Abstract
Background People with stroke often have increased bone loss and fracture rate. Increasing evidence has demonstrated a link between cardiovascular health and bone loss in other patient populations.
Objective The study objectives were: (1) to compare the bone density and geometry of the radius diaphysis on the left and right sides in people with chronic stroke and people who were matched for age (control participants) and (2) to examine the relationship between the bone strength index at the hemiparetic radius diaphysis and vascular health in people with chronic stroke.
Design This was a case-control study.
Methods The radius diaphysis on both sides was scanned with peripheral quantitative computed tomography in 65 participants with chronic stroke and 34 control participants. Large-artery and small-artery elasticity indexes were evaluated with a cardiovascular profiling system.
Results The paretic radius diaphysis had significantly lower values for cortical bone mineral density, cortical thickness, cortical area, and the bone strength index but a larger marrow cavity area than the nonparetic radius diaphysis in participants with chronic stroke, whereas no bone measurement showed a significant side-to-side difference in control participants. Multiple regression analyses showed that the large-artery elasticity index and grip strength remained significantly associated with the bone strength index at the hemiparetic radius diaphysis after controlling for age, sex, time since stroke diagnosis, body mass index, and physical activity (R2=.790).
Limitations This study was cross-sectional and could not establish causality. The radius diaphysis is not the most common site of fracture after stroke.
Conclusions Both the integrity of the vasculature and muscle strength were significantly associated with the bone strength index at the hemiparetic radius diaphysis in participants with chronic stroke. The results may be useful in guiding rehabilitative programs for enhancing bone health in the paretic arm after stroke.
Stroke is one of the most common global health problems. Stroke not only causes impairment in sensorimotor, cognitive, and social domains but also leads to serious complications, such as fragility fractures (ie, fractures that occur because of a fall from standing height or less),1 more than 20% of which occur in upper-extremity skeletal sites.2,3
One of the major contributing factors to fragility fractures after stroke is accelerated bone loss on the affected side, which is more substantial in the upper extremity than in the lower extremity.4 Therefore, studying bone health in the paretic upper extremity in people with stroke is of high clinical relevance. Bone mineral density (BMD) and bone geometry also are important factors in determining bone strength.5 Studying both of these factors may provide a more comprehensive understanding of bone status.
Peripheral quantitative computed tomography (pQCT) was used recently to study the volumetric BMD (vBMD, in mg/cm3) and geometric properties of upper-extremity long bones after stroke.6–9 Several cross-sectional studies with small samples of participants examined the radius diaphysis in people with chronic stroke and found significant side-to-side differences in cortical vBMD, cortical bone area, and cortical thickness.6–9 Later, Lazoura et al10 used a prospective study design to evaluate bone changes in people with subacute stroke. However, changes in bone geometry and recovery of various aspects of physical function were not reported.
The influence of neuromuscular impairments on bone changes after stroke has been well studied. For example, more severe muscle weakness was shown to be significantly associated with lower areal BMD (in g/cm2, measured by dual-energy X-ray absorptiometry) in the paretic upper extremity.11–13 In a pQCT study, muscle strength, functional recovery, and spasticity were found to be associated with side-to-side differences in the cortical thickness and cortical bone mineral content (BMC) of the radius shaft in people with chronic stroke.8 A common limitation of these studies, however, was that the integrity of the cardiovascular system—a factor that can have an influence on bone health—was overlooked.
Increasing evidence suggests that cardiovascular disease and osteoporosis are separate, but related, entities.14 Impairment in blood flow may have a negative impact on bone tissue because bone is a highly vascularized structure. Research in this area has gained tremendous momentum in the past decade, and several studies have provided evidence of an association between low BMD and compromised vascular health.15–20 For example, Collins et al17 found that after adjustment for potential confounders, peripheral artery disease remained significantly associated with a higher rate of hip bone loss and an increased risk of nonspine fractures in older men. Using magnetic resonance imaging technology, Griffith et al21 found that bone marrow perfusion was positively correlated with hip BMD in patients with osteoporosis and osteopenia. Considering the link between vascular health and bone health in other populations, it is possible that vascular mechanisms also play a role in bone loss in people after stroke.
