The low-dose ACTH stimulation test (LDST) is widely used to evaluate hypothalamic–pituitary–adrenal (HPA) axis function in children; however, optimal cortisol sampling times and interpretation strategies remain controversial. Reliance on early or single time-point measurements may lead to false-positive diagnoses of adrenal insufficiency (AI). The aim of the study is. to characterize the timing of peak cortisol responses during LDST in children without AI and to assess the incremental diagnostic contribution of extended sampling and time-specific cortisol thresholds. We retrospectively analyzed 177 pediatric patients who underwent LDST for suspected central adrenal insufficiency at a single tertiary center. Serum cortisol was measured at baseline and at 15, 30, 45, and 60 min following intravenous administration of 1 µg ACTH. Adrenal sufficiency was defined as a peak cortisol ≥ 18 µg/dL(497 nmol/L). Peak timing distribution, basal predictors, incremental diagnostic contribution of additional time points, number needed to test (NNT), and false-positive rates using fixed versus time-specific cutoffs were evaluated. Peak cortisol occurred most frequently at 15 min (48.6%), followed by 30 min (28.8%), baseline (10.7%), 45 min (9.0%), and 60 min (2.8%). Termination of testing at 30 min would have misclassified 11.9% of patients as insufficient despite normal later responses. Extension to 45 min provided meaningful diagnostic improvement, whereas routine extension to 60 min presented only marginal additional benefit (NNT = 30). Higher basal cortisol levels were independently associated with earlier peak responses (p = 0.021), while demographic and auxological factors showed no association. Application of time-specific, percentile-based cortisol thresholds reduced false-positive classifications nearly fivefold at 30 min compared with a uniform 18 µg/dL cut-off. Conclusion: LDST cortisol responses in children show substantial interindividual variability in peak timing. Extension of sampling to 45 min and use of time-specific interpretation thresholds significantly improve diagnostic accuracy and reduce false-positive AI diagnoses in pediatric practice.<Table Float="No" ID="Taba"> <tgroup cols="2"> <colspec align="left" colname="c1" colnum="1" /> <colspec align="left" colname="c2" colnum="2" /> <tbody> <row> <entry nameend="c2" namest="c1"> What is Known: • The low-dose ACTH stimulation test (LDST) is widely used to evaluate HPA axis function in children, but optimalcortisol sampling times and interpretation strategies remain controversial. • Reliance on early or single time-point cortisol measurements can lead to false-positive diagnoses of adrenal in sufficiency. </entry> </row> <row> <entry nameend="c2" namest="c1"> What is New: • Peak cortisol timing varies substantially between children, and terminating the LDST at 30 min would misclassify~12% of patients as insufficient despite normal later responses; higher basal cortisol independently predicts earlier peaks. • Extending sampling to 45 min and applying time-specific, percentile-based cortisol thresholds markedly improve diagnostic accuracy, reducing false-positive classifications nearly fivefold at 30 min compared with a uniform 18μg/dL cut-off. </entry> </row> </tbody> </tgroup> </Table>

Avoiding false-positive adrenal insufficiency diagnoses in children: insights from cortisol kinetics during pediatric low-dose ACTH stimulation test

期刊论文
发表日期：

2026年6月2日

查看全文

Derya Tepe,
Sirmen Kızılcan Çetin,
İrem Gökdemir,
Pınar Kocaay,
Duygu Deligozoglu,
Fatih Gürbüz

摘要

The low-dose ACTH stimulation test (LDST) is widely used to evaluate hypothalamic–pituitary–adrenal (HPA) axis function in children; however, optimal cortisol sampling times and interpretation strategies remain controversial. Reliance on early or single time-point measurements may lead to false-positive diagnoses of adrenal insufficiency (AI). The aim of the study is. to characterize the timing of peak cortisol responses during LDST in children without AI and to assess the incremental diagnostic contribution of extended sampling and time-specific cortisol thresholds. We retrospectively analyzed 177 pediatric patients who underwent LDST for suspected central adrenal insufficiency at a single tertiary center. Serum cortisol was measured at baseline and at 15, 30, 45, and 60 min following intravenous administration of 1 µg ACTH. Adrenal sufficiency was defined as a peak cortisol ≥ 18 µg/dL(497 nmol/L). Peak timing distribution, basal predictors, incremental diagnostic contribution of additional time points, number needed to test (NNT), and false-positive rates using fixed versus time-specific cutoffs were evaluated. Peak cortisol occurred most frequently at 15 min (48.6%), followed by 30 min (28.8%), baseline (10.7%), 45 min (9.0%), and 60 min (2.8%). Termination of testing at 30 min would have misclassified 11.9% of patients as insufficient despite normal later responses. Extension to 45 min provided meaningful diagnostic improvement, whereas routine extension to 60 min presented only marginal additional benefit (NNT = 30). Higher basal cortisol levels were independently associated with earlier peak responses (p = 0.021), while demographic and auxological factors showed no association. Application of time-specific, percentile-based cortisol thresholds reduced false-positive classifications nearly fivefold at 30 min compared with a uniform 18 µg/dL cut-off.

Conclusion: LDST cortisol responses in children show substantial interindividual variability in peak timing. Extension of sampling to 45 min and use of time-specific interpretation thresholds significantly improve diagnostic accuracy and reduce false-positive AI diagnoses in pediatric practice.

What is Known:

• The low-dose ACTH stimulation test (LDST) is widely used to evaluate HPA axis function in children, but optimalcortisol sampling times and interpretation strategies remain controversial.

• Reliance on early or single time-point cortisol measurements can lead to false-positive diagnoses of adrenal in sufficiency.

What is New:

• Peak cortisol timing varies substantially between children, and terminating the LDST at 30 min would misclassify~12% of patients as insufficient despite normal later responses; higher basal cortisol independently predicts earlier peaks.

• Extending sampling to 45 min and applying time-specific, percentile-based cortisol thresholds markedly improve diagnostic accuracy, reducing false-positive classifications nearly fivefold at 30 min compared with a uniform 18μg/dL cut-off.