April 2013

Poststroke bone changes in patients who use AFOs

The medical literature suggests that changes in bone density and other bone characteristics after stroke persist after patients have regained ambulatory status. Whether ankle foot orthoses have a shielding effect on bone remodeling, however, remains unclear.

By Kyle Sherk, MS, CPO

Bones have come to be known as dynamic living structures within the body of the individual they support. The attachment points for muscles and ligaments, once thought to be static and lifeless, are now known to adapt continually to certain loads that stress them sufficiently. They weaken when unloaded, as in bed rest and space flight, and can reinforce themselves when loaded again.

Wolff’s Law, developed more than a century ago, gives some insight into the mechanisms of bone remodeling under various conditions that are being investigated today in universities around the world. In the sterile environment of the laboratory, young and otherwise healthy individuals have been shown to regain the bone mineral density (BMD) lost during 90 days of bed rest in about one year.1 Debate continues as to whether the effect of gravity or muscular pull upon the bones in response to gravity provides most of the strain needed to induce remodeling to increase bone strength.2 It can be surmised from these BMD studies that returning to ambulatory status after any bed rest due to a diseased state would be beneficial for the recovery of or, at the very least, attenuation of bone loss during the recovery period.

The patients seen by orthotists and prosthetists suffer not only from the ailment (i.e., fractures or amputation) for which they require an assistive device, or that ailment’s after-effects (as in stroke and polio), but also from some period of bed rest associated with their initial recovery. Many prosthetic socket designs used after amputation do not load the end of the transected bone or bones, and this has been shown to lead to continued loss of bone mineral content (BMC) and BMD at the cut end of the bone.3 Several orthotic designs also unload part or all of the entire limb they surround, such as ischeal weight-bearing knee ankle foot orthoses. The long-term effects of using such orthoses to maintain ambulatory status are not known, and their rarity makes them difficult to study in numbers sufficient to generalize results.

In stroke survivors, bone densities have been shown to differ significantly between the affected and sound sides, in both the upper or lower extremities. This discrepancy has been attributed to the bed rest following the stroke,4,5 a difference in loading of the limbs following the recovery period,6-9 and lack of sun exposure to convert vitamin D to its biosynthetic form.10-12  Since most lasting effects of stroke are seen in gait changes and muscle weakness or paralysis, the bones of survivors of stroke were often not studied until the 1990s and 2000s. Advances in radiologic imaging allowed for accurate in vitro assessment of changes in BMD and BMC over time. An improved understanding of how bones function within the body and an anecdotally noted increase in fractures in survivors of stroke likely contributed to the increased interest in the study of bones of stroke survivors.


Figure 1: Percentage losses in areal BMD at the primary hip sites as assessed by DXA. This information was compiled from studies by Demirbag, Paker, and Sherk.4,23,25

Anecdotal evidence has been confirmed by several studies showing an increase in hip fractures as well as other common fall-related fractures in stroke survivors.13,14 Studies of poststroke populations have reported losses in hip and spine BMD and an increase in diagnosis of osteopenia and osteoporosis.4,15-24 It is not known whether the use of an ankle foot orthosis (AFO) might accelerate or mediate the loss of bone density in the calcaneus,15 as the orthosis could shield the calcaneus from the ground reaction forces (GRFs) seen in normal walking or decrease the efficacy of shock absorbing mechanisms, potentially increasing impulse of the GRFs at the calcaneus.

A limited number of longitudinal studies have shown continued BMD loss (assessed with dual energy x-ray absorptiometry [DXA]) during the first year poststroke at the femoral neck and greater trochanter even with a return to ambulatory status.6,18 Individuals who were unable to return to an ambulatory status had even greater areal BMD losses, as would be expected. The study by Jørgenson did not describe what gait aids, if any, were used by patients to regain ambulatory status, instead classifying them by the degree of assistance provided by another individual to achieve ambulation.18

Generalized results from several longitudinal studies are compiled in Table 1; they show significant losses on the affected side over as little as three months (Yavuzer et al).10,15,18,20 Women also appear to have an increased risk of bone loss following a stroke.

Several studies have used peripheral quantitative computed tomography (pQCT) to garner a 3D view of bone changes. DXA, while providing a more global view of how and where BMD and BMC are changing, is limited to assessment of two dimensions. DXA’s primary advantage is its speed and ease in assessing losses of areal BMD in the key clinical areas of the hips and lumbar spine. From this information, the individual’s BMD is assigned a T-score relative to the age-adjusted average for his or her sex and weight. The T-score is a standard deviation in which positive scores are better than the average and negative scores are worse. The World Health Organization defines osteopenia as a T-score between -1 and -2.4 and osteoporosis as a T-score of -2.5 or lower.

