Scientific Reports (Mar 2025)

Risk factors and prediction model of metabolic disorders in adult patients with pituitary stalk interruption syndrome

  • Deyue Jiang,
  • Shengjie Wang,
  • Yan Xiao,
  • Peng Zhi,
  • Erhan Zheng,
  • Zhaohui Lyu,
  • Qinghua Guo

DOI
https://doi.org/10.1038/s41598-025-91461-7
Journal volume & issue
Vol. 15, no. 1
pp. 1 – 13

Abstract

Read online

Abstract Pituitary stalk interruption syndrome (PSIS) is an infrequently occurring congenital condition, and there exists a dearth of systematic investigative work focusing on the clinical features and long-term outcomes in adult patients. Individuals who have reached adulthood with PSIS are at an increased risk of developing metabolic disorders, including metabolic syndrome (MS) and non-alcoholic fatty liver disease (NAFLD) or metabolic dysfunction associated fatty liver disease (MAFLD), which are also one of the main factors for the poor prognosis of these patients. An analysis was conducted on the clinical data of adult PSIS patients who visited the endocrinology department of the First Medical Center of the People’s Liberation Army General Hospital from January 2005 to August 2023. Patients were grouped based on their MAFLD and MS status, and the differences in clinical characteristics and risk factors between the groups were analyzed. Machine learning models were used to construct a prediction model for the occurrence of MAFLD in adult PSIS patients and to analyze high-risk predictors. Out of 136 PSIS adult patients, 93.3% were male. The prevalence of MAFLD was 55.5%, and MS was 22.3%. Patients with a history of growth hormone (GH) treatment were less likely to develop MAFLD (P = 0.032). MAFLD patients exhibited higher rates of hypertension, hyperuricemia, obesity, MS, and dyslipidemia. Multiple risk factors may contribute to MS, while no significant link was found between MS and hormone replacement. However, GH non-treatment may serve as the notable predictor of MAFLD in PSIS patients revealed by the Ridge regression model of machine learning model with the highest predictive performance of a mean area under the curve (AUC) of 0.82. The prevalence of MS and MAFLD is high among adult patients with PSIS. Multiple risk factors may contribute to these two diseases, and after constructing a predictive model, we found that MAFLD may be closely linked to the previous lack of GH treatment.

Keywords