Because of prolonged periods of immobilization, secondary conditions of low bone mineral density (BMD) and increased occurrence of nontraumatic fractures are prevalent in children with cerebral palsy (CP),1 with approximately 20% sustaining a femoral fracture during their lifetime.2 Maximizing peak bone mass before puberty is essential in minimizing the risk of fractures.3
Mechanical loading has been demonstrated to have a critical role in bone health in animals,4–6 adults,7–9 and children.10–18 In the lower extremities, mechanical loading is applied to the bone through muscle tension and weight-bearing.19 However, the specific anabolic signals induced by musculoskeletal loading are not yet understood. Subsequently, the optimal parameters of mechanical loading (ie, mode, frequency, intensity, and duration) have yet to be reported.19 In recent decades, Frost's20,21 mechanostat has provided insight into the underlying mechanisms, suggesting that repetitive, reciprocal loading, which produces oscillating stresses and strains on bone cells, is paramount in maximizing bone mass.19
Investigations into the effectiveness of loading interventions support the principles of Frost's mechanostat. While few controlled studies investigating the effect of passive standing interventions (continuous, stationary loading) have demonstrated ambiguity with respect to efficacy for increasing BMD in the long bones of children who are nonambulatory,11–13,18 dynamic standing interventions have demonstrated potential to increase BMD and bone mineral content (BMC). Physical and high-impact activities such as running and jumping, which apply large-magnitude forces and induce muscle contractions, demonstrated increases in BMC in children who are active compared with children who are inactive.16,17 Low-magnitude, high-frequency vibration, simulating muscle tension, induced increases in trabecular BMD in the long bones of children who are ambulatory, but not in children who are not ambulatory, suggesting the higher-magnitude forces experienced during ambulation are essential.13,14,15 Gudjonsdottir and Mercer22 designed a dynamic stander that applied reciprocal loading to the long bones in children with CP. Two children used the dynamic stander for 8 weeks, while 2 stood passively. Increases in BMD were observed in both children in the dynamic stander and 1 child in the passive stander.18
With previous studies demonstrating the importance of dynamic loading (such as the loading applied during ambulation, exercise, or low-magnitude, high-frequency vibration), the purpose of this study was to investigate the effectiveness of a novel dynamic standing intervention compared with a conventional passive standing intervention on bone health in children with CP who are nonambulatory, through assessments of BMD, BMC, and 2-dimensional area. The information gained from comparing bone health in a dynamic standing intervention, which mimics the forces applied during the natural walking gait, to that in a passive standing intervention may contribute in determining the optimal loading parameters for therapeutic interventions and understanding the underlying anabolic signals induced by mechanical loading.
A 15-month longitudinal study was conducted through the Rehabilitation Engineering Research Center on Technology for Children with Orthopedic Disabilities, a collaboration between the New Jersey Institute of Technology (NJIT) and Children's Specialized Hospital (CSH). All children were recruited through CSH and attended either First Children School (Fanwood, New Jersey) or Passaic County Elks Cerebral Palsy Center (Clifton, New Jersey). All standing sessions were conducted in the educational settings and followed the classrooms’ prior protocol of 5 d/wk for 30 min/d. Children were placed in either commercially available standers from Prospect Designs (New Hartford, Connecticut), Rifton Equipment (Rifton, New York), and EasyStand (Altimate Medical, Morton, Minnesota) or a novel dynamic stander. In each dynamic stander, novel footplates were incorporated into the existing standing frames to provide reciprocal loading that mimics the forces applied to the lower limbs during the natural walking gait23 (Figure 1). The pneumatic footplates were controlled through a custom MATLAB program (Mathworks, Natick, Massachusetts). Load cells were incorporated to measure the forces being applied during the standing sessions. Minimal modifications were made to the existing frame body.
The study consisted of 3 phases: (1) during months 1 through 6, the children stood in their current conventional passive standers (n = 4) or the novel dynamic stander (n = 5); (2) during months 6 through 9, all children stood passively; and (3) during months 9 through 15, the children returned to the standing intervention of the first phase.
Fourteen prepubescent children who were nonambulatory were recruited. Nine children completed the full 15-month intervention. Four children did not complete the 15-month intervention because of logistical and health reasons unrelated to the intervention, and 1 child began to ambulate, thereby being excluded from the current analysis. The children were between 4 and 9 years of age, currently participating in a daily passive standing intervention. Gross Motor Function Classification System levels included 2 children at level III, 2 at level IV, and 5 at level V. Two children received enteral feedings, and 6 were on anticonvulsant medications. Institutional review board approval was obtained through the host institute, and informed parental consent and subject assent were obtained.
