Results of lumbosacral distraction spondylodesis for the treatment of spondylolisthesis, failed-back syndrome, and lumbar instability

Various methods of lumbosacral fusions for the treatment of degenerative spinal diseases are used clinically. Results vary greatly depending on indication, type of fusion, implants, and method of evaluation. In a retrospective clinical and radiological examination after an average follow-up time of 3.9 years this study reports on the outcome of lumbosacral distraction spondylodesis (LSDS) in a consecutive series of 147 patients being fused for the treatment of spondylolisthesis, failed-back syndrome, or lumbar instability. LSDS consists of a posterolateral fusion together with an autologous corticocancellous H-graft wedged under distraction between the spinous processes of L4 and S1. With 81.0% good and excellent results this noninstrumented fusion technique showed the best outcome in patients with spondylolisthesis, while in cases with a failed-back syndrome or lumbar instability only 62.3% excellent to satisfying outcomes were noted. The rate of pseudarthrosis was 13.6% in the whole patient group; no major complications such as nerve root damage, postoperative neurological deficits, or spinal stenosis were found.

Morphometric parameters of the radial neck: an anatomical study

INTRODUCTION

The purpose of this study was to describe the anatomy of the intramedullary canal of the radial neck for stem design of radial head prostheses.

METHODS

Radial neck measurement was performed from the proximal edge of the radial tuberosity to the distal border of the articular surface on 40 macerated proximal radii using X-rays and Optosil imprints of the intramedullary canal. Radial neck to shaft angle was measured under image intensification.

RESULTS

The neck length was 13.3 mm (range 8.0-20.0 mm) measured on X-rays and 13.5 mm (range 9.9-18.8 mm) on Optosil imprints. Concerning the neck's intramedullary canal we obtained a minimum diameter of 9.7 mm (range 6.3-16.4 mm) on X-ray. Corresponding Optosil imprints had a mean diameter of 11.6 mm (range 6.6-18.6 mm) at their proximal end, 10.5 mm (range 5.5-16.3 mm) in the middle and a minimum of 9.8 mm (range 6.0-16.9 mm) at the distal end. Differences between left and right sides were not significant between X-ray and Optosil imprints, nor between left X-ray and right imprints and vice versa at a P-level of 0.001. Radial neck to shaft angle was 167.8 degrees (range 160.5-178 degrees ) measured under image intensification at a rotation angle of 58.6 degrees (range 50-70 degrees ) supination.

CONCLUSION

Geometry of the radial neck's intramedullary canal has been described for stem design. The minimum diameter was seen at the distal end, just proximal to the radial tuberosity. A forearm rotation of 58.6 degrees to full supination should be considered during implantation of monopolar prostheses. The wide variety of lengths and diameters complicates standard prosthetic design, but ipsilateral and contralateral X-rays can be useful for preoperative stem size calculation.

Metastasis to the pancreas--an indication for pancreatic resection?

BACKGROUND AND AIMS

Isolated metastatic involvement of the pancreas is very rare. To evaluate the possible benefit of surgery, we retrospectively analyzed patients that underwent pancreatic resection for metastases into the pancreas.

PATIENTS/METHODS

In 12 patients (8 men and 4 women), metastatic disease was treated by pancreatic resection (two total pancreatectomies, nine pylorus-sparing duodenopancreatectomies, and one left-side pancreatic resection) between 1993 and 2005 at our institution. Primary malignomas were renal cell carcinoma (RCC; n = 7), malignant melanoma (n = 4), and colon cancer (n = 1). All patients were followed-up until November 2006 or until death.

RESULTS

Complications requiring relaparotomy were found in two patients (retroperitoneal abscess and bile fistula), whereas one patient with pancreatic fistula could be treated by conservative measures. There was no perioperative mortality. Median survival time was 51 months (5-105 months). At the end of follow-up, seven patients were alive at 12 to 86 months, whereas five died between 5 and 105 months: four died of the disease, and one patient died of cardiac failure without evidence of recurrent cancer.

CONCLUSION

Patients with isolated pancreatic metastasis particularly of RCC benefit from surgery. Pancreatic resection may achieve long-term survival or good palliation in selected cases of other primaries as well.

