Short-term unilateral leg immobilization alters peripheral but not central arterial structure and function in healthy young humans
Short-term leg immobilization is an acute model of inactivity, which induces vascular deconditioning. The present study was conducted to determine if short-term leg immobilization induced alterations in central and peripheral conduit artery structure (diameter and compliance), function (resting blood flow and mean wall shear rate), and peripheral flow-mediated dilation. Healthy participants (n = 7 women and n = 8 men) were studied before and after 12 days of unilateral leg immobilization. Carotid artery structure and function were unaltered with immobilization indicating that the unilateral immobilization did not have a detectable effect on this representative central artery. In contrast, peripheral measures of arterial structure at the common femoral and popliteal arteries showed significant reductions in both the immobilized and non-immobilized limbs but to a greater extent in the immobilized limbs. Specifically, femoral and popliteal artery compliance and femoral artery diameter were reduced in both the immobilized and the non-immobilized limb (p < 0.05) while popliteal artery diameter was reduced only in the immobilized leg. Popliteal artery flow-mediated dilation, an indicator of peripheral artery function, was increased in the immobilized limb, which parallels reports in paralyzed limbs of spinal-cord-injured individuals. The time course of vascular alterations with inactivity likely follows a sequence of adaptations in arterial structure and function reflecting differing initial flow patterns, and arterial wall composition, and diverse hemodynamic stimuli within different blood vessels.
KeywordsBlood flow Endothelial function Deconditioning
Many studies of induced inactivity (i.e., bed rest) or local inactivity/unloading (i.e., immobilization) leading to ‘deconditioning’ have focused on skeletal muscle adaptations (Dudley et al. 1992; Urso et al. 2006; Yasuda et al. 2005). The confounding effects of microgravity (Bleeker et al. 2005a, b), spaceflight (Norsk et al. 2006) or traumatic injury (de Groot et al. 2003) have received limited previous mechanistic research examining the potential vascular adaptations associated with inactivity-induced deconditioning. Studies evaluating peripheral vascular adaptations to immobilization have shown both unaltered (Bleeker et al. 2005a, b), and decreased resting blood flow coinciding with decreased peripheral artery cross-sectional area (Bleeker et al. 2005a; de Groot et al. 2003; Norsk et al. 2006). Both decreases (de Groot et al. 2003) and maintenance (Olive et al. 2003), in resting blood flow have also been observed in long-term models of muscular inactivity, such as spinal cord injury.
Resting blood flow is influenced by both blood vessel structure and function. Unilateral lower limb suspension does not appear to alter vessel responsiveness to nitric oxide (Bleeker et al. 2005a, b, 2006); however, a trend to enhanced vasodilation in response to a flow-mediated shear stimulus was observed (Bleeker et al. 2005a). Arterial compliance, an index of arterial structure, is reduced in immobilized upper limbs as a result of fracture (Giannattasio et al. 1998); however, there may be a confounding effect of tissue damage and repair in this model. In a non-injury model (Sugawara et al. 2004), femoral artery distension, an index of arterial distensibility, tended to decrease in the immobilized leg after 7 days. Examinations of the vascular effects of immobilization have not included measurements of both central and peripheral vascular changes in combination with direct assessment of muscle atrophy in an effort to assess whether the observed effects are systemic or local and if they are related to muscle morphological changes.
The aim of the present study was to evaluate the effects of unilateral lower leg immobilization on structural and functional properties of both central and peripheral conduit arteries. A full characterization of vascular adaptations requires examinations of both central and peripheral blood vessels and determination of local versus systemic regulatory mechanisms. We hypothesized that 12 days of unilateral immobilization would induce reductions in peripheral artery compliance and diameter and increases in flow-mediated dilation (FMD) in the immobilized limb, but not in the non-immobilized limb, due to reductions in vascular shear stress in the immobilized limb. Further, we hypothesized that there would be no change in the vascular properties of the carotid (i.e., central) artery since unilateral limb immobilization is a model of local inactivity (Tanaka et al. 2000).
