Non-invasive measurements of bone mass in adult renal osteodystrophy

Mixed uremic osteodystrophy: an ill-described common bone pathology in patients with chronic kidney disease

Renal osteodystrophy (ROD) starts early and progresses with further loss of kidney function in patients with chronic kidney disease (CKD). There are four distinct types of ROD based on undecalcified bone biopsy results. Adynamic bone disease and osteomalacia are the predominant forms of low bone turnover, while hyperparathyroid bone disease and mixed uremic osteodystrophy (MUO) are typically associated with high bone turnover. MUO is a prevalent but poorly described pathology that demonstrates evidence of osteomalacia on top of the high bone formation/resorption. The prevalence of MUO ranges from 5 to 63% among different studies. The pathogenesis of MUO is multi-factorial. Altered phosphate homeostasis, hypocalcemia, vitamin D deficiency, increased FGF-23, interleukins 1 and 6, TNF-α, amyloid, and heavy metal accumulation are the main inducers of MUO. The clinical findings of MUO are usually non-specific. The use of non-invasive testing such as bone turnover markers and imaging techniques might help to suspect MUO. However, it is usually impossible to precisely diagnose this condition without performing bone biopsy. The principal management of MUO is to control the maladaptive hyperparathyroidism along with correcting any nutritional mineral deficiencies that may induce mineralization defect. MUO is a common but still poorly understood bone pathology category; it demonstrates the complexity and difficulty in understanding ROD. A large prospective bone biopsy-based studies are needed for better identification as proper diagnosis and management would improve the outcome of patients with MUO.

Bone turnover markers are associated with bone density, but not with fracture in end stage kidney disease: a cross-sectional study

BACKGROUND

Fracture risk is increased in chronic kidney disease (CKD), but assessment of bone fragility remains controversial in these patients. This study investigated the associations between bone turnover markers, bone mineral density (BMD), and prevalent fragility fracture in a cohort of kidney transplantation candidates.

METHODS

Volumetric BMD of spine and hip was measured by quantitative computed tomography. Parathyroid hormone (PTH), bone-specific alkaline phosphatase, procollagen type-1 N-terminal propeptide, tartrate resistant alkaline phosphatase, and C- and N-terminal telopeptides of type 1 collagen were analyzed from fasting morning blood samples. Fragility fractures included prevalent vertebral fractures and previous low-trauma clinical fractures.

RESULTS

The fracture prevalence was 18% in 157 adult kidney transplant candidates. Fractured patients had reduced BMD and Z-score at both spine and hip. Levels of bone turnover markers were significantly higher in patients on maintenance dialysis than in pre-dialysis patients; but did not differ between patients with and without fracture. There were strong, positive correlations between PTH and all bone turnover markers. PTH was negatively associated with Z-score at lumbar spine and total hip; in contrast, bone turnover markers were only negatively associated with total hip Z-score.

CONCLUSIONS

Bone turnover markers were negatively associated with bone density, but not associated with prevalent fracture in kidney transplantation candidates. The role of bone turnover markers in assessing bone fragility in CKD will require further investigation.

TRIAL REGISTRATION

This study was registered at ClinicalTrials.gov with identifier NCT01344434 .

Micro-CT in the Assessment of Pediatric Renal Osteodystrophy by Bone Histomorphometry

BACKGROUND AND OBJECTIVES

Computed tomography (CT) measurements can distinguish between cortical and trabecular bone density in vivo. High-resolution CTs assess both bone volume and density in the same compartment, thus potentially yielding information regarding bone mineralization as well. The relationship between bone histomorphometric parameters of skeletal mineralization and bone density from microcomputed tomography (μCT) measurements of bone cores from patients on dialysis has not been assessed.

DESIGN, SETTING, PARTICIPANTS, & MEASUREMENTS

Bone cores from 68 patients with ESRD (age =13.9±0.5 years old; 50% men) and 14 controls (age =15.3±3.8 years old; 50% men) obtained as part of research protocols between 1983 and 2006 were analyzed by bone histomorphometry and μCT.

RESULTS

Bone histomorphometric diagnoses in the patients were normal to high bone turnover in 76%, adynamic bone in 13%, and osteomalacia in 11%. Bone formation rate did not correlate with any μCT determinations. Bone volume measurements were highly correlated between bone histomorphometry and μCT (bone volume/tissue volume between the two techniques: r=0.70; P<0.001, trabecular thickness and trabecular separation: r=0.71; P<0.001, and r=0.56; P<0.001, respectively). Osteoid accumulation as determined by bone histomorphometry correlated inversely with bone mineral density as assessed by μCT (osteoid thickness: r=-0.32; P=0.01 and osteoid volume: r=-0.28; P=0.05). By multivariable analysis, the combination of bone mineral density and bone volume (as assessed by μCT) along with parathyroid hormone and calcium levels accounted for 38% of the variability in osteoid volume (by histomorphometry).

