Spinal cord injury and vascular function: evidence from diameter-matched vessels

Abstract

The effect of a spinal cord injury (SCI) on vascular function has been clouded by both the physiological and mathematical bias of assessing vasodilation in arteries with differing diameters both above and below the lesion and when comparing with healthy, nondisabled controls (CTRL). Thus, we measured vascular function, with flow-mediated vasodilation (FMD), in 10 SCI and 10 CTRL with all arteries matched for diameter (≈0.5 cm): brachial artery (BA, arm, functional limb in both groups) and popliteal artery (PA, leg, disused limb in SCI, functional limb in CTRL). PA %FMD was significantly attenuated in SCI (5.6 ± 0.6%) compared with CTRL (8.4 ± 1.3%), with no difference in the BA (SCI: 8.6 ± 0.9%; CTRL: 8.7 ± 0.7%). However, unlike the arm, where muscle mass was preserved, the legs of the SCI were significantly smaller than CTRL (∼70%). Thus, reactive hyperemia (RH), which is heavily dependent on the volume of muscle occluded, in the PA was attenuated in the SCI (144 ± 22 mL) compared with CTRL (258 ± 16 mL) but not different in the BA. Consequently, shear rate was significantly diminished in the PA of the SCI, such that %FMD/shear rate (vascular responsiveness) was actually greater in the SCI (1.5 ± 0.1% · s⁻¹) than CTRL (1.2 ± 0.1% · s⁻¹). Of note, this was significantly greater than both their own BA (0.9 ± 0.1% · s⁻¹) and that of the CTRL (0.9 ± 0.1% · s⁻¹). Therefore, examining vessels of similar size, this study reveals normal vascular function above the lesion and vascular dysfunction below the lesion. However, below the lesion there was, actually, evidence of increased vascular responsiveness in this population.

NEW & NOTEWORTHY This study examined the effect of a spinal cord injury (SCI) and subsequent limb disuse on vascular function, assessed by %FMD, in diameter-matched vessels above and below the lesion in subjects with SCI and controls. The results reveal normal vascular function above the lesion and vascular dysfunction below the lesion (%FMD). However, below the lesion there was, actually, evidence of increased vascular responsiveness (%FMD/shear rate) in this population.

INTRODUCTION

A spinal cord injury (SCI), resulting in either paraplegia or tetraplegia, is one of most devastating traumatic events that a person can experience, with a worldwide annual incidence of up to 500,000 people (1–3). Cardiovascular sequelae are a common consequence of an SCI, and these cardiovascular conditions are the primary cause of morbidity and mortality, both acutely and chronically, following such an injury (4–7). Cardiovascular complications with an SCI often emanate from a severe impairment in the control of the autonomic nervous system, resulting in arterial hypotension, bradycardia, and autonomic dysreflexia. Moreover, specific vascular complications, such as deep vein thrombosis, coronary artery disease, and systemic atherosclerosis, are common findings in individuals with an SCI (8).

With such clear evidence of an increased risk and prevalence of cardiovascular disease with an SCI, an understanding of the effect of an SCI and subsequent limb disuse on vascular dysfunction (9), a prodromal symptom of hypertension, and cardiovascular disease would be beneficial. Flow-mediated vasodilation (FMD), which assesses the change in artery diameter in response to reactive hyperemia (RH), subsequent to a brief period of ischemia, provides an index of conduit artery vascular function (10, 11). When appropriately performed (11), this noninvasive method of assessing vascular function can be applied to the upper or lower limbs and can provide an indication of both local and systemic vascular function (12). With an SCI, resulting in paraplegia, the assessment of FMD in both a functional arm and a disused leg, in combination with such measures in healthy, nondisabled controls (CTRL), may provide insight into the local and systemic consequences of an SCI on vascular function.

