Nine male chronic (>1 yr), complete (ASIA A) spinal cord injured (SCI) subjects and eight male able-bodied (AB) subjects volunteered to participate in the study. The AB group was recruited from a university community, whereas the SCI group was recruited from the Shepherd Center in Atlanta, GA. The physical characteristics of the SCI and AB subjects are shown in Table 1. The time since injury and level of injury for the SCI individuals are shown in Table 2. The AB subjects, who served as controls were included if they were similar to the SCI group on age, height, and weight and if they did not report exercising more than 3 d a week and were not engaged in any formal exercise-training program for 6 months before the study. Exercise history in the AB subjects was obtained by self-report. Subjects were excluded if they reported that they were smokers. None of the subjects had any history of disease or other confounding factors. Medications were recorded and the only medication that was used in either group was antispasticity medication (Baclofen) in three of the SCI individuals. Baclofen has had no known reported vascular effects. Magnetic resonance (MR) image collection for all subjects occurred at Shepherd Center, Atlanta, GA, or Health South, Athens, GA. The study was conducted with the approval of the Institutional Review Board at the University of Georgia, and all subjects provided written informed consent.
Subjects were asked to abstain from fatty foods, caffeine, and alcohol for at least 12 h before testing. Subjects were placed in a supine position on a table for at least 10 min before testing. Blood pressure was measured in the arm throughout the entire testing period using an automated blood pressure machine (Datascope, Mahwah, NJ). Resting artery diameter and blood flow were measured in the femoral artery using Doppler ultrasound. Two trials of leg ischemia were performed by inflating a 12-mm–wide cuff just above the knee. Ischemia was induced in the distal thigh and leg by inflation of a blood pressure cuff distal to the probe to a pressure 100 Torr above systolic pressure. A minimum of 5 min of recovery was allowed between trials. The durations of cuff ischemia were 4 and 10 min. Preliminary studies in our laboratory indicated that there is no order effect for duration of cuff ischemia. Thus, longer durations of ischemia were done last to minimize discomfort to the subject. Cuff inflation and deflation were rapid (1–2 s) and performed using a Hokanson (Bellevue, WA) device.
Blood flow was measured in the femoral artery using quantitative Doppler ultrasound (General Electric LogiQ 400CL, Rainbow City, AL). A linear array transducer was used at a frequency of 6–9 MHz. The imaging site was located on the upper third of the thigh and was marked to ensure replication of probe placement. Doppler measurements were made proximal to the cuff to ensure that the vessel placement was maintained throughout cuff occlusion. Resting diameter was measured in the axial view during diastole. Pulsed Doppler ultrasound was recorded in the longitudinal view using an insonation angle between 45° and 60°. The velocity gate was set to include the entire arterial diameter.
Velocity measurements were auto calculated every heartbeat by General Electric’s advanced vascular program software for the LogiQ 400 CL. The minimum, maximum, and time-averaged maximum velocity measurements were saved directly to a computer allowing data acquisition on a beat-by-beat basis. Images were saved to magnetic optical disks for measurement of vessel diameter by custom-made software (Labview 6i, Austin, TX). Blood flow was calculated by the product of vessel CSA and the time average maximum velocity. The resting vessel diameter was used for calculation of blood flow, as no dilation was found in the femoral artery during cuff occlusion. The observation that the femoral artery does not significantly dilate during cuff occlusion has been reported previously (20) and has been seen in our previous studies with exercise (18) and cuff occlusion (17). Conductance was calculated by the blood flow divided by the mean arterial pressure (MAP).
Maximum blood flow was determined as the highest blood flow for each ischemic test. The halftime to recovery was determined as the point at which blood flow dropped to one half the difference between maximum flow and resting flow. This was used as an index of vascular reactivity as it reflects the ability of the arteries to return blood flow to resting values (19).
Near infrared spectroscopy (NIRS).
Muscle oxygen delivery was determined by the use of NIRS halftime to recovery (15). NIRS measurements were made using a continuous light source, dual wavelength spectrophotometer (Runman, NIM, Inc., Philadelphia, PA). The probe (2 × 6 cm) contained two small tungsten filament lamps 6 cm apart, which emit white light, and two photo detectors with filters for 760- and 850-nm light located between the lights (effective separation distance between lamps and detectors of 3 cm). The light sources were a pair of small tungsten flashlight bulbs, which gave brief flashes of light every 1.2 s with a power level of less than 1 W. The light photons migrated through the tissue and were collected by the detectors at wavelengths set by two optical filters. The difference in signal between light-on and light-off was used to correct for the presence of ambient light. Oxyhemoglobin has a greater absorbance at 850 nm compared with 750 nm, with deoxyhemoglobin absorbing more at 760 than 850 nm. The difference signal between 760 and 850 nm was used as the index of relative oxygen saturation.