CONCLUSIONS

Measures of bone volume can be accurately assessed with μCT. Bone mineral density is lower in patients with excessive osteoid accumulation and higher in patients with adynamic, well mineralized bone. Thus, bone mineralization may be accurately assessed by μCT of bone biopsy cores. Additional studies are warranted to define the value of high-resolution CT in the prediction of bone mineralization in vivo.

Bone histomorphometry in renal osteodystrophy

On bone biopsies from patients with chronic kidney disease, measurements are made of the turnover, mineralization, and volume. Turnover depends on the bone formation rate and bone resorption rate; the former can be measured using tetracycline labelling. The osteoid width and bone apposition rate determine the mineralization rates. Bone volume includes both mineralized and unmineralized bone and is directly related to the porosity. Using these measurements, biopsies can be separated into the classic types of renal osteodystrophy: normal, adynamic, high-turnover, mixed, and osteomalacia. Fracture rates among these types are not consistent, but several studies have found high fracture rates with adynamic or osteomalacia. The bone density tests cannot distinguish between different types of bone histology.

Osteoporosis in the elderly with chronic kidney disease

Considering the aging dialysis population of today, increasing our knowledge about the nature, diagnosis and the treatment of bone mineral density (BMD) problems in end-stage renal disease (ESRD) patients deserves more attention. Osteoporosis is basicly defined as a decrease in bone mass. Large epidemiological studies in general population have identified several risk factors for osteoporosis including advancing age, female gender, white race, decreased calcium intake, gastric acid suppression therapy, sedentary lifestyle, premature loss of gonadal function, decreased estrogen secretion, thin body habitus, decreased physical activity, cigarette smoking, alcohol abuse, excess glucocorticoid exposure, and possibly some genetic factors. Osteoporosis in ESRD patients is only a part of a wider spectrum of metabolic bone problems, namely uremic osteodystrophy. Therefore, its diagnosis, management and follow-up may differ from the general population and an individualization of diagnosis and definition for dialysis population may be necessary. However, standard diagnostic tools such as dual energy X-ray absorptiometry (DEXA) have been widely used for the assessment of bone mineral deficiency status in ESRD patients. Regardless of the methods, most of the studies are in concordance with a reduced BMD in HD and PD patients. Dialysis patients are known to be at increased risk for low-trauma fractures. Thinning of cortical bone, which is responsible for the largest contribution toward reduced bone mineral content in chronic renal failure results in increased fracture risk. In either normal population and dialysis patients, fracture risk is increased with age. But in dialysis patients, besides age, several other factors may also affect the degree of bone mineral deficiency, and age-BMD relationship may be blunted. Female sex, in hemodialysis patients is negatively associated with total hip BMD. While several studies have been unable to demonstrate any association between BMD and PTH levels, larger body size has been shown to have a significant positive effect on BMD in both hemodialysis and peritoneal dialysis patients. Although they have been used in small groups of chronic kidney disease (CKD) and ESRD patients, because of their potential nephrotoxicity and hypocalcemic effects, use of biphosphonates in renal patients is questionable. Currently, bone biopsy, in order to exclude adynamic bone disease is recommended before beginning treatment with bisphosphonates in chronic kidney disease and dialysis patients.

Calcium phosphate metabolism and bone mineral density with nocturnal hemodialysis