Previous studies evaluating the impact of SCI on FMD below the spinal cord lesion have been equivocal. For example, De Groot et al. (13, 14) described preserved vascular function in the superficial femoral artery of subjects with an SCI, resulting in an 11%–15% flow-induced increase in artery diameter. The same group also reported a significantly greater FMD response in subjects with an SCI (8%) compared with CTRL (4%) (15, 16). Interestingly, in contrast, Stoner et al. (17) reported that FMD was not different with an SCI compared with CTRL, when performed in the posterior tibial artery. However, such studies, and their conclusions, are complicated by the reduction in vessel size in the legs of subjects with an SCI, which influences this vascular function assessment both in terms of the varying intrinsic reactivity of differing sized vessels and the shear stimulus that is also altered (15, 16). Indeed, below the spinal lesion, there is consistent evidence of significant arterial remodeling, with a ∼22% smaller baseline diameter in the superficial femoral and popliteal arteries in subjects with an SCI compared with healthy CTRL (18). Adding further complexity to these changes in vessel size, in terms of FMD, is the impact of the diminished leg muscle mass with an SCI, which greatly attenuates the reactive hyperemia (RH) following cuff release. Therefore, interestingly, when the FMD is normalized by shear rate, most of the previous investigations report similar results in subjects with an SCI and CTRL (15, 16), suggesting preserved vascular responsiveness in the disused limbs (11). However, it is likely that greater clarity, with respect to the effect of an SCI and subsequent limb disuse on vascular responsiveness, can be achieved by the assessment of diameter-matched vessels above and below the lesion, but this has yet to be performed.

Therefore, in diameter-matched vessels above (brachial artery, BA) and below (popliteal, artery, PA) the lesion in subjects with SCI-induced paraplegia and in healthy, nondisabled CTRL, the aim of this study was to determine the effect of an SCI and the multiple sequelae of such an injury on vascular function. We tested two hypotheses: first, an SCI does not systemically attenuate vascular function, whereas, second, the many consequences of an SCI, including limb disuse and denervation, do attenuate vascular function.

METHODS

Subjects

Ten subjects with an SCI (8 men and 2 women) were initially recruited for this study. The 10 healthy, nondisabled CTRL (8 men and 2 women) were then recruited to match the subjects with an SCI in terms of age, sex, and BA and PA diameters. All subjects with an SCI had clinically confirmed complete spinal cord lesions between the sixth thoracic vertebra (T-6) and the 12th thoracic vertebra (T-12) (American Spinal Injury Association class A) (19). The level of lesion was also tested in the laboratory by neurological exam to confirm the absence of feedback (cutaneous pinprick and cold perception) from the legs and the maintenance of sensation in the arms. For the SCI group, the average time post injury was 16 ± 13 yr (2–36 yr). None of the subjects were smokers, and most of the subjects with an SCI were physically active performing endurance-type exercise for 6.5 ± 3.5 h a week, whereas the CTRL were less active, performing 2.0 ± 1.5 h of exercise a week. It is important to note that all the subjects with an SCI were involved in endurance activities, performed only with the arms, and none of these subjects regularly engaged in either active or passive stimulation of the limbs below the spinal cord lesion. Descriptive characteristics of the subjects are presented in Table 1 (20). All procedures conformed to the standards set by the Declaration of Helsinki, and the Institutional Review Boards of the University of Utah and the Salt Lake City VA Medical Center approved the study. Written informed consent was obtained from all subjects before their participation. Subjects reported to the laboratory in a fasted state and had not performed exercise within the past 24 h.

Table 1.

Subject characteristics

Subjects	SCI (n = 10)	CTRL (n = 10)
Age, yr	42 ± 9	41 ± 2
Body mass, kg	74 ± 10	61 ± 5*
Height, m	1.83 ± 0.12	1.72 ± 0.16*
Lower leg muscle volume, cm3	456 ± 78	1435 ± 108*
Forearm muscle volume, cm3	793 ± 24	828 ± 18
Popliteal artery diameter, cm	0.52 ± 0.02	0.55 ± 0.03
Brachial artery diameter, cm	0.50 ± 0.03	0.49 ± 0.02
Carotid artery diameter, cm	0.65 ± 0.03	0.67 ± 0.03
ASIA grade	A	–
Years post injury	16 ± 13	–
Quadriceps spasticity Ashworth scale (0–4)	0	–
Glucose, mg/dL	97 ± 11	62 ± 10
Cholesterol, mg/dL	160 ± 8	184 ± 6*
HDL, mg/dL	43 ± 5	65 ± 2*
LDL, mg/dL	98 ± 5	108 ± 6
Triglycerides, mg/dL	96 ± 15	78 ± 19
Hemoglobin, g/dL	14.6 ± 0.3	14.8 ± 0.8
WBC, K/μL	4.4 ± 0.7	5.1 ± 0.6
Neutrophil, K/μL	2.6 ± 0.3	2.9 ± 0.4
Lymphocyte, K/μL	1.7 ± 0.2	1.6 ± 0.2
Monocyte, K/μL	0.43 ± 0.2	0.41–0.03

Data are presented as means ± SE. All subjects with an SCI had a spinal lesion between T-6 and T12. American Spinal Injury Association (ASIA) (19) score is used to classify the severity of the lesion: a = sensory and motor complete. The quadriceps spasticity Ashworth scale (20) is used to classify the severity of muscular spasms during passive movements in subjects with an SCI: 0 = no spasms. CTRL, controls; HDL, high-density lipoprotein; LDL, low-density lipoprotein; SCI, spinal cord injury; WBC, white blood cells. *Represents values significantly different between groups.

Upper- and Lower-Limb Volume Measurements

Forearm and lower leg circumferences (distal, proximal end, and one-third distal to the proximal end) and length (joint to joint) were measured to calculate limb volume (21). In addition, ventral (forearm) and dorsal (lower leg) skinfold measurements were taken to assess subcutaneous fat and allow the calculation of muscle volume assessed anthropometrically (Vol_anthropo) (22–24), as recently validated, by our group, in people with an SCI (25), Vol_anthropo in the lower legs was corrected, based upon magnetic resonance imaging (MRI) data, using the following equation:

Muscle volume from MRI = 1.077 · Vol_anthropo – 250

Flow-Mediated Dilation

FMD measurements of the BA and PA were performed according to recommended procedures (11). Briefly, subjects reported to the laboratory on the morning of the study in a rested and fasted state, and, after a blood sample, lay supine for ∼20 min before testing. A blood pressure cuff was placed on the upper right arm proximal to the elbow and distal to the placement of the ultrasound Doppler probe on the BA (GE Medical Systems, Milwaukee, WI). After 20 min, the same procedure was performed in the leg, with a blood pressure cuff placed on the right lower-leg proximal to the knee and distal to the placement of the ultrasound Doppler probe on the PA (23). For both BA and PA FMD measurements, baseline vessel diameter and blood velocity were obtained, and the cuff was inflated to a suprasystolic pressure (250 mmHg) for 5 min. Continuous measurements of vessel diameter and blood velocities were made for 30 s before cuff inflation (baseline), 10 s before cuff release, and 2 min following cuff release.

Ultrasound Doppler Brachial, Popliteal, and Carotid Artery Blood Flow Assessments

BA, PA, and carotid artery diameters and blood velocities were measured with ultrasound Doppler using a General Electric (GE) LogiQ-7 ultrasound system (GE Medical Systems, Milwaukee, WI), as previously described (26). BA and PA diameters were then analyzed using commercially available software (Medical Imaging Applications, LLC, Coralville, IA). Using mean blood velocity (V_mean) and arterial diameter, blood flow was calculated as:

Blood flow (mL/min) = V_mean · π (vessel diameter/2)²·60

Shear Rate and Reactive Hyperemia Calculation

Post cuff release shear rate was calculated using the following equation: shear rate (s⁻¹) = 8 V_mean/vessel diameter. Both the cumulative shear rate (s⁻¹ · s) (area under the curve, AUC up to the time of peak FMD) and the RH post cuff release (total blood flow for 2 min) were integrated using the trapezoidal rule and calculated as:

$${\displaystyle \sum \text{[}{y}_i\text{(}{x}_{\text{(}i\text{+1)}}-x_{\text{(}i\text{)}}\text{)+(1/2)(}{y}_{\text{(}i\text{+1)}}-y_{\text{(}i\text{)}}\text{)(}{x}_{\text{(}i\text{+1)}}-x_{\text{(}i\text{)}}\text{)]}}$$

Pulse Wave Velocity

After 20 min of quiet rest in the supine position, duplex-mode ultrasound Doppler measurements were taken at the carotid, radial, and popliteal arteries to assess arterial stiffness of the upper (carotid-radial pulse wave velocity, c-rPWV) and lower (carotid-popliteal pulse wave velocity, c-pPWV) extremities. Pulse wave velocity was calculated using the foot-to-foot ECG gated method, as described previously (27).

Statistical Analysis

A two-way ANOVA with Tukey’s post hoc test was used to determine significant differences between vascular function in the groups and limbs. In addition to the ANOVA of %FMD/shear rate, an analysis of covariance (ANCOVA) was used to covary for shear rate for the analysis of %FMD. Pearson’s correlations were used to assess relationships between variables. Significance was set at an α level of 0.05, and data are presented as means ± SE.

RESULTS

Subject Characteristics

All subjects, 10 SCI and 10 CTRL, successfully completed the experimental protocol without problems or discomfort. In all the subjects with an SCI, a neurological exam confirmed that there was no neural feedback from the legs, while sensation was intact in the arms. By experimental design, there was no difference in PA and BA diameters, within or between groups (Fig. 1). However, partly a consequence of this matching of vessel size, body mass, and height was significantly different between the two groups, with the CTRL being, on average, shorter and lighter. In addition, due to disuse, lower-limb muscle volume in the subjects with an SCI was significantly smaller than the CTRL (Table 1).