Materials and methods
Healthy participants (8 males, 7 females; males: 20.8 ± 0.7 years, females: 20.4 ± 0.8 years) were recruited for the study, which was approved by the Hamilton Health Sciences Research Ethics Board in accordance with the revised 1983 Declaration of Helsinki for research involving human subjects. All subjects were screened using a medical history questionnaire and none exhibited any medical problems that would interfere with their participation in this study. Exclusion criteria were tobacco smoking; a recent fracture of the limb, cardiovascular disease, diabetes, and certain medications (i.e. oral contraceptives) known to interfere with safety of the subjects and of the measures being examined. None of the subjects were endurance-trained and subjects were excluded if they exercised more than 5 h per week. Informed written consent was obtained from each participant if they met the inclusion criteria.
The immobilization protocol has been previously described by Yasuda et al. (2005). Participants had one leg randomly assigned and immobilized using a standard knee brace (epX Knee Control Plus, Smith Orthopedics, Topeka, KS). Subjects were permitted to remove the brace once daily under the supervision of the research staff to allow daily sanitation to take place at the McMaster Exercise Metabolism Research Laboratory. The knee brace itself restricted knee extension and flexion, resulting in a complete removal of weight bearing for the immobilized limb, and the immobilized knee joint was maintained at 130 degrees of extension at all times. Ambulation and complete unloading were facilitated with the use of crutches throughout the protocol. Subjects completed a series of cardiovascular measurements prior to and following 12 days of immobilization while measurements of quadriceps cross-sectional area of the non-immobilized limb only were made following 14 days of immobilization. Thigh lean cross-sectional area was determined using Magnetic Resonance Imaging (MRI). A 1.5T scanner (Symphony Quantum, Siemens, Erlangen, Germany) obtained a cross-section of the thigh at 70% of distal distance between the greater trochanter and the lateral joint space. CSA of the quadriceps femoris was determined from these images using manual planemetry image analysis (Image Pro Plus V4.0, Media Cybernetics, Silver Springs, MD).
Identical cardiovascular testing procedures were completed at PRE and POST. Before each testing session, participants were familiarized with the testing procedures. Participants were instructed to abide by several control measures at all testing sessions. These control measures included abstinence from alcohol, caffeine, and exercise during the 24-h preceding testing. All testing was conducted 4 h postprandial following the consumption of identical meals. Time of day was also within a 2-h window at all time-points. Vascular measurements were conducted in the same temperature-controlled (22–24°C), dimly lit environment in the supine position. A 10-min instrumentation period was followed by 10 min of rest after which resting blood pressure measurements were obtained using an automated oscillometric device in triplicate (model CBM-7000, Colin Medical Instruments, San Antonio, USA).
Measurements of heart rate were obtained via 2 sets of 3 electrode electrocardiograph (ECG). The ECG signal from one set was obtained using a Cardiomatic device (Model MSC 71233, Medical Systems Corp., Miami, FL,USA) which subsequently output this signal to a data acquisition board (model ML795, ADInstruments, Colorado Springs, USA) interfaced with Chart 5.0 software. The second set of ECG electrodes was used for gating of the Doppler ultrasound (Model System Five, GE Medical Systems, Horten, Norway) and vessel images with the heart cycle.
Resting arterial diameters
B-mode Ultrasound imaging was used to determine minimum, maximum and calculated mean (1/3max + 2/3min) arterial diameter measurements at the common carotid, and right and left common femoral, and popliteal arteries. The sonographer had access to the PRE immobilization images to ensure identical positioning at POST using anatomical landmarks. Landmarking consisted of either localization relative to a bifurcation or collateral circulation. The carotid artery images were consistently acquired ~2 cm proximal to the carotid bulb, while common femoral images were acquired ~2 cm proximal to the bifurcation dividing this artery into the superficial and profundus branches. Images acquired at the popliteal artery were consistently acquired at 2–3 cm proximal to the branching of the tibial artery. Three full heart cycle digital video clips were acquired of each vessel at each time-point and an average was determined from these clips. Resting arterial diameters of the popliteal artery were used in the calculation of normalized and absolute vessel dilation responses to the flow-mediated dilation protocol.
Central arterial compliance measurements were determined using the method described by Rakobowchuk et al. (2005). Briefly, a handheld pen-like tonometer (model SPT-301, Millar Instruments Inc., Texas, USA) was positioned over the point of greatest pulsation of the right carotid artery. From this tonometer, continuous measurements of carotid blood pressure were obtained. Due to the effects of manually applied hold-down pressure, the carotid waveform was equated to simultaneous, continuous mean and diastolic blood pressure obtained at the radial artery, since diastolic and mean arterial pressures are essentially identical in the supine position (Kelly et al. 1989). This device equates the radial tonometrically acquired pressure waveform to an automated brachial blood pressure measurement made with the internal brachial cuff system. This procedure is fully completed prior to the measurement of pressure at the carotid artery; therefore, this radial derived blood pressure waveform is essentially identical to the one that exists in the brachial artery allowing beat-specific measurements and subsequent equating of blood pressure waveforms in the carotid artery.
B-mode Ultrasound imaging using a 10-MHz linear array probe of the opposite carotid artery was acquired simultaneous to blood pressure measurements. The synchronization of blood pressure and video clips was ensured via a triggered square-wave pulse indicating image acquisition time in the Powerlab Chart 5 software. Similar procedures were completed for both the popliteal and common femoral arteries; however, blood pressure was obtained from the radial artery measurement already described. This method was used because tonometrically obtained blood pressures at both the femoral and popliteal arteries are likely confounded by the inability to depress these arteries against solid structures.
∆CSA is the difference between systolic and diastolic cross-sectional area of the artery derived from diameter measurements, ∆P is the difference between systolic and diastolic pressure.
Mean wall shear rate
Popliteal flow mediated dilation
The popliteal artery was visualized from the posterior portion of the knee while the subject was supine, and the leg was elevated 25 cm with a fixed positioning device, both at rest and following 9 min of ischemia (de Groot et al. 2005, 2004). The occlusion cuff was placed distal to the knee surrounding the maximum diameter of the leg. Images were acquired at 30, 60, 90, 120, and 180 s following ischemia. Upon evaluation of these images using the described automated edge-detection software, the largest diameter determined from the post ischemia images acquired at end diastole were used to calculate absolute and normalized flow-mediated dilation.
Repeatability of measurements
We expected the design of the current study to permit an internal control leg for assessment of day-to-day reproducibility. However, several vessel parameters were altered in both the immobilized and non-immobilized limbs making this impossible. We have, however, analyzed the test re-test reliability in 20 individuals using the same techniques described in the current manuscript. The repeatability of the measurements was determined in young healthy (10 females and 10 males) participants by repeating all measurements 5–7 days apart. Carotid diameter and compliance showed coefficients of variation of 2 and 8%, respectively. Popliteal diameter and distensibility also showed coefficients of variation of 2 and 18%, respectively. The relative popliteal FMD coefficient of variation was 28% in 16 participants who were assessed (Rakobowchuk et al. 2008).
Data are presented as means ± SD and were analyzed by two-way repeated measures ANOVA (condition by time) for measures of the peripheral arteries while central measures at the carotid artery, heart rate, blood pressure, and quadriceps CSA were analyzed using a one-way repeated measures ANOVA. Significant differences between experimental conditions (immobilized leg and non-immobilized leg), or time points expressed as PRE and POST, were further analyzed by Tukey honestly significant post hoc analysis when appropriate. A p value of <0.05 was used to determine statistical significance.
Heart rate and blood pressure
Resting heart rate (PRE 60 ± 10, POST 64 ± 9, p = 0.37) and systolic blood pressure (PRE 121 ± 10, POST 121 ± 10, mmHg p = 0.40) were similar at all testing time-points.
Effectiveness of immobilization protocol
Quadriceps femoris CSA decreased significantly following immobilization (PRE 68.4 ± 15.8, POST 65.7 ± 15.9 cm2, p < 0.01). The mean reduction of quadriceps CSA with immobilization was 2.75 ± 5.68 cm2.
Resting arterial diameter
Carotid arterial compliance was unaltered (p = 0.34) with immobilization (Fig. 1f), whereas peripheral arterial compliance was affected by the intervention. Specifically, common femoral artery compliance was reduced in both legs (immobilized and non-immobilized) over time (Fig. 1d). Although the reduction in compliance appeared greater in the immobilized leg, the interaction did not reach statistical significance (p = 0.07). Popliteal artery compliance was also reduced over time in both legs (p < 0.001) (immobilized and non-immobilized legs of −48.0 and −23.9%, respectively) and although the immobilized limb reduced to a greater extent, this was not significantly different (limb × time interaction p = 0.44; Fig. 1e). Popliteal artery compliance was reduced in the immobilized legs at all time points (p < 0.05).
Resting mean blood velocity and blood flow
Resting mean wall shear rate
Relative flow-mediated dilation showed an interaction (p < 0.02), which was a result of a significant increase of FMD in the immobilized leg (PRE 6.4 ± 1.3%, POST 13.5 ± 2.7%) while the non-immobilized leg (PRE 6.3 ± 1.4%, POST 8.9 ± 1.7%) did not increase significantly. When normalized to the delta peak shear rate, the immobilized leg (PRE 0.017 ± 0.007% s, POST 0.032 ± 0.008% s) still showed a significant (p = 0.01) increase in dilation while the non-immobilized leg did not change (PRE 0.016 ± 0.003% s−1, POST 0.022 ± 0.004% s−1).
Local inactivity, in the absence of acute injury, caused changes to the vasculature with peripheral artery compliance impacted to the largest degree in the immobilized limb, followed by the non-immobilized limb, while no changes were observed in a large central elastic vessel (carotid artery). This is the first instance where immobilization has been shown to influence both the structure and function of conduit vessels in the inactive limb and to a lesser degree in the non-immobilized control limb. Apart from structural vessel changes, FMD was increased in the immobilized lower limb while it remained unaltered in the non-immobilized lower limb. The fact that no changes were observed in the carotid artery, likely indicates that more substantial, rather than local limb, decreases in activity are required to alter properties of large central elastic arteries. We observed an increase in femoral and popliteal mean blood velocity and vessel shear rate in the immobilized leg, which was accompanied by reductions of mean arterial diameter and enhanced arterial reactivity in response to shear. This is the first instance where this immobilization-induced enhancement of endothelial-dependent dilation has been observed when normalized to shear. In previous studies, this observation was not statistically significant (Bleeker et al. 2005a) and was not compared with the weight-bearing limb. Combined, these alterations likely ensure maintenance of conduit artery blood flow, which did not change with short-term immobilization and show that the effect of immobilization appears to have the greatest impact in the peripheral vessels.
The strength of the current study lies in the combined assessment of both structural and functional vessel attributes measured bilaterally and in the central arterial system, which emphasize the progressive, and non-uniform effects of immobilization on arterial properties. This is the first instance where this comprehensive approach has been undertaken as previous studies either focused on arterial structure (Sugawara et al. 2004) or arterial function (flow mediated dilation or hyperaemic blood flow) (Bleeker et al. 2005a).This model of immobilization effectively reduced the cross-sectional area of the quadriceps in the immobilized leg (Yasuda et al. 2005). Cross-sectional fiber area, and fiber type changes, for this study have been previously reported (Yasuda et al. 2005). These data showed reductions of fiber cross-sectional in the immobilized leg while the non-immobilized leg showed no significant changes. However, the lack of assessment of cross-sectional area of the quadriceps in the non-immobilized limb in the current study makes it impossible to normalize peripheral artery changes to muscle cross-sectional area.
The reduction in peripheral arterial compliance noted both in the femoral and popliteal arteries mirrored previous limb fracture and repair studies (Giannattasio et al. 1998) and this was rapid (within 12 days). A similar time-course for changes in the immobilized limb was suggested by trends in previous leg immobilization studies (Sugawara et al. 2004); however, the current study is the first to also document arterial adaptations in the contralateral limb.