Skeletal muscle average CSA was determined using proton-weighted MR imaging. This method has been shown to be highly reproducible and reliable for determination of skeletal muscle composition. MR images of the both legs and thighs were collected with a 1.5-T magnet (TR/TE 500/14, 40-cm field of view, 1 NEX, 256 × 256 matrix, General Electric, Milwaukee, WI). Transaxial images, 1-cm thick and 0.5 cm apart, were taken from the knee joint to the ankle joint (leg) and from the hip joint to the knee joint (thigh) using a whole-body coil. Each of the subject’s feet was strapped to a brace to maintain the ankle joint at approximately 90° and the knee and hip joint extended.
Data were downloaded to a disk and analyzed on specifically designed software (X-Vessel, East Lansing, MI). The methodology for the image data analyses has been reported previously (11). Briefly, images were automatically segmented into fat (high-intensity), muscle (mid-intensity), and background/bone (low-intensity) regions as follows. A preliminary segmentation of each two-dimensional slice into fat and nonfat regions was obtained by simplex optimization of the correlation between a Sobel-gradient image computed from the original image and the corresponding gradient images computed from single-threshold fat-segmented images. This first-pass segmentation was used to correct for intensity variations across the original image caused by RF heterogeneity. The corrected original image was then re-segmented into the three intensity components using a fuzzy c-mean clustering algorithm. Manual selection of a pixel of muscle subsequently highlights all muscle pixels in the region, and provides a total number of muscle pixels exclusive of any fat or low pixels. The individual collecting the data performed two trials on sample images to determine test retest reliability (r = 0.99). Pixel number was converted to area by multiplying by the field of view and dividing by the total number of pixels in the entire image.
For each subject, leg CSA was determined with the first distal slice containing the head of the soleus and continuing proximally toward the knee until the end of the tibia was reached. Thigh CSA was determined with the first slice not containing gluteal muscle and continuing distally toward the knee until the patella could be seen. Muscle volume was calculated by determining the CSA for each slice, multiplying it by the thickness of the slice taking into account interslice distance and then summing them together. Leg muscle volume was used to normalize peak blood flow measurements from the cuff occlusions. Limb muscle volume (leg + thigh muscle volume) was used to normalize femoral artery diameter to total muscle volume of the lower limb.
Independent samples t-tests (SPSS version 10.0) were conducted to compare for differences in subject characteristics and femoral artery diameters between the two groups. The data were analyzed to verify normality and to test for any outliers. Levene’s test was conducted to determine equality of variances and was corrected for if inequality was found. A mixed-model repeated-measures analysis was used to determine differences between AB and SCI groups across cuff durations for blood flow, conductance, and halftime to recovery. Mauchly’s test was conducted to determine if sphericity was violated. If sphericity was violated, the repeated measures ANOVA was corrected using the Greenhouse-Geiser correction factor. Femoral artery diameter and peak blood flow were expressed as a ratio to limb and leg muscle volume, respectively, to determine whether differences between groups in these variables were explained by muscle volume differences. All analyses were conducted at a significance level of 0.05.
The AB and the SCI groups were not significantly different in age, height, weight, resting heart rate, or blood pressure (Table 1). Individual data for the duration of injury and level of injury in the SCI subjects can be found in Table 2. Limb muscle volume was significantly lower (≈ 40%) in SCI compared with AB individuals (t (15) = 4.116, P = 0.001, d = 2.0) (Fig. 1). Similarly, SCI individuals had smaller volumes for the thigh and leg.
SCI individuals had significantly smaller (37%) femoral artery diameters than the AB individuals (t (12) = 4.965, P < 0.001, d = 2.53). However, the ratio of femoral artery diameter to limb volume was not significantly different between groups (P = 0.418) (Fig. 2). Absolute resting blood flow was not significantly different between AB (200 ± 80 mL·min−1) and SCI (280 ± 110 mL·min−1) individuals. However, the ratio of resting blood flow to lower leg volume was approximately twice as high in SCI (11 ± 6 mL·100 cm−3·min−1) than AB (5 ± 2 mL·100 cm−3·min−1) individuals (t (10.120) = 2.946, P = 0.014, d = 1.4).
