regul Integr Comp Physiol. 2013;305(7):R811–21.
34. Christensen PM, Jacobs RA, Bonne T, Flück D, Bangsbo J, Lundby C. A short period of high-intensity interval training improves skeletal muscle mitochondrial function and pulmonary oxygen uptake kinetics. J Appl Physiol. 2016;120(11):1319–27.
35. Dogra S, Spencer MD, Murias JM, Paterson DH. Oxygen uptake kinetics in endurance-trained and untrained postmenopausal women. Appl Physiol Nutr Metab. 2013;38(2):154–60.
36. Stathokostas L, Kowalchuk JH, Paterson DH. Influence of hormone replacement therapy and aerobic exercise training on oxygen uptake kinetics in postmenopausal women. Appl Physiol Nutr Metab. 2013;38(6):657–65.
37. Proctor DN, Miller JD, Dietz NM, Minson CT, Joyner MJ. Reduced submaximal leg blood flow after high-intensity aerobic training. J Appl Physiol. 2001;91(6):2619–27.
38. Zumstein A, Mathieu O, Howald H, Hoppeler H. Morphometric analysis of the capillary supply in skeletal muscles of trained and untrained subjects—its limitations in muscle biopsies. Pflugers Arch. 1983;397(4):277–83.
39. Pollock RD, O’Brien KA, Daniels LJ, et al. Properties of the vastus lateralis muscle in relation to age and physiological function in master cyclists aged 55–79 years. Aging Cell. 2018;17(2):e12735.
40. Proctor DN, Sinning WE, Walro JM, Sieck GC, Lemon PW. Oxidative capacity of human muscle fiber types: effects of age and training status. J Appl Physiol. 1995;78(6):2033–8.
41. Ingjer F, Brodal P. Capillary supply of skeletal muscle fibers in untrained and endurance-trained women. Eur J Appl Physiol Occup Physiol. 1978;38(4):291–9.
42. Brodal P, Ingjer F, Hermansen L. Capillary supply of skeletal muscle fibers in untrained and endurance-trained men. Am J Physiol. 1977;232(6):H705–12.
43. Gliemann L, Olesen J, Biensø RS, et al. Resveratrol modulates the angiogenic response to exercise training in skeletal muscles of aged men. Am J Physiol Heart Circ Physiol. 2014;307(8):H1111–9.
44. Barnouin Y, McPhee JS, Butler-Browne G, et al. Coupling between skeletal muscle fiber size and capillarization is maintained during healthy aging. J Cachexia Sarcopenia Muscle. 2017;8(4):647–59.
45. Coggan AR, Spina RJ, King DS, et al. Skeletal muscle adaptations to endurance training in 60- to 70-yr-old men and women. J Appl Physiol. 1992;72(5):1780–6.
46. Hoier B, Olsen K, Hanskov DJA, Jorgensen M, Norup LR, Hellsten Y. Early time course of change in angiogenic proteins in human skeletal muscle and vascular cells with endurance training. Scand J Med Sci Sports. 2020;30(7):1117–31.
47. Tanaka H, DeSouza CA, Seals DR. Absence of age-related increase in central arterial stiffness in physically active women. Arterioscler Thromb Vasc Biol. 1998;18(1):127–32.


Supplemental Digital Content

Copyright © 2021 by the American College of Sports Medicine