An elevated calcium x phosphate product (Ca x P) is an independent risk factor for vascular calcification and cardiovascular death in dialysis patients. More physiological dialysis in patients undergoing nocturnal hemodialysis (NHD) has been shown to produce biochemical advantages compared with conventional hemodialysis (CHD) including superior phosphate (P) control. Benefits of dialysate with greater calcium (Ca) concentration are also reported in NHD to prevent Ca depletion and subsequent hyperparathyroidism, but there are concerns that a higher dialysate Ca concentration may contribute to raised serum Ca levels and greater Ca x P and vascular disease. The NHD program at our unit has been established for 4 years, and we retrospectively analyzed Ca and P metabolism in patients undergoing NHD (8-9 h/night, 6 nights/week). Our cohort consists of 11 patients, mean age 49.3 years, who had been on NHD for a minimum of 12 months, mean 34.3 months. Commencement was with low-flux (LF) NHD and 1.5 mmol/L Ca dialysate concentration, with conversion to high-flux (HF) dialyzers after a period (mean duration 18.7 months). We compared predialysis serum albumin, intact parathyroid hormone, P, total corrected Ca, and Ca x P at baseline on CHD, after conversion to LF NHD and during HF NHD. We also prospectively measured bone mineral density (BMD) on all patients entering the NHD program. Bone densitometry (DEXA) scans were performed at baseline (on CHD) and yearly after commencement of NHD. With the introduction of HF dialyzers, the Ca dialysate concentration was concurrently raised to 1.75 mmol/L after demonstration on DEXA scans of worsening osteopenia. Analysis of BMD, for all parameters, revealed a decrease over the first 12 to 24 months (N = 11). When the dialysate Ca bath was increased, the median T and Z scores subsequently increased (data at 3 years, N = 6). The mean predialysis P levels were significantly lower on LF NHD vs. CHD (1.51 vs. 1.77 mmol/L, p = 0.014), while on HF NHD P was lower again (1.33 mmol/L, p = 0.001 vs. CHD). Predialysis Ca levels decreased with conversion from CHD to LF NHD (2.58 vs. 2.47 mmol/L, p = 0.018) using a 1.5 mmol/L dialysate Ca concentration. The mean Ca x P on CHD was 4.56 compared with a significant reduction of 3.74 on LF NHD (p = 0.006) and 3.28 on HF NHD (p = 0.001 vs. CHD), despite the higher dialysate Ca in the latter. We conclude that an elevated dialysate Ca concentration is required to prevent osteopenia. With concerns that prolonged higher Ca levels contribute to increased cardiovascular mortality, the optimal Ca dialysate bath is still unknown. Better P control on NHD, however, reduces the overall Ca x P, despite the increased Ca concentration, therefore reducing the risk of vascular calcification.

Disordered Mineral Metabolism and Vascular Calcification in Nondialyzed Chronic Kidney Disease Patients

It is well established that abnormalities in mineral metabolism are apparent early in the course of chronic kidney disease (CKD) and result in clinically relevant consequences such as renal osteodystrophy. Furthermore, there is emerging evidence linking some of these abnormalities (hyperphosphatemia) to the high cardiovascular morbidity and mortality experienced by nondialyzed patients with CKD. Most studies have evaluated vascular calcification in patients with stage 5 CKD. Reports published over the last 2 years show that the process begins rather early in CKD and is particularly severe among elderly and type 2 diabetic patients. Furthermore, "calcium begets calcium", such that the calcification burden in early CKD is an important predictor of subsequent progression, including the rapid increase seen in stage 5 CKD. There is an increasing body of evidence that supports the thesis that elevated serum levels of phosphorus and calcium and deficiency of inhibitors of calcification (for example, fetuin-A) are important in the progression of vascular calcification in patients with end-stage renal disease. However, the concentrations of calcium and phosphorus shown to induce mineralization in cell culture studies are not observed in most patients until late in stage 4 or stage 5 CKD. Cross-sectional and longitudinal studies have also been unable to show a correlation between serum levels of markers of disordered mineral metabolism and severity of vascular calcification. Future studies should evaluate the pathogenetic role of phosphorus retention, which occurs early in the course of CKD, in the induction and/or progression of vascular calcification. Finally, there is a need to identify alternative pathogenetic mechanisms that may be important causes of the high calcification burden observed early in CKD.

Bone mineral density and its correlation with clinical and laboratory factors in chronic peritoneal dialysis patients

