Bone mineral density changes within six months of renal transplantation : Transplantation

Journal Logo

Clinical Transplantation

Bone mineral density changes within six months of renal transplantation

Mikuls, Ted R.1; Julian, Bruce A.2; Bartolucci, Al3; Saag, Kenneth G.4

Author Information
  • Free


Osteoporosis leading to fracture represents a major source of morbidity complicating solid organ transplantation (1–5). In prior studies of renal allograft recipients, vertebral bone mineral density (BMD) decreased by approximately 3% to 7% in the first 6 to 12 months after transplantation (6–8), with significant declines evident as early as 1 month (8). Highly consistent with these data, incident fractures occur at least five times more frequently in renal transplant patients than otherwise healthy adults (5).

Declines in BMD after kidney transplantation are significantly correlated with higher daily and cumulative glucocorticoid doses (9). Improvements in effective immunosuppressive regimens, including the use of mycophenolate mofetil in place of azathioprine, have led to improved allograft survival and lower cumulative glucocorticoid doses. Although the negative impact of glucocorticoids on bone is well known, the effect of other immunosuppressive agents and other disease and demographic factors in transplant-associated bone loss is less well defined. We characterized bone health and determined predictors of BMD change in an inception cohort of renal transplant patients receiving an immunosuppressive regimen consisting primarily of glucocorticoids, cyclosporine A (CsA), and mycophenolate mofetil. We hypothesized that renal transplant recipients would lose significant bone mass within 6 months of engraftment and that such bone loss would be primarily mediated by cumulative glucocorticoid use.



From September 1999 to November 2001, we studied unselected recipients of renal allografts from either cadaveric or living donors at the University of Alabama at Birmingham. All study participants (n=45) were undergoing their first transplantation and were older than 18 years of age. Patients undergoing simultaneous kidney-pancreas transplants were excluded from participation. The University of Alabama at Birmingham Institutional Review Board for Human Use approved the study, and each patient provided informed written consent.


Immunosuppressive therapy consisted of initial high-dose glucocorticoids and anti-T-lymphocyte antibody (daclizumab) for induction and mycophenolate mofetil (2 g/day), CsA, and low- to intermediate-dose glucocorticoids for maintenance. Anti-T-lymphocyte antibody was administered intraoperatively (1 mg/kg) and redosed (1 mg/kg) on postoperative day 5. Glucocorticoids were administered initially as methylprednisolone 500 mg intraoperatively, 250 mg on the next day, and then 100 mg on the second postoperative day. In the absence of acute rejection, the methylprednisolone/prednisone was tapered to 30 mg per day by postoperative day 10. For most patients, prednisone was gradually reduced during the next 8 to 12 weeks to a maintenance dose of 7.5 to 10 mg per day. CsA treatment was begun at a dose of approximately 6 to 8 mg/kg in divided doses when satisfactory renal function was established and subsequently adjusted according to serial trough drug levels. Tacrolimus was substituted for CsA in four patients during the course of follow-up (three patients initially received CsA and were switched to tacrolimus secondary to CsA toxicity or rejection episodes). All episodes of acute graft rejection were treated with high-dose glucocorticoids (methylprednisolone 500 mg/day for 3–4 days). Primary graft dysfunction was defined as the need for hemodialysis within 2 weeks of engraftment. To determine cumulative glucocorticoid dose, methylprednisolone was converted to prednisone equivalents by multiplying the dose by a factor of 1.25.