It is well known that most people with stroke have a history of cardiovascular disease or suboptimally controlled cardiovascular risk factors.22,23 Poor cardiovascular fitness (as indicated by low peak oxygen consumption [V̇o2peak]) also is prevalent in people after stroke.24 Previous studies of people with chronic stroke showed that lower V̇o2peak was associated with lower hip BMD25 and tibial bone strength index.26 However, V̇o2peak is not a specific measure of vascular health. Additionally, because weight-bearing bones may have different responses to stroke impairments than non–weight-bearing bones,7,26,27 it is important to study the relationship between vascular function and upper-extremity bone health. To date, no study has examined this relationship.
This study was undertaken to address the knowledge gap and the limitations of previous studies in this research area. The objectives were: (1) to compare the side-to-side differences in pQCT variables measured at the radius diaphysis in people with chronic stroke and people who were matched for age (control participants) and (2) to determine the association between vascular health and the bone strength index of the hemiparetic radius diaphysis in people with chronic stroke.
We hypothesized that, compared with control participants, people with stroke would have significantly more compromised bone status on the paretic side than on the nonparetic side and that vascular health would be significantly associated with the bone strength index of the hemiparetic radius diaphysis in people with chronic stroke.
Method
Study Design and Setting
This was a case-control study. Because the second aim of the study was to examine the associations between the bone strength index and other factors only in people with stroke, the study was also cross-sectional in nature. Bone imaging was conducted by an experienced technician at the Centre for Osteoporosis Care and Control in Hong Kong, and other clinical assessments were performed by a physical therapist in a rehabilitation research laboratory at Hong Kong Polytechnic University. The technician who performed the bone imaging was unaware of the results obtained from the physical therapist's assessment, and the physical therapist was unaware of the results of the bone imaging. Data were collected from February 2009 through May 2011.
Participants
We chose to study people with chronic stroke because they are particularly susceptible to functional decline and physical inactivity28,29 and may have accumulated more substantial bone loss because of the chronic nature of the disease. The sample size calculation was based on an alpha of .05 and a power of .80. A previous study showed a significant side-to-side difference in cortical BMC in the radius diaphysis in people with chronic stroke,7 with an effect size of 0.7. We assumed a similar effect size (0.7) for the within-group and between-group comparisons of pQCT variables. The estimated number of participants required to detect a significant difference in cortical BMC was 68 (34 participants with stroke and 34 control participants).
Another aim of the study was to identify the associations between the bone strength index and cardiovascular parameters in people with stroke. In a previous pQCT study of the tibial diaphysis, V̇o2peak was shown to be associated with a side-to-side difference in the tibial bone strength index (R2=.20–.36).26 As determined by multiple regression analyses, the minimum sample required was 65 participants with stroke (effect size=0.25, 7 predictors).
We recruited participants with chronic stroke from local stroke self-help groups. The inclusion criteria were: diagnosis of stroke (onset of ≥6 months), age of 18 years or older, ability to follow simple verbal instructions, medically stable, and of Chinese origin. The exclusion criteria were: recurrent stroke, other neurological disorders (eg, spinal cord injury), serious musculoskeletal injuries or diseases (eg, amputations), metal implants or previous fracture in the upper limb, taking medications for treating osteoporosis before or after stroke (eg, bisphosphonates), and other serious diseases that prevented participation. Control participants (age matched) were recruited from the community. The eligibility criteria used for the participants with stroke applied to the control participants as well, except that the control participants could not have a history of stroke. Written informed consent was obtained from each participant before the study commenced. The study was conducted in accordance with the Declaration of Helsinki.
Measurements
Demographic and clinical factors.
All enrolled participants underwent a thorough interview, in which relevant information (eg, age, medical history) was obtained. The body mass index (BMI, in kg/m2) was calculated after height (in meters) and weight (in kilograms) were measured (Mechanical Beam Scale; Health O Meter, Alsip, Illinois). Cognitive function was assessed with the Abbreviated Mental Test.30 For assessment of the level of physical activity, the Physical Activity Scale for the Elderly was administered.31 From this questionnaire, a physical activity score was generated; a higher score represented a higher level of physical activity.