There were no significant differences between the SCI and AB groups in heart rate (SCI −75 ± 14 bpm; AB −71 ± 14 bpm;t (15) = 0.684, P = 0.504) or blood pressure during or immediately after cuff occlusion (systolic: SCI −124 ± 16 mm Hg; AB −125 ± 8 mm Hg;t (15) = 0.234, P = 0.818; diastolic: SCI −79 ± 13 mm Hg; AB −76 ± 5 mm Hg;t (15) = 0.656, P = 0.522; MAP: SCI −97 ± 14 mm Hg; AB −96 ± 5 mm Hg;t (15) = 0.148, P = 0.884). During cuff ischemia, the proximal placement of the Doppler probe relative to the cuff reduced blood flow to ≈25% of resting flow. Upon release of the cuff, there was an initial burst of blood flow with the first heartbeat and then a large hyperemic response. Ignoring the first heart beat, the flow response increased to a peak value approximately 2–15 s after release of the cuff and then returned to the resting value (Fig. 3). The peak blood flow response was dependent upon duration of ischemia (10 min > 4 min, F (1,15) = 44.062, P < 0.001, η2 = 0.746). Peak blood flow was significantly lower (≈ 40%) in the SCI compared with AB individuals for both cuff durations (F (1,15) = 21.832, P < 0.001, η2 = 0.593). However, the ratios of peak blood flow to leg volume were not significantly different between the SCI and AB groups (P = 0.891) (Fig. 4). Conductance was significantly different between cuff durations (F (1,15) = 50.855, P < 0.001, η2 = 0.772) and was significantly reduced in the SCI (0.17 ± 0.03 mL·s−1·mm Hg−1 and 0.22 ± 0.03 mL·s−1·mm Hg−1 for 4 and 10 min cuff, respectively) compared with AB individuals (0.31 ± 0.07 mL·s−1·mm Hg−1 and 0.39 ± 0.09 mL·s−1·mm Hg−1 for 4 and 10 min cuff, respectively) (F (1,15) = 32.618, P < 0.001, η2 = 0.685). The conductance followed the trends of blood flow as would be expected, as there were no changes in MAP throughout cuff occlusion.
The halftime for blood flow to return to resting values also increased with increasing duration of cuff ischemia for each group (10 min > 4 min, F (1,15) = 70.490, P < 0.001, η2 = 0.825). The SCI group took significantly longer compared with the AB group for blood flow to return to baseline from peak blood flow across cuff durations (F (1,15) = 4.635, P = 0.048, η2 = 0.236) (Figs. 3 and 5). SCI individuals showed a similar pattern of oxygen desaturation with cuff ischemia to those of normal subjects (Fig. 6). Halftime to recovery of oxygen saturation as measured with NIRS was not significantly different between the SCI (14 ± 6 s and 16 ± 10 s for 4 and 10 min cuff, respectively) and the AB (12 ± 5 s and 15 ± 6 s for 4 and 10 min cuff, respectively) groups for either cuff duration. These findings indicate that there was no difference in oxygen delivery between the two groups.
The primary findings of this study were that femoral artery diameter and peak blood flow response after cuff occlusion were reduced in the SCI compared with able-bodied individuals. However, these reductions were evident only for the absolute values, as correcting for the reduced muscle volume in SCI individuals eliminated the differences in diameter and blood flow. To our knowledge, this is the first study to investigate vascular changes in SCI individuals in relation to changes in muscle volume. This result suggests that adaptations in muscle vascular function closely parallel skeletal muscle atrophy in SCI individuals. Furthermore, we observed a prolonged blood flow recovery after cuff occlusion in SCI individuals possibly indicating altered vascular reactivity.
We found approximately a 35% reduction in lower limb muscle CSA and volume of chronic, complete SCI compared with AB subjects. It is well documented that significant muscle atrophy occurs after complete SCI due to extreme inactivity. Mean CSA of single fibers from SCI individuals is two thirds the size of able-bodied muscle (14). Muscle CSA as determined by MRI, is reduced by one third in the muscles of the lower limb (3) and in the quadriceps femoris (6) compared with age-matched control subjects. Other studies have indicated that muscle CSA decreases from 1 to 17 months after injury and is roughly 50% of able-bodied controls (24). It should be noted that our measurements of muscle volume underestimated the total leg muscle volume. This was done to conform to the sensitive volume of the magnetic resonance magnet and to make use of reproducible landmarks. The muscle volume measures should be relatively consistent between SCI and able-bodied individuals. Interestingly, by excluding the thigh muscles above the start of the gluteal muscle, we may be more accurately matching muscle mass perfusion to the femoral artery at the level where vessel diameter measurements were made.
