Cortical and Trabecular Bone Mineral Loss From the Spine and Hip in Long-Duration Spaceflight


  • The authors have no conflict of interest


We measured cortical and trabecular bone loss using QCT of the spine and hip in 14 crewmembers making 4- to 6-month flights on the International Space Station. There was no compartment-specific loss of bone in the spine. Cortical bone mineral loss in the hip occurred primarily by endocortical thinning.

Introduction: In an earlier study, areal BMD (aBMD) measurements by DXA showed that cosmonauts making flights of 4- to 12-month duration on the Soviet/Russian MIR spacecraft lost bone at an average rate of 1%/month from the spine and 1.5%/month from the hip. However, because DXA measurements represent the sum of the cortical and trabecular compartments, there is no direct information on how these bone envelopes are affected by spaceflight.

Materials and Methods: To address this, we performed a study of crewmembers (13 males and 1 female; age range, 40-55 years) on long-duration missions (4-6 months) on the International Space Station (ISS). We used DXA to obtain aBMD of the hip and spine and volumetric QCT (vQCT) to assess integral, cortical, and trabecular volumetric BMD (vBMD) in the hip and spine. In the heel, DXA was used to measure aBMD, and quantitative ultrasound (QUS) was used to measure speed of sound (SOS) and broadband ultrasound attenuation (BUA).

Results and Conclusions: aBMD was lost at rates of 0.9%/month at the spine (p < 0.001) and 1.4-1.5%/month at the hip (p < 0.001). Spinal integral vBMD was lost at a rate of 0.9%/month (p < 0.001), and trabecular vBMD was lost at 0.7%/month (p < 0.05). In contrast to earlier reports, these changes were generalized across the vertebrae and not focused in the posterior elements. In the hip, integral, cortical, and trabecular vBMD was lost at rates of 1.2-1.5%/month (p < 0.0001), 0.4-0.5%/month (p < 0.01), and 2.2-2.7%/month (p < 0.001), respectively. The cortical bone loss in the hip occurred primarily by cortical thinning. Calcaneal aBMD measurements by DXA showed smaller mean losses (0.4%/month) than hip or spine measurements, with SOS and BUA showing no change. In summary, our results show that ISS crewmembers, on average, experience substantial loss of both trabecular and cortical bone in the hip and somewhat smaller losses in the spine. These results do not support the use of calcaneal aBMD or QUS measurements as surrogate measures to estimate changes in the central skeleton.


LOSS OF SKELETAL mineral is an important medical complication of long-duration spaceflight. Areal BMD (aBMD) measurements by DXA showed in an earlier study that, despite prescription of an intense exercise regimen designed to maintain mechanical loading of the skeleton, crewmembers on the Soviet/Russian MIR space station lost aBMD at an average monthly rate 0.3% from the total skeleton, with 97% of that loss from the pelvis and legs.(1) Regional DXA aBMD measurements showed a loss of 1.06%/month from the spine and 1.15-1.56%/month from the hip.(2,3) These measurements varied extensively between crewmembers; some individuals incurred losses equivalent to one-half the bone mineral they would lose in a lifetime of normal aging, whereas others showed minimal to no change.(2) Although the bone loss may not be sufficient to increase the risk of fracture immediately on return, it may result in increased fracture risk at more advanced ages because of onset of age-related osteoporosis.

While previous studies have defined the magnitude of total, or integral, bone loss at several skeletal sites, there is little information regarding the spatial distribution of these changes within specific skeletal sites. Bone structure is adapted to provide maximum strength with respect to the spectrum of mechanical loads sustained during normal human activities. Because these loads change in extended weightlessness, both the total mass and the spatial distribution of bone mineral may change as well, with different rates of cortical and trabecular bone loss and variation between anatomic subregions of the same bone. Because bone strength depends on loading directions, material properties, and bone geometry, a redistribution of bone mineral may have a potential impact on bone strength at the conclusion of the flight. Such spatial heterogeneity has been observed in the tibias and femorae of rats undergoing hindlimb unloading,(4) and Vico et al.(5) observed differential cortical and trabecular tibial bone loss in MIR cosmonauts. Additionally, data gathered by Oganov et al.(6) from a small group of cosmonauts who flew on the Salyut missions indicated accelerated loss of spinal bone from the posterior elements of the vertebral body, which serve as muscle attachment points and protection for the spinal cord. However, aside from the limited Oganov et al.(6) data, there has been no systematic investigation using modern bone densitometry procedures of subregional effects of spaceflight in the vertebrae and hip, the skeletal sites associated with the most serious osteoporotic fractures in the elderly and presumably of greatest relevance to the future health of crewmembers.