Age-related changes in cortical bone mass: data from a German female cohort

To describe data from digital radiogrammetry (DXR) in an unselected German female cohort over a wide age range. Using a retrospective study design we analyzed radiographs of the hand from 540 German women (aged 5-96 years) using an automated assessment of cortical thickness, metacarpal index (MCI), and estimated cortical bone mineral density (DXR-BMD) on digitized radiographs. Both hands were radiographed in 97 women. In this group DXR-BMD and cortical thickness were significantly higher in the right metacarpals while there was no significant difference in MCI. To study the association with age we differentiated young (<20 years), middle-aged (20-45 years), and an older patients (>45 years). In young women all parameters increased significantly with age in a linear fashion (r=0.8 for DXR-BMD, r=0.7 for MCI). In those aged 25-45 years DXR-BMD and MCI were highest (peak bone mass). In women aged 45 or older all parameters decreased with age in an almost linear fashion with an annual change ranging from 0.7% to 0.9%. Our results for an unselected German female cohort indicate that DXR is a reliable, widely available osteodensitometric technique based on the refinement of conventional radiogrammetry. These findings are comparable to those from other studies and represent a valid resource for clinical application and for comparisons with other ethnic groups.

Assessment of prevalent and incident vertebral fractures in osteoporosis research

Vertebral fractures are the hallmark of osteoporosis, and it has been shown that the presence of vertebral fractures adds to the risk of future osteoporotic fractures. Thus, the evaluation of spinal radiographs for prevalent or incident vertebral fractures is important in the clinical evaluation of patients with osteoporosis, in clinical drug trials for osteoporosis treatment and in the epidemiologic evaluation of populations at risk for osteoporosis. Traditionally, conventional lateral radiographs of the thoracolumbar spine have been analyzed visually by radiologists or experienced clinicians to identify vertebral fractures in patients with clinical indications. Epidemiologic studies or clinical drug trials in osteoporosis research have considerably different requirements and expectations. To reduce potential subjectivity in qualitative readings and to provide definable, reproducible and objective methods to detect vertebral fracture, and to accommodate the assessment of large numbers of radiographs, various morphometric approaches have been explored and employed. However, potential deficiencies in morphometry have led to a renewed interest in standardized qualitative visual approaches for defining vertebral fractures and visual criteria. Numerous studies have shown that semiquantitative interpretation, after careful centralized training and standardization, can produce results with excellent intra- and interobserver reproducibility. We firmly believe that the experience from several studies has shown that there is a major role for radiologists and clinicians alike to carefully assess and diagnose vertebral fractures using standardized grading schemes such as the one proposed in this review. In the context of epidemiologic studies and clinical drug trials, quantitative morphometry may be used; however, the studies would be flawed if quantitative morphometry were to be performed in isolation without additional adjudication by a trained and highly experienced radiologist or clinician.

[Lumbar intraspinal juxtafacet cysts: MR imaging and CT-arthrography]

PURPOSE

To present data on the MR imaging appearance of lumbar intraspinal juxtafacet cysts (JFC) and to assess the importance of additional CT arthrography.

MATERIAL AND METHODS

Twenty-eight patients (16 women, 12 men) with a mean age of 64 years (range: 43 - 82), who underwent MR imaging because of radicular pain or spinal claudication, were found to have an intraspinal cyst associated with the facet joint. In 14 patients, additional CT-arthrography was performed to determine whether a communication exists between the cyst and the facet joint and to try to rupture the cyst.

RESULTS

In T(2)-weighted images, juxtafacet cysts show a typical pattern consisting of a hyperintense center and hypointense rim. The center is likely to be inhomogeneous because of recurrent hemorrhage in the cyst. In T1-weighted images, the cysts are hypo/isointense. Irregular hyperintensity may indicate subacute hemorrhage, which may aggravate the clinical symptoms. MR allows superior visualization of the cyst in all anatomical planes. It also enables assessment of typical accompanying changes, such as degenerative spondylolisthesis and facet hypertrophy. All patients, who had CT-arthrography, were found to have a direct communication between joint space and cyst. Transarticular rupture of the cyst was possible in five patients. Two of these five patients had good to excellent improvement, and the remaining three patients underwent surgery.

CONCLUSION

MR imaging is the method of choice for diagnosing lumbar intraspinal juxtafacet cysts. CT-arthrography of the facet joint is helpful in cases with difficult differential diagnosis, and in the preoperative planning. Furthermore, it assists in the primary interventional treatment.

