Prevalence of Hypercortisolism in Difficult-to-Control Type 2 Diabetes

Abstract

Objective: Despite the use of multiple glucose-lowering medications, glycemic targets are not met in a significant fraction of people with type 2 diabetes. In this prospective, observational study we assessed the prevalence of hypercortisolism, a potential contributing factor to inadequate glucose control.

Research design and methods: Individuals with type 2 diabetes and HbA1c 7.5%-11.5% (58-102 mmol/mol) on two or more glucose-lowering medications with or without micro-/macrovascular complications or taking multiple blood pressure-lowering medications were screened with a 1-mg dexamethasone suppression test. Common causes of false-positive DSTs were excluded. The primary end point was the prevalence of hypercortisolism, defined as post-DST cortisol >1.8 μg/dL (50 nmol/L). Characteristics associated with hypercortisolism were assessed with multiple logistic regression. The percentage and characteristics of participants with hypercortisolism and adrenal imaging abnormalities were also assessed.

Results: Post-DST cortisol was unsuppressed in 252 of 1,057 participants (prevalence 23.8%; 95% CI 21.3, 26.5). Hypercortisolism prevalence was 33.3% among participants with cardiac disorders and 36.6% among those taking three or more blood pressure-lowering medications. Adrenal imaging abnormalities were reported in 34.7% of participants with hypercortisolism. Use of sodium-glucose cotransporter 2 inhibitors (odds ratio 1.558), maximum-dose glucagon-like peptide 1 receptor agonists (1.544), tirzepatide (1.981), or a higher number of blood pressure-lowering medications (1.390); older age (1.316); BMI <30 kg/m2 (1.639); non-Latino/Hispanic ethnicity (3.718); and use of fibrates (2.676) or analgesics (1.457) were associated with higher prevalence (all P < 0.03).

Conclusions: Hypercortisolism was associated with hyperglycemia in approximately one-quarter of individuals with inadequately controlled type 2 diabetes despite multiple medications.

Graphical abstract

Figure 1

Multiple logistic regression model for hypercortisolism vs. no hypercortisolism. Error bars: 95% CIs.

Figure 2

Results of adrenal CT scans in participants with hypercortisolism. Abdominal CT scan results were available in 219 of 252 (86.9%) participants. Reasons for not completing the CT scan included withdrawal of consent (n = 13), lost to follow-up (n = 6), physician decision (n = 2), and other (n = 12).