To address these issues, we carried out a longitudinal study of BMD in crewmembers making 4- to 6-month flights on the International Space Station (ISS), performing pre- and postflight measurements. The principal goal of the study was to characterize the subregional distribution of vertebral and proximal femoral bone loss. To accomplish this, we used vQCT, which provides a series of cross-sectional images at the L1 and L2 vertebrae and the hip. These images can be calibrated to measure trabecular and cortical volumetric BMD (vBMD), BMC, and bone volume. Another goal of the study was to compare aBMD of the spine, hip, and calcaneus and calcaneal QUS to historical data from the MIR and other long-duration missions.


Study subjects

We performed pre- and postflight measurements in 14 subjects (average age, 44.6 ± 4.0 years; 13 males and 1 female) who were crewmembers of the second to the sixth ISS missions. The number of subjects studied per mission and the mission lengths are summarized in Table 1. All participants were informed of the study and provided informed consent according to procedures mandated by the Internal Review Boards of the collaborating institutions (NASA Johnson Space Center, Russian Space Agency, Baylor College of Medicine, and University of California San Francisco).

Table Table 1.. Number of Subjects per Mission and Mission Length
original image


The ISS crewmembers underwent pre- and postflight DXA scanning of the hip, lumbar spine, and heel (fast array fanbeam, QDR 4500; Hologic, Waltham, MA, USA) as part of the normal medical monitoring schedule at NASA Johnson Space Center. Preflight procedures were usually performed an average of 30-60 days before launch, and postflight measurements were normally performed within 7-10 days of landing. Within a day of the DXA procedures, the subjects were transported to the Department of Radiology at Methodist Hospital, Baylor College of Medicine, where they underwent vQCT of the hip and spine and QUS of the heel (Sahara; Hologic). Helical CT images (GE Hispeed Advantage; GE Medical Systems, Milwaukee, WI, USA) were acquired at the L1 and L2 vertebrae and the hips (3-mm-thick sections; 80 kVp; 280 mA). To relate the CT image units to equivalent concentration of calcium hydroxyapatite and to provide simultaneous calibration, a bone mineral reference standard (four-sample calibration phantom; Image Analysis, Columbia, KY, USA) was placed under the participants' spine and hips and scanned simultaneously. QUS measurements were performed according to the manufacturer's specifications and included speed of sound (SOS) and broadband ultrasound attenuation (BUA).

vQCT mineral analysis

CT images were transferred to a computer workstation and processed to extract measures of vBMD and bone size using analysis techniques described previously.(7,8) The processing task included calibration of the CT images from the native scanner Hounsfield Units to equivalent concentration (g/cm3) of calcium hydroxyapatite (HA) and determination of trabecular, cortical, and integral regions of interest from vQCT scans of the spine (Fig. 1A) and proximal femur (Figs. 1B and 1C). For each region, we computed vBMD (g/cm3), BMC (g), and bone volume (cm3). The spinal trabecular region of interest (Fig. 1A2) was a semicircular region encompassing the anterior vertebral body, centered on the midvertebral level but encompassing 70% of the volume between the vertebral endplates. The spinal integral regions of interest included a region encompassing the total vertebra except for the transverse processes (Fig. 1A3), a region containing the vertebral centrum (Fig. 1A1), and a region containing the posterior elements, which was obtained by subtracting region A1 from region A3. The proximal femoral regions of interest included volumes of trabecular, cortical, and integral bone in the femoral neck in a region encompassing the greater and lesser trochanters and in an overall region comprising both the femoral neck and trochanter. The trochanteric region was computed by subtracting the femoral neck from the overall proximal femoral region. Trabecular bone regions (Figs. 1B3 and 1C3) were determined by eroding the integral bone regions to produce regions of the same shape but fully contained within the medullary volumes. Regions of cortical bone (Figs. 1B1 and 1C1) were determined by applying a threshold of 0.35 g/cm3 to the voxels falling outside the trabecular regions but within the integral regions (Figs. 1B2 and 1C2). The trabecular and cortical vBMD values represent the equivalent HA concentration averaged over the voxels contained in the different trabecular and cortical regions of interest. Because the spatial resolution of the CT system is larger than the thickness of a trabeculum or a very thin cortex (such as the vertebral cortex or antero- and superomedial femoral neck cortices), the cortical and trabecular regions contain nonbone components. Thus, the cortical and trabecular vBMD values should be interpreted as equivalent cortical and medullary tissue concentrations of HA, respectively.

