Multi-Patient Analysis of Steroid-Induced Hyperglycemia in Diabetic Patients Using Continuous Glucose Monitoring.

Introduction: Steroid-induced hyperglycemia (SIH) is a frequent complication of glucocorticoid therapy, yet most prior studies have relied on sparse glucose measurements that limit understanding of timing and trajectory. Continuous glucose monitoring (CGM) offers high-resolution insights into these fluctuations with continuous temporal resolution that enables identification of transient and delayed glycemic responses. Methods: In this prospective, single-arm observational study, CGM data were analyzed from 58 adults with diabetes who received a single standard-of-care steroid injection for interventional pain management. Participants received dexamethasone (n=38), methylprednisolone (n=15), or triamcinolone (n=5). Dexcom G7 CGM recorded glucose every 5 minutes for up to 10 days post-injection. Glucose trajectories were summarized and stratified by steroid type, gender, and age (<64.6 vs ≥64.6 years). Results: Across steroid types, glucose increased within ~2 hours after injection. Dexamethasone produced the most consistent excursion, peaking at ~220-225 mg/dL and returning toward baseline by 24-36 hours. Methylprednisolone showed a more moderate delayed increase (~175-185 mg/dL) with sustained elevation for several days. Triamcinolone demonstrated similar peak levels (~220 mg/dL) but marked variability, limiting interpretability. Females generally exhibited higher and earlier peaks than males, most notably with dexamethasone. Younger patients showed more dynamic excursions and less uniform recovery, whereas older patients demonstrated flatter trajectories and more stable return toward baseline. Data completeness was highest through 48-72 hours post-injection. Discussion: This study provides one of the most detailed CGM characterizations of SIH, demonstrating steroid-specific and demographic-modified responses. Dexamethasone displayed a reproducible excursion phase, methylprednisolone a delayed pattern, and triamcinolone inconsistent variability. These findings highlight the utility of CGM for identifying temporal SIH features, with the first 48 to 72 hours yielding clinically reliable monitoring window.