Our results indicate that femoral artery diameter size was significantly reduced by 40% in SCI compared with able-bodied individuals. However, this difference was no longer significant when the diameter was expressed relative to muscle volume. Our finding without normalization to muscle volume is consistent with previous literature in chronic tetraplegics (16) and paraplegics (1,7,8). These results are significant as they suggest that decreases in arterial diameter size may be matched to muscle atrophy that occurs in SCI individuals.
Peak blood flow response to reactive hyperemia was significantly lower in the SCI compared with the able-bodied group. This finding is consistent with previous literature that has shown blood flow is lower in SCI compared with able-bodied individuals after 5 min of cuff occlusion (16) and with exercise (8). However, when we expressed blood flow relative to lower leg muscle volume there was no longer a significant difference in response to cuff occlusion (Fig. 4). These findings indicate muscle blood flow per unit muscle mass is not impaired in SCI individuals after reactive hyperemia.
NIRS was used as an independent measure of blood flow. A previous study has used the rate of recovery of oxygen saturation after cuff ischemia as a marker of the ability to deliver oxygen (15). With short cuff durations (up to 4–5 min), there is little or no depletion of phosphocreatine, and thus the hyperemic flow response after cuff release should reflect the inflow of oxygenated blood. The lack of difference in rate of reoxygenation between SCI and able-bodied individuals supports the hypothesis that normalized blood flow is not impaired in SCI individuals. In this study, we used 10 min of cuff ischemia, which should have resulted in some depletion of phosphocreatine (15) and thus an increase in postcuff oxygen consumption. As SCI patients have reduced mitochondrial density (2) and impaired muscle oxidative enzymes (23), we expected the recovery from the 10 min cuff to be slower in SCI than for able-bodied individuals. This was not the case, and we can only suggest that the amount of phosphocreatine depletion with 10 min of cuff ischemia (not measured) was not enough to demonstrate the expected differences in oxidative capacity.
We did not find a significant difference in resting blood flow between groups. This finding is in contrast to several studies in which a lower resting blood flow has been reported in SCI individuals (8,29). The difference between results could be due to methodological differences between studies. By normalizing our resting data by muscle mass, we actually found that resting blood flow was higher in the SCI group than the able-bodied group. It is possible that the high resting blood flows in the SCI group in our study were a result of our not obtaining a “true” resting condition. Some of our SCI subjects had muscle spasms when they were transferred to the examination table. This increase blood flow coupled with a slow rate of recovery of blood flow may have elevated our resting values. Our study design included waiting 10 min after transferring to try to account for this, but it is possible that resting blood flow was still elevated. Another possibility is that blood flow at rest may be in excess after disuse due to poor coupling between metabolic demand and blood flow (30). This would be consistent with a deconditioned vascular system or the decrease in sympathetic activity found below the level of injury in SCI subjects (9).
A prolonged halftime to recovery of blood flow was found in the SCI compared with the able-bodied group. The prolonged halftime to recovery of blood flow is in agreement with our previous results in incomplete SCI (17) and complete SCI individuals (18). Halftime to recovery of blood flow has been used previously as an index of vascular reactivity (22,25), and this would suggest that SCI individuals have reduced vascular reactivity. Activity status has been related to vascular reactivity as inactivity is associated with reduced vascular reactivity (26) and exercise training improves vascular reactivity (19). Furthermore, reduced vascular reactivity has been associated with many diseases including diabetes (27) and heart disease (4). This study was not designed to address the mechanism behind the reduced vascular reactivity. It could indicate a greater buildup of metabolic factors/vasoactive substances after cuff occlusion and/or a diminished ability to remove them. Reduced sympathetic tone may also play a role. The lack of any differences between SCI and able-bodied individuals in the pattern of desaturation during cuff ischemia or the rate of recovery after ischemia suggest that the reduced vascular reactivity is not due to differences in metabolic rate or in oxygen delivery. Regardless of the exact mechanism, the prolonged halftime to recovery would suggest that vascular reactivity is altered in the SCI subjects. Future studies could be designed to look at the exact mechanism behind this finding. Consequences of reduced vascular reactivity are abnormal oxidative metabolism, increased insulin insensitivity, or hypertension (31); all of which are prevalent in SCI individuals (12).