Figure FIG. 1..

vQCT regions of interest in vertebra and hip. Regions of interest are white pixels superimposed on image data. (A) Vertebral regions. Lateral views of these regions are in the top row and axial views are in the bottom row: (A1) vertebral centrum integral, (A2) vertebral trabecular, and (A3) total vertebral integral. A posterior elements region was comprised of bone within the total vertebral integral region (A3) but outside the vertebral centrum integral region (A1). (B) Overall proximal femoral regions of interest: (B1) cortical, (B2) integral, and (B3) trabecular. (C) Femoral neck regions of interest: (C1) cortical, (C2) integral, and (C3) trabecular. The trochanteric region is comprised of the bone outside of the femoral neck region of interest but within the overall proximal femoral region.

Figure FIG. 2..

Definitions of proximal femoral and vertebral planes for geometric measurements and strength estimates. (A) Proximal femur: middle image is coronal projection through hip with arrows pointing to (left) femoral neck slice of minimum cross-sectional area (fnCSA) and (right) trochanteric slice of maximum cross-sectional area (trochCSA). (B) Left image is sagittal view showing location from which vertebral midplane is reconstructed (right). Region of interest used for mvCSA and mvCSI is outlined in black.

vQCT geometric and strength analysis

We measured the cross-sectional areas (CSA) of the femoral neck, mid-trochanter, and mid-vertebrae (Fig. 2). To determine whether periosteal expansion occurred, the volumes of the femoral neck, trochanteric, and total femur integral regions of interest, which included all voxels contained within the outer bone margin, were used as measures of total bone volume. The volumes of femoral neck, trochanteric, and total femur cortical regions of interest were used as measures of cortical bone volume. As estimates of periosteal area, we computed the positions along the femoral neck axis of minimum and maximum CSA, which corresponded to the smallest CSA of the femoral neck (fnCSA) and the largest CSA through the trochanteric region (trochCSA), respectively (Fig. 2A). To estimate the effect of the spaceflight on mechanical competence of the femoral neck, we calculated indices of bending(9) and compressive strength from the scans. To calculate the femoral neck bending strength index (fnBSI), we reconstructed from the volumetric scan a 2-mm-thick section centered at the fnCSA position, and for each voxel within the outer bone boundary, computed an equivalent elastic modulus (ei) using parametric relationships applied by Keyak et al.(10) to finite element modeling of the hip. The strength index was computed as an effective polar moment of inertia (Ix + Iy) divided by a measure of the femoral neck width (W):

equation image

Ix and Iy are the effective moments of inertia computed along the principal axes of the cross-section and are normalized by eb, the elastic modulus of cortical bone. (x,y) are the elastic modulus weighted centroid of the cross-section. W is the diameter of a circular cross-section of equivalent area. The compressive strength indices (CSI) for the femoral neck and mid-vertebra were performed according to methods described by Sievanen.(11) The compressive strength index of the femoral neck (fnCSI) was computed as:

equation image

where fnvBMD is the integral BMD of the femoral neck region. The vertebral compressive strength index (mvCSI) was computed from the cross-sectional area (mvCSA) and integral BMD of the mid-vertebral region (mvvBMD):

equation image

Data analysis

For each subject, we computed absolute and percentage differences between the pre- and postflight measurements and normalized these differences by the length of the mission in months. All calculations were performed using SAS Version 8 (SAS Institute, Cary, NC, USA). Statistical significance of the changes between the bone measurements pre- and postflight was determined using the Wilcoxson signed-rank test. Random effects models were used to compare cortical, trabecular, and integral BMD changes for spine and hip. In these analyses, the rank of the percentage change per month was used as the dependent variable, and bone compartments and measured sites were the independent variables, with subject identification as a random effect to control for multiple bone measurements from the same subjects. Significance levels for pairwise comparisons between bone types and measurement sites were adjusted by the Tukey-Kramer procedure. Correlations between changes measured at different skeletal sites were quantified using the Pearson correlation coefficient.


Spine measurements

Results of spine measurements are summarized in Table 2. aBMD and integral vBMD were lost at an average rate of 0.8%/month and 0.9%/month, respectively. Interestingly, the rate of trabecular bone loss tended to be smaller (0.7%/month) than any of the integral regions, although these rates of loss were not significantly different (p > 0.05). In contrast to the earlier findings by Oganov et al.,(6) the changes observed in the posterior element regions were identical to those observed in the vertebral body region. One-half of the subjects experienced losses of posterior element vBMD exceeding 8% over the course of the mission, but these changes were accompanied by equivalent losses in the vertebral body. No statistically significant correlations were observed between changes measured by DXA or QCT at the spine and those measured by DXA or QCT at other skeletal sites.