In contrast, pQCT provides a detailed view of specific cross-sections of the scanned limb. This adds to the time a patient needs to be able to remain still for capture of the detailed scan but also adds more structural information about the bones and muscles within the given cross-section. The radiation dose associated with either device is very low, about 1.5 mrem (5.38 μSv) for a full-body DXA scan with bilateral hip detail scans and around .12 mrem (.43 μSv) for each pQCT scan. By comparison, the average radiation dose associated with natural sources is estimated at .82 mrem per day (300 mrem/year or 1076 μSv/year).

Bones and braces

The original intent of our study was to compare the differences between the affected and sound sides of poststroke patients who were users of different styles of AFOs (solid ankle and hinged ankle). Due to recruiting difficulty, only nine individuals were included. Of the nine AFOs used in this cohort, five were hinged, three were dynamic, and only one was a solid ankle design. The participants were aged 64.2 ± 1.9 years (range, 60 to 74 years), 172.7 ± 3.2 cm tall, and weighed 85.1 ± 7.1 kg. The five men were significantly taller than the four women (p < .05) but did not weigh more and were of similar age.


Participants were 13.5 ± 4.4 years poststroke (range, 4 to 47.5 years), had taken 11.8 ± 5.7 weeks to return to ambulation after the stroke (range, 2 to 52 weeks), and had used an AFO for six months to 12 years (average, 6.5 ± 1.4 years). The outlier for time since stroke (47.5 years) was also the outlier for AFO use (six months) and used the AFO only for improved mediolateral stability. No differences in poststroke demographics were found between men and women. There was an approximately even distribution of gait style between step-to (n = 4) and step-over (n = 5) with the affected limb.25 Analyzing the participant pool by gait style did not result in any greater differences between the sides, but the small sample size may have been a limiting factor.

The data from our participants’ DXA scans did not reveal any new or controversial findings compared with previously published research (Figure 1).5,9,23,25 The total hip BMC was 12% lower on the affected side than the unaffected side (p < .05). The trochanteric BMC was 19% lower on the affected side (p < .05). The areal BMD values were also lower on the affected side: total hip 10%, trochanter 12%, and femoral neck 7% (all p < .05). These areal BMD values equated to T-scores from 0 to -2.8 on the affected-side hip sites and 1 to -2.5 on the sound-side hip sites. This suggests returning to ambulatory status did not ameliorate losses in hip BMD; however, our cross-sectional study design prevented us from ascertaining the timeline of the losses at the hip to see if any improvements occurred after participants’ return to ambulatory status. Interestingly, there were no differences in the incidence of osteopenia or osteoporosis between sides for the different areas of the hip.25

poststroke-tbl2Our pQCT results did provide interesting insight into the remodeling scheme in the tibia on the affected side. We defined the distal end of the tibia (ankle mortise) as 0% and measured proximally from this origin. Significant differences are highlighted in Table 2. We did find that the total BMD at each of the three sites (4%, 38%, and 66% sites), trabecular BMD at the 4% site, and cortical BMD at the 38% site were lower on the affected side than the sound side. By comparison, Talla et al found significant differences in the total BMDs and total BMCs at the 4% and 66% sites of the tibia only; the 38% site was not assessed in their study.26 In our study, tibial remodeling was definitively shown to occur on the endosteal surface of the bones. The endosteal circumference was significantly greater at the 38% and 66% sites on the affected side, indicating that the endosteal surface had undergone more resorption than the periosteal surface had undergone deposition (no difference between the sound and affected sides). This combination led to a significant difference between the sides in the cortical area at the 38% and 66% tibial sites. Cortical thickness was similarly disparate between the sides.25-27 The 4% site, which is predominantly trabecular bone, did not undergo the same analysis for cortical area and thickness. The 4% site did not reveal a difference in the size of the bones (total area), but the total and trabecular BMDs were significantly lower on the affected side. No discrepancy between the respective areas has been reported previously.26,28