Bone Densitometry and Laboratory Measurements

Baseline BMD and laboratory measurements were obtained a median of 16 days posttransplant (range 9–33 days) and repeated a median of 5.7 months posttransplant (range 4.8–9.3 months). BMD measurements of the femoral neck and lumbar spine were measured using dual-energy x-ray absorptiometry (DXA) with a Hologic QDR 4500 (Hologic Inc., Bedford, MA) and the National Health and Nutrition Examination Survey reference database. Osteoporosis was defined by the World Health Organization criterion as a BMD value of 2.5 or more SD below the race- and gender-matched young adult mean. Osteopenia was defined as a BMD value of 1 or more SD and less than 2.5 SD below this mean. DXA precision error estimates were 0.98% and 1.39% for the spine and femoral neck, respectively. Intact parathyroid hormone (iPTH) and calcitriol (1,25[OH]2-Vitamin D) levels were measured using a commercially available immunometric assay (iPTH; Immunolite, DPL, Los Angeles, CA) and radioimmunoassay (calcitriol; Quest, Atlanta, GA). The reference range for iPTH was 10 to 65 pg/mL and for calcitriol was 24 to 65 pg/mL. Urinary N-telopeptide (NTX) levels were measured with a commercially available kit (Specialty Laboratories, Los Angeles, CA; reference range <65 nM BCE/mM).

Statistical Analysis

Baseline BMD and laboratory values were compared with 6-month follow-up values using nonparametric statistics (signed-rank test). Three patients had no baseline calcitriol or urinary NTX values. Follow-up iPTH, urinary NTX, and calcitriol levels were not available for 7, 8, and 17 patients, respectively. Dialysis duration, baseline urinary NTX, iPTH, calcitriol levels, and cumulative prednisone and CsA dose were categorized into tertiles to satisfy the linearity assumption of the logit. To examine the potential impact of menopausal status on bone loss, we categorized gender into the following groups: male, premenopausal female, and postmenopausal female. Pearson correlation coefficients were calculated to assess the association between cumulative prednisone and CsA dosages. Odds ratios (ORs) and 95% confidence intervals (CIs) were used to describe associations between predictor variables and significant spinal BMD declines (≥3%) and were determined with logistic regression. A 3% BMD decline at the lumbar spine was chosen as the minimally significant detectable difference at this site based on the precision error of our DXA. All analyses were adjusted for follow-up duration. Variables with at least borderline associations (P ≤0.25) were included in a multivariable model and subsequently eliminated in a stepwise fashion until all remaining variables had associated P values of 1.0 or less. In addition, to assess for a potential effect modification of iPTH on cumulative prednisone dose, one-way interactions were examined in the final multivariable model. Goodness of fit and model calibration were assessed using Hosmer-Lemeshow and c statistics, respectively (10,11). All analyses were performed with SAS for Windows, version 8.0 (SAS Inc., Cary, NC).


Patient Characteristics

A summary of baseline patient characteristics is shown in Table 1. Of 19 women, 12 were postmenopausal with a median age at onset of 47 years (range 32–52). Two postmenopausal women were taking hormone replacement therapy at the time of transplantation. In addition, a minority of patients reported pretransplant use of bisphosphonates (n=2), thyroid replacement medications (n=5), calcium (n=9), or vitamin D supplements (n=3). Nine patients received calcium supplements between the time of transplantation and the 6-month study visit. In regard to sex hormone status, three men reported a history of infertility, and three women reported a history of amenorrhea.

Table 1:
Table 1. Baseline characteristics of renal transplant recipients (N=45)

Ten patients (22%) underwent preemptive transplantation, receiving a kidney allograft before requiring dialysis therapy. Of the 35 patients who received pretransplant dialysis, 24 received only hemodialysis, five received only peritoneal dialysis, and six received both hemodialysis and peritoneal dialysis. The duration of pretransplant dialysis therapy was longer for those receiving an allograft from a cadaveric donor compared with those receiving a kidney from a living donor (38.9±29.5 vs. 11.0±16.4 months;P =0.001). Causes of end-stage renal disease included hypertension (n=12), diabetes mellitus (n=2), the combination of hypertension and diabetes (n=4), polycystic kidney disease (n=10), immunoglobulin A nephropathy (n=3), systemic lupus erythematosus (n=2), congenital disease (n=1), reflux nephropathy (n=1), focal segmental glomerulosclerosis (n=1), unspecified glomerulonephritis (n=4), and idiopathic reasons (n=5). There were no episodes of primary graft dysfunction postengraftment.