Mechanisms of altered vascular parameters
Carotid artery compliance did not show significant reductions with immobilization. Given that carotid blood flow was not altered, greater or prolonged reductions in global physical activity were likely needed to alter central or carotid parameters. As described, our findings suggest vascular changes with decreases in activity are not simply the opposite of those associated with increases in activity (Jasperse and Laughlin 2006). The non-uniform adaptations are interesting since increases in activity (whole body aerobic exercise) have been shown to improve central arterial compliance in the absence of peripheral arterial compliance changes (Dart and Kingwell 2001; Tanaka et al. 2000). This dissimilarity of effect location, combined with the enhanced endothelial function observed with immobilization (Bleeker et al. 2005a, b) and spinal cord injury (de Groot et al. 2003, 2004), suggests that the effects of decreased activity are not simply the reverse of increased activity.
As outlined earlier, femoral and popliteal mean arterial diameters in the immobilized leg were reduced to a greater extent than their non-immobilized counterparts. This may be a structural reorganization although altered sympathetic nervous activity and basal NO release may also have been altered and cannot be ruled out in this study. Previous studies have noted reductions of arterial diameter with unilateral leg suspension, which were apparent by 7 (Sugawara et al. 2004) and 14 days (Bleeker et al. 2005a), respectively. Studies monitoring artery diameter with exercise training have noted increases that are likely mediated by exposure to augmented shear stress (Dinenno et al. 2001; Miyachi et al. 1998). Conversely, the current study likely reduced basal shear and NO production leading to inward remodeling of the vessel, while the sensitivity of the vessel to NO may have been enhanced (Bleeker et al. 2005a). Enhanced NO sensitivity likely ensures the vessel can dilate and normalize shear stresses when high conduit blood flows are required (Bleeker et al. 2005a).
The FMD protocol used in the current study involved a 9-min occlusion period and an attempt to normalize the resultant popliteal artery dilation to the peak shear stimulus evoked by the cuff release. This 9-min duration of occlusion was based on previous studies involving similar measures (de Groot et al. 2005, 2004) thus facilitating comparisons, and recent work indicates that the duration of the occlusion can alter the mechanisms responsible for stimulating dilation. The most appropriate method for the normalization of FMD data is an area of debate requiring further clarification (Harris et al. 2010, Atkinson et al. 2009). Further, our randomization procedure may have resulted in a high number of dominant limbs being immobilized. This may explain the high compliance in the immobilized limb throughout the study at the popliteal artery. However, this difference persisted throughout the intervention suggesting it was a true anatomical difference in this population.
Implications of this study
In summary, we have found alterations of arterial structure in both femoral and popliteal arteries in both limbs following unilateral limb immobilization, whereas the carotid artery was unaltered. These changes were not simply the opposite to those of exercise training suggesting the mechanisms responsible for these changes with immobilization are not identical to those that accompany increased activity. Further studies into the mechanisms responsible for these alterations are required.
The main source of funding for the collection of data was from grant support in aid of research from Proctor and Gamble to MAT. We would like to acknowledge a discovery grant from NSERC to M.J MacDonald in aiding in the completion of this research. M. Rakobowchuk and E. Glover were the recipients of Canadian Institutes for Health Research Canada Graduate Scholarships and acknowledge this funding.
Conflicts of interest
The authors do not have any conflict of interests.
- Atkinson G, Batterham AM, Black MA, Cable NT, Hopkins ND, Dawson EA, Thijssen DH, Jones H, Tinken TM, Green DJ (2009) Is the ratio of flow-mediated dilation and shear rate a statistically sound approach to normalization in cross-sectional studies on endothelial function? J Appl Physiol 107:1893–1899CrossRefPubMedGoogle Scholar
- Rakobowchuk M, Tanguay S, Burgomaster KA, Howarth KR, Gibala MJ, Macdonald MJ (2008) Sprint interval and traditional endurance training induce similar improvements in peripheral arterial stiffness and flow mediated dilation in healthy humans. Am J Physiol Regul Integr Comp Physiol 295:R236–R242PubMedGoogle Scholar