ESPE Abstracts (2016) 86 P-P2-653

ESPE2016 Poster Presentations Growth P2 (47 abstracts)

Improving the ‘Gold Standard’: The Insulin Tolerance Test Revisited

Nikolaos Daskas , John Barton , Christine Burren & Elizabeth Crowne


University Hospitals Bristol, Bristol, UK


Background: The optimal method to assess GH status remains controversial. GH provocation tests are used and the Insulin Tolerance Test (ITT) is regarded as the ‘gold standard’ to diagnose GH deficiency (GHD). The original selection of 0, 20, 30, 60, 90 and 120 min time points is still used in many protocols worldwide, but variations have evolved.

Objective and hypotheses: Comparing standard ITT (StdITT) to a revised (RevITT) protocol.

Method: ITT was performed according to local protocol. StdITT measured GH at −30, 0, 30, 60, 90 and 120 min. RevITT added three additional samples (15, 45 and 75 min). Data are presented as mean±S.D. unless specified otherwise.

Results: ITT where hypoglycaemia was achieved were included (376/392). 65 were done using RevITT. Median GH levels were highest at 60 min on StdITT and 45 min on RevITT. Peak GH levels were measured (in decreasing frequency) at 60, 30, 0, −30 min on StdITT and 45, 0, 60, 30 min on RevITT but did not necessarily preclude GHD. Using peak GH cut-offs of 7 μg/l for prepubertal children and 5 μg/l for adolescents, 214/311 (68%) of StdITT and 38/65 (58%) of RevITT were abnormal. Analysis of 124 normal tests (97 StdITT and 27 RevITT) identified 51/124 (42 StdITT and 9 RevITT), where a single GH level was above the diagnostic cut-off. The 45 min sample represented the peak GH level in 8/27 normal RevITT and in two tests it was the only level precluding GHD. The 75min sample did not preclude GHD. The 15 min sample represented the peak GH level in 3/27 normal RevITT, but did not preclude GHD on its own in any of these three tests. The −30 min sample precluded GHD in 5/124.

Conclusion: Early and more frequent sampling are key to diagnosing GHD accurately, precluding GHD in an additional 7.5% (45 min sample) and 4% (−30 min sample). The 120 min can be removed as it does not contribute to GHD diagnostic yield.

Article tools

My recent searches

No recent searches.