MRI of bone marrow in the distal radius: in vivo precision of effective transverse relaxation times

The effective transverse relaxation time T2* is influenced by the presence of trabecular bone, and can potentially provide a measure of bone density as well as bone structure. We determined the in vivo precision of T2* in repeated bone marrow measurements. The T2* measurements of the bone marrow of the distal radius were performed twice within 2 weeks in six healthy young volunteers using a modified water-presaturated 3D Gradient-Recalled Acquisition at Steady State (GRASS) sequence with TE 7, 10, 12, 20, and 30; TR 67; flip angle (FA) 90 degrees. An axial volume covering a length of 5.6 cm in the distal radius was measured. Regions of interest (ROIs) were determined manually and consisted of the entire trabecular bone cross-section extending proximally from the radial subchondral endplate. Reproducibility of T2* and area measurements was expressed as the absolute precision error (standard deviation [SD] in ms or mm2) or as the relative precision error (SD/mean x 100, or coefficient of variation [CV] in %) between the two-point measurements. Short-term precision of T2* and area measurements varied depending on section thickness and location of the ROI in the distal radius. Absolute precision errors for T2* times were between 1.3 and 2.9 ms (relative precision errors 3.8-9.5 %) and for area measurements between 20 and 55 mm2 (relative precision errors 5.1-16.4%). This MR technique for quantitative assessment of trabecular bone density showed reasonable reproducibility in vivo and is a promising future tool for the assessment of osteoporosis.

Basal ganglia alterations and brain atrophy in Huntington's disease depicted by transcranial real time sonography

OBJECTIVES AND METHODS

Transcranial real time sonography (TCS) was applied to 49 patients with Huntington's disease and 39 control subjects to visualise alterations in the echotexture of the basal ganglia. For comparison T1 weighted, T2 weighted, and fast spin echo MRI was performed in 12 patients with Huntington's disease with and in nine patients without alterations of the basal ganglia echotexture as detected by TCS and T1 weighted, T2 weighted, and fast spin echo MRI. Furthermore, the widths of the frontal horns, third ventricle, and the lateral ventricles were depicted in TCS examinations and correlations examined with corresponding CT slices.

RESULTS

Eighteen out of 45 (40%) of the patients with Huntington's disease with adequate insonation conditions showed hyperechogenic lesions of at least one basal ganglia region. In 12 patients TCS depicted hyperechogenic lesions of the substantia nigra; in six patients the head of the caudate nucleus was affected. The lentiform nucleus (n=3) and the thalamus (n=0) were less often affected or spared. Hyperechogenic lesions were significantly more frequent in patients with Huntington's disease than in 39 control subjects, who had alterations of the echotexture in 12.8% (4/39) of the examinations. The number of CAG repeats and the clinical status correlated with the identification of hyperechogenic lesions of the substantia nigra (p<0.01). Hyperechogenic lesions of the caudate nucleus were associated with an increased signal intensity in T2 weighted MR images (p<0.05). All TCS parameters indicating brain atrophy correlated with CT findings (p<0.0001).

CONCLUSIONS

TCS detects primarily abnormalities of the caudate nucleus and substantia nigra in Huntington's disease. These changes in the echotexture may represent degenerative changes in the basal ganglia matrix and are partially associated with CAG repeat expansion and the severity of clinical findings.

[Conventional radiology of osteoporosis and radiographic absorptiometry]

Radiographic findings suggestive of osteopenia and osteoporosis are frequently encountered in daily medical practice and can result from a wide spectrum of diseases ranging from highly prevalent causes such as postmenopausal and involutional osteoporosis to very rare endocrinologic and hereditary or acquired disorders. As bone mass is lost, changes in the trabecular and cortical bone structure occur. The changes that are visible on conventional radiographs may give specific information on the etiology of the underlying disease. Even though conventional radiography is indispensable in the differential diagnosis of osteopenia and osteoporosis, visual assessment of bone mass and its changes is subjective and strongly depends on the radiographic technique. With radiogrammetry and radiographic absorptiometry two methods exist that rely on conventional radiography and allow for a quantification of bone mass.

Spinal and femoral DXA for the assessment of spinal osteoporosis

The objective was to determine the diagnostic sensitivity of spinal and femoral dual x-ray absorptiometry (DXA) and to study whether a combination of both sites may enhance discriminatory capability in regard to the presence of vertebral fractures. Spinal and femoral DXA were obtained in 324 postmenopausal women, of whom 90 had at least one vertebral fracture. Age-adjusted logistic regression analyses, ROC analyses, and sensitivity-specificity statistics were used to assess the discriminatory ability of spinal and femoral bone density (BMD) alone and in combination. The age-adjusted odds ratios per standard deviation decrease in BMD (OR) for spinal and femoral measurements were comparable (Ward's triangle: OR = 1.62; femoral neck: OR = 1.51; total hip: OR = 1.47; spine: OR = 1.34). Combining spinal and femoral bone density measurements did not improve diagnostic sensitivity of DXA considerably as compared to using BMD of a single site and adjusting the "fracture threshold." The conclusion drawn is that spinal and femoral BMD measurements using DXA have a comparable diagnostic sensitivity for vertebral fracture discrimination. Different individuals at risk for osteoporosis may be identified using both methods. The clinical usefulness of a combination of two bone density measurements needs further study in a prospective setting.