We found tibial strength was significantly lower at the 4% and 66% sites on the affected side than on the sound side. The 4% site had a lower bone strength index (BSI) on the affected side, which is an indication of loss of BMD, area, or both.25 Because BSI is calculated from the significantly different total BMD and the similar total areas (BSI = BMD2 x total area), this should not be surprising. This significant difference was also found by Pang et al.28 However, since trabecular bone relies on the strength of individual thin struts to maintain shape and support, a minor loss of BMD can lead to large losses in overall strength. The clinical significance of this difference in BSI for the 4% site is not yet known. At the 38% site, which was encompassed by each subjects’ AFO, there were no differences in the bone strength measures. This is contrary to Pang’s 2008 paper, which reported a significant difference in the 30% site’s BSI between sides of ambulatory stroke survivors, but not the incidence of AFO use.27

In our study, at the 66% site, the minimum rotated area moment of inertia (Imin) and strength strain index (SSI) were significantly lower (30.6% and 10.3%, respectively, both p < .05) on the affected side compared with the sound side. The decrease in theImin at the 66% site could lead to an increased fracture risk in this area of the tibia. Interestingly, the same result for SSI was not found by Talla.26 In our study, the maximum rotated area moment of inertia (Imax)for 66% site was not different between sides. These results indicate that a shift in loading of the proximal tibia may be occurring. The source of this potential shift has not been examined. Since none of the AFOs were proximal weight-bearing designs, an unloading mechanism involving the AFO is not possible for the proximal tibia site. The alignment of the limb within the AFO or a forced shift in the manner of weight-bearing with the AFO donned may contribute to this decrease in the proximal Imin, but one would think this would also lead to a difference at the 38% site.

While a shielding mechanism does not appear to be present based on the 38% tibial site results, it cannot be fully ruled out because of the 4% tibial site results. The higher turnover of trabecular bone may indicate an initial loss of BMD that is not recovered once the individual is ambulatory. Differentiating between a change in gait following a stroke and any effect of using an orthosis could be difficult in a patient population that relies on AFOs for safe and effective ambulation.


The use of an AFO to regain ambulatory status may have greater global affects on the body than were seen in our study. The changes in bony structure may have occurred prior to returning to ambulatory status. Many questions remain about how orthotic devices affect the bones they encompass. Our study was able to provide only a glimpse into the complex loading mechanisms that may affect the limbs of survivors of stroke and, potentially, other AFO users. A longitudinal study covering a majority of the recovery period following a stroke should answer more of these questions. Gait analysis combined with radiologic bone analysis could also answer more questions regarding lower limb loading mechanisms in all AFO users.

Due to large variations in physiologic variables, tight control of participant grouping is important when analyzing the poststroke population. Other radiologic methods, such as measuring calcaneal BMD, may assist in determining how much shielding, if any, occurs within custom AFOs. We can surmise that many off-the-shelf designs, which do not typically encompass the heel, would not have a shielding effect on the calcaneus. The effects of an AFO on the lower extremity may also be studied in people with dropfoot as a result of radiculopathy, as their gait may be affected in ways that are similar to the effects of stroke, but without the life-threatening consequences and resultant bed rest associated with the latter diagnosis.

Kyle Sherk, MS, CPO, is an orthotist and prosthetist with Hanger Clinic in Englewood, CO.