Grip strength (in kilograms) was measured with a Jamar dynamometer (Sammons Preston, Mississauga, Ontario, Canada). The testing position was in accordance with the recommendation of the American Society of Hand Therapists.32 With the participant in an upright sitting position, the arm was adducted against the side of the trunk and neutrally rotated. The elbow was flexed at 90 degrees with neutral supination and pronation and the wrist in a neutral position. Three trials were performed to obtain the mean grip strength on each side, with a 1-minute rest between trials. On the basis of the data obtained from these 3 trials, the test-retest reliability of the grip strength test was found to be excellent (intraclass correlation coefficient [3,1]=.943–.984). For evaluation of the severity of spasticity at the elbow joint, the Modified Ashworth Scale was used (0=no spasticity, 4=joint rigid). This scale has been shown to have acceptable reliability (Kendall tau correlation=.847).33
Bone imaging.
We used pQCT (XCT 3000, Stratec Medizintechnik GmbH, Pforzheim, Germany) to scan the radius diaphysis on each side (voxel size=0.5 mm, scan speed=25 mm/s). A scout view was obtained after the forearm was positioned properly and the anatomical reference line was positioned at the cortical end plate of the distal radius. Scans with a thickness of 2.3 mm were obtained at the radius diaphysis (at 33% of the total bone length proximal to the reference line). The data were analyzed with customized software (Stratec software, version 6.0). Cortical bone analysis was performed with CORTBD (mode 1) at a threshold of 710 mg/cm3; thus, densities of greater than 710 mg/cm3 were defined as cortical bone by the system.
The variables of interest included cortical BMC (in mg/mm; mineral content of pure cortical bone in a 1-mm slice), cortical BMD (in mg/cm3; mean density of pure cortical bone in a 1-mm slice), total area (in mm2; cross-sectional area of the bone after the soft tissue has been peeled off), cortical bone area (in mm2; area assigned as pure cortical), marrow cavity area (in mm2; cortical area subtracted from total area), cortical thickness (in mm; difference between the outer radius and the inner radius of the cortical shell), and polar stress–strain index (p-SSI, in mm3). The p-SSI is a bone strength index reflecting the torsional rigidity of the long bone shaft; it was shown to be highly correlated with the failure load when the bone was loaded in 3-point bending.34 The precision of our pQCT scanner was determined by twice examining 30 people who were healthy, with repositioning after the first scan. The coefficients of variation of the aforementioned variables were 0.69%, 0.54%, 1.97%, 1.01%, 6.48%, 2.05%, and 2.36%, respectively.
Vascular elasticity.
For the measurement of vascular elasticity, an HDI/PulseWave CR-2000 Research CardioVascular Profiling System (Hypertension Diagnostics Inc, Eagan, Minnesota) was used.35 The system provides large-artery and small-artery elasticity indexes.36,37 The former represents the elasticity of the aorta and large arteries, and the latter represents the elasticity of the small arteries and arterioles; higher values indicate a healthier vascular system.
A participant was placed in the supine position and asked to relax. A blood pressure cuff was placed on the nonparetic upper arm. To minimize possible movement of the radial artery during the assessment procedures, we applied a rigid plastic wrist stabilizer to the wrist on the other side, with the forearm in a supinated position. A piezoelectric acoustic sensor was placed over the radial artery just beside the styloid process of the radius. Blood pressure was measured with a linear dynamic deflation method. The sensor was adjusted so that the highest relative signal strength was acquired. When the radial arterial blood pressure waveform became stable, the waveform data over a 30-second period were recorded for arterial compliance analysis.36 Two trials were performed to obtain the mean large-artery and small-artery elasticity indexes. Using the reference large-artery and small-artery elasticity indexes provided by the manufacturer, we classified the participants as having or not having normal values.35 For assessment of the test-retest reliability of the system for measuring arterial compliance in participants with stroke, the experiment outlined above was repeated after a short rest period, and excellent reliability was found (intraclass correlation coefficient [3,2]=.90).