Spinal trabecular bone loss and fracture in American and Japanese women

This study examined trabecular bone mineral density (BMD) in Japanese women with and without spinal fracture, and compared the results to American women with and without fracture. The quantitative computed tomography (QCT) systems used at the University of California, San Francisco (UCSF) and at Nagasaki University were cross-calibrated. Normative BMD was assessed with the K2HPO4 liquid phantom in 538 Americans aged 20-85 years, and with the B-MAS200 phantom in 577 Japanese aged 20-83 years. These BMD were adjusted for use with the Image Analysis solid phantom using the result of cross-calibration. The trabecular BMD in 111 postmenopausal American women (55 with fracture), and in 185 postmenopausal Japanese women (67 with fracture) were compared for investigation of the difference in BMD values relative to fracture status. The absolute BMD values in Japanese were lower than those in Americans, and the differences were greater with advancing age. The magnitude of the BMD difference was 8.6, 20.5, 38.1 mg/cm3 in women aged 20-24 years, 40-44 years, 60-64 years, respectively. In premenopausal women, BMD began to decrease at the age of 20 in Japanese, whereas the peak bone mass was maintained until the age of 35 in the American women. In immediate postmenopausal women, BMD significantly decreased in both populations. In later postmenopausal women, BMD significantly decreased with age in the Japanese women but decreased less rapidly in the American women. The aging decrease of BMD was 1.4% and 2.2% per year in the later postmenopausal American and Japanese women, respectively. The fracture threshold is considered to be lower in Japanese women. However, the BMD difference between American and Japanese women with fracture was similar to that without fracture. The Z-scores of fracture subjects versus controls were 2.9 in American and 1.8 in Japanese women. In conclusion, Japanese women were found to have a lower BMD and lower fracture threshold than American women. The significant decrease of spinal trabecular BMD in late postmenopause is potentially responsible for the higher prevalence of spinal fracture in Japanese women.

Peripheral measurement techniques for the assessment of osteoporosis

Peripheral measurement techniques have been the first to be developed for the assessment of osteoporosis, and they remain useful. Besides traditional approaches such as radiographic absorptiometry (RA), radiogrammetry, and single-photon absorptiometry (SPA), new peripheral approaches have been developed that offer powerful ways to assess skeletal status in osteoporosis. These include single x-ray absorptiometry (SXA), peripheral dual x-ray absorptiometry (pDXA), peripheral quantitative computed tomography (pQCT), quantitative ultrasound (QUS) techniques, and magnetic resonance imaging (MRI) approaches. This review describes the current role of peripheral imaging techniques vis-à-vis their central imaging counterparts. Peripheral measurement techniques are attractive because equipment cost is substantially lower, radiation exposure is small, and the devices require less space and sometimes are even portable. Additionally, QUS and MRI offer the potential to measure aspects of bone status beyond the limits of bone densitometry. Peripheral techniques represent important diagnostic methods for the assessment of osteoporosis.

Assessment of fracture risk by bone density measurements

Bone densitometry in its various applications has become an established tool for the diagnosis of osteoporosis. Bone density has been shown to be significantly associated with the risk of future fracture in many prospective studies. From long-term prospective studies, it can be concluded that peak bone density and bone loss are important predictors of subsequent fracture, and that fracture can be predicted over a longer period. Bone density predicts fracture even in elderly persons aged 80 years and older. However, in this population some fractures, such as the cervical hip fracture, may be more strongly influenced by other risk factors. The differences between the various densitometric techniques in predicting future osteoporotic fracture of any type is marginal. However, it seems that bone density measurements at the site of fracture do perform better than measurements at other sites. There is no evidence that measuring a second site improves the diagnostic capability of bone densitometry. The association between bone density and future fracture is partly independent of age and other significant predictors of fracture such as falls, cognizance, and mobility. Quantitative ultrasonic measures of bone quality have been shown to have a predictive capability that is comparable to that of bone density. From the perspective that bone densitometry and quantitative ultrasound independently predict fractures, these measures actually seem complementary rather than competitive. Simple geometric measures of the bones such as hip axis length and vertebral depth may be derived from images of bone densitometry scans and are also predictive of hip fracture or vertebral fracture independently of bone density. Using the current knowledge of the association between bone density, quantitative ultrasound, geometric properties, and fractures as well as clinical risk factors, new models for fracture prediction can be developed for future application in clinical practice for the benefit of the individual patient.

Comparisons of noninvasive bone mineral measurements in assessing age-related loss, fracture discrimination, and diagnostic classification

The purpose of this study was to examine the commonly available methods of noninvasively assessing bone mineral status across three defined female populations to examine their interrelationships, compare their respective abilities to reflect age- and menopause-related bone loss, discriminate osteoporotic fractures, and classify patients diagnostically. A total of 47 healthy premenopausal (age 33 +/- 7 years), 41 healthy postmenopausal (age 64 +/- 9 years), and 36 osteoporotic postmenopausal (age 70 +/- 6 years) women were examined with the following techniques: (1) quantitative computed tomography of the L1-L4 lumbar spine for trabecular (QCT TRAB BMD) and integral (QCT INTG BMD) bone mineral density (BMD); (2) dual X-ray absorptiometry of the L1-L4 posterior-anterior (DXA PA BMD) and L2-L4 lateral (DXA LAT BMD) lumbar spine, of the femoral neck (DXA NECK BMD) and trochanter (DXA TROC BMD), and of the ultradistal radius (DXA UD BMD) for integral BMD; (3) peripheral QCT of the distal radius for trabecular BMD (pQCT TRAB BMD) and cortical bone mineral content (BMC) (pQCT CORT BMC); (4) two radiographic absorptiometric techniques of the metacarpal (RA METC BMD) and phalanges (RA PHAL BMD) for integral BMD; and (5) two quantitative ultrasound devices (QUS) of the calcaneus for speed of sound (SOS CALC) and broadband ultrasound attenuation (BUA CALC). In general, correlations ranged from (r = 0.10-0.93) among different sites and techniques. We found that pQCT TRAB BMD correlated poorly (r < or = 0.46) with all other measurements except DXA UD BMD (r = 0.62,p < or = 0.0001) and RA PHAL BMD (r = 0.52, p < or = 0.0001). The strongest correlation across techniques was between QCT INT BMD and DXA LAT BMD (r = 0.87, p < or = 0.0001), and the weakest correlation within a technique was between pQCT TRAB BMD and pQCT CORT BMC (r = 0.25,p < or = 0.05). Techniques showing the highest correlations with age in the healthy groups also showed the greatest differences among groups. They also showed the best discrimination (as measured by the odds ratios) for the distinction between healthy postmenopausal and osteoporotic postmenopausal groups based on age-adjusted logistic regression analysis. For each anatomic site, the techniques providing the best results were: (1) spine, QCT TRAB BMD (annual loss, -1.2% [healthy premenopausal and healthy postmenopausal]); Student's t-value [not the T score], 5.4 [healthy postmenopausal vs. osteoporotic postmenopausal]; odds ratio, 43 [age-adjusted logistic regression for healthy postmenopausal vs. osteoporotic postmenopausal]); (2) hip, DXA TROC BMD (-0.46; 3.5; 2.2); (3) radius, DXA UD BMD (-0.44; 3.3; 1.9) and pQCT, CORT BMC (-0.72; 2.9; 1.7); (4) hand, RA PHAL (-0.51; 3.6; 2.0); and (5) calcaneus, SOS (-0.09; 3.4; 2.1) and BUA (-0.52; 2.6; 1.7). Despite these performance trends, the differences among sites and techniques were statistically insignificant (p > 0.05) using age-adjusted receiver operating characteristic (ROC) curve analysis. Nevertheless, kappa score analysis (using -2.0 T score as the cut-off value for osteopenia and -2.5 T score for osteoporosis) showed that in general the diagnostic agreement among these measurements in classifying women as osteopenic or osteoporotic was poor, with kappa scores averaging about 0.4 (exceptions were QCT TRAB/INTG BMD, DXA LAT BMD, and RA PHAL BMD, with kappa scores ranging from 0.63 to 0.89). Often different patients were estimated at risk by using different measurement sites or techniques.

