Authors:
Whitehouse AJ, Maybery MT, Hart R, Sloboda DM, Newnham JP, Stanley FJ, et al.
Authors notes:
Developmental Medicine and Child Neurology. 2014;56(10):1025
Keywords:
Western Australian Pregnancy Cohort (Raine) Study, testosterone, umbilical cord blood, radioimmunoassay (RIA), head circumference
Abstract:
In 2010, we presented data from a female subsample (n=82) of the Western Australian Pregnancy Cohort (Raine) Study that demonstrated a link between high levels of free testosterone from mixed umbilical cord blood samples derived by radioimmunoassay (RIA) and reduced head circumference growth during the first year of life.
Subsequent to the publication of these data, we validated a liquid chromatography-tandem mass spectrometry (LCMS/MS) assay for measuring a number of androgens from the umbilical cord blood samples.
The re-analysis of the 82 female samples revealed no statistically significant association between concentrations of free testosterone from umbilical cord blood (derived by LCMS/MS) and percentage head circumference growth during the first year of life.
Similarly, when we investigated the broader cohort, the regression models incorporating confounding variables revealed no significant association between free-testosterone levels in cord blood (derived by LCMS/MS) and percentage head circumference growth during the first year of life for both females and males.
The discrepancy between our initial and re-analyzed findings is because of the different platforms we used to derive free-testosterone levels from the mixed umbilical cord blood.
While our original RIA method had been thoroughly validated for use on female serum, its performance on cord blood had not been specifically evaluated and it appears that it unexpectedly detected a number of undefined cross-reacting steroids.
Our LCMS/MS method, on the other hand, was rigorously optimized to ensure that interfering and cross-reacting substances were not detected and that cord serum testosterone concentrations were accurately and precisely measured.
For this reason the LCMS/MS values reported were much lower than detected by the RIA and we consider the more recent LCMS/MS data more reliable.
We urge caution in interpreting our original finding and recommend that future studies employ LCMS/MS techniques to measure sex-steroids from umbilical cord blood.