The aim of this study was to assess the clinical and laboratory correlations of bone mineral density (BMD) measurements among a large population of patients on chronic peritoneal dialysis (PD). This cross-sectional, multicenter study was carried out in 292 PD patients with a mean age of 56 +/- 16 years and mean duration of PD 3.1 +/- 2.1 years. Altogether, 129 female and 163 male patients from 24 centers in Canada, Greece, and Turkey were included in the study. BMD findings, obtained by dual-energy X-ray absorptiometry (DEXA) and some other major clinical and laboratory indices of bone mineral deposition as well as uremic osteodystrophy were investigated. In the 292 patients included in the study, the mean lumbar spine T-score was -1.04 +/- 1.68, the lumbar spine Z-score was -0.31 +/- 1.68, the femoral neck T-score was -1.38 +/- 1.39, and the femoral neck Z score was -0.66 +/- 1.23. According to the WHO criteria based on lumbar spine T-scores, 19.2% of 292 patients were osteoporotic, 36.3% had osteopenia, and 44.4% had lumbar spine T-scores within the normal range. In the femoral neck area, the prevalence of osteoporosis was slightly higher (26%). The prevalence of osteoporosis was 23.3% in female patients and 16.6% in male patients with no statistically significant difference between the sexes. Agreements of lumbar spine and femoral neck T-scores for the diagnosis of osteoporosis were 66.7% and 27.3% and 83.3% for osteopenia and normal BMD values, respectively. Among the clinical and laboratory parameters we investigated in this study, the body mass index (BMI) (P < 0.001), daily urine output, and urea clearance time x dialysis time/volume (Kt/V) (P < 0.05) were statistically significantly positive and Ca x PO(4) had a negative correlation (P < 0.05) with the lumbar spine T scores. Femoral neck T scores were also positively correlated with BMI, daily urine output, and KT/V; and they were negatively correlated with age. Intact parathyroid hormone levels did not correlate with any of the BMD parameters. Femoral neck Z scores were correlated with BMI (P < 0.001), and ionized calcium (P < 0.05) positively and negatively with age, total alkaline phosphatase (P < 0.05), and Ca x P (P < 0.01). The overall prevalence of fractures since the initiation of PD was 10%. Our results indicated that, considering their DEXA-based BMD values, 55% of chronic PD patients have subnormal bone mass-19% within the osteoporotic range and 36% within the osteopenic range. Our findings also indicate that low body weight is the most important risk factor for osteoporosis in chronic PD patients. An insufficient dialysis dose (expressed as KT/V) and older age may also be important risk factors for osteoporosis of PD patients.

Impaired muscle strength is associated with fractures in hemodialysis patients

INTRODUCTION

Fractures are extremely common among hemodialysis (HD) patients.

METHODS

To assess if bone mineral density (BMD) and/or tests of muscle strength were associated with fractures, we studied 37 men and 15 women, 50 years and older, on HD for at least 1 year. We excluded subjects with prior renal transplants and women taking hormone replacement therapy. We inquired about low-trauma fractures since starting dialysis. Subjects underwent BMD testing with a Lunar DPX-L densitometer. Tests of muscle strength included: timed up and go (TUG), 6-min walk, functional reach, and grip strength. Lateral and thoracic radiographs of the spine were obtained and reviewed for prevalent vertebral fractures. We used logistic regression to examine associations between fracture (prevalent vertebral, self-reported low trauma since starting dialysis and/or both) and BMD, and fracture and muscle-strength tests. Analyses were adjusted for age, weight, and gender.

RESULTS

Mean age was 66+/-9.0 years, mean weight was 72.9+/-15.2 kg, and most (35 of 52) participants were Caucasian. Average duration of dialysis was 40.2 (interquartile range: 24-61.2) months. The most common cause of renal failure was diabetes (16 subjects). There were no differences by gender or fracture. Of the 52 subjects, 27 had either a vertebral fracture or low trauma fracture. There was no association between fractures, hip or spine BMD, or grip strength. In contrast, greater functional reach [odds ratio (OR) per standard deviation (SD) increase: 0.29; 95% CI: 0.13-0.69), quicker TUG (OR per SD decrease: 0.14; 95% CI: 0.11-0.23), and a greater distance walked in 6 min (OR per SD increase: 0.10; 95% CI: 0.03-0.36) were all associated with a reduced risk of fracture.

CONCLUSIONS

Impaired neuromuscular function is associated with fracture in hemodialysis patients.

Bone density and heel ultrasound testing do not identify patients with dialysis-dependent renal failure who have had fractures

Patients with dialysis-dependent renal failure are at increased risk for low-trauma fractures. However, the optimal means of identifying patients at high risk for fracture is not known. We assessed the association between fracture history and two tests of bone mineral density (dual x-ray absorptiometry [DEXA] and calcaneal ultrasound) among patients with hemodialysis-dependent renal failure. We evaluated 71 men and 33 women aged 55 years or older who had been receiving hemodialysis for at least 1 year. All patients underwent spinal radiography, DEXA of the hip and lumbar spine, and calcaneal ultrasonography. We assessed risk factors for low-trauma fractures by questionnaire and medical chart review. Of patients, 52% had a fracture on spinal radiographs or a history of a low-trauma fracture, 69% had osteopenia by DEXA, and 26% had a low heel ultrasound measurement. Neither DEXA nor calcaneal ultrasound was associated with fracture history, however. Our findings indicate that fractures among patients with dialysis-dependent renal failure are common. Tests of bone strength do not adequately identify patients with a history of fractures. Prospective studies to determine the optimal method of identifying patients with dialysis-dependent renal failure at high risk for fracture are needed.

