Glucocorticoid (GC) excess drives obesity, insulin resistance and type 2 diabetes. Obstructive sleep apnoea (OSA) is a prevalent condition associated with both activation of the hypothalamic-pituitary-adrenal (HPA) axis and an adverse metabolic phenotype. However, a causal link between these two features has not been established. We designed a novel human model of intermittent hypoxia (IH) aimed at replicating the systemic insulin resistance associated with OSA to determine whether limiting GC availability has a beneficial metabolic effect.
Healthy male participants (n=17, age: 25±6 yrs; BMI:24.1±1.7 kg/m2) underwent a hyperinsulinaemic-euglycaemic clamp incorporating stable isotopes with adipose tissue microdialysis and biopsy, under conditions of normoxia. Volunteers were then randomized to receive either the GC receptor antagonist, RU486 (600 mg od, 1 week) or no treatment. All investigations were then repeated under conditions of IH. IH was achieved by alternating the FiO2 of inspired gases between air (FiO2 21%) and a hypoxic gas mix (FiO2 5%; balance nitrogen) to achieve 12 desaturations/hr.
In the fasted-state under normoxia, RU486 improved adipose tissue insulin sensitivity (significant reductions circulating non-esterified fatty acids (NEFA), glycerol and Adipo-IR all P<0.05). Our model of IH successfully replicated features of OSA (11.4+/−0.8 desaturation/hr. 87+/−4 %O2 saturations). IH induced insulin resistance as measured by reduced insulin-mediated suppression of circulating triglyceride (TAG) and VLDL TAG (P<0.05). In the fed, hyperinsulinaemic-state, and under IH, RU486 worsened the adverse metabolic features (reduced insulin-mediated suppression of TAG; VLDL-TAG; βOH-Butyrate, all P<0.05). These data endorse our previous observations demonstrating that GCs enhance insulin action in vitro and in vivo in the fed-state.
In conclusion, we have demonstrated a differential effect of GC receptor antagonism in the fed and fasted-state. Limiting GC action under conditions of IH worsened insulin resistance suggesting that activation of the HPA axis in OSA does not drive the development of an adverse metabolic phenotype.