Bone Density Measurements
Dual-energy x-ray absorptiometry (DXA) scans were obtained at 3-month intervals (0, 3, 6, 9, 12, and 15 months) by using the lateral distal femoral method, which has been developed to obtain measures in children with CP.24 The General Electric Pediatric Lunar Prodigy Advanced DXA densitometer was used to acquire all scans, and analysis was completed with General Electric enCORE software version 9.30.044 (Madison, Wisconsin). The software's ⅓ forearm mode was employed to allow for multiple custom regions of interest (ROIs) to be set. The same technician and research team performed all scans and placed all ROIs for analysis. To determine the precision of this study (reproducibility and repeatability of our measures), 2 scans of each leg were obtained at each session. The BMD measures for the same-day, same-leg scans of each child were analyzed to determine that the precision of this study was 4.5%, within the published ±5% for the femur.25,26
Eleven ROIs were investigated. Three ROIs (ROIs 10 through 12, hereby referred to as “anatomical ROIs”) were placed as described by Henderson and colleagues24 (Figure 2). The ROIs were developed with the intention that ROI 10 would consist primarily of trabecular bone, ROI 11 would consist of trabecular and cortical bone, and ROI 12 would consist primarily of cortical bone. These ROIs are increasingly being used in children as their height is computed as a function of the diameter of the midshaft of the femur, thereby accounting for differences in size between children and accommodating for a degree of growth in longitudinal studies.24 However, as it is known that the ratio of length and width of the femur changes throughout childhood27 and children with CP have varying degrees of diminished longitudinal growth,28 an additional 8 ROIs with a constant height of 1 cm were investigated (ROIs 2 through 9, hereby referred to as the “1-cm ROIs”) (Figure 2).
The normalized BMD, BMC, and ROI area within each ROI were calculated for each participant by subtracting the respective baseline measure (0 month) from the measure of the elapsed period of interest and dividing by the baseline. Percentage changes were calculated by multiplying the normalized values by 100 and pooling the right and left leg data for each child for each standing intervention3,14,15 at each elapsed time.
The primary outcome measures were BMD, BMC, and ROI area. The 3 anatomical ROIs and the eight 1-cm ROIs were separated for statistical analysis. Mean percentage changes for 5 children in the dynamic standing intervention and 4 in the passive standing intervention (total n = 9) at each elapsed time interval and within each ROI compared with baseline were analyzed by using Wilcoxon and Kruskal-Wallis tests with a .05 level of significance (α) (PASW Statistics 18, SPSS Inc, Quarry Bay, Hong Kong).
Although true typical walking gait produces reciprocal loading from 0% body weight (during swing) to more than 100% of body weight (during stance), the risk of fractures within this study's sample led researchers to apply conservative forces. The conservative application of force, coupled with the partial support of the child's body weight, resulted in dynamic loading that reciprocated between 5% and 25% body weight on each leg at a frequency in the order of 1 Hz, mimicking the cadence of slow walking. Children in the passive loading intervention group showed modest variation in loading on each leg of 17% to 20% because of natural shifting of body weight. The frequency of the passive weight shifting was random and infrequent throughout the sessions. Overall compliance for the children receiving the passive standing intervention was 85% and for those receiving the dynamic standing intervention was 83%.
Three Anatomical ROIs
In the children in the dynamic standing intervention, BMD was determined to significantly increase after 9 months (P < .044; Figure 3A). Increases within the cortical bone were greater than those in the trabecular bone (P = .043). No significant increases were found throughout the duration of the study for the passive standing intervention. However, the baseline BMD was maintained during passive standing (Figure 3B).
Bone mineral content in the children in the dynamic standing intervention group increased significantly after 6 months (P = .010; Figure 4A). Children in the passive standing intervention group experienced a significant increase in their BMC after 9 months (P = .007), but this was not sustained throughout the duration of the study (Figure 4B).
The ROI area increased in both standing interventions at similar magnitudes after 6 months (Figures 5A and 5B).
Significant increases in BMD were observed after 3 months within the dynamic standing intervention. Similar to the anatomical ROIs, the increases occurred in the proximal cortical bone (Figure 3C). Bone mineral density was maintained throughout the duration of the study for the passive standing intervention (Figure 3D).
Bone mineral content within the dynamic standing intervention (Figure 4C) increased significantly after 6 months (P = .010). Within the passive standing intervention, significant increases occurred after 9 months (Figure 4D).
As in the 3 anatomical ROIs, the ROI area increased in both standing interventions at similar magnitudes after 6 months (Figures 5C and 5D).
