Relationship of Sex Hormones to Bone Geometric Properties and Mineral Density in Early Pubertal Girls: Use of Correlation Analyses

Stefano Mora

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 0.2–0.3 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

  1. Mora S, Gilsanz V 2003 Establishment of peak bone mass. Endocrinol Metab Clin North Am 32:39–63[Medline]
  2. Wang Q, Nicholson PHF, Suuriniemi M, Lyytikäinen A, Helkala E, Alen M, Suominen H, Cheng S 2004 Relationship of sex hormones to bone geometric properties and mineral density in early pubertal girls. J Clin Endocrinol Metab 89:1698–1703[Abstract/Free Full Text]
  3. Healy MJ 1992 Statistics from the inside. 7. Regression and correlation. Arch Dis Child 67:1306–1309[Medline]




This Article
Full Text (PDF)
Submit a related Letter to the Editor
Purchase Article
View Shopping Cart
Alert me when this article is cited
Alert me when eLetters are posted
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in PubMed
Alert me to new issues of the journal
Download to citation manager
Request Copyright Permission
Google Scholar
Articles by Mora, S.
Articles citing this Article
PubMed
PubMed Citation
Articles by Mora, S.


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