Table Table 2.. Spinal Bone Mineral Changes for DXA and vQCT
original image

Hip BMD measurements

DXA femoral neck and total femur regions measurements showed overall losses of 1.4%/month and 1.5%/month, respectively, similar to mean values for the MIR crewmembers.(2,3) vQCT results, which are displayed in Table 3, showed differing patterns of change for integral, cortical, and trabecular bone. Integral vBMD and BMC from vQCT were lost at rates similar to the DXA measures (1.2-1.6%/month). In both percentage and absolute units, cortical BMC showed statistically significant declines in the femoral neck, trochanteric, and total femur regions (1.6-1.7%/month). Cortical vBMD showed a smaller average decline of 0.3-0.4%/month. Although the amount of trabecular bone loss was much smaller than cortical loss in absolute units, trabecular BMC and vBMD tended to show the largest percentage losses uniformly across femur regions. The Tukey-Kramer's adjusted p value for difference between trabecular and cortical bone loss, and between trabecular and integral bone loss, was less than 0.0001. No statistically significant correlations were observed between changes measured by DXA or QCT at the hip and those measured by DXA or vQCT at other skeletal sites.

Table Table 3.. Preflight and Postflight Values as Well as Changes in BMC and vBMD in Integral, Cortical, and Trabecular Components of Femoral Neck, Trochanteric, and Total Femur Hip Regions as Measured by vQCT
original image

Geometric measures and strength estimates

Changes in geometric parameters and strength estimates are shown in Table 4. Vertebral cross-sectional area was unaffected by the flight, although mvCSI, which scales with the square of integral vBMD, declined significantly. The total bone volumes in the femoral neck, trochanteric, and total femur regions did not change, nor did fnCSA or trochCSA. However, the volume of cortical bone in all of the proximal femoral regions showed statistically significant declines in the femoral neck and total femur regions. fnBSI and fnCSI declined significantly.

Table Table 4.. Table of Bone Geometry Measures and Strength Indices, Containing Preflight Values, Postflight Values, and Percentage Changes per Month (SD)
original image

Calcaneal measurements

Calcaneal DXA and ultrasound results are shown in Table 5. Calcaneal aBMD showed a statistically significant decline (p < 0.01) at 0.4%/month. These changes were not correlated to changes measured by DXA or vQCT at the spine or hip. Neither QUS-SOS nor QUS-BUA showed statistically significant changes. No correlations were observed between changes in either ultrasound parameter and changes in calcaneal aBMD.

Table Table 5.. Calcaneal DXA and QUS Measurements Containing Preflight Values, Postflight Values, and Percentage Changes per Month (SD)
original image


As measured with DXA, ISS crewmembers members lost bone from the lumbar spine and hip at rates of 0.8-0.9%/month and 1.2-1.5%/month, respectively. These values are similar to previous observations of the MIR cosmonauts. Thus, the array of musculoskeletal conditioning exercises, as they are currently implemented on the ISS, may be insufficient to prevent bone mineral loss from the central skeleton in long-duration spaceflight.

In our spinal vQCT measurements, we observed comparable rates of bone loss between the vertebral body integral region of interest and the posterior element region. In a spinal CT study of four cosmonauts making flights of up to 8 months on the Salyut Spacecraft, Oganov et al.(6) reported minimal loss of vertebral body trabecular vBMD, but losses on the order of 8% from the posterior elements, with up to 4% declines in the volume of the adjoining paraspinal muscles. In this study, we did not observe a focal loss of bone in the posterior elements; those subjects losing bone in the posterior elements had similar changes in the vertebral body. The absence of focal posterior element bone loss in our ISS subjects is potentially caused by differences in physical activity between the Salyut and ISS crewmembers. The latter have access to a newly developed exercise device designed to preserve the strength of the lower back muscles. Such a device would tend to maintain muscle loading on the posterior elements.

Consistent with the MIR studies and bedrest studies, our DXA data showed the largest losses, 1.2-1.5% per month, in the hip. In terms of mass, over 90% of the mineral loss was from cortical bone, which declined at a rate of 1.6-1.7% per month. Trabecular vBMD was lost at a rate of 2.2-2.7% per month. Although the loss of femoral trabecular bone density was small in absolute terms, it was larger in percentage terms than the loss of integral or cortical bone, tending to be largest in the femoral neck. The higher percentage loss of trabecular bone in the proximal femur is consistent with previous findings in spaceflight,(5) in hindlimb unloading,(4) and in spinal cord injury,(12) which show accelerated loss of trabecular bone at distal femoral and tibial sites.