Immunosuppressive Therapies

During the follow-up period, patients used a mean cumulative dose of prednisone 4.2±1.1 g (median 3.9 g, range 2.9–9.9 g). During the study period, only one patient was diagnosed with graft rejection and given pulse glucocorticoid therapy. The mean cumulative CsA dose was 59.9±22.6 g (median 58.7 g, range 0–120.1 g). After excluding the four patients who received tacrolimus, there was a modest inverse association between cumulative prednisone and CsA dose (r =−0.22). For the four patients in whom tacrolimus was substituted for CsA, the mean cumulative tacrolimus dose was 1.1±1.0 g (median 0.7 g, range 0.4–2.5 g).

Laboratory Measurements and Bone Densitometry

Baseline and 6-month BMD, calcitriol, iPTH, and urinary NTX values are summarized in Table 2. At baseline, a majority (56%) had BMD-defined osteopenia at the femoral neck, whereas a smaller proportion (9%) had osteoporosis. In contrast, a majority of patients (64%) had normal BMD values at the lumbar spine compared with 29% with osteoporosis and 7% with osteopenia. At follow-up, subjects had lost a mean of 2.4% BMD at the lumbar spine (P =0.003) but did not experience significant declines at the femoral neck. Nineteen subjects (42%) lost more than 3% lumbar BMD during the 6-month follow-up period.

Table 2:
Table 2. Mean (SD) bone mineral density, intact parathyroid hormone calcitriol, and urine N-telopeptide values for patients undergoing renal transplantation at baseline and at 6-month follow-upa

A high proportion of patients (84%) demonstrated elevated iPTH levels at baseline. In fact, more than one half demonstrated baseline iPTH values exceeding twice the upper limit of normal. In contrast, baseline calcitriol levels were below the lower limit of normal in 16 patients (36%). Limiting the analysis to those in whom both baseline and follow-up measures were available, there was a 35% decrease in the mean iPTH (P <0.001), a 15% decrease in mean urine NTX (P =0.08), and a 30% increase in mean calcitriol (P =0.02) posttransplantation (Table 2).

Associations with Spinal Bone Loss

We observed no association of follow-up time (either time between DXA scans or time from engraftment to the initial DXA scan) with bone loss. In the initial univariate regression analyses (all analyses adjusted for duration between DXA scans), only the highest tertiles of cumulative prednisone dose (OR=24.8; 95% CI 2.5–249 and OR=16.7; 95% CI 1.6–171) and prior alcohol use (OR=8.4; 95% CI 1.6–44.5) were significantly associated with spinal BMD loss (Table 3). Calcium supplement use, either before or after engraftment, did not seem to mitigate against bone loss. Greater cumulative CsA dose, receipt of a kidney from a living donor, female gender (more pronounced for premenopausal women), and a history of diabetes trended toward associations with less bone loss. In contrast, there were nonsignificant trends toward more bone loss associated with non-white race, past smoking, longer duration of pretransplant dialysis, and lower baseline calcitriol levels. However, these factors were not significant in the final multivariable model, which included only cumulative prednisone dose and past alcohol use (Table 4). These results were the same when patients receiving tacrolimus were excluded from the analysis. There was no evidence of a significant effect modification of iPTH on cumulative prednisone dose (data not shown).

Table 3:
Table 3. Univariate associations of patient characteristics and treatment variables with significant reductions (≥3%) in lumbar bone mineral density after kidney transplantation
Table 4:
Table 4. Multivariable predictors of lumbar bone mineral density decline (≥3%) after kidney transplantation


In our prospective cohort of 45 new renal transplant recipients, more than 40% of patients showed significant lumbar BMD loss within 6 months of transplantation. Lumbar bone loss was most strongly associated with cumulative glucocorticoid dose and history of alcohol use. We did not, however, observe significant changes in femoral neck BMD in the first 6 months after kidney transplantation. Although baseline levels were not predictive of spinal bone loss, we observed significant declines in iPTH, significant increases in calcitriol, and a trend toward decreased urine NTX posttransplantation.

Our findings corroborate those of several previous studies showing significant spinal BMD reductions after renal transplantation (6–9,12–20). Many of these studies were cross-sectional (9,15–20), and of the available prospective investigations, some examined patients at various time intervals after transplantation (13,14). A prospective study design with near uniform follow-up, such as ours, substantially increases the ability to detect causal associations between transplant-related factors and BMD loss.