Figure 1.

Representative ultrasound images of brachial and popliteal arteries. A and B: representative images of a nondisabled control (CTRL). C and D: representative images of a subject with a spinal cord injury (SCI). Note, by experimental design, artery diameters were not different either between the brachial and popliteal arteries or between subjects with an SCI and CTRL.

Baseline Blood Flow

Baseline blood flow in the PA was lower compared with the BA in subjects with an SCI (PA, 24 ± 13 mL/min; BA, 37 ± 11; P < 0.05). In contrast, baseline blood flow was higher in the PA compared with the BA in the CTRL (PA, 58 ± 13 mL/min; BA, 36 ± 9 mL/min; P < 0.05) (Fig. 2A). Between groups, baseline blood flow was lower in the PA of the subjects with an SCI compared with that of the CTRL (∼138%; P < 0.05), while baseline BA blood flow was not different between groups (∼3%; P = 0.8). Baseline BA and PA blood flow normalized for the muscle volume that is fed by these vessels was not different, between groups, in either the forearm (CTRL: 0.044 ± 0.003, SCI: 0.045 ± 0.004 mL/min/cm³) or the lower leg (CTRL: 0.043 ± 0.01, SCI: 0.047 ± 0.03 mL/min/cm³, P = 0.8) (Fig. 2B).

Figure 2.

Post-cuff reactive hyperemia (RH). A: represents absolute popliteal and brachial artery blood flow at baseline and after cuff release in subjects with a spinal cord injury (SCI) and healthy, nondisabled controls (CTRL). B: represents reactive hyperemia (RH) relative to muscle volume below the cuff. Data are means presented as ± SE; *significant difference between groups in popliteal artery; §significant difference between popliteal and brachial in SCI; †significant difference between SCI popliteal and CTRL brachial; ‡significant difference between popliteal and brachial in CTRL; #significant difference between CTRL popliteal and SCI brachial.

Reactive Hyperemia

Absolute RH (AUC), from cuff release until 120 s post-cuff release, in the subjects with an SCI was significantly attenuated in the PA (144 ± 22 mL) compared with their BA (334 ± 16 mL) as well as the PA (258 ± 16 mL) and BA (356 ± 55 mL) of the CTRL. In addition, the RH AUC of the PA in the CTRL was significantly lower than that of both the BA in the CTRL and SCI (Fig. 2A). These results were the consequence of differences in post-cuff release hyperemia in the PA and BA of the subjects with an SCI and the CTRL. Specifically, blood flow was significantly attenuated in the PA of the subjects with the SCI compared with the PA of the CTRL from the 4th to the 30th s post-cuff release, and this significant difference was maintained for another 30 s when compared with the RH in the BA of both the subjects with an SCI and the CTRL. Blood flow was attenuated in the PA of the CTRL compared with the BA of both subjects with an SCI and the CTRL from the 4th to the 20th s post-cuff release (Fig. 2A). In contrast, BA post-cuff RH was not different over time in the subjects with an SCI and the CTRL (Fig. 2A). Here, it should be noted that unlike the arm, where muscle mass was preserved, the lower legs of the subjects with an SCI were significantly smaller than the CTRL (Table 1), with RH being heavily dependent on the volume of muscle occluded. Indeed, there was evidence of a strong relationship between RH and muscle volume in the lower leg (r = 0.77), but not in the forearm (r = 0.01), helping to explain the attenuated RH in the PA of the subjects with an SCI compared with CTRL and the lack of a difference in the BA (Fig. 2A). Consequently, when post-cuff RH was normalized per cm³ of muscle volume, there was now no difference between the groups in terms of RH in either the PA (CTRL: 0.12 ± 0.2, SCI: 0.13 ± 0.2 mL/cm³) or the BA (CTRL: 0.67 ± 0.1, SCI: 0.57 ± 0.9 mL/cm³). Interestingly, despite normalizing for muscle volume, post-cuff release RH in the PA, of both groups, was significantly attenuated compared with the BA (Fig. 2B).