We chose to use large-artery and small-artery elasticity indexes in this study because they have been shown to be associated with cardiovascular disease.38–41 For example, Panaich et al39 showed that both large-artery and small-artery elasticity indexes were independently associated with subclinical coronary atherosclerosis in a group of 6,814 adults who were apparently healthy and who were 45 years of age or older. Duprez et al40 showed that the small-artery elasticity index was significantly related to future coronary heart disease, stroke, and heart failure in a sample of 6,235 patients with atherosclerosis. Grey et al41 showed that for every 2 units of reduction in the small-artery elasticity index, there was a 50% increase in the risk of cardiac events. The advantage of using arterial elasticity indexes derived from pulse-wave contour analysis was that they could shed light on the functional and structural abnormalities in both large and small arteries.42 This property is important because endothelial dysfunction, which adversely affects small-artery function, is a major contributor to progressive structural changes in vasculature. Other methods used to measure vascular stiffness, such as aortic pulse-wave velocity and ultrasonography of conduit arteries, only provide information on structural changes in large arteries.42
Data Analysis
Differences in the characteristics of the participants with stroke and the control participants were compared with the independent t test (for interval or ratio data), the Mann-Whitney U test (for ordinal data), and the chi-square test (for nominal data). The Kolmogorov-Smirnov test was used to check the normality of data. An analysis of variance was performed for each bone variable. In each analysis-of-variance model, a mixed design was used, with side as the within-participant factor (paretic/nondominant versus nonparetic/dominant) and group as the between-participant factor (stroke versus control). Post hoc paired t tests were used to compare the variables between the 2 sides in both participants with stroke and control participants. In a secondary analysis, the above-described procedures were repeated after separation of the data collected from men and women.
To determine the degree of association of the p-SSI with other variables in the participants with stroke, we used the Pearson r or the Spearman rho, depending on the level of data. Hierarchical multiple regression analyses were then performed, with the p-SSI of the paretic radius diaphysis as the dependent variable. The selection of independent (predictor) variables was based on both physiological relevance and the results of a bivariate correlation analysis. In addition, to guard against possible multicollinearity, we checked bivariate correlations among the independent variables. If a strong correlation existed (r>.5), then the 2 independent variables would not be entered into the same regression model.43 Potentially confounding factors, such as age, sex, BMI, time since stroke diagnosis, and physical activity, were entered into the regression model, followed by grip strength. Vascular variables that were found to be significantly related to the p-SSI in the bivariate correlation analysis were then entered into the regression model. The same analysis was repeated for the nonparetic side.
All entered independent variables were included in the regression equation, regardless of whether they were statistically significant.44 This method would enable us to determine whether grip strength and vascular elasticity were associated with the p-SSI after adjusting for the effects of other relevant factors. The variance inflation factor statistic was used to determine whether mutlicollinearity was present in each regression model. A variance inflation factor value of greater than 10 was indicative of multicollinearity.45
Statistical analyses were performed with SPSS 17.0 software (SPSS Inc, Chicago, Illinois). The significance level was set at .05 (2-tailed). For post hoc analysis, the alpha level was adjusted according to the number of comparisons made (Bonferroni correction).
Role of the Funding Source
This study was supported by the Research Grants Council (General Research Fund no. 526708).
Results
All 65 participants with stroke (stroke group) and 34 control participants (control group) fulfilled the eligibility criteria and completed the assessments. We were able to obtain from each participant a complete data set, including the independent variables used in the subsequent multiple regression analyses. The stroke group was more sedentary (lower score on the Physical Activity Scale for the Elderly; P<.001) and had more comorbidities (P≤.001) than the control group (Tab. 1). The grip strength on the paretic side was significantly lower (by 44%) than that on the nonparetic side, indicating substantial paresis. There was a tendency for the stroke group to have a higher proportion of participants with abnormal large-artery elasticity indexes (26.2%) than the control group (11.8%), but the difference did not reach statistical significance (P=.09). When the data collected from men and women were analyzed separately, the small-artery elasticity index for women in the stroke group was found to be significantly lower than that for women in the control group (P=.02).
Characteristics of Participants With Stroke (Stroke Group) and Participants Matched for Age (Control Group)a
pQCT Variables
The Figure shows the pQCT images of the radius diaphysis of a representative participant with stroke and a control participant. The paretic side (image A) had a substantially thinner cortical bone shell and a larger marrow cavity area than the nonparetic side (image B). In contrast, no such difference between the 2 sides was seen for the control participant (images C and D).