In contrast with our study, older renal transplant inception cohorts (6–8,12) used maintenance immunosuppressive regimens that included azathioprine rather than mycophenolate mofetil. This may be salient because newer more effective immunosuppressive regimens may reduce cumulative glucocorticoid requirements. Although its potential effect on BMD in humans is not well defined, animal studies indicate that azathioprine therapy may impair osteoblast function and increase the number of osteoclasts at sites of bone resorption (21). Similar findings have not been reported with mycophenolate mofetil (22). Although we observed a protective trend associated with a higher cumulative CsA dose, the role of CsA in transplant-associated bone loss remains unclear (23,24). Although CsA greatly accelerates bone turnover in experimental animals (25), in vitro studies indicate that CsA may inhibit bone resorption (26). Moreover, CsA use in the transplant setting may reduce glucocorticoid requirements, thus having a net positive effect on bone.

The 6-month vertebral BMD decrement in our study (2.4%) approximates that reported by Kwan et al. (2.8%) in their small prospective cohort of renal allograft recipients (6). Recognizing substantial overlap of confidence limits and significant differences in design when comparing studies, other inception cohorts have shown mean spinal BMD losses as high as 1.6% per month (8). A cumulative 6-month loss as great as 6.8% was reported by a member of our study team in a prior prospective investigation also performed at the University of Alabama at Birmingham in 1991 (7). It is notable that in Julian et al.’s study, patients received a 6-month cumulative prednisone dose (3.8±1.1 g) slightly less than that received by our study subjects who received a mean cumulative dose of 4.2±1.1 g. This indicates that a changing pattern of cumulative glucocorticoid exposure may not alone explain the apparent improvement in bone outcome of our newer cohort compared with the former Alabama study.

Differences in our results compared with those of other studies may relate to other variations in immunosuppressive regimens, study populations, and methods of BMD measurement. Patients in the study by Julian et al. (7) were substantially younger (including only one postmenopausal woman) and demonstrated higher baseline spinal BMD values (1.17±1.11 g/cm2). In addition, BMD was measured in this early investigation using dual-photon absorptiometry. In recent years, this technique has largely been replaced by DXA, a technique with better precision and accuracy (27). Consistent with prior investigations, it is possible that our results may be influenced by disparities in the duration of follow-up, both in terms of time from engraftment to the initial DXA and time between study evaluations. However, it is unlikely that this variability represents a significant source of bias because neither interval was associated with BMD loss at a univariate level.

Although strongly associated with bone loss, glucocorticoid use explains only a portion of the bone loss that occurs posttransplant (20). Patients with end-stage renal disease have low bone mass and show increased bone loss caused by other factors such as secondary hyperparathyroidism and osteomalacia (28,29). We observed only a nonsignificant trend suggesting, but not confirming, an association between baseline calcitriol levels and vertebral bone loss. Similar to our work, a recent prospective cohort study (n=47) also found no association between decreased vertebral BMD and baseline iPTH levels (12). In contrast, a larger, longitudinal study involving 115 renal graft recipients followed at various time intervals posttransplant reported modest correlations between initial iPTH and spinal BMD declines (r =0.33, P <0.001) (14). These data, coupled with results from our study, indicate that pretransplant iPTH and calcitriol levels are, at best, only modest predictors of the vertebral bone loss that characterizes the early posttransplant period.

Although chronic alcohol use has been implicated as a risk factor for osteopenia, fracture (27), and osteonecrosis (30), this is the first study to our knowledge suggesting an association between prior and current alcohol consumption and bone loss immediately after renal transplantation. Although a novel finding, we cannot conclude with certainty that the association between alcohol consumption and vertebral BMD loss in this study results directly from this exposure. Although we controlled for a number of potential confounders, a history of alcohol use may be a surrogate marker for a more primary association closely associated with alcohol use. It is possible, for example, that reduced weight-bearing activity or suboptimal nutritional status posttransplant (factors that we did not measure) could explain this perceived association. In addition, because we did not anticipate this association a priori or collect detailed quantitative information on alcohol exposure, it should be conservatively viewed as a hypothesis-generating finding.

