Comparison of the Humoral Markers of Bone Turnover and Bone Mineral Density in Patients on Haemodialysis and Continuous Ambulatory Peritoneal Dialysis

Transcription

1 Original Paper Nephron 2002;91: Accepted: July 5, 2001 Comparison of the Humoral Markers of Bone Turnover and Bone Mineral Density in Patients on Haemodialysis and Continuous Ambulatory Peritoneal Dialysis G. Hampson a S. Vaja a C. Evans b C.A. Chesters c R. Pettit d W. Evans d D. Thomas e P.T. Seed f W.D. Fraser c a Department of Chemical Pathology, St Thomas Hospital, London, Departments of b Medical Biochemistry, d Medical Physics, and e Renal Medicine, University Hospital of Wales, Cardiff, c Department of Clinical Chemistry, Royal Liverpool University Hospital, Liverpool, and f Department of Public Health Medicine, Guy s, Kings, and St Thomas Medical and Dental School, London, UK Key Words Dialysis W Renal osteodystrophy W Bone markers W Bone mineral density Abstract Renal osteodystrophy is an important complication in patients with end-stage renal disease on maintenance dialysis. The aim of this study was to compare the biochemical markers of bone formation (serum collagen type I C-terminal propeptide) and resorption (serum deoxypyridinoline DPD and pyridinoline PYR) with the bone mineral density (BMD) at lumbar spine, femoral neck, and forearm in patients with end-stage renal disease on haemodialysis (HD) versus continuous ambulatory peritoneal dialysis (CAPD). Fifty-nine adult patients, 45 on CAPD (18 females, 27 males) and 14 on HD (2 females, 12 males), were studied. The mean age was 44 B SEM 1.6 and 54.4 B 4.8 years, respectively. No significant differences in serum calcium, phosphorus, creatinine, and parathyroid hormone were found between patients on HD and CAPD in predialysis samples. Serum urea was significantly lower (p = 0.02) in the CAPD group. Serum PYR (nmol/l) and DPD (nmol/l) were significantly higher in patients on HD as compared with those on CAPD: 105 B 23.3 versus 43.7 B 3.47 (p = 0.007) and 31.0 B 2.4 versus 24.4 B 1.4 (p = 0.027), respectively. The results were still significantly higher in the HD patients following correction for serum creatinine and body mass index. There was a close correlation between dialysate DPD and creatinine in both dialysis modalities (HD r = 0.9, CAPD r = 0.76). The clearance of DPD did not differ significantly between the CAPD membrane and the HD membrane (p = 0.22). Serum collagen type I C-terminal propeptide was not significantly different between the HD and CAPD patients. The results were unaffected following correction for age and gender. The BMD was measured in 38 (65%) of the patients (HD n = 8, CAPD n = 30) by dual-energy X-ray absorptiometry and expressed as Z scores. This was reduced at all sites in the patients with end-stage renal disease. The BMD was significantly lower at the ultradistal forearm (mostly trabecular bone) in HD patients as compared with CAPD patients (n = 0.02). A similar trend was observed at the lumbar spine, although the results failed to reach significance. In the whole population (n = 38), linear regression analysis revealed a significant negative correlation between BMD at the ultradistal forearm and serum PYR (r = 0.35, p = ABC Fax karger@karger.ch S. Karger AG, Basel /02/ $18.50/0 Accessible online at: Dr. Geeta Hampson Department of Chemical Pathology St Thomas Hospital Lambeth Palace Road, London SE1 7EH (UK) Tel (ext. 2881), Fax

2 0.04) and DPD (r = 0.33, p = 0.049). Combined measurements of BMD and biochemical markers of bone resorption may have potential in the identification of patients at high risk of bone loss who may require further evaluation of bone remodeling by bone histomorphometry. Introduction Copyright 2002 S. Karger AG, Basel Renal osteodystrophy is associated with substantial morbidity and mortality [1]. A substantial increase in the fracture rate has been documented in dialysis patients [2]. Schober et al. [3] have demonstrated that a reduction in mineralized cortical bone in dialysis patients with clinically significant renal osteodystrophy correlated with generalized thinning of the cortical bone as measured by single-photon absorptiometry. Recently it has also been shown that modes of dialysis may have different effects on bone mass. Continuous ambulatory peritoneal dialysis (CAPD) may have a more favourable influence on bone mineral density (BMD) than haemodialysis (HD) [4]. Measurement of the BMD by dual-energy X-ray absorptiometry scan is widely used to quantify the BMD at various sites and is a non-invasive procedure [5]. This technique may be a useful method for evaluating quantitative bone changes among patients on CAPD and HD, although more research needs to be carried out to investigate whether this technique is clinically useful in dialysis patients [6]. A high bone turnover has been shown to be an independent risk factor for fractures [7]. The humoral markers of bone collagen synthesis and degradation may have an additional role in the assessment of bone turnover in patients with end-stage renal disease (ESRD) [8]. Although bone histomorphometry remains, to date, the gold standard for the evaluation of renal osteodystrophy [9], the procedure is painful, invasive, and costly. Because of its invasiveness, many patients are reluctant to undergo this investigation. Measurement of the BMD and the humoral markers of bone turnover may, therefore, provide a useful non-invasive approach in identifying those patients with ESRD who may be at increased risk of bone loss and may, therefore, require further evaluation of bone remodeling by bone histomorphometry. The combination of these two parameters has been found to be useful in other patient populations such as in women with postmenopausal osteoporosis [7]. There is still a paucity of studies into the assessment and correlation of bone turnover using the serum pyridinium (PYR) cross-links and BMD at the different skeletal sites in patients on CAPD and HD. Serum pyridinium cross-links [11] are released into blood during the resorptive process and because of their low molecular weight ( Da) are normally excreted in urine, although assessment of their clearance by the dialysis membranes, in particular CAPD membranes, in patients with ESRD is lacking. The aim of this cross-sectional study was firstly to determine the serum concentrations of the biochemical markers of bone turnover and BMD in patients on the two dialysis modalities and secondly to compare the bone markers with densitometric measurements in this population. In addition, we also examined the clearance of the serum cross-links by the two dialysis membranes in order to investigate whether assessments of circulating concentrations of these markers are equally valid in HD and CAPD patients. Patients and Methods Patients All 180 patients who were regularly dialyzed at the renal unit in Cardiff between February and December 1996 were invited to take part in the study. Of these, 59 patients (33%; 20 women and 39 men) with ESRD on maintenance dialysis agreed to participate in this prospective study and were, therefore, enrolled in the study following informed consent. The study was performed in accordance with the Helsinki declaration and approved by the local ethics committee. Forty-five patients were on maintenance CAPD and 14 on HD. The mean age (B SEM) for the CAPD and HD patients was 44 B 1.6 and 54.4 B 4.8 years, respectively. The mean duration of CAPD and HD treatments was 23.4 B 3.2 and 19.8 B 4.7 months, respectively. The causes for ESRD were as follows: diabetic nephropathy (n = 22), interstitial nephropathy (n = 10), chronic glomerulonephritis (n = 20), polycystic kidney disease (n = 3), nephrectomy (n = 2), and unknown (n = 2). All patients were offered a choice as to the preferred dialysis modality after demonstration of both methods. One patient who was partially sighted due to diabetic retinopathy was started on HD as he was deemed incapable of carrying out CAPD. Patients on HD were treated 8 12 h/week using standard dialysis membranes with constant blood flow ( ml/min) and a dialysate flow rate of 500 ml/min. Patients on CAPD were dialyzed using 1.5- to 2.0-liter bag exchanges four times/day. The regular dialysate calcium concentration was 1.25 mmol/l. At the time of the study, all patients were also receiving concomitant drugs, diuretics (CAPD: n = 11, HD: n = 2), antihypertensives (CAPD: n = 39, HD: n = 4), iron (CAPD: n = 10, HD: n = 7), calcitriol (CAPD: n = 34, HD: n = 9), calcium carbonate (CAPD: n = 34, HD: n = 11), vitamins B/C (CAPD: n = 18, HD: n = 10), and recombinant human erythropoietin (CAPD: n = 11, HD: n = 2). Three patients on CAPD were taking aluminium hydroxide instead of calcium carbonate as phosphate binder because of hypercalcaemia. The majority of diabetic patients were on insulin (CAPD: n = 13, HD: n = 2). The rest were either on oral hypoglycaemic agents or were diet controlled. None of the patients had received other drugs known to affect bone metabolism. None of the patients had any his- Bone Turnover in Patients on Dialysis Nephron 2002;91:

3 Table 1. Summary of the clinical characteristics of the patients on HD and CAPD (mean values and differences) HD patients CAPD patients Differences (95% CI) n (male/female ratio) 14 (12/2) 45 (27/18) Age, years Body mass index, kg/m Time on dialysis, months Underlying nephropathy (%) Diabetic nephropathy 4 (28.6) 18 (40) Interstitial nephropathy 3 (21.4) 7 (15.5) Chronic glomerulonephritis 5 (35.7) 15 (33.3) Polycystic kidney disease 1 (7) 2 (4.4) Nephrectomy 0 2 (4.4) 10 2 Unknown 1 (7) 1 (2.2) 9 19 tory of fractures. The proportions of patients taking drugs involved in calcium homeostasis were similar in both groups at the time of inclusion in the study (calcium carbonate: HD 78%, CAPD 75%, calcitriol: HD 64%, CAPD 75%). All patients were Caucasians. Sixteen of the 18 female patients on CAPD were post-menopausal. None of them were receiving therapy with oestrogen. The clinical characteristics of the patients are summarised in table 1. Clearance of Serum Pyridinium Cross-Links by HD and CAPD Membranes In order to assess the clearance of the cross-links by the HD membrane, 7 patients on HD had pre- and post-hd samples for urea, creatinine, deoxypyridinoline (DPD) and pyridinoline (PYR) analysis. In all patients dialysate samples were collected during the middle of dialysis for the above-mentioned analyses. Dialysis was carried out in the HD patients using standard haemophan membranes. Six patients on CAPD underwent a modified peritoneal equilibration test. Blood samples were taken before and after a 4-hour exchange (2.5-liter bag). Dialysate samples were collected at the end of the 4- hour exchange. The clearance for DPD was expressed as the ratio of dialysate (D) DPD/D creatinine divided by the ratio of serum (S) DPD/S creatinine. Biochemical Assays Predialysis blood samples were taken from all patients. In addition, 14 of the 45 patients on CAPD had a repeat blood sample taken 3 months later. Serum concentrations of urea, creatinine, electrolytes, total calcium, albumin, phosphate, alkaline phosphatase, and aspartate transaminase were measured by standard laboratory techniques on a Hitachi 747 (Boehringer Mannheim, Germany). The serum intact parathyroid hormone (PTH) concentrations were measured by an immunochemiluminometric assay (Ciba Corning Diagnostics, Halstead, UK). The assay reference range was pmol/l. The interassay coefficient of variation (CV) was 12.4%. Serum 25-hydroxyvitamin D 3 was determined by radioimmunoassay (Diasorin, Wokingham, UK). The interassay CVs were 15 and 11% at serum concentrations of 13.8 and 53.7 ng/ml, respectively. The normal range for the assay was 8 50 ng/ml. Serum samples for the determination of the markers of bone turnover were frozen at 20 C until analysis. Bone turnover was determined by measuring serum collagen C- terminal propeptide (CICP), and the collagen cross-links PYR and DPD. Serum CICP was measured in duplicate using an immunometric assay (Prolagen C kit; Metra Biosystems, Great Haseley, UK). The reference ranges for adult women and men of serum CICP were and ng/ml, respectively. The interassay CVs were 6.4 and 3.0% at CICP concentrations of 4.6 and 27.1 ng/ml, respectively. Serum free PYR was measured using a modification of the urine pyrilinks assay (Metra Biosystems). Serial dilutions (1:2, 1:5, and 1:10) of serum from 2 patients on HD and CAPD were carried out in assay buffer. PYR measurements were carried out on the undiluted and diluted serum samples. Parallelism on dilution was shown on both patients samples. Mean linear recovery over 1:2 and 1:5 dilutions was 105% (range %). A dilution factor of 1:5 instead of the recommended 1:10 dilution for urine was chosen, as it allowed