Diagnosis and monitoring of renal osteodystrophy

A variety of biochemical investigations and radiological techniques are available to assist in the diagnosis and monitoring of renal osteodystrophy. Measurement of serum parathyroid hormone remains the single most useful biochemical test in predicting bone histology in an individual patient. Newer biochemical markers of bone turnover are unlikely to supplant this in everyday practice, but may provide useful supplementary information in the future. The present review discusses the role of radiological investigation, including bone densitometry and quantitative ultrasound. Bone biopsy remains the 'gold standard' investigation. Its invasive nature and the need for specialized processing and interpretation limits its use in clinical practice, although it still has a role particularly in the investigation of low turnover states. Also, as molecular biological techniques are increasingly being used, the evaluation of biopsy specimens will in the future provide new insights into the disordered bone cell function that occurs in renal osteodystrophy.

Bone mineral density and biochemical markers of bone turnover in patients with predialysis chronic renal failure

BACKGROUND

Metabolic bone disease might commence early in the course of renal failure. This study therefore examined the frequency and severity of the skeletal changes in predialysis chronic renal failure by measurements of bone mineral density (BMD), biochemical markers of bone turnover (osteocalcin, bone-specific alkaline phosphatase, carboxy terminal propeptide of type I collagen, and carboxy-terminal telopeptide of type I collagen), parathyroid hormone (PTH), ionized calcium (Ca++), phosphate (P), and vitamin D metabolites.

METHODS

The study was performed in 113 patients (male/female: 82/31) with chronic renal diseases [mean glomerular filtration rate (GFR) of 37 ml/min] and in 89 matched, normal control subjects.

RESULTS

The patients had significantly (P<0.05) reduced BMD in the spine (-6.3%), the femur (-12.1%), the forearm (-5.7%), and the total body (-4.2%) as compared with the control subjects. Dividing the patients into quartiles according to GFR revealed that BMD decreased with the gradual decline in renal function at all the measured skeletal sites, but was most pronounced in the femur: 0.63+/-0.03, 0.74+/-0.02, 0.77+/-0.02, and 0.82+/-0.03 g/cm2 in each quartile from lowest to highest GFR compared with 0.82+/-0.02 g/cm2 in the control group (P<0.0001). All of the measured bone markers showed increasing plasma levels with the more advanced stages of renal failure. Serum PTH and serum P levels increased, whereas serum Ca++ and 1,25-dihydroxyvitamin D decreased. BMD Z-scores of the femur and of the forearm correlated to the biochemical markers and to PTH (P<0.05 to P<0.0001). The biochemical markers all showed strong correlations to PTH, also when corrected for the effect of the decline in GFR (r = 0.40 to 0.92, P<0.01 to P< 0.0001).

CONCLUSION

Skeletal changes are initiated at an early stage of chronic renal failure, as estimated from reduced BMD and elevated levels of PTH and from the biochemical markers of both bone formation and bone resorption.

Racial differences in bone mineral density and bone loss among end-stage renal disease patients

Although black patients without end-stage renal disease (ESRD) have a greater bone mineral density (BMD) than whites, the impact of race on BMD among patients with ESRD who are likely to have varying degrees of renal osteodystrophy is not known. We undertook a cohort study of 106 hemodialysis patients comparing BMD and bone loss between black and white patients with ESRD to determine if black patients have a greater BMD and less bone loss than white patients with ESRD. BMD was determined by dual-energy radiograph absorptiometry (DEXA). Osteopenia was defined as greater than 1 standard deviation (SD) less than the mean of peak bone mass (T score <-1), and osteoporosis was defined as greater than 2.5 SDs less than the mean of peak bone mass (T score <-2.5). The association between BMD and race was estimated using linear regression. The risk for osteopenia among black compared with white patients was calculated using logistic regression. Black patients were similar to white patients with respect to all characteristics noted, except black patients were less likely to be men (69.7% v 49. 4%) and tended to have greater intact parathyroid hormone (PTH) values (mean, 403.2 +/- 384.5 pg/mL v 161.4 +/- 129.0 pg/mL). Compared with whites, the BMD of blacks was a mean of 1.15 (95% confidence interval [CI], 0.54 to 1.78) SDs greater at the femoral neck after adjusting for age, PTH level, and sex. The percentage of bone loss per year was similar between blacks and whites. The risk for osteopenia among blacks was significantly less than that among whites (odds ratio = 0.15; 95% CI, 0.04 to 0.59) after adjusting for age, sex, and PTH level. Black patients with ESRD have a greater BMD and are at decreased risk for osteopenia compared with whites, independent of renal osteodystrophy. When considering bone disease among patients with ESRD, physicians should also consider osteoporosis and the impact of race on BMD.