Lastly, our results indicate that a maximal blood flow response is not elicited with 4 min of cuff occlusion, as there was a significant increase in peak blood flow from 4 to 10 min of cuff occlusion. Furthermore, the groups were similar in their responses to both 4 and 10 min of cuff occlusion. This finding is contrary to others that suggest using different cuff durations to obtain maximal flow for various populations (5). Our results suggest that similar cuff durations should be used when comparing between able-bodied and SCI subjects and that at least a 10-min cuff is needed to elicit a maximal response. Previous results in our laboratory have indicated that flow after 10 min of cuff occlusion has been sufficient to elicit a maximal response in able-bodied subjects (Olive, J. L., unpublished observations, July 2001).
In conclusion, this study found that lower-limb muscle volume, femoral artery diameter size, and blood flow were significantly reduced in complete, chronic SCI individuals compared with able-bodied individuals. The significant reductions in femoral artery size and blood flow in SCI individuals were no longer different when expressed per unit muscle volume. This study is one of the first studies that have demonstrated that vascular remodeling and muscular atrophy are closely linked. Furthermore, we found a significant prolonged recovery of blood flow to baseline after cuff occlusion, which may suggest reduced vascular reactivity in complete SCI subjects. Future studies need to investigate the mechanisms and the time course for vascular remodeling and muscle atrophy to determine whether they are dependent upon each other.
The authors would like to acknowledge Ron Meyer, Ph.D., for his assistance in magnetic resonance image analysis by use of X-Vessel, and Scott Bickel and Jill Slade for their assistance with subject recruitment. We would also like to thank Chris Black, Vanessa Castellano, Lee Stoner, Allison DeVan, Chris Elder, Kristen Dudley, and Michelle Layton for their assistance in data collection.
Financial support was provided by the Paralyzed Veterans Association and NIH grants HL65179 and HD39676.
1. Boot, C. R. L., J. T. Groothuis, H. van Langen, and M. T. E. Hopman. Shear stress levels in paralyzed legs of spinal cord-injured individuals with and without nerve degeneration. J. Appl. Physiol. 92: 2335–2340, 2002.
2. Borisov, A. B., S. K. Huang, and B. M. Carlson. Remodeling of the vascular bed and progressive loss of capillaries in denervated skeletal muscle. Anat. Rec. 258: 292–304, 2000.
3. Castro, M., D. Apple, E. Hillegass, and G. Dudley. Influence of complete spinal cord injury on skeletal muscle cross-sectional area within six months of injury. Eur. J. Appl. Physiol. 80: 373–378, 1999.
4. Celermajer, D. S., K. Sorensen, V. Gooch, et al. Non-invasive detection of endothelial dysfunction in children and adults at risk of atherosclerosis. Lancet 340: 1111–1115, 1992.
5. Gerrits, H., A. de Haan, A. Sargeant, H. van Langen, and M. Hopman. Peripheral vascular changes after electrically stimulated cycle training in people with spinal cord injury. Arch Phys. Med. Rehabil. 82: 832–839, 2001.
6. Hillegass, E. A., and G. A. Dudley. Surface electrical stimulation of skeletal muscle after spinal cord injury. Spinal Cord 37: 251–257, 1999.
7. Hopman, M. T. E., E. Nommensen, W. N. J. van Asten, B. Oeseburg, and R. A. Binkhorst. Properties of the venous vascular system in the lower extremities of individuals with paraplegia. Paraplegia 32: 810–816, 1994.
8. Hopman, M. T. E., W. van Asten, and B. Oeseburg. Changes in blood flow in the common femoral artery related to inactivity and muscle atrophy in individuals with long-standing paraplegia. Adv. Exp Med. Biol. 388: 379–383, 1996.
9. Houtman, S., W. Colier, B. Oeseburg, and M. Hopman. Systemic circulation and cerebral oxygenation during head-up tilt in spinal cord injured individuals. Spinal Cord 38: 158–163, 2000.
10. Kamiya, A., S. Iwase, D. Michikami, Q. Fu, and T. Mano. Head-down bed rest alters sympathetic and cardiovascular responses to mental stress. Am. J. Physiol. Regul. Integr. Comp Physiol. 279: R440–R447, 2000.
11. Kanaley, J. A., C. Sames, L. Swisher, et al. Abdominal fat distribution in pre-and postmenopausal women: the impact of physical activity, age, and menopausal status. Metabolism 50: 976–982, 2001.