Peripheral quantitative computed tomography (pQCT) images of the radius diaphysis at the 33% site in a participant with stroke and a participant who was matched for age (control participant). The pQCT images were from the paretic side (A) and the nonparetic side (B) of a woman with stroke and the nondominant side (C) and the dominant side (D) of a woman who was a control participant. The paretic side had a considerably thinner cortical bone shell and a larger marrow cavity area than the nonparetic side. The side-to-side difference in the control participant was unremarkable.
When the data from all of the participants were analyzed (Tab. 2), it was found that the magnitude of the side-to-side differences in cortical BMC (P=.02), cortical vBMD (P≤.001), and p-SSI (P=.02) was group dependent (ie, a significant side × group interaction). Post hoc analysis indeed showed that in the stroke group, the paretic side had significantly lower cortical BMC (P≤.001), cortical vBMD (P=.01), cortical area (P≤.001), cortical thickness (P≤.001), and p-SSI (P≤.001) but a larger marrow cavity (P≤.001) than the nonparetic side. In the control group, there was no significant side-to-side difference in any of the pQCT variables.
Comparison of Peripheral Quantitative Computed Tomography Variablesa
In the sex-specific analysis, the paretic side in women in the stroke group had significantly lower cortical BMC (P≤.001), cortical vBMD (P≤.001), cortical area (P≤.001), cortical thickness (P≤.001), and p-SSI (P≤.001) but a larger marrow cavity area (P≤.001) than the nonparetic side (Tab. 2). The side-to-side differences in the same bone parameters in men were less pronounced than those in women in the stroke group, with significant findings only in cortical BMC (P=.01), cortical vBMD (P=.01), and p-SSI (P=.01).
Associations With Bone Strength Index
Correlation analyses were performed for the stroke group only. The p-SSI of the paretic radius diaphysis showed significant associations with the large-artery elasticity index (r=.417, P≤.001) and the small-artery elasticity index (r=.497, P≤.001) in a bivariate correlation analysis. Because spasticity and grip strength were strongly correlated with each other (ρ=−.643, P≤.001) and because the p-SSI showed a stronger association with grip strength (r=.712, P≤.001) than spasticity (ρ≤−.356, P=.01), only grip strength was entered as one of the independent variables in the regression analysis to avoid multicollinearity.
In multiple regression analyses, after accounting for age, sex, BMI, time since stroke diagnosis, physical activity, and grip strength, the large-artery elasticity index remained significantly associated with the p-SSI of the radius diaphysis on the paretic side (Tab. 3, model 1). Overall, this regression model explained a total of 79.0% of the variance in the p-SSI (P≤.001). Of the 2 modifiable factors, the effect of the large-artery elasticity index was more modest than that of grip strength, as reflected by the difference in the β coefficients (0.139 versus 0.234). On the other hand, the small-artery elasticity index was no longer significantly related to the p-SSI after other relevant factors were entered into the regression model (P=.60). Further analysis was done to determine whether similar associations of the large-artery elasticity index and grip strength with the p-SSI existed on the nonparetic side (Tab. 3, model 2). In this regression model, only grip strength (P=.03), not the large-artery elasticity index (P=.36), was significantly associated with the p-SSI after controlling for relevant factors. The variance inflation factor values were well below 5.0 in both regression models, indicating that multicollinearity was not a problem.
Multiple Regression Analyses for Predicting the Polar Stress–Strain Index (p-SSI) of the Radius Diaphysisa
Discussion
The results showed that people with stroke had significantly more compromised bone status measured at the radius diaphysis on the paretic side than on the nonparetic side. Additionally, both the large-artery elasticity index and grip strength remained independently associated with the p-SSI of the hemiparetic distal radius diaphysis after accounting for the effects of other relevant factors.
Comparison of pQCT Variables
Our analysis revealed significantly lower cortical BMC (8%) and cortical vBMD (2%) on the paretic side than on the nonparetic side in the stroke group. Such side-to-side differences were not observed in the control group. The results thus indicated a possible decline in bone mass in the hemiparetic radius diaphysis after stroke, with increased cortical bone porosity. These findings are in agreement with the findings of previous studies with smaller samples of people with stroke and without a control group.6,7 The total area showed no significant side-to-side difference in people after chronic stroke, indicating a relative preservation of overall bone size. On the other hand, the marrow cavity area on the paretic side was significantly larger than that on the nonparetic side. The combination of preservation of total area and larger marrow cavity area indicated possible endosteal resorption after stroke. However, because of the cross-sectional nature of the present study, we cannot prove that endosteal resorption was actually taking place in our stroke group. Generally, the side-to-side differences in pQCT parameters were more prominent in women with stroke than in men with stroke. This finding may be related to the more severe motor impairment and lower vascular elasticity indexes in women with stroke (Tab. 1) as well as factors that were not measured (eg, level of endogenous sex hormones).
Vascular Health and Bone Strength
Both grip strength and the large-artery elasticity index were independently associated with the p-SSI of the paretic radius diaphysis. The relationship between muscle strength and bone mass in people with stroke as well as other populations has been quite well established.6–8,46–48 The link between the large-artery elasticity index and bone strength, albeit modest, is a new finding for people with stroke. Nevertheless, the findings extended those demonstrating a link between bone health and vascular health in other populations.14–21,49 For example, Collins et al17 found that peripheral arterial disease was associated with high rates of hip bone loss in older men. The progression of aortic calcification was associated with the rate of decline in lumbar spine BMD and osteoporotic fractures in older adults.19,20
There are several potential explanations for the relationship between arterial elasticity and bone strength in our participants. First, various common factors (eg, aging, physical exercise) may affect bone arterial compliance and the integrity of bone tissue. For example, Kingwell et al38 showed that systemic arterial compliance is significantly related to exercise tolerance in people who are not disabled. A lower BMD also was found to be significantly associated with a lower physical activity level in older adults.50,51 However, it is unlikely that the link between bone health and cardiovascular function was merely attributable to shared etiological factors. For example, Farhat et al15 showed that a lower vBMD in the spine was related to more severe aortic calcification, independent of age and risk factors (eg, physical activity, weight) shared by osteoporosis and cardiovascular disease, in a sample of women who were middle-aged and healthy. Similarly, in the present study, we found a significant relationship between the p-SSI and the large-artery elasticity index, independent of age, BMI, physical activity, and muscle strength (Tab. 3, model 1).
Second, endothelial dysfunction is associated with multiple cardiovascular risk factors.14 Because bone is a highly vascularized structure, impaired blood flow attributable to endothelial dysfunction could have a negative impact on bone properties and vice versa.14,51 The pathophysiology underlying the link is not entirely understood, but mechanisms involving cytokines, oxidized lipids, and homocysteine have been implicated.14,16,52 More research is needed to decipher the pathophysiological links between arterial elasticity and bone health. Although a bivariate correlation analysis revealed a significant, positive association between the small-artery elasticity index and the p-SSI of the paretic radius diaphysis, their effects were diminished in a multivariate analysis.
Why was the large-artery elasticity index more powerful than the small-artery elasticity index in determining the p-SSI of the radius diaphysis? Vascular disease involves the dysfunction of the endothelial lining of small arteries and arterioles and is initiated by one or more factors, such as aging, inactivity, diabetes, high cholesterol, and high blood pressure, all of which are prevalent in patients after stroke.23,29 This early change in endothelial function may be detected as a decline in arterial elasticity.35 Typically, an alteration in endothelial function is easier to detect in very small arteries and arterioles, as reflected by a decline in the small-artery elasticity index. As the vascular disease progresses, large arteries are affected as well, leading to a reduction in the large-artery elasticity index.35 Thus, a reduction in the large-artery elasticity index is indicative of more advanced cardiovascular dysfunction. These factors may partially explain why the large-artery elasticity index, rather than the small-artery elasticity index, was significantly associated with the p-SSI of the hemiparetic radius.
Grip strength was significantly associated with the p-SSI on both sides, whereas the large-artery elasticity index was significantly related to the p-SSI only on the paretic side (Tab. 3). This finding may be indicative of the stronger influence of muscle function on bone health overall. Perhaps the substantial muscle weakness on the paretic side made the bone tissue more susceptible to further detrimental changes arising from the effects of vascular health. A previous study showed that only grip strength, not the large-artery elasticity index, was associated with the bone strength index in the distal epiphyseal region of the radius (mostly trabecular bone) in people with chronic stroke.8 This finding may suggest that the effects of muscle strength and vascular health on bone are site specific. Site-specific effects also were observed in the lower limbs. For example, in people with chronic stroke, gait speed was significantly related to the bone strength index at the tibial distal epiphysis but not the diaphysis.26,53 In people with spinal cord injury, muscle function was less strongly related to bone strength at the tibial diaphysis than at the distal epiphysis.54 Perhaps the radius epiphyseal region was more sensitive to the degree of paresis and the diaphyseal region was responsive to both muscle weakness and compromised vascular health.