With current research in bone health demonstrating that physical activity and low-magnitude, high-frequency vibration decrease the progression of osteoporosis, primarily in trabecular bone, we hypothesized that incorporating reciprocal loading, which mimics the natural gait, into the therapeutic protocols of children with CP who are nonambulatory would improve BMD at a greater rate than conventional passive standing. The results of this study conclude that dynamic standing does significantly increase BMD, primarily in cortical bone, whereas passive standing appears to maintain the baseline BMD.
The relatively fast scan-acquisition time and low radiation dose make DXA the current reference standard for determining BMD in children with CP. However, the 2-dimensional estimation of the parameters decreases the precision of the imaging modality.26,27 Therefore, increases in BMD, BMC, and area are considered of clinical significance only if they exceed 5% precision. The increases induced in the BMD of cortical bone do exceed the 5% precision within the dynamic standing intervention group, whereas within the passive standing intervention group, they do not (Figure 3). Figure 4 illustrates that whereas both interventions induce BMC increases greater than 5%, the increases induced during dynamic standing are greater, verifying that the larger, consistent 1-Hz reciprocal loading provided by dynamic standing is more effective than the infrequent shifting during passive standing.
Ethical and logistical reasons prevented collection of longitudinal data on children who are nonambulatory and participating in a no-standing intervention. Therefore, one limitation of this study is the inability to determine the effectiveness of passive standing compared with no intervention. Passive standing did maintain the baseline measures of BMD within this study, but it cannot be concluded whether this has increased benefit to the baseline BMD for a no-standing intervention.
Investigation of the 1-cm ROIs revealed 2 trends not apparent in the anatomical ROIs. The first trend was the suggested progression of increases in BMD in the cortical bone. Although the increases did not reach significance in the mid-ROIs within our small sample size, the increases in the midregions (ROIs 6 and 7) did begin to approach significance, whereas ROI 11 did not. This suggests that the modeling is greatest at the cortical bone region nearing the midshaft of the diaphysis, with a temporal shift distally from the diaphyseal region through the metaphyseal region. As the increases approach significance in the 1-cm ROIs and not the anatomical ROIs, it suggests that the anatomical ROIs are too large to detect the subtle increase in the ratio of the cortical to trabecular bone within the metaphyseal region. Future studies with larger sample sizes should be conducted to further investigate whether the 1-cm ROIs can provide additional information on the trends associated with loading interventions.
The second trend apparent in the 1-cm ROIs is a plateau within the dynamic standing intervention. Although a temporal distal shift begins to appear in the trends of the BMD, the increase in BMD from months 12 to 15 in ROIs 8 and 9 is sustained, suggesting a potential plateau may have been reached. This plateau could be a result of the conservative nature of the dynamic loading. A longer-duration study should be completed to determine whether this plateau persists and whether the distal ROIs experience a similar plateau. The plateau is not apparent in the anatomical ROIs, suggesting that the overall size of the anatomical ROIs is masking the underlying trends.
Each standing intervention demonstrated increases in BMC. Significant increases within dynamic standing, however, occurred before increases in passive standing and at a greater magnitude (13%, SD 5% vs 8%, SD 5%, respectively). These increases are similar to trends previously reported in children who were active versus in children who were inactive over a 6-year span.17 In addition, increases within the dynamic standing intervention are maintained and continued to reach significance for the duration of the study, whereas within the passive standing intervention, they do not. This suggests that the increase found with passive standing could be attributed to growth of the children. Future studies with larger sample sizes and anatomical measures of the length of the femur should be completed to determine the effect of growth.
The similar magnitudes of increase in the ROI area between the intervention groups suggest that the loading paradigm is not the influential factor stimulating cross-sectional growth of the femur. Further studies, with larger sample sizes and control groups consisting of children who are ambulatory and children who are nonambulatory and not participating in a standing intervention are required to determine whether the increases are due to loading in general or the natural growth spurts experienced throughout childhood.
Growth is currently an unresolved issue among longitudinal studies using assessment with DXA. Whereas growth remains unresolved in this study, it is known that 70% of femoral growth occurs at the distal growth plate.29 As children in this study experienced 2 to 4 cm of growth in total height, the approximate longitudinal growth of the femur equates to 0.5 to 1 cm30 and therefore the growth within the ROIs would range from 0.35 to 0.7 cm. This magnitude of growth could temporally shift the 1-cm ROIs proximally almost 1 full ROI between 0 and 15 months. This limitation is not considered detrimental to the study as the rate of growth between each group is not significantly different and therefore the overall effect of growth should be the same within each intervention. In addition, if the ROIs did shift proximally, ROI 2 at 0 months should be compared with ROI 3 at 15 months, ROI 3 with ROI 4, and so on. Examination of the data reveals that this would currently lead to an underestimation of the significance of the dynamic standing intervention on BMD. As still unresolved, growth is acknowledged to be a constant confounding factor in the study, and the conservative analysis (ie, comparing ROI 2 at 0 months with ROI 2 at 15 months, ROI 3 with 3) is maintained for this study.
