Bone mineral content (BMC) was measured in the nondominant arm of 18 children (aged 37/12 to 17½ years) for a total 783 months after renal transplantation. Using photon absorptiometry, 89 measurements were made; 17 of the 18 patients had a functioning graft and one patient died. Significant demineralization, a BMC greater than -2 SD below appropriate control volumes, was found in 11 of 18 patients (62%) and 55 of 89 measurements (61%). Bone loss was progressive; among the 16 patients followed for more than 6 months, ten showed a decline of more than 0.5 SD in BMC, five had no change, and only one showed improvement. No relationship was found between BMC and the use of furosemide, type of transplant (15 living, seven cadaver), prior renal disease (six with glomerulonephritis, 11 tubulointerstitial), need for a second graft (five patients), chronic anticonvulsant therapy, or serum calcium and phosphate values. BMC was slightly correlated (P < .05) with alkaline phosphatase values. BMC was more strongly correlated with serum creatinine (y = -0.48x + 1.25, r = -.042, P < .001) and prednisone dose (mg/kg/d) (y = -0.65x + 0.481, r = -.543, P < .001) in an inverse relationship. Patients whose serum creatinine value was less than 1 mg/dL had a BMC of -0.71 ± 0.34 SD; those with serum creatinine value greater than 2 mg/dL had BMC of -3.32 ± 0.31 SD, different at P < .001. Patients receiving daily prednisone therapy had a significantly lower BMC than those receiving alternate-day therapy (-3.11 ± 1.23 SD v -1.72 ± 1.29 SD, P < .005).