Another potential limitation of our study is that significant BMD losses occurred only in the lumbar spine. This finding differs from that of a small prospective study, which in addition to vertebral BMD loss showed significant declines in femoral neck BMD within 6 months after renal transplantation (6). However, the more rapid decline in vertebral BMD in our study is consistent with other investigations of glucocorticoid-induced bone loss. It is well known that glucocorticoid administration leads to preferential loss of trabecular bone followed only later by losses in cortical bone at sites such as the femoral neck (31–34). Similar to our study, Julian et al. observed significant BMD declines in the lumbar spine (an area rich in trabecular bone) with simultaneous increases in radial BMD (a region of primarily cortical bone) by 6 months after engraftment (7). Although our sample size is only moderate and our study is thus underpowered to assess subtle associations, it remains one of the larger prospective evaluations of renal transplant-associated bone disease.


Renal transplant patients showed a rapid and significant decline in vertebral BMD within 6 months after transplantation. Consistent with other investigations, the association of cumulative prednisone dose with spinal bone loss underscores the importance of glucocorticoid-sparing regimens in posttransplant management. The novel association of alcohol use with spinal BMD reduction, if confirmed by other investigators, indicates that abstinence from alcohol may represent an important, potentially modifiable cofactor in the pretransplant period. Given the rapidity with which bone loss occurs after renal transplantation, particularly in the lumbar spine, measures to prevent osteoporosis should be considered in the immediate postoperative setting.


The authors thank Ms. Mary Elkins and Ms. Amy Mudano for their assistance with data collection and data entry.