At first, the potential benefit of standing interventions on BMD may appear to be conservative. However, in the context of the overall goal of creating a therapeutic intervention that would induce significant changes in bone health without modifications to current protocols and strict control of a child's daily environments, the modest increases induced by dynamic standing suggest a clinical application, especially as they occur within the fracture region of the distal femur—the site in which the majority of nontraumatic fractures occur. The dynamic standing intervention produced significant increases in the BMD and BMC of the children within the first 6 months, despite the fact that diet and activities outside school could not be strictly controlled, growth was not directly accounted for, and dynamic loading was applied with conservative precaution and short dosage (30 min/d, 5 d/wk).
With the potential efficacy demonstrated in this pilot study, future studies regarding the dynamic standing intervention will include device modifications to the dynamic stander to make it more compact, inclusion of longitudinal measures of BMD and BMC in children who are ambulatory and children who are not ambulatory and not participating in a standing intervention to determine the full spectrum of the effect of loading, a longer study duration, a greater sample size, longitudinal measures of anthropometric data to account for growth, and outcome measures such as heart rate variability, bowel and bladder function, respiratory function, behavior, muscle tone, and range of motion.
Dynamic standing was determined to significantly increase BMD and BMC in the cortical bone of children with CP who are nonambulatory, whereas passive standing appeared to maintain the BMD. As previous interventions that provide loading primarily increased BMD and BMC in trabecular bone, the findings from this pilot study hold the potential to provide critical insight into the mechanisms of bone health induced through mechanical loading and the optimal parameters of loading interventions. Future studies with a larger sample size, comparisons of measures of BMD and BMC in children who are ambulatory and children who are nonambulatory and not participating in standing interventions, additional outcome measures, and a longer study duration should be conducted to fully investigate the efficacy of the novel dynamic standing intervention in this study.
The authors thank Bruno Mantilla (NJIT) for his assistance in study design; John Hoinowski (NJIT) and Timothy Feurey (CSH) for their assistance in the fabrication of the dynamic standers; John Stasil (CSH) and Amanda Irving and Katharine Swift (NJIT) for their assistance in data collection; Sherif Tadros (NJIT) for his assistance with data analysis; and Camila de Oliveira, Amy Boos, and Olga Hizkiyahu (NJIT), as well as the teachers, aides, and therapists at First Children School and Passaic County Elks Cerebral Palsy Center for their assistance in the daily standing interventions.
1. Rittweger J, Felsenberg D. Recovery of muscle atrophy and bone loss from 90 days bed rest: results from a one-year follow-up. Bone. 2009;44:214–224.
2. Ko C, Tse P, Chan A. Risk factors of long bone fractures in non-ambulatory cerebral palsy children. Hong Kong Med J. 2006;12:426–431.
3. Bass S, Pearce G, Bradney M, et al. Exercise before puberty may confer residual benefits in bone density in adulthood: studies in active prepubertal and retired female gymnasts. J Bone Miner Res. 1998;13(3):500–507.
4. Bell RR, Tzeng DY, Draper HH. Long-term effects of calcium, phosphorus and forced exercise on the bones of mature mice. J Nutr. 1980;110:1161–1167.
5. Rubin CT, Lanyon MR. Regulation of bone formation by applied dynamic loads. J Bone Joint Surg. 1984;66:397–402.
6. Lanyon LE, Rubin CT. Static vs dynamic loads as an influence on bone remodeling. J Biomechanics. 1984;17:897–906.
7. Rubin C, Recker R, Cullen D, Ryaby J, McCabe J, McLeod K. Prevention of postmenopausal bone loss by a low-magnitude, high-frequency mechanical stimuli: a clinical trial assessing compliance, efficacy and safety. J Bone Miner Res. 2004;19(3):343–351.
8. Gilsanz V, Wren T, Sanchez M, Dorey F, Judex S, Rubin C. Low-level, high-frequency mechanical signals enhance musculoskeletal development of young women with low BMD. J Bone Miner Res. 2006;21(9):1464–1474.