Flow-Mediated Dilation

FMD, expressed as relative change in vessel diameter (%FMD), was significantly attenuated in the PA of subjects with an SCI compared with their BA, as well as both the PA and BA of the CTRL (Fig. 3A). It is important to note that this result was not affected by physiological and mathematical bias, in favor of the smaller arteries, because, by experimental design, the vessels (BA and PA in both SCI and CTRL) were not different in terms of baseline diameter. However, post-cuff shear rate (AUC) was greater in the BA of both the subjects with an SCI and the CTRL compared with the PA (P < 0.05) (Fig. 3B). There was no significant difference between the groups in terms of BA shear rate (CTRL = 10 ± 0.7 · s⁻¹; SCI = 9 ± 0.3 · s⁻¹; P = 0.8), but shear rate in the PA was significantly attenuated in the subjects with an SCI compared with the CTRL (CTRL = 7 ± 0.3 · s⁻¹; SCI = 4 ± 0.3 · s⁻¹; P < 0.05, Fig. 3B). Again, it should be noted that the lower legs of the subjects with an SCI were significantly smaller than the CTRL (Table 1), resulting in an attenuated RH in the PA of this group (Fig. 2A). In addition, RH in the BA and PA correlated significantly (P < 0.05) with BA and PA shear rate (r = 0.78 and r = 0.94, respectively). Furthermore, as with RH, there was evidence of a strong relationship between shear rate and muscle volume in the lower leg (r = 0.81, P < 0.05), but not in forearm (r = 0.05), resulting in an attenuated shear rate in the PA of the subjects with an SCI compared with the CTRL (P < 0.05) and a lack of a difference in the BA (Fig. 3B). Consequently, when %FMD was normalized for the significantly diminished shear rate in the PA of the subjects with an SCI, %FMD/shear was actually significantly greater in the SCI (1.6 ± 0.1% ·s⁻¹) than the CTRL (1.2 ± 0.1% · s⁻¹) and significantly greater than their own BA (0.9 ± 0.1% · s⁻¹) and that of the CTRL (0.9 ± 0.1% · s⁻¹, Fig. 3C). The ANCOVA for %FMD, with shear rate as the covariate, did not support the conclusion of augmented vascular responsiveness in the PA of subjects with an SCI but, also, did not identify any dysfunction compared with the controls. However, it should be note that this more complex statistical analysis was greatly underpowered (β = 0.2).

Figure 3.

Flow-mediated dilation (FMD). A: represents percent change in FMD of popliteal and brachial arteries in subjects with a spinal cord injury (SCI) and healthy, nondisabled controls (CTRL). B: represents shear rate area under the curve for popliteal and brachial arteries in subjects with an SCI and (CTRL). C: represents FMD normalized by shear rate in popliteal and brachial arteries in subjects with an SCI and (CTRL). Individual data, the range, and the median values are presented as darkened circles and whisker plots, respectively. *Significant difference between groups in popliteal artery; §significant difference between popliteal and brachial in SCI; †significant difference between SCI popliteal and CTRL brachial; ‡significant difference between popliteal and brachial in CTRL; #significant difference between CTRL popliteal and SCI brachial.

Arterial Stiffness

For both the subjects with an SCI and the CTRL, c-r PWV was significantly faster than c-p PWV (Fig. 4). However, neither c-r PWV (CTRL = 4.3 ± 0.3 m · s⁻¹; SCI = 4.2 ± 0.3 m · s⁻¹; P = 0.7) nor c-p PWV (CTRL = 7.1 ± 0.3 m · s⁻¹; SCI = 6.9 ± 0.5 m · s⁻¹; P = 0.6) was different between groups (Fig. 4).

Figure 4.

Pulse wave velocity (PWV). PWV for the carotid-radial (c-r) and the carotid-popliteal (c-p) in subjects with a spinal cord injury (SCI) and healthy, nondisabled controls (CTRL). Individual data, the range, and the median values are presented as darkened circles and whisker plots, respectively. §Significant difference between c-p and c-r in SCI; †significant difference between SCI c-p and CTRL c-r; ‡significant difference between c-p and c-r in CTRL; #significant difference between CTRL c-p and SCI c-r.

DISCUSSION

This study examined the effect of a spinal cord injury and subsequent chronic limb disuse on vascular function, assessed by FMD, in diameter-matched vessels above (BA) and below (PA) the lesion in subjects with SCI-induced paraplegia and in nondisabled CTRL. In agreement with our first hypothesis, an SCI does not appear to have a systemic effect on vascular function. This is based on the lack of a difference in %FMD in the upper limbs of the subjects with an SCI and the CTRL.

