Endocrine Unit, Division of Pediatrics, Department of Reproductive Medicine and Pediatrics, University of Pisa, I-56125 Pisa, Italy
Address correspondence to: Giampiero I. Baroncelli, Endocrine Unit, Division of Pediatrics, Department of Reproductive Medicine and Pediatrics, University of Pisa, Via Roma 67, I-56125 Pisa, Italy. E-mail: g.baroncelli{at}med.unipi.it.
To the editor:
In their recent article in JCEM, Schweizer et al. (1) reported that children with GH deficiency have normal cortical bone density (CD) and total area, reduced cortical thickness (CT) and cortical area (CA), and increased bone marrow area at diagnosis by peripheral quantitative computed tomography (pQCT) measurement at proximal radius (65% site). After 12 months of GH treatment, CD decreased, total area increased, and CT, CA, and bone marrow area did not change. These results markedly differ from previous data showing reduced radial and lumbar bone mineral density (BMD) at diagnosis, which improved during treatment, by using two-dimensional densitometric techniques such as single-photon absorptiometry and dual energy x-ray absorptiometry (DXA) (2).
Schweizer et al. (1) suggested that the reduced CT and CA are likely the cause of the low BMD values reported in children with GH deficiency at diagnosis. However, DXA does not give information on bone structure, but only projected area can be assessed by this method, which reflects the subjects bone size (3). Moreover, on the basis of their hypothesis, the increased BMD values by DXA reported in children with GH deficiency during treatment (2) should be due to increased CT and CA, but both of these parameters of bone structure did not vary during treatment in their patients. Do the authors have an explanation of this discrepancy?
In patients with Turner syndrome, by using pQCT at 65% site of the proximal radius, it has been demonstrated that decreased CT and CA were the main cause of reduced volumetric density (4). Which may be the explanation of the fact that the patients with GH deficiency have a reduced CT and CA but a normal CD at diagnosis?
Schweizer et al. (1), on the basis of the increased values of biochemical bone markers during treatment, suggested that the decreased CD observed during therapy could be explained by the strong increase in bone remodeling and remodeling space, as reported in adults with GH deficiency. How can this hypothesis be sustained considering the fact that no significant correlations were found between changes in biochemical bone markers and changes in bone structure? It must also be kept in mind that in children with GH deficiency during GH treatment, biochemical bone markers reflect not only the remodeling process, but also the process of linear growth and modeling (5).
Schweizer et al. (1) reported that although their patients had reduced CD and increased bone turnover associated with rapid growth rate during the first year of GH treatment, there were no data indicating that the rate of fractures was increased in GH-deficient children. The authors did not mention that children with GH deficiency have approximately a 4-fold decreased fracture frequency during treatment (including the first year of therapy) compared with healthy controls (6), a likely consequence of a protective effect of GH treatment on bone strength.
Schweizer et al. (1) stated that the measurements of the densitometric parameters were transformed into Z-score values based on the age-specific references (7). The age of the examined patients ranged from 3.3 to 14.4 yr, but the age of normal subjects used for comparison ranged from 6 to 40 yr. How did the authors calculate Z-score values in patients less than 6 yr old? Moreover, some important methodological aspects should be also considered. Children under 6 yr of age usually are not able to keep their arms in a fixed position for the time of measurement (7). How did the authors assess pQCT in the youngest patients? The age of 67 yr also appears to be the lowest at which cortical bone analysis is technically feasible at the 65% site of the radius (7). In addition, one limitation of the pQCT equipment is the partial volume effect, which interferes mostly with the analysis of CD, because this effect is mainly evident in children with short stature who have very thin CT (3). The authors did not discuss these important methodological aspects.
In addition to height age, do the authors have information about the correction of the densitometric data for bone age? Moreover, the authors did not mention whether some of the older patients had pubertal development during GH treatment, because the changes in bone structure occurring during this period of life may be an additional factor influencing their results.
Received December 11, 2003.
References