Radiographic absorptiometry of the phalanges in healthy children and in girls with turner syndrome

Although bone mineral status in children has been measured with various techniques, information about development of the actual bone mass density during childhood and adolescent growth is scarce. Our modified radiographic absorptiometry (RA) determines bone mass density (BMaD) three dimensionally at the diaphyseal and metaphyseal site of the middle phalanx of the left second digit, representing predominantly cortical (50% site) and trabecular bone compartments (25% site), respectively. The objectives of this study were to establish reference curves with 95% prediction intervals of BMaD in relation to bone age (BA) during childhood and adolescence (N = 303) determined by RA. The specific effects of female puberty on BMaD were studied comparing the values of 110 untreated girls with Turner syndrome (TS) with those of the female reference group. For either sex, a piecewise linear model with one inflection point (IP) was postulated for the relationship of both the 25% and 50% site with BA. The IPs appeared at exactly the same BA (11.5 “years”) for both the 25% and 50% site in boys and for the 25% site in girls. However, in girls the 50% site IP appeared 0.25 “years” later. All BMaD values to the left of the IPs showed little increase with age. In contrast, the slopes to the right of the IN showed in both genders regression coefficients of approximately 0.05 for the 25% site. For the 50% site, the regression coefficient in girls was markedly higher (0.075) than in boys (0.058), resulting only in girls in a significant difference between the 25% and the 50% site to the right of the IP (p = 0.03). The 95% prediction interval for both measurements is somewhat larger in girls than in boys. Gender differences for the onset of the pubertal BMD increase in relation to chronological age (CA) reported with single and dual photon absorptiometry and dual energy X-ray absorptiometry could not be confirmed; however, these are subjective to changes in size and shape of bone during growth. Our data are in agreement with quantitative computed tomography measurements. Untreated girls with TS, lacking spontaneous puberty, show a slight increase with BA without an IP and confirm the absent influence of puberty. Modified RA of the hand proves to be a convenient, accurate, and reproducible method to measure actual BMaD in the pediatric clinic. Since BA reflects biological maturation better than CA, relating the BMaD values to BA should be helpful in deciding whether or not a child has an abnormal BMaD.