Assessment of osteoporosis: comparison of radiographic absorptiometry of the phalanges and dual X-ray absorptiometry of the radius and lumbar spine

PURPOSE

To evaluate radiographic absorptiometry (RA) of the phalanges in healthy women and in women with osteoporosis and to compare the results of RA with those of dual x-ray absorptiometry (DXA) of the radius and spine.

MATERIALS AND METHODS

Thirty-two healthy premenopausal women, 39 healthy postmenopausal women, and 35 postmenopausal women with osteoporosis underwent RA of the phalanges and DXA of the radius and lumbar spine. Pairwise comparisons, age-related bone losses, and percentage decrements and Student t values for intergroup discrimination were calculated. The ability to identify patients with osteoporotic fractures was evaluated by using receiver operating characteristic and age-adjusted logistic regression analyses. The diagnostic agreement for osteoporosis was assessed with kappa statistics.

RESULTS

Findings from RA were correlated with those from spinal DXA (r = .56). The annual bone losses in healthy women, as measured with RA, radial DXA, and spinal DXA, were 0.47%, 0.47%, and 0.32%, respectively. Intergroup percentage decrements and t values obtained with RA were comparable to those obtained with radial and with spinal DXA. Receiver operating characteristic analysis showed no statistically significant differences. The odds ratios for RA, radial DXA, and spinal DXA were 2.1, 1.9, and 2.4, respectively. The kappa scores were 0.44 for both RA versus radial DXA and RA versus spinal DXA, and the score was 0.22 for radial DXA versus spinal DXA.

CONCLUSION

RA appears to be a useful technique for assessing age- and menopause-related bone loss and for identifying women with osteoporosis.

Verification of the position of pedicle screws in lumbar spinal fusion

Medial or lateral pedicle screw penetration with the potential to affect neural structures in a wellknown and frequent problem associated with posterior spinal fusion. We evaluated the placement of pedicle screws (n = 141) in 36 patients following posterior lumbar spinal fusion with Socon or Kluger instrumentation via a lateral transpedicular approach. The examination was based on CT and MR images performed after removal of the instrumentation, on average 1 year after implantation. We found seven pedicle screws with lateral cortical penetration of the pedicle and five screws with medial cortical penetration of the pedicle (8.5% pedicle penetration overall). No severe radicular complications accompanied these pedicle penetrations. The mean insertion angles of the pedicle screws at the L4 level were 22.6 degrees and 23.1 degrees for the left and the right side, respectively. At the L5 level the mean insertion angle was 20.5 degrees on the left side and 21.5 degrees on the right, and at the S1 level the mean angle was 16.2 degrees on the left and 15.2 degrees on the right. The results of this study indicate that the lateral transpedicular approach is a safe procedure for pedicle screw insertion.

Predictors of ankle and foot fractures in older women. The Study of Osteoporotic Fractures Research Group

To determine risk factors for ankle and foot fractures, data collected from 9704 women 65 years of age or older from four areas of the United States were analyzed. Self-reported baseline questionnaires covered areas such as lifestyle factors (physical activity, diet, and smoking habits) and functional impairment (history of fracture, falling, and other diseases). Bone mineral density (BMD) and performance on neuromuscular tests were also measured. During 5.9 years of follow-up, 191 women fractured an ankle and 204 women fractured a foot. Proportional hazard models were used to estimate relative risks. In multivariable models, factors associated with ankle fracture included one or more falls in the year prior to baseline (relative risk [RR] 1.5; 95% confidence interval [CI] 1.1-2.1), greater vigorous physical activity (RR per 2 times/week, 1.2; CI 1.1-1.3), weight gain since age 25 (RR per 20% gain, 1.4; CI 1.2-1.5), self-reported osteoarthritis (RR 0.5; CI 0.3-0.8), a sister's history of hip fracture after age 50 (RR 1.7; CI 1.0-3.0), out of house < or = 1 per week (RR 3.0; CI 1.4-6.6), and low distal radius BMD (RR per -0.1 g/cm2, 1.2; CI 1.0-1.4). Factors associated with foot fracture included insulin-dependent diabetes (RR 2.9; CI 1.2-7.2), use of seizure medications (RR 2.3; CI 1.0-5.7) or of benzodiazepines (RR 1.5; CI 1.1-2.2), history of hyperthyroidism (RR 0.5; CI 0.3-1.0), poor far depth perception (RR 0.7; CI 0.5-1.0), and low distal radius BMD (RR per -0.1 g/cm2, 1.3; CI 1.1-1.5). Ankle and foot fractures have different profiles of risk factors that are largely independent of low bone mass.