Also, in agreement with our second hypothesis, %FMD in the PA was significantly attenuated in the subjects with an SCI compared with the CTRL. However, below the lesion there was, actually, evidence of increased vascular responsiveness in this population. Specifically, although %FMD in the PA was significantly attenuated in the subjects with an SCI compared with the CTRL, this can be explained by the significantly attenuated RH in the leg of the subjects with an SCI. Of note, FMD and RH in the legs were highly correlated, and when RH was normalized for muscle volume, there was no difference in RH between the groups. In fact, with this attenuated shear stimulus to dilate taken into account, subjects with an SCI exhibited a significantly greater PA %FMD/shear rate compared with the CTRL. In combination, these novel findings, unbiased by differences in baseline vessel diameter, reveal evidence of vascular dysfunction only below the lesion of people with an SCI, which, with the attenuated shear stimulus taken into account, actually indicates augmented vascular responsiveness compared with CTRL. Thus, it is concluded that although vascular function below the SCI lesion can present as vascular dysfunction (%FMD), partially, but not only, driven by the atrophied muscle volume occluded during the leg FMD and the, subsequently, attenuated RH and shear rate, there is some evidence of augmented vascular responsiveness (%FMD/shear rate) below the lesion in this population.

Reactive Hyperemia in Active and Disused Limbs

To better understand the peripheral and systemic effects of an SCI on vascular responsiveness, we assessed RH responses following 5 min of suprasystolic cuff occlusion in the arm and the leg of subjects with an SCI and CTRL. The absolute RH response in the PA was significantly lower (∼75%) in the legs of subjects with an SCI compared with the CTRL (Fig. 2A), which is in agreement with previous studies (14, 28). However, our previous work in this population (29) and that of others (28, 30) has underlined the importance of muscle volume in terms of the peripheral vascular adaptations and vasoreactivity subsequent to an SCI. In addition, both animal and human models (31, 32) have revealed that after denervation and subsequent muscle atrophy, the correlation between capillary length-to-fiber ratio and muscle fiber radius is preserved. Thus, vascular atrophy after an SCI appears to be closely linked to muscle atrophy.

Therefore, as reported in a previous study by our group (23), involving populations that exhibited differing volumes of limb muscle, RH response should be expressed relative to the volume of muscle occluded by the cuff. Indeed, this was, again, evident in the current dataset, with a strong positive relationship evident between the widely varying lower leg muscle volume and RH (r = 0.77), which was not apparent in the more uniformly sized arm (r = 0.01). Importantly, taking into account the atrophic leg muscle volume of the subjects with an SCI, the RH response in the PA of these subjects was no longer different from the CTRL (Fig. 2B). This normal RH response, relative to muscle mass, is an interesting observation, as RH is commonly recognized as an index of microvascular function (33). Thus, this lack of a change in muscle mass-specific RH following an SCI appears to be, at least in part, a consequence of limb atrophy, and likely vascular rarefaction, with no evidence of functional vascular maladaptation. Indeed, the spinal cord injury itself, the associated deficits, and secondary complications, coupled with this traumatic event, could also contribute to the aforementioned changes in RH (34).

Interestingly, several previous studies have reported an increased absolute RH response in upper limbs of subjects with an SCI compared with control subjects, likely related to the greater levels of arm physical activity required in this population (13, 14, 28). However, as already recognized, RH should be expressed relative to the muscle mass occluded by the cuff, otherwise differences in the volume of arm muscle, not reported in the previous investigations, may confound such comparisons. Therefore, in contrast to previous work, in the present study, where the volume of muscle in the forearm, distal to the cuff, was measured, and was not different between the groups, both absolute and relative RH in the arm of the subjects with an SCI were not different from the CTRL (Fig. 2, A and B). This indicates an SCI does not appear to negatively impact the microvascular function in the functional limbs in this population. Finally, as the subjects with an SCI were likely more active with their arms, based on time engaged in arm-focused activity, these observations suggest that perhaps muscle mass, rather than the level of physical activity, is an important factor in determining RH (23).

FMD in Active and Disused Limbs

The interpretation of vascular function studies in people with an SCI has been difficult because the %FMD data were influenced by different baseline vessel diameters. In fact, several previous studies have highlighted the inherent inverse relationship between vessel size and FMD (10, 35). Therefore, the FMD response should be anticipated to be higher in subjects with an SCI, who have smaller vessel diameters than CTRL. In the current study, to bypass this confounding arithmetical factor, by experimental design, subjects with an SCI and CTRL were recruited who exhibited similar vessel diameters for both the PA and BA (0.48–0.56 cm range). Consequently, the subsequent FMD results can be more simply interpreted as an indicator of vascular function both within subjects and between subjects with an SCI and the CTRL.