In addition to losses of trabecular and cortical BMD and BMC, we also observed statistically significant declines in the volumes of the cortical regions of interest at the femoral neck, trochanter, and total femur (1.2-1.3%/month, p < 0.05). Although the cortical measures declined, the parameters describing the outer bone perimeter (volumes of integral bone regions, fnCSA and trochCSA) remained unchanged, suggesting that the cortical bone loss occurred along the endosteal margin without periosteal apposition. In recent cross-sectional studies of aging using DXA imaging in conjunction with special analysis techniques, Beck et al.(13) noted that periosteal apposition protects the strength of the femoral neck against bone mineral loss by increasing the cross-sectional moment of inertia. In our study, the loss of cortical bone volume, in the absence of periosteal apposition, was consistent with the observed (2.55%/month) reduction in fnBSI.

Our results do not support the use of heel measurements, particularly calcaneal QUS, to act as surrogates for central bone mass in long-duration spaceflight. As measured by DXA, the amount of bone loss from the heel was strikingly smaller (0.4%/month versus 0.8-1.5%/month) than that observed from the central skeleton. The lower responsiveness of the calcaneal DXA measurements compared with the central DXA measurements are consistent with a previous report(3) and may reflect the fact that the heel is subject to ground reaction forces during in-flight treadmill exercises, which are highly attenuated at the central skeletal sites such as the spine and hip. Interestingly, the rate of bone loss from the heel in this study tended to be smaller than the loss observed in bedrest studies, where treadmill countermeasures have not been used.(14) No QUS parameter changed significantly over the mission, a finding inconsistent with a previously published bedrest study.(15) The poor responsiveness of QUS in spaceflight is at variance with results of ground-based studies(16) and may reflect changes in the calcaneal soft tissue caused by sustained weightlessness as well as the potential protective effect of treadmill exercise against calcaneal bone loss.

This study has both fundamental strengths and weaknesses. It is the first systematic study of bone mineral loss in spaceflight to use vQCT to map subregional bone loss at the two most critical skeletal sites: the spine and hip. It is the first such study to show differential patterns of cortical and trabecular bone mineral loss at the hip and to combine these measures to estimate changes in indices of femoral neck strength. The weaknesses of our study include the small sample size (inherent to almost all human spaceflight studies) and the effect of partial volume averaging on our cortical bone measurements. Partial volume averaging errors affect measurement of structures that are small compared with the spatial resolution of the imaging device. For thin cortices such as the superomedial aspect of the femoral neck, this results in underestimation of BMD and overestimation of volume or thickness measurements, with smaller errors for estimation of mass.(17) Our hip cortical region of interest is primarily composed of the inferomedial cortex of the proximal femur, for which thickness values exceeding 3-4 mm have been reported in humans, but also includes the thin superomedial cortex (thickness ≈0.3 mm).(18) In a study based on a helical CT scanner similar to our own, Prevrhal et al.(19) reported that detection of changes in cortical thickness is possible for cortices with thicknesses >1.2 mm. Thus, it is likely that the cortical volume changes observed in this study primarily reflected changes occurring in the thick inferomedial cortex. Another limitation in our study is that the various bone strength indices have not been validated as predictors of fracture risk and should not be taken as such. In this study, we have used these indices as interpretive tools to relate changes in bone density and geometry to estimate the potential impact of the spaceflight on whole bone strength in bending and compression.

In conclusion, our study show that bone mineral loss, especially from the central skeleton, occurs in the ISS crewmembers at a rate comparable with that of previous MIR crewmembers. In the hip, the changes in cortical and trabecular bone mass, when combined in the form of compressive and bending strength indices, predict that long-duration spaceflight may result in a substantial reduction of bone strength. This underlines the importance of continuing to improve countermeasures to preserve musculoskeletal conditioning during long-duration space missions.


The authors acknowledge the organizational and logistic assistance of Karin Bergh, David Baumann, Christine Miles, Dr Lakshmi Putcha, Dr Todd Schlegel, and Dr Michael Greenisen. The authors thank Terry Schuelke and Dr Nahid Rianon for excellent technical assistance. This study was supported by NASA Contract NAS-9-99055 from NASA Johnson Spaceflight Center.