Laboratory of Pediatric Endocrinology, Scientific Institute H. San Raffaele, Milan, Italy 20132
Address correspondence to: Stefano Mora, M.D., Laboratory of Pediatric Endocrinology, Scientific Institute H S. Raffaele, Via Olgettina 60, 20132 Milano, Italy. E-mail: mora.stefano{at}hsr.it.
To the editor:
Sex hormones play a crucial role in skeletal development, as well as in bone mass accrual (1). Nevertheless, the relationship among sex hormone concentrations, bone dimensions, and bone mineral density is still largely unknown.
In a recent issue of JCEM, Wang et al. (2) presented data on a large group of pubertal girls who underwent bone mass studies. The population was carefully selected, and the technique for bone density measurement allowed the simultaneous analysis of geometric properties of the bone analyzed. The authors found that some bone measurements were significantly correlated with 17ß-estradiol (E2), after controlling for age and body size. Direct correlations were found with cortical proportion and cortical thickness (r = 0.26 and r = 0.25, respectively). Bone mineral density was also positively correlated with E2 levels (r = 0.23). Marrow cavity proportion correlated negatively with E2 (r = 0.19) and positively with sex hormone-binding globulin (r = 0.17). The authors conclude that E2 exerts a positive effect on bone geometric and densitometric variables during pubertal development.
I believe that the conclusions are not supported by the data presented. The authors performed correlation analyses and correctly controlled for the confounding effects of age and body size. However, the correlations they found are modest at most, albeit significant. The large number of observations is responsible for the low probability values found by the authors (3). The strength of a relationship is given by the correlation coefficient. Values of 0.20.3 indicate weak correlations between variables (3).
Building the conclusions of correlation studies on probability values is a common mistake that can heavily affect the conclusions of well-planned and well-conducted clinical studies.
Received April 13, 2004.
References
HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |