Background
Use of high dose glucocorticoids (GC) in patients with diabetes can exacerbate glycaemic control. However, in patients without diabetes, this risk is uncertain. This study aims to examine the incidence of GC-induced diabetes, and attendant risk factors in a cohort of non-diabetic neurosurgical patients receiving high dose dexamethasone perioperatively.
Methods
Adult non-diabetic neurosurgical patients receiving high dose dexamethasone were prospectively studied. Exclusion criteria included pregnancy, HbA1c>6.0%, and anti-diabetes therapies. Data collected were: family history of diabetes, BMI, fasting glucose, insulin, C-peptide, HbA1c (prior to surgery and 6-weeks after last dose of GC). Estimates of beta cell function and insulin sensitivities using HOMA-B, HOMA-IR and HOMA-S were done. Perioperative glucose readings were recorded and 75g OGTTs performed at the end of 6 weeks. Paired student t-tests and multiple linear regression analyses were used (SPSS 24). P value ≤ 0.05 was considered as significant.
Results
Data from sixteen patients (7 males) were available (median age 61.5 years, and mean BMI 28.1+4.6). Prior to GC, eight had impaired fasting glucose. Mean total dexamethasone dose was 90+29 mg with mean treatment duration 17+8 days. Perioperative mean BGL was 7+1.5 mmol/L. At six weeks after stopping dexamethasone, all 16 patients had normal 75g OGTTs. There was statistically significant increase in mean HOMA-B (83+8.0% to 105.6+11.2%, p=0.044) and decrease in mean fasting glucose (5.7+0.8 mmol/L to 4.8+0.5 mmol/L, p=0.005). There were no significant changes to fasting insulin, C-peptide, HOMA-IR and HOMA-S.
Conclusion
Even though some patients had pre-existing risk factors for diabetes, use of high dose dexamethasone in this group of non-diabetic neurosurgical patients did not result in diabetes. Compensatory improvements in beta cell function limited hyperglycaemia during GC treatment. Beta cell function was normal at 6 weeks after stopping dexamethasone.