Quantitative ultrasound in the assessment of skeletal status in uremic patients

Renal osteodystrophy (ROD) can be characterized by both high (HT) and low (LT) bone turnover states. Although bone biopsy remains the "gold standard" to diagnose ROD, noninvasive tools for the diagnosis and follow-up of such bone disease are desirable. Recently, ultrasound (US) techniques, proposed to assess skeletal status, have been shown to be correlated not only with bone density but also with bone quality. We have investigated 98 patients on chronic hemodyalisis (HD) and 98 healthy, sex- and age-matched subjects. Amplitude-dependent speed of sound (AD-SOS) and ultrasound bone profile score (UBPS) at phalanxes and speed of sound (SOS), broadband ultrasound attenuation (BUA), and a quantitative ultrasound index (QUI/stiffness) at the heel were performed in both groups. In all subjects intact parathyroid hormone (PTH), total alkaline phosphatase (T-ALP), bone isoenzyme alkaline phosphatase (B-ALP), and carboxy-terminal telopeptide of type I collagen (ICTP) were assessed. All US parameters were significantly lower in the hemodialysis group than in control subjects. Moreover, among US parameters only AD-SOS and UBPS showed a significant correlation with PTH, T-ALP, and B-ALP. Dialytic age showed a modest, but significant correlation only with US parameters at the phalanxes. On the basis of bone biochemical markers, we considered a group with high and a group with normal to low bone turnover. AD-SOS and UBPS, but not SOS, BUA, and stiffness were significantly (p < 0.01) lower in the high bone turnover than in low bone turnover group. Furthermore, in the high bone turnover group, parameters of the US phalanxes strongly correlated with B-ALP. Our results seem to demonstrate that US parameters are a useful tool in the assessment of skeletal status in patients on maintenance dialysis.

Prevalence and risk factors for osteopenia in dialysis patients

Dialysis patients are at risk for low bone mineral density (BMD) consequent of hyperparathyroidism, 1,25-dihydroxyvitamin D deficiency, previous immunosuppression, chronic acidosis, secondary amenorrhea, and chronic heparin and aluminum exposure. We wanted to determine the prevalence and distribution of osteopenia and the influence of risk factors for osteopenia in dialysis patients. Dual energy x-ray absorptiometry was used to record BMD at the lumbar spine (LS), hip, and nondominant forearm. Results were expressed as Z-scores (standard deviations from the mean of a healthy age- and gender-matched reference population). Osteopenia was defined as a Z-score worse than -2. In the 250 dialysis patients studied, the prevalence of osteopenia at the LS, femoral neck (FN) and ultradistal radius (UD) was 8%, 13% and 20%, respectively. The median Z-scores at these sites were all significantly different from the healthy reference population median of 0 and were 0.29 (P = 0.008), -0.67 (P < 0.001), and -1.01 (P < 0.001), respectively. Previous transplantation was associated with as much as a one Z-score lower BMD at the FN (P = 0.0069) and UD (P = 0.0011) and a marginally significant reduction at the LS (P = 0.0777). Previous parathyroidectomy was associated with a markedly higher LS BMD (P = 0.0001) and a higher BMD at the FN (P = 0.0017) but not the UD (P = 0.3691). A history of secondary amenorrhea was associated with a lower FN BMD (P = 0.0047) but not a significantly lower BMD at the LS (P = 0.0978) or UD (P = 0.2327). In hemodialysis patients without a history of transplantation, parathyroidectomy, or secondary amenorrhea, there was no correlation between Z-score at any site and duration of dialysis. Thus, osteopenia in dialysis patients occurs in both axial and appendicular sites and sites of compact and cancellous bone. It is more common with previous transplantation and secondary amenorrhea, whereas a history of parathyroidectomy is associated with increased BMD. No relationship was found between BMD and duration of hemodialysis, which suggests that important changes in BMD occur during the predialysis stage of chronic renal failure.