Efficacy of Low Insulin-Like Growth Factor-1 Levels and a Single Insulin Tolerance Test in Diagnosing Growth Hormone Deficiency

ABSTRACT Pediatric growth hormone deficiency requires dynamic testing for accurate diagnosis because of the pulsatile secretion of growth hormone. The insulin tolerance test (ITT) is regarded as one of the most robust individual provocative tests for growth hormone deficiency, although the broader pediatric diagnostic standard requires a composite of auxological, clinical, and biochemical criteria together with an inadequate response to two independent stimulation tests, except in specific high-probability contexts; insulin-like growth factor-1 (IGF-1) alone has limited diagnostic accuracy. This study evaluated whether combining low IGF-1 levels with a single failed ITT could improve diagnostic efficiency for pediatric growth hormone deficiency. We conducted a retrospective cohort study of 214 children with short stature (aged 2–16 y) evaluated at King Chulalongkorn Memorial Hospital between August 2019 and January 2025. Growth hormone deficiency was confirmed in 63 patients using two growth hormone stimulation tests (peak growth hormone level: < 7 ng/mL). IGF-1 levels were expressed as standard deviation scores and analyzed at cutoff values of ≤ –1, –1.5, –2, and –2.5 in combination with failed ITT results. Diagnostic performance indices were calculated at each predefined cutoff/test combination, including sensitivity, specificity, positive predictive values, negative predictive values, area under the receiver operating characteristic curve, odds ratios, and accuracy. An IGF-1 standard deviation score of ≤ –2.5 combined with a single failed ITT showed a sensitivity of 38.6% and a specificity of 92.6%, with a positive predictive value and a negative predictive value of 73.9% and 73.5%, respectively (p<0.001). Because the reference standard for the growth hormone deficiency required failure of two stimulation tests, one of which was the ITT, we assessed this circularity directly: among 100 patients with both the ITT and clonidine results, findings were discordant in 37 (exact McNemar p=0.008), and IGF-1 standard deviation scores did not differ significantly across concordant and discordant subgroups (Kruskal–Wallis p=0.407). An IGF-1 standard deviation score cutoff of ≤ –2.5 combined with a single failed ITT identifies a subgroup of children at higher probability of growth hormone deficiency who may be prioritized for confirmatory evaluation. Given the modest sensitivity and positive predictive values of this criterion and the circularity inherent in a reference standard that itself incorporates the ITT, these findings should be regarded as hypothesis-generating rather than as a basis for changing current diagnostic practice, which continues to require two independent stimulation tests except in specific high-probability contexts; prospective, multicenter validation is required.