  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(2):214-224.
  2. Kohrt WM, Barry DW, Schartz RS. Muscle forces or gravity: what predominates mechanical loading on bone? Med Sci Sport Exerc 2009;41(11):2050-2055.
  3. Sherk VD, Bemben MG, Bemben DA. BMD and bone geometry in transtibial and transfemoral amputees. J Bone Miner Res 2008;23(9):1449-1457.
  4. Demirbag D, Ozdemir F, Kokino S, Berkarda S. The relationship between bone mineral density and immobilization duration in hemiplegic limbs. Ann Nucl Med 2005;19(8):695-700.
  5. Pang MY, Eng JJ. Muscle strength is a determinant of bone mineral content in the hemiparetic upper extremity: implications for stroke rehabilitation. Bone 2005;37(1):103-111.
  6. Jørgensen L, Crabtree NJ, Reeve J, Jacobsen BK. Ambulatory level and asymmetrical weight bearing after stroke affects bone loss in the upper and lower part of the femoral neck differently: bone adaption after decreased mechanical loading. Bone 2000;27(5):701-707.
  7. Pang MYC, Ashe MC, Eng JJ. Muscle weakness, spasticity and disuse contribute to demineralization and geometric changes in the radius following chronic stroke. Osteoporos Int 2007;18(9):1243-1252.
  8. Sato Y, Kuno H, Kaji M, et al. Increased bone resorption during the first year after stroke. Stroke 1998;29(7):1373-1377.
  9. Worthen LC, Kim CM, Kautz SA, et al. Key characteristics of walking correlate with bone density in individuals with chronic stroke. J Rehabil Res Dev 2005;42(6):761-768.
  10. Yavuzer G, Ataman S, Süldür N, Atay M. Bone mineral density in patients with stroke. Int J Rehabil Res 2002;25(3):235-239.
  11. Sato Y, Metoki N, Iwamoto J, Satoh K. Amelioration of osteoporosis and hypovitaminosis D by sunlight exposure in stroke patients. Neurology 2003;61(3):338-342.
  12. Sato Y, Fujimatsu Y, Honda Y, et al. Accelerated bone remodeling in patients with poststroke hemiplegia. J Stroke Cerebrovasc Dis 1998;7(1):58-62.
  13. Dennis MS, Lo KM, McDowall M, West T. Fractures after stroke: frequency, types and associations. Stroke 2002;33(3):728-734.
  14. Pouwels S, Lalmohamed A, Leufkens B, et al. Risk of hip/femur fracture after stroke: A population-based case-control study. Stroke 2009;40(10):3281-3285.
  15. Bainbridge NJ, Davie MW, Haddaway MJ. Bone loss after stroke over 52 weeks at os calcis: influence of sex, mobility and relation to bone density at other sites. Age Ageing 2006;35(2):127-132.
  16. Pang MYC, Eng JJ, McKay HA, Dawson AS. Reduced hip bone mineral density is related to physical fitness and leg lean mass in ambulatory individuals with chronic stroke. Osteoporos Int 2005;16(12):1769-1779.
  17. Haddaway MJ, Bainbridge NJ, Powell DE, Davie MW. Bone resorption in stroke and institutionalized subjects. Calcif Tissue Int 2009;84(2):118-125.
  18. Jørgensen L, Jacobsen BK, Wilsgaard T, Magnus JH. Walking after stroke: does it matter? Changes in bone mineral density within the first 12 months after stroke. A longitudinal study. Osteoporos Int 2000;11(5):381-387.
  19. Jørgensen L, Jacobsen BK. Changes in muscle mass, fat mass, and bone mineral content in the legs after stroke: a 1 year prospective study. Bone 2001;28(6):655-659.
  20. Lazoura O, Groumas N, Antoniadou E, et al. Bone mineral density alterations in the upper and lower extermities 12 months after stroke measured by peripheral quantitative computed tomograpghy and DXA. J Clin Densitom 2008;11(4):511-517.
  21. Liu M, Tsuji T, Higuchi Y, et al. Osteoporosis in hemiplegic stroke patients as studied with dual-energy x-ray absorptiometry. Arch Phys Med Rehabil 1999;80(10):1219-1226.
  22. Mussolino ME, Madans JH, Gillum RF. Bone mineral denisty and stroke. Stroke 2003;34(5):e20-e22.
  23. Paker N, Bugdayc D, Tekdos D, et al. Relationship between bone turnover and bone density at the proximal femur in stroke patients. J Stroke Cerebrovasc Dis 2009;18(2):139-143.
  24. Şahin L, Özoran K, Gündüz OH, et al. Bone mineral density in patients with stroke. Am J Phys Med Rehabil 2001;80(8):592-596.
  25. Sherk KA, Sherk VD, Anderson MA, et al. Differences in tibia morphology between the sound and affected sides in ankle-foot orthosis-using survivors of stroke. Arch Phys Med Rehabil 2012 Oct 30. [Epub ahead of print]
  26. Talla R, Galea M, Lythgo N, et al. Contralateral comparison of bone geometery, BMD and muscle function in the lower leg and forearm after stroke. J Musculoskelet Neuronal Interact 2011;11(4):306-313.
  27. Pang MY, Ashe MC, Eng JJ. Tibial bone geometry in chronic stroke patients: influence of sex, cardiovascular health, and muscle mass. J Bone Miner Res 2008;23(7):1023-1030.
  28. Pang MY, Ashe MC, Eng JJ. ­­­ Compromised bone strength index in the hemiparetic distal tibia epiphysis among chronic stroke patients: the association with cardiovascular function, muscle atrophy, mobility, and spasticity. Osteoporos Int 2010;21(6):997-1007.

Leave a Reply

Your email address will not be published.

This site uses Akismet to reduce spam. Learn how your comment data is processed.