Here, again, it should be noted that, unlike the arm, where muscle mass was preserved, the lower legs of the subjects with an SCI were significantly smaller than the CTRL (Table 1), resulting in an attenuated RH, which, with all vessels matched for diameter, is the major determinant of shear rate (BA and PA RH were significantly and strongly, positively correlated with BA and PA shear rate r = 0.78 and r = 0.94, respectively). Although, somewhat, likely aided by the marked difference in lower leg muscle volume between the subjects with an SCI and the CTRL, which was not the case in the arm, there was evidence of a strong and significant, positive relationship between shear rate and muscle volume in the lower leg (r = 0.81) but not in the forearm (r = 0.05). These relationships, and lack thereof, help to explain the attenuated shear rate in the PA of the subjects with an SCI compared with CTRL and that there was no such difference, between groups, in the BA (Fig. 3B). Thus, in agreement with the recognized stimulus-response pattern of shear stress and vasodilation, the lower shear exhibited in the PA by the subjects with an SCI elicited a smaller FMD response (Fig. 3, A and B). However, when this muscle mass-dependent attenuation in the shear stimulus was taken into account, by normalizing for shear rate, PA FMD was no longer attenuated compared with the CTRL (Fig. 3C). In fact, without bias from differences in baseline vessel diameter, this study supports the concept of vascular hyperactivity in the disused leg. This is in agreement with our previous work with the passive leg movement (PLM) assessment of vascular responsiveness (29) and other previous studies utilizing the FMD test (13, 14, 36–38). Together, these results indicate a preserved, or even augmented, vascular responsiveness in the inactive legs of subjects with an SCI, who are otherwise healthy, compared with nondisabled controls (Table 1). It should be noted that the ANCOVA for %FMD, with shear rate as the covariate, did not support the conclusion of augmented vascular responsiveness in the PA of the subjects with an SCI but, also, did not identify any dysfunction compared with the controls. However, this more complex statistical analysis was greatly underpowered (β = 0.2), even when using a large effect size for the power calculation, as exhibited by the mathematical normalization for the shear rate approach (Cohen’s d = 0.95, Fig. 3C).

In terms of the upper limbs, not directly affected by the SCI, this study documents that FMD was similar between subjects with an SCI and the CTRL. In fact, the BA of SCI group was not different from the CTRL in terms of hyperemic stimulus, %FMD, and, therefore, %FMD/shear rate. These findings support the concept that there is no SCI-related systemic effect that alters vascular function of the arteries in the functional limbs above the lesion. Furthermore, as with other studies in patients with an SCI (38), in the current study, there was no evidence of improved vascular function in the arms of subjects with an SCI, who tend to use their arms more than their nondisabled counterparts during exercise, and, likely, during activities of daily living.

PWV in the Active and Disused Limbs

PWV is a relatively simple, noninvasive assessment of arterial stiffness, the development of which is a significant risk factor for cardiovascular disease. Indeed, a decrease in arterial compliance reduces a vessel’s buffering capacity, which is associated with an increased pulse pressure, elevated left ventricular wall tension, and augmented work of the heart. In a previous investigation, Miyatani et al. (39) utilized PWV to assess arterial stiffness in the arms and legs of subjects with an SCI and CTRL and found no difference in PWV in the functional and disused limbs or between groups (∼1,000 cm/s). In agreement with this previous investigation, the current results reveal that arterial stiffness was not systemically influenced by an SCI (as assessed in the arm) or locally influenced by the lack of use (as assessed in the leg). In fact, c-p PWV and c-r PWV were remarkably similar between the SCI group and CTRL (Fig. 4). Although, in a previous study, PWV below the spinal cord lesion was measured in a distal artery (posterior tibial artery) (39), the assessment of carotid-popliteal PWV, in the current study, still provided a robust measure of vascular stiffness. However, future studies might explore the PWV of more distal sites in SCI. In agreement with the other vascular assessments performed in this investigation, the current data document that arterial stiffness is not an obligatory cardiovascular risk factor in subjects with an SCI.