12. Kocina, P. Body composition of spinal cord injured adults. Sports Med. 23: 48–60, 1997.
13. Kroese, A. The effect of inactivity on reactive hyperemia in the human calf: a study with strain gauge plethysmography. Scand. J. Clin. Lab Invest. 37: 53–58, 1977.
14. Martin, T., R. Stein, P. Hoeppner, and D. Reid. Influence of electrical stimulation on the morphological and metabolic properties of paralyzed muscle. J. Appl. Physiol. 72: 1393–1400, 1992.
15. Mccully, K., and B. Natelson. Impaired oxygen delivery in chronic fatigue syndrome. Clin. Sci. 97: 603–608, 1999.
16. Nash, M. S., B. M. Montalvo, and B. Applegate. Lower extremity blood flow and responses to occlusion ischemia differ in exercise-trained and sedentary tetraplegic persons. Arch. Phys. Med. Rehabil. 77: 1260–1265, 1996.
17. Olive, J. L., K. K. Mccully, and G. A. Dudley. Blood flow response in individuals with incomplete spinal cord injuries. Spinal Cord 40: 640–646, 2002.
18. Olive, J. L., J. Slade, G. Dudley, and K. Mccully. Is muscle fatigue related to blood flow: a comparison between able-bodied and spinal cord injured individuals. J. Appl. Physiol. 94: 701–708, 2003.
19. Parker, J., C. Oltman, J. Muller, P. Myers, H. Adams, and M. H. Laughlin. Effects of exercise training on regulation of tone in coronary arteries and arterioles. Med. Sci. Sports Exerc. 26: 1252–1261, 1994.
20. Radegran, G. Ultrasound Doppler estimates of femoral artery blood flow during dynamic knee extensor exercise in humans. J. Appl. Physiol. 83: 1383–1388, 1997.
21. Radegran, G., and B. Saltin. Human femoral artery diameter in relation to knee extensor muscle mass, peak blood flow, and oxygen uptake. Am. J. Physiol. Heart Circ. Physiol. 278: H162–H167, 2000.
22. Radegran, G., and B. Saltin. Nitric oxide in the regulation of vasomotor tone in human skeletal muscle. Am. J. Physiol. Heart Circ. Physiol. 276: H1951–H1960, 1999.
23. Rochester, L., M. Barron, C. Chandler, R. Sutton, S. Miller, and M. Johnson. Influence of electrical stimulation of the tibialis anterior muscle in paraplegic subjects: 2. Morphological and histochemical properties. Paraplegia 33: 514–522, 1995.
24. Scelsi, R., C. Marchetti, P. Poggi, S. Lotta, and G. Lommi. Muscle fiber type morphology and distribution in paraplegic patients with traumatic cord lesion. Acta Neuropathol. (Berl.) 57: 243–248, 1982.
25. Shoemaker, J. K., J. R. Halliwill, R. L. Hughson, and M. J. Joyner. Contributions of acetylcholine and nitric oxide to forearm blood flow at exercise onset and recovery. Am. J. Physiol. Heart Circ. Physiol. 273: H2388–H2395, 1997.
26. Shoemaker, J. K., C. S. Hogeman, D. H. Silber, K. Gray, M. Herr, and L. I. Sinoway. Head-down-tilt bed rest alters forearm vasodilator and vasoconstrictor responses. J. Appl. Physiol. 84: 1756–1762, 1998.
27. Steinberg, H., H. Chaker, R. Leaming, A. Johnson, G. Brechtel, and A. Baron. Obesity/insulin resistance is associated with endothelial dysfunction. J. Clin. Invest. 97: 2601–2610, 1996.
28. Steinberg, H., G. Paradisi, J. Cronin, et al. Type II diabetes abrogates sex differences in endothelial function in premenopausal women. Circulation 101: 2040–2046, 2000.
29. Taylor, P., D. Ewins, B. Fox, D. Grundy, and I. Swain. Limb blood flow, cardiac output and quadriceps muscle bulk following spinal cord injury and the effect of training for the Odstock functional electrical stimulation standing system. Paraplegia 31: 303–310, 1993.
30. Tyml, K., and O. Mathieu-Costello. Structural and functional changes in the microvasculature of disused skeletal muscle. Front Biosci. 6: d45–d52, 2001.
31. Villa, E., O. Gonzalez-Albarran, A. Rabano, and R. Garcia-Robles. Effects of hyperinsulinemia on vascular blood flow in experimental obesity. J. Steroid Biochem. Mol. Biol. 69: 273–279, 1999.