Clinical Implications
The significant association between the large-artery elasticity index and the p-SSI of the paretic radius diaphysis in the present study may help open the field for clinical applications and research in cardiovascular function and bone health in people with stroke. For example, there may be potential for common therapeutic interventions.15 What are the effects of the medications used to treat cardiovascular health on bone tissue in patients with stroke? Would noninvasive interventions, such as aerobic exercise training, improve both vascular and bone outcomes? Because both the large-artery elasticity index and grip strength are related to the bone strength index on the paretic side, can bone strength be enhanced by adding an aerobic component to a strengthening program?
It is known that certain aerobic exercise modes, such as cycling and swimming, are not effective in improving bone health in young adults.55–57 On the other hand, Chien et al58 and Welsh and Rutherford59 showed that training through treadmill walking and stepping and skipping exercises was effective in improving hip BMD in older adults. However, it was not known whether the improvement in BMD was attributable to the aerobic aspect or the impact nature of the exercises. Nevertheless, our results suggest the potential of using aerobic exercise to improve upper-limb bone outcomes in people with stroke, in whom poor cardiovascular health is common. A combined leg muscle strengthening and aerobic exercise program was shown to have positive effects on tibial bone mass and geometry in people with chronic stroke.60 In another study of people with chronic stroke, aerobic treadmill exercise was found to enhance tibial cortical thickness.61 No study has examined the effects of cardiovascular and muscle strength training on upper-limb bone outcomes. Research in this important area is needed.
Limitations
We acknowledge that the present study has limitations. First, the results can be generalized only to people who dwell in the community and participate in community activities. The participants with stroke were recruited from community stroke self-help groups and thus may have been more physically able than their peers who were not members of these self-help groups. Second, the present study was cross-sectional and thus could not establish causality. Although we found a significant association between vascular elasticity and the p-SSI, we were not certain whether changes in vascular elasticity after stroke would lead to worse bone outcomes over time, nor did we know the cardiovascular and bone status of the participants before stroke. Future studies should use a prospective design to examine the relationship between changes in cardiovascular health indicators and bone parameters as time progresses. Third, increasing evidence suggests that bone metabolism is regulated by the central nervous system and that the activation of sensory, autonomic, and peripheral nervous systems may have distinct effects on bone mass.62 However, the present study did not determine whether the compromised bone status on the paretic side was attributable to disuse of the affected limb or stroke-induced alterations in certain neural signals that influenced bone health. Finally, the radius diaphysis is not the most common site of fracture after stroke. However, the experimental protocol allowed us to examine the relationship between vascular health and the integrity of cortical bone tissue.
Conclusions
The present study showed that in people with chronic stroke, the radius diaphysis on the paretic side had lower bone mass and more unfavorable bone geometric properties than that on the nonparetic side. Such side-to-side differences in bone properties were not observed in control participants. This also was the first study to show that the large-artery elasticity index and grip strength were independently associated with the bone strength index of the hemiparetic radius diaphysis. Further study should address the pathophysiological mechanisms underlying the link between bone strength and vascular health and should explore the potential for common therapeutic interventions in patients with stroke.
Footnotes
Dr Pang and Dr Jones provided concept/idea/research design, writing, facilities/equipment, and consultation (including review of manuscript before submission). Mr Yang provided data collection. Dr Pang provided data analysis, project management, and study participants.
Ethical approval for the study was granted by the ethics committee of Hong Kong Polytechnic University.
An abstract of the preliminary results of this study was presented at the International Osteoporosis Foundation Regionals 1st Asia-Pacific Osteoporosis Meeting; December 10–13, 2010; Singapore.
This study was supported by the Research Grants Council (General Research Fund no. 526708).
- Received September 24, 2012.
- Accepted February 14, 2013.
- © 2013 American Physical Therapy Association
References
Issue highlights