the PYR concentrations in serum to fall within the linear part of the assay standard curve. The dilution factor was taken into account in the final calculation. The interassay CV was less than 10%. The mean serum free PYR in normal subjects has been previously reported as 1.9 B 0.4 nmol/l [11]. Serum free DPD was measured by an in-house method using a polyethylene glycol assisted double antibody radioimmunoassay consisting of a primary sheep polyclonal antibody against DPD and a donkey antisheep secondary antibody. An iodinated DPD conjugate was used as tracer. The reference range for the assay using 120 healthy volunteers (46 men and 74 premenopausal women) was nmol/l. The interassay CVs were 10.0, 8.4, and 9.7% at 1.6, 3.5, and 6.2 nmol/l, respectively. Bone Densitometry Only 38 of the 59 patients (64.4%; HD: n = 8, CAPD: n = 30) attended for BMD measurements. BMD was determined by dualenergy X-ray absorptiometry using the Hologic QDR 1000/W scanner (Waltham, Mass., USA) at the three skeletal sites of functional importance: neck of femur, lumbar spine, and forearm (ultradistal and distal one third). These skeletal sites have a different proportion of cortical and trabecular bone. The measurement CV was 1% at forearm and spine and 1.8% at the femoral neck. To take into account the age- and sex-related changes in BMD, the results were expressed as Z scores, the number of standard deviations of the measured value from the mean value of age- and sex-matched normal 96 Nephron 2002;91: Hampson/Vaja/Evans/Chesters/Pettit/ Evans/Thomas/Seed/Fraser

4 Table 2. Laboratory data in patients on HD and CAPD (mean B SEM) HD patients CAPD patients p Serum urea, mmol/l 29.4B B Serum creatinine, Ìmol/l 942B82 802B33 NS Serum calcium, mmol/l 2.31B B0.03 NS Serum phosphate, mmol/l 1.88B B0.09 NS Serum PTH, pmol/l 25.7B8.5 33B5.1 NS Serum 25-hydroxyvitamin D 3, ng/ml 10B B1.4 NS Serum PYR, nmol/l 105B B Serum DPD, nmol/l 31.0B2.4 24B Serum CICP, ng/ml 170B B9.68 NS CICP/PYR ratio 2.2B B The differences between groups were analyzed by regression with robust standard errors (corrected for non-normality). NS = Not significant. subjects, established by the manufacturers (Hologic database), and included in the scanner s software. In addition, to assess whether the Hologic normal data were similar to those of our normal population, 40 subjects (20 male, 20 female) from our population in Cardiff with no evidence of renal disease, aged 50 B 0.85 years, were chosen as normal controls and had BMD measured at lumbar spine and femoral neck. Statistical Analyses Statistical analyses were made using the statistical software Stata 5 (Stata, College Station, Tex., USA). Comparisons were made between dialysis types (HD or CAPD) for age, gender, and underlying nephropathy. Differences in the biochemical markers of bone turnover and BMD by dialysis type were estimated separately and following correction for age and gender using multiple regression. For continuous variables, differences in the mean values were analyzed using robust linear regression with dummy variables. Linear regression analysis was performed to assess the relationship between the biochemical markers of bone turnover and BMD. Results Patient Characteristics Although the mean age was higher in patients on HD than in those on CAPD, this difference was not statistically significant, as shown by the confidence intervals (table 1). There was no significant difference in the prevalence of the causes of ESRD and in the time on dialysis between the two groups. This was also maintained in the 38 patients who had BMD measurements Biochemistry No significant differences in the blood concentrations of total calcium, albumin, phosphorus, creatinine, and 25- hydroxyvitamin D 3 were seen between the HD and CAPD groups (table 2). Serum urea was significantly higher in the HD subjects (p = 0.02). There was no significant difference in PTH concentrations between the patients on the two dialysis modalities. The serum CICP concentrations did not differ between the two groups. The serum CICP levels were higher than the upper range of normal in only 30 and 17% of the patients on HD and CAPD, respectively. The mean serum PYR concentration was significantly higher in HD as compared with CAPD patients (p = 0.007). The serum DPD level was also significantly higher in the HD patients as compared with the CAPD group (p = 0.027). The ratio of serum CICP/PYR was significantly lower in the HD patients (p = 0.002). The results were unaffected following correction for age and gender. There was no significant difference in any of the measurements (1) between diabetic and non-diabetic subjects and (2) between male and female patients in the CAPD group, as shown in table 3. Clearance of Serum Cross-Links There was no significant difference in the predialysis serum concentrations of DPD and creatinine between HD and CAPD patients participating in the clearance study (mean B SEM): DPD HD 23.9 B 1.5 nmol/l, CAPD 21.5 B 2.25 nmol/l; creatinine HD 950 B 299 Ìmol/l, CAPD 1,091 B 120 Ìmol/l. In the HD patients, serum DPD fell significantly after a dialysis session in parallel with creatinine (fig. 1). The mean (B SEM) percentage decreases in serum DPD and creatinine were 55.5 B 3.5 and 59.2 B 2.3, respectively. The same pattern was observed with PYR. In the CAPD group, as expected, Bone Turnover in Patients on Dialysis Nephron 2002;91:

5 Fig. 1. Clearance of serum DPD and creatinine by HD (a, c) and CAPD (b, d). there was no significant drop in serum DPD and creatinine following the peritoneal equilibration test (fig. 1). However, DPD did cross the peritoneal dialysis membrane, as the concentration of DPD in the dialysate fluid at 4 h was B 1.95 nmol/l. There was a close correlation between dialysate DPD and creatinine concentrations in the HD and CAPD patients (r = 0.90 and 0.76, respectively). The dialysate DPD/creatinine ratio corrected for their initial serum concentrations did not differ between the HD and CAPD patients: HD 0.85 B 0.066, CAPD 0.74 B (p = 0.22). Bone Mineral Density Figure 2 illustrates the mean Z score of the subjects according to dialysis modalities. Both groups displayed a negative Z score which was more evident at femoral neck and forearm when compared to the normal population using the Hologic database. This was confirmed when the Z scores at lumbar spine and femoral neck from the patients with ESRD were compared with our own control population. No significant difference was observed in BMD values at femoral neck and lumbar spine, when HD patients were compared with CAPD patients, although Table 3. Statistical analysis of any difference in the biochemical bone markers by gender and between the diabetic and non-diabetic subjects Male/female ratio PYR (ß-coefficient) 12.2 (8.4) p = Diabetes mellitus 6.7 (7.9) p = DPD (SE) 6.0 (3.7) p = (2.6) p = CICP (SE) 16.2 (17.5) p = (17) p = the mean Z score was lower at the lumbar spine. At the ultradistal forearm, the BMD was significantly reduced in the HD patients (p = 0.02). However, at the distal one third of the forearm, a compartment consisting of mostly cortical bone [4], there was no difference in BMD between the two groups. No significant differences in the Z scores at the three skeletal sites were observed between male and female and diabetic and non-diabetic subjects (table 4). 98 Nephron 2002;91: Hampson/Vaja/Evans/Chesters/Pettit/ Evans/Thomas/Seed/Fraser

6 Fig. 2. a BMD at lumbar spine and femoral neck expressed as Z scores in patients on HD and CAPD and in control subjects. b The BMD was significantly lower at the ultradistal forearm in patients on HD as compared with those on CAPD. * p = Table 4. Statistical analysis of any difference in BMD expressed as Z scores at lumbar spine, neck of femur, and forearm by gender and diabetic status BMD lumbar spine L1 L4 (ß-coefficient) neck of femur (SE) forearm (ultradistal) (SE) Male/female ratio (0.728) p = (0.53) p = (0.45) p = Diabetes mellitus 1.01 (0.67) p = (0.53) p = (0.48) p = Regression Analysis Linear regression analysis showed a positive correlation between serum PYR and DPD (r = 0.78, p = ; fig. 3). There was a weaker correlation of serum PYR with serum CICP (r = 0.32; p = 0.02). A weak but significant correlation was observed between PTH and serum PYR (r = 0.27, p = 0.05). The serum PTH did not correlate with DPD or CICP. In addition, 5 of the 14 (36%) CAPD patients who had follow-up blood samples at 3 months had more than a 30% change in serum PTH concentra- Bone Turnover in Patients on Dialysis Nephron 2002;91:

7 Fig. 3. Correlation between serum PYR and DPD (r = 0.78; p = ). tions: 61 B 12.2%. We observed no corresponding changes in their serum PYR and DPD concentrations. Regression analysis of Z scores at the skeletal sites with the biochemical markers of bone resorption revealed a negative correlation between Z scores at the ultradistal forearm with serum PYR and DPD (r = 0.35, p = 0.041, and r = 0.33, p = 0.049, respectively). No significant correlation was found between PYR and DPD at any other sites (DPD/PYR vs. BMD in lumbar spine and neck of femur r = and 0.1, respectively). However, a negative correlation was noted between serum CICP and BMD at the femoral neck (r = 0.33, p = 0.049). No significant correlation was seen between PTH and BMD at any site. Discussion This study demonstrates that the serum concentrations of PYR and DPD are negatively correlated with BMD at the ultradistal forearm in patients with ESRD. Patients on HD had a lower BMD at this site and higher serum PYR and DPD levels than patients maintained on CAPD. Measurement of the biochemical markers of bone resorption may be useful in patients on dialysis. Bone-specific alkaline phosphatase has previously been shown to be a good marker of the bone formation rate in dialysis patients [12], and its diagnostic efficiency in the diagnosis of low turnover bone disease has been previously evaluated [8]. In the present study, however, we wanted to examine the specific markers of bone resorption, as these have been less well studied than alkaline phosphatase. Higher serum PYR levels, a marker of bone resorption, have previously been shown to be associated with high-turnover renal bone disease in HD subjects [11]. These authors also observed that serum PYR correlated better with bone histology than PTH. Circulating PYR and DPD concentrations have also been shown to increase progressively with the duration of renal failure and dialysis in patients on HD and CAPD [14]. In the present study, the HD patients had significantly increased levels of the biochemical markers of bone resorption as compared with the CAPD patients despite no significant differences in serum PTH levels and proportion of patients on calcitriol and calcium carbonate between the two groups. In contrast to other studies [13, 15], serum CICP, another marker of bone turnover, was elevated in only a small proportion of patients on the two dialysis modalities. This may be explained by concomitant calcitriol therapy in the majority of our patients, as the concentrations 100 Nephron 2002;91: Hampson/Vaja/Evans/Chesters/Pettit/ Evans/Thomas/Seed/Fraser

8 of CICP have been shown to be modulated by calcitriol [15]. Although serum CICP was not different between the two groups, the ratio serum CICP/PYR was significantly lower in the HD patients, suggestive of an imbalance of bone turnover in this group of patients. The higher serum PYR and DPD concentrations may be related to increased bone resorption in this patient group, as previously documented [11, 14]. Increased bone breakdown may be multifactorial in its pathogenesis and merits further investigation. One possible factor which could be implicated is immune system alterations. Cytokines such as interleukin 6 and tumour necrosis factors, which act directly on bone turnover, have been shown to have a more severe effect in patients on HD as compared with those on CAPD, as these inflammatory mediators may be activated during HD sessions [16]. Other possible contributory factors for the difference in the biochemical markers of bone resorption between the two groups are that firstly residual renal function may be better preserved in CAPD patients [4] or secondly they are in a less catabolic state as suggested by the lower serum urea in the CAPD group. The observed difference in bone resorption markers is unlikely to be due to the differences in underlying nephropathy between the groups, as there was no significant difference in the proportion of subjects with the different causes of ESRD. Diabetes mellitus has been shown to be an important risk factor for the development of adynamic bone disease [10]. Although 40% of the CAPD patients were diabetics as compared with 29% in the HD group, this is unlikely to have accounted for the lower bone resorption in the CAPD patients, as we observed no significant difference in bone turnover between the diabetic and non-diabetic subjects. Although retention of the markers of bone resorption as a result of impaired excretion may contribute in part to the measured concentrations [11], this cannot explain the marked difference in the biochemical markers between the two groups of patients in our study, as there were no differences in serum creatinine concentrations between the CAPD and HD subjects. Moreover, the difference in serum PYR and DPD between the two groups was still significant following correction for serum creatinine and body mass index. There are very few studies on the handling characteristics of PYR and DPD by the two dialysis modalities. PYR and DPD as small molecules (molecular weight Da) should be effectively removed by both HD and CAPD. Our results show firstly that the clearance of these markers is similar to that of creatinine as previously described [14]; secondly, the type of dialysis membrane is unlikely to have a major influence on the differences in the circulating levels of these resorption markers between HD and CAPD patients. We observed no significant difference in the clearance of DPD by the CAPD membrane and the membrane used for HD, although the time of sampling may be important, particularly in HD patients. Measurement of serum PYR and DPD may thus be useful in the assessment of bone resorption in chronically dialyzed patients. The biochemical marker of bone resorption DPD has previously been shown to be a good predictor of BMD in other populations such as postmenopausal women [7]. Our study demonstrates that elevated circulating concentrations of the markers of bone resorption have a negative influence on BMD at the forearm, implying that these biochemical markers may be predictive of bone density in dialysis patients too. Indeed, low bone mass and generalized loss of mineralized bone have been reported in patients on dialysis [3, 17]. This may account for the increased fracture risk observed in this population. Appendicular cortical bone measured by single-photon absorptiometry has been shown to be reduced in patients with clinically significant renal osteodystrophy on maintenance HD, irrespective of the histological classification (osteitis fibrosa or osteomalacia) [3]. Although bone densitometry cannot discriminate between the different types of renal osteodystrophy, it is useful in detecting a reduction in bone mass or a lack of mineralization at the different skeletal sites [18]. Both these factors are associated with an increased fracture risk. This present study shows a reduction in BMD at all sites in both groups of patients. The BMD at the distal forearm (ultradistal site), which contains mostly trabecular bone, was significantly lower in the HD patients. A similar trend was observed at the lumbar spine, although the results failed to reach significance. This may be partly due to artefactual increases in the BMD at the spine because of the presence of aortic calcifications and/or osteosclerosis which is very often seen in patients with hyperparathyroidism [2]. This is in keeping with previous reports that the trabecular bone compartment is better studied at the ultradistal site of the forearm [19, 20]. In addition, female patients with ESRD tend to have a lower BMD in the trabecular bone compartment due to oestrogen deficiency. However, in the present study, we observed no significant differences in BMD expressed as Z scores and bone turnover between the male and female patients. In contrast to other studies [4], we also observed no difference in BMD at the distal third of the forearm (mostly cortical bone) between the two groups. However, our results indicate that preservation of cortical bone in CAPD patients, as previously Bone Turnover in Patients on Dialysis Nephron 2002;91:

9 reported [4], may be due to less bone degradation in this group. Despite its limitations due to the small sample size, this cross-sectional study demonstrates that the negative relationship between the biochemical markers of bone resorption and BMD is also observed in dialysis patients. HD patients have lower BMD than CAPD patients, and this may be attributed to increased bone breakdown in HD patients. We suggest that measurement of these two non-invasive parameters, in addition to PTH, may have a potential value in the identification of a subgroup of patients with increased bone loss and low BMD and who may, therefore, require bone biopsies for a precise diagnosis of the bone remodeling status and thus the application of the correct therapeutic regimen. However, the use of PYR and DPD for this clinical purpose requires further studies, in particular longitudinal ones, to evaluate the appropriate cutoff values. Acknowledgment This work was supported by a grant from the Welsh Scheme for the Development of Health and Social Research awarded to Dr. G. Hampson. References 1 Malluche HH, Langub MC, Monier-Faugère MC: Pathogenesis and histology of renal osteodystrophy. Osteoporos Int 1997;7(suppl 3): Piraino B, Chen T, Cooperstein L, Segré G, Puschett J: Fractures and vertebral bone mineral density in patients with renal osteodystrophy. Clin Nephrol 1988;30: Schober HC, Han ZH, Foldes AJ, Shih M, Rao DS, Balena R, Parfitt AM: Mineralized bone loss at different sites in dialysis patients: Implications for prevention. J Am Soc Nephrol 1998;9: Mottet JJ, Horber FF, Casez JP, Descœudres C, Jaeger P: Evidence for preservation of cortical bone mineral density in patients on continuous ambulatory dialysis. J Bone Miner Res 1996;11: Cummings SR, Black DM, Nevitt MC, Browner W, Cauley J, Ensrud K, Genant HK, Palermo L, Scott J, Vogt TM: Bone density at various sites for prediction of hip fractures. Lancet 1993;341: Hruska KA: Renal osteodystrophy. Baillières Clin Endocrinol Metab 1997;11: Melton LJ, Khosla S, Atkinson EJ, O Fallon WM, Riggs BL: Relationship of bone turnover to bone density and fractures. J Bone Miner Res 1997;12: Urena P, De Vernejoul MC: Circulating biochemical markers of bone remodelling in uremic patients. Kidney Int 1999;55: Malluche HH, Langub MC, Monier-Faugère MC: The role of bone biopsy in clinical practice and research. Kidney Int Suppl 1999;73: Sherrard D, Hercz G, Pei Y, Maloney N, Greenwood C, Manuel A, Saiphoo C, Fenton S, Segré G: The spectrum of bone disease in endstage renal failure: An evolving disorder. Kidney Int 1993;43: Urena P, Ferreira A, Kung VT, Morieux C, Simon P, Ang KS, Souberbielle JC, Segré GV, Drüeke TB, De Vernejoul MC: Serum pyridinoline as a specific marker of collagen breakdown and bone metabolism in haemodialysis patients. J Bone Miner Res 1995;10: Urena P, Hruby M, Ferreira A, Ang KS, De Vernejoul MC: Plasma total versus bone alkaline phosphatase as markers of bone turnover in haemodialysis patients. J Am Soc Nephrol 1996;7: Mazzaferro S, Pasquali M, Ballanti P, Bonucci E, Costantini S, Chicca S, De Meo S, Perruza I, Sardella D, Taggi F, Coen G: Diagnostic value of serum peptides of collagen synthesis and degradation in dialysis renal osteodystrophy. Nephrol Dial Transplant 1995;10: Ibrahim S, Mojiminiyi S, Barron JL: Pyridinium crosslinks in patients on haemodialysis and continuous ambulatory peritoneal dialysis. Nephrol Dial Transplant 1995;10: Coen G, Ballanti P, Mazzaferro S, Pasquali M, Bonucci E: Procollagen type 1 C-terminal extension peptide, PTH and 1,25(OH) 2 D 3 in chronic renal failure. Bone 1993;14: Monier-Faugère MC, Malluche HH: Role of cytokines in renal osteodystrophy. Curr Opin Nephrol Hypertens 1997;6: Gabay C, Ruedin P, Slosman D, Bonjour JP, Leski M, Rizzoli R: Bone mineral density in patients with end-stage renal failure. Am J Nephrol 1993;13: Johnson DW, McIntyre HD, Brown A, Freeman J, Rigby RJ: The role of DEXA bone densitometry in evaluating renal osteodystrophy in continuous ambulatory peritoneal dialysis patients. Perit Dial Int 1996;16: Fletcher S, Jones RG, Rayner HC, Harnden P, Hordon LD, Aaron JE, Oldroyd B, Brownjohn AA, Turney JH, Smith MA: Assessment of renal osteodystrophy in dialysis patients: Use of bone alkaline phosphatase, bone mineral density and parathyroid ultrasound in comparison with bone histology. Nephron 1997;75: Joffe P, Heaf JG, Jensen C: Can bone histomorphometry be predicted by clinical assessment and non-invasive techniques in peritoneal dialysis? Miner Electrolyte Metab 1996;22: Nephron 2002;91: Hampson/Vaja/Evans/Chesters/Pettit/ Evans/Thomas/Seed/Fraser