Vascular Adaptations, Disease, and Assessment Following an SCI

There are structural adaptations of vessels below the spinal cord lesion, including a reduction in diameter and an increase in intima-media thickness, and these are associated with a reduced ankle-brachial index (18), indicative of a relationship between this arterial remodeling, potentially vascular function, and the prevalence of cardiovascular disease in the SCI population. Therefore, clinically it would seem reasonable to use some form of vascular responsiveness assessment in subjects with an SCI to monitor this cardiovascular disease risk factor. However, the current findings in terms of vascular responsiveness, our previous work using the PLM model (29), and the majority of other studies in this area (13, 14, 36–38), using an array of differing approaches, document a preservation or even an augmentation of vascular responsiveness in the paralyzed limbs of subjects with an SCI (13, 14, 36–38). The current findings encapsulate this conundrum in a single study, with evidence of diminished vascular function below the lesion, which agrees with an increased risk of CVD in people with an SCI, but greater vascular responsiveness in this location when the diminished shear stimulus is take into account. Furthermore, adding to this conundrum, systemic vascular function/responsiveness, above the lesion, not affected by disuse, was unaltered by the SCI. Clearly, additional studies are warranted that continue to unravel the link between CVD and SCI and how best to assess vascular health in this population. Finally, in terms of vascular adaptation, it is interesting to note that despite the vast differences in leg muscle mass between the subjects with an SCI and the CTRL, popliteal artery diameters were able to be matched between groups. This clear disconnect between vessel diameter and both the distal muscle mass and the metabolic capacity of the muscle also warrants further investigation.

Experimental Considerations

Although not a focus of this study, which explains the, somewhat, rudimentary assessment (self-reported hours of exercise/week), the subjects with an SCI appeared to be more physically active than the CTRL (6.5 ± 3.5 vs. 2.0 ± 1.5 h, respectively). However, due to the contrasting modalities of exercise between the two groups and no additional information regarding the intensity of the physical activity, it is difficult to assign any of the current findings to this observation. As already recognized, perhaps the only, potentially, meaningful inference that can be made is with regard to arm exercise, the most likely form of exercise to be performed by the subjects with an SCI. Specifically, it is recognized that physical activity can translate into improved peripheral vascular responsiveness, but the current data, that are in line with previous publications (38), suggest that the vascular responsiveness of the BA is not different between subjects with SCI and CTRL and, consequently, not associated with the level of physical activity executed with the arms in people with an SCI. This potential disconnect is deserving of additional investigation.

This study was novel in that all vessels examined were of a similar diameter, removing the historically confounding fact that people with an SCI have smaller conduit vessels below the lesion, compared with nondisabled controls. However, the shear stimulus was not controlled and, therefore, varied dependent, somewhat, on the volume of muscle mass occluded. There are several potential approaches to actively control the shear stimulus; however, each has limitations and complications that may alter the FMD response. For example, the length of time the distal cuff is inflated can be varied, but the length of occlusion time can vary the dependency of FMD on nitric oxide (NO) (40). Warming of the extremities, in addition to or instead of distal cuffing, alters the response to the shear stimulus (40, 41). Proximal occlusion to limit the hyperemic response, upon distal cuff release, results in dilation that is primarily not NO mediated (42), but in addition, limiting venous outflow results in congestion within the muscle bed and alters blood flow patterns and, therefore, stimulus in the artery (43). For these reasons, the current study opted to use a standard 5-min distal cuff occlusion to stimulate hyperemia and then normalize the %FMD for shear rate.

Interestingly, it would appear that the knowledge gained from the vascular adaptations associated with SCI-induced limb disuse cannot be universally applied to other clinical populations impacted by diminished mobility. For instance, in a recent investigation, we documented that both local circulation and vascular responsiveness, normalized for limb muscle volume, are actually severely impacted by long-term bed rest (44). It will be interesting to determine whether potential countermeasures to offset disuse-mediated changes in vascular responsiveness, such as passive stretching of the immobile limbs, the efficacy of which we recently demonstrated in healthy individuals (45), might temper the deleterious effects of inactivity across clinical populations, including SCI.

GRANTS

This study was partially supported by the National Heart, Lung, and Blood Institute at the National Institute of Health (PO1 HL1091830) and the Veterans Administration Rehabilitation Research and Development Service (E6910-R, E1697-R, E3207-R, E9275-L, and E1572-P).

DISCLOSURES

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

AUTHOR CONTRIBUTIONS

M.V., J.D.T., and R.S.R. conceived and designed research; M.V., J.D.T., and S.J.I. performed experiments; M.V. and S.J.I. analyzed data; M.V., M.A., and S.J.I. interpreted results of experiments; M.V. prepared figures; M.V. drafted manuscript; M.V., M.A., J.D.T., S.J.I., and R.S.R. edited and revised manuscript; M.V., M.A., J.D.T., S.J.I., and R.S.R. approved final version of manuscript.