Objective To investigate the relationship between fasting plasma glucose (FPG) and age/sex, to further analyze the appropriate cutoff age for detecting fasting hyperglycemia (FHG), and to provide a reference for blood glucose screening and health management.

Methods In this cross-sectional study, a total of 50 607 examinees from the Health Management Center of a Shanghai-based Medical Consortium in 2023 were included. The age and sex distributions of FPG levels and the detection rate of FHG (FPG≥6.1 mmol·L⁻¹) were analyzed. A non-parametric method was used to plot the receiver operating characteristic (ROC) curve of age for predicting hyperglycemia. The area under the ROC curve (AUC) reflected the predictive value of age for FHG, and the appropriate cutoff age for hyperglycemia detection was determined by the maximum Youden index.

Results The median and the 25th and 75th percentiles of FPG in the study population was 5.00 (4.68, 5.41) mmol·L⁻¹, with levels systematically higher in males 5.06 (4.73, 5.53) mmol·L⁻¹ than in females 4.93 (4.64, 5.28) mmol·L⁻¹. Both FPG levels and the abnormal detection rate increased non-linearly with age, plateauing approximately at 80 years of age. ROC curve analyses showed that age had good discriminatory value for hyperglycemia status (AUC=0.778 for the total population), with the appropriate predictive age threshold being 48 years old (48 years old for males, 49 years old for females).

Conclusion The present study found that both FPG levels and the risk of abnormal glucose regulation were increased at 48‒49 years old among the hospital-based physical examination population, with risk remaining consistently higher in males than in females. The appropriate cutoff age for FPG was 48 years old. Specifically, the period of 45‒50 years of age should be set as an intensified screening phase for blood glucose screening, with particular emphasis on earlier intervention for males. Furthermore, glycemic management targets for the elderly population require individualized consideration.