Send to

Choose Destination
See comment in PubMed Commons below
Pediatr Endocrinol Rev. 2011 Sep;9 Suppl 1:535-7.

Growth hormone deficiency: new approaches to the diagnosis.

Author information

University-Children's Hospital, Pediatric Endocrinology, Tuebingen, Germany.


International consensus statements based on expert experience recommended guide lines how to diagnose GHD. Most recommendations reached only a low level of evidence. Cut-offs for GH were central part of these recommendations, their definition however was arbitrary. Evidence based cut-offs are needed. Using newborn screening cards from healthy and affected newborns, the GH cut-off to detect severe congenital GHD was reassessed and redefined. A GH cut-off level of 7 microg/L confirmed the diagnosis of severe GHD with 100% sensitivity and 98% specificity on the basis of our assay method, if clinical evidence was present. The previous cut-off of 20 microg/L cited in the international consensus statements was based on old GH assays methods not used anymore. For the calculation of an non-arbitrary GH cut-off for biochemical testing in children, we defined an auxological gold standard for GH deficiency: non-familial short stature due to catch-down growth during the childhood phase of growth in combination with an effective catch-up growth in response to low-dose GH therapy, after exclusion of alternative growth disorders and other potential confounders of growth velocity (true positives). Reference cohorts were normally growing children with Turner syndrome or SGA short stature having the same age (true negatives). Using our in-house GH RIA, highest diagnostic accuracy was provided at a peak GH cutoff during spontaneous secretion at night of 7.3 microg/L (sensitivity, 96.8%; specificity, 82.4%; AUC = 0.93). For arginine, cut-off with the highest number of true test results was 6.6 microg/L (sensitivity, 84.3%; specificity, 75.5%; AUC = 0.83). Importantly, children diagnosed GHD in the past with GH test values above the new cut-offs showed a lower response to GH. In conclusion, by use of retrospective and prospective cohort studies evidence-based cut-offs for GH levels measured in newborns and children can be calculated. By use of these cut-offs, tests can be improved. Because of the well known intrinsic diagnostic inaccuracy of any GH test, the correct selection of the child to be tested remains of utmost importance. The diagnosis of growth hormone deficiency (GHD) in childhood is guided by recommendations of national and international consensus statements which are based on the experience of experts. Most of these recommendations reach only a low level of evidence. Research on two central topics of these guidelines has recently been published by us and will be reviewed here.

[Indexed for MEDLINE]
PubMed Commons home

PubMed Commons

How to join PubMed Commons

    Supplemental Content

    Loading ...
    Support Center