1. Lee A, Mull R, Keenan G, et al. Osteoporosis and bone morbidity in cardiac transplant recipients. Am J Med 1994; 96: 35–41.
2. Shane E, Rivas M, Staron R, et al. Fracture after cardiac transplantation: a prospective longitudinal study. J Clin Endocrinol Metab 1996; 81: 1740–1746.
3. Shane E, Rivas M, Silverberg S, et al. Osteoporosis after cardiac transplantation. Am J Med 1993; 94: 257–264.
4. Riemens S, Oostdijk A, van Doormaal J, et al. Bone loss after liver transplantation is not prevented by cyclical etidronate, calcium and alfacalcidol. Osteoporosis Int 1996; 6: 213–218.
5. Ramsey-Goldman R, Dunn J, Dunlop D, et al. Increased risk of fracture in patients receiving solid organ transplants. J Bone Miner Res 1999; 14: 456–463.
6. Kwan J, Almond M, Evans K, et al. Changes in total body bone mineral content and regional bone mineral density in renal patients following renal transplantation. Miner Electrolyte Metab 1992; 18: 166–168.
7. Julian B, Laskow D, Dubovsky J, et al. Rapid loss of vertebral mineral density after renal transplantation. N Engl J Med 1991; 325: 544–550.
8. Horber F, Casez J, Steiger U, et al. Changes in bone mass early after kidney transplantation. J Bone Miner Res 1994; 9: 1–9.
9. Grotz W, Mundinger F, Gugel B, et al. Bone mineral density after kidney transplantation. Transplantation 1995; 59: 982–986.
10. Harrell F, Lee K, Califf R, et al. Regression modeling strategies for improved prognostic prediction. Stat Med 1984; 3: 143–152.
11. Lemeshow S, Hosmer J. A review of goodness of fit statistics for use in the development of logistic regression models. Am J Epidemiol 1982; 115: 92–106.
12. Masse M, Girardin C, Ouimet D, et al. Initial bone loss in kidney transplant recipients: a prospective study. Transplant Proc 2001; 33: 1211.
13. Pichette V, Bonnardeaux A, Prudhomme L, et al. Long-term bone loss in kidney transplant recipients: a cross-sectional and longitudinal study. Am J Kid Dis 1996; 28: 105–114.
14. Grotz W, Mundinger F, Rasenack J, et al. Bone loss after kidney transplantation: a longitudinal study in 115 graft recipients. Nephrol Dial Transplant 1995; 10: 2096–2100.
15. Parker C, Freemont A, Blackwell P, et al. Cross-sectional analysis of renal transplantation osteoporosis. J Bone Miner Res 1999; 14: 1943–1951.
16. Braun W, Richmond B, Protiva D, et al. The incidence and management of osteoporosis, gout, and avascular necrosis in recipients of renal allografts functioning more than 20 years (level 5A) treated with prednisone and azathioprine. Transplant Proc 1999; 31: 1366–1369.
17. Kakado Y, Takahara S, Ichimaru N, et al. Factors influencing vertebral bone density after renal transplantation. Transpl Int 2000; 13 (Suppl 1): S431–435.
18. Heaf J, Tvedegaard E, Kanstrup I, et al. Bone loss after renal transplantation: role of hyperparathyroidism, acidosis, cyclosporine, and systemic disease. Clin Transplant 2000; 14 ( 5): 457–463.
19. Cayco A, Wysolmerski J, Simpson C, et al. Posttransplant bone disease: evidence for a high bone resorption rate. Transplantation 2000; 70 ( 12): 1722–1728.
20. Caglar M, Adeera L. Factors affecting bone mineral density in renal transplant patients. Ann Nucl Med 1999; 13 ( 3): 141–145.
21. Bryer H, Isserow J, Armstrong E, et al. Azathioprine alone is bone sparing and does not alter cyclosporin A-induced osteopenia in the rat. J Bone Miner Res 1995; 10: 132–138.
22. Dissanayake I, Goodman C, Bowman A, et al. Mycophenolate mofetil: a promising new immunosuppressant that does not cause bone loss in the rat. Transplantation 1998; 65: 275–278.
23. Callegari P. Is there cyclosporine-induced bone disease? J Clin Rheumatol 1997; 3: S93–S96.
24. Eptein S, Shane E, Bilezikian J. Organ transplantation and osteoporosis. Curr Opin Rheumatol 1995; 7: 255–261.
25. Movsowitz C, Epstein S, Fallon M, et al. Cyclosporin-A in vivo produces severe osteopenia in the rat: effect of dose and duration of administration. Endocrinology 1988; 123: 2571–2577.
26. Stewart P, Green O, Stern P. Cyclosporine A inhibits calcemic hormone-induced bone resorption in vitro. J Bone Miner Res 1986; 1: 285–291.
27. Morgan S, Saag K, Julian B, et al. Osteopenic bone diseases. In: Koopman W, ed. Arthritis and allied conditions, vol. 2 [ed. 14]. Philadelphia: Lippincott Williams & Wilkins, 2001, pp 2449–2513.
28. Rix M, Andreassen H, Eskildsen P, et al. Bone mineral density and biochemical markers of bone turnover in patients with predialysis chronic renal failure. Kidney Int 1999; 56: 1084–1093.
29. Stehman-Breen C, Sherrard D, Walker A, et al. Racial differences in bone mineral density and bone loss among end-stage renal disease patients. Am J Kidney Dis 1999; 33: 941–946.
30. Matuso K, Hirohata T, Sugioka Y, et al. Influence of alcohol intake, cigarette smoking and occupational status on idiopathic osteonecrosis of the femoral head. Clin Orthop 1988; 234: 115–123.
31. Dykman T, Gluck O, Murphy W, et al. Evaluation of factors associated with glucocorticoid-induced osteopenia in patients with rheumatic diseases. Arthritis Rheum 1985; 28: 361–368.
32. Laan RFJM, van Riel PLCM, van de Putte LBA, et al. Low-dose prednisone induces rapid reversible axial bone loss in patients with rheumatoid arthritis. Ann Intern Med 1993; 119: 963–968.
33. Laan R, Buijs W, van Erning L, et al. Differential effects of glucocorticoids on cortical appendicular and cortical vertebral bone mineral content. Calcif Tissue Int 1993; 52: 5–9.
34. Hansen M, Podenphant J, Florescu A, et al. A randomised trial of differentiated prednisolone treatment in active rheumatoid arthritis. Clinical benefits and skeletal side effects. Ann Rheum Dis 1999; 58: 713–718.
© 2003 Lippincott Williams & Wilkins, Inc.