9. Verschueren S, Roelants M, Delecluse C, Swinnen S, Vanderschueren D, Boonen S. Effect of a 6-month whole body vibration training on hip density, muscle strength, and postural control in postmenopausal women: a randomized controlled pilot study. J Bone Miner Res. 2004;19(3):352–359.
10. Stuberg W. Considerations related to weight-bearing programs in children with developmental disabilities. Phys Ther. 1992;72(1):35–40.
11. Herman D, May R, Vogel L, Johnson J, Henderson R. Quantifying weight-bearing by children with cerebral palsy while in passive standers. Pediatr Phys Ther. 2007;19:283–287.
12. Hough JP, Boyd RN, Keating JL. Systematic review of interventions for low bone mineral density in children with cerebral palsy. Pediatrics. 2010;125(3):e670–e678.
13. Caulton JM, Ward KA, Alsop CW, Dunn G, Adams JE, Mughal MZ. A randomised controlled trial of standing programme on bone mineral density in non-ambulant children with cerebral palsy. Arch Dis Child. 2004;89:131–135.
14. Ward K, Alsop C, Caulton J, Rubin C, Adams J, Mughal Z. Low magnitude mechanical loading is osteogenic in children with disabling conditions. J Bone Miner Res. 2004;19(3):360–368.
15. Wren T, Lee D, Hara R, et al. Effect of high-frequency, low-magnitude vibration on bone and muscle in children with cerebral palsy. J Pediatr Ortho. 2010;30:732–738.
16. Chad K, Bailey D, McKay H, Zello G, Snyder R. The effect of a weight-bearing physical activity program on bone mineral content and estimated volumetric density in children with spastic cerebral palsy. J Pediatr. 1999;135:115–117.
17. Bailey D, Mckay H, Mirwald R, Crocker P, Faulkner R. A six-year longitudinal study of the relationship of physical activity to bone mineral accrual in growing children: the University of Saskatchewan Bone Mineral Accrual Study. J Bone Miner Res. 1999;14(10):1671–1679.
18. Gudjonsdottir B, Mercer V. Effects of a dynamic stander versus a static prone stander on bone mineral density and behavior in four children with severe cerebral palsy. Pediatr Phys Ther. 2002;14:38–46.
19. Rubin CT, Lanyon LE. Osteoregulatory nature of mechanical stimuli: function as a determinant for adaptive remodeling in bone. J Ortho Res. 1987;5:300–310.
20. Frost H. Bone's mechanostat: a 2003 update. Anat Rec A Discov Mol Cell Evol Biol. 2003;275:1081–1101.
21. Frost H. A 2003 update on bone physiology and Wolff's Law for clinicians. Angle Orthod. 2004;74(1):3–15.
22. Gudjonsdottir B, Mercer V. A motorized dynamic stander. Pediatr Phys Ther. 2002;14:49–51.
23. Damcott MD. The Effect of Dynamic Standing on the Bone Mineral Density of Non-ambulatory Children: A Pilot Study [master's thesis]. Newark, NJ: New Jersey Institute of Technology, Biomedical Engineering Department; 2011.
24. Henderson R, Lark R, Newman J, et al. Pediatric reference data for dual x-ray absorptiometric measures of normal bone density in the distal femur. AJR Am J Roentgenol. 2002;178:439–443.
25. Bloomfield SA, Mysiw WJ, Jackson RD. Bone mass and endocrine adaptations to training in spinal cord injured individuals. Bone. 1996;19(1):61–68.
26. Gordon C, Bachrach L, Carpenter T, et al. Dual energy x-ray absorptiometry interpretation and reporting in children and adolescents: the 2007 ISCD pediatric official positions. J Clin Densitom. 2008;11(1):43–58.
27. Goulding A, Gold E, Cannan R, Williams S, Lewis-Barned N. Changing femoral geometry in growing girls: a cross-sectional DEXA study. Bone. 1996;19:645–649.
28. Henderson R, Gilbert S, Clement M, Abbas A, Worley G, Stevenson R. Altered skeletal maturation in moderate to severe cerebral palsy. Dev Med Child Neuro. 2005;47:229–236.
29. Pritchett JW. Longitudinal growth and growth-plate activity in the lower extremity. Clin Ortho Rel Res. 1992;275:274–279.
30. Winters D. Biomechanics and Motor Control of Human Movement. 2nd ed. New York, NY: John Wiley & Sons Inc; 1990.
assistive devices/therapeutic use; bone mineral content; bone mineral density; cerebral palsy; child; disabled children/rehabilitation; dynamic weight-bearing; physical therapy modalities; weight-bearing