0005 Time-Restricted Eating Improves Intermittent Hypoxia-Induced Dysglycemia

Introduction Obstructive sleep apnea (OSA) is associated with dysglycemia, insulin resistance, and type 2 diabetes. Continuous positive airway pressure has not been shown to improve metabolic outcomes in OSA. Intermittent hypoxia (IH) modeling OSA causes dysglycemia in rodent models. Time restricted eating (TRE), which limits daily caloric intake to a fixed window, improves metabolic profiles in mice and humans. We hypothesized that TRE would improve IH-induced dysglycemia. Methods C57BL/6J mice were fed a high-fat diet for four weeks, and then divided into four groups (n=16/group), with additional exposures lasting four weeks. Group 1 was exposed to room air (RA), with ad libitum access to HFD. Group 2 was exposed to RA, but on a TRE protocol, restricting HFD intake to 9h/d, from 9:00 PM to 6:00 AM. Group 3 was exposed to IH (FiO2 0.21 to 0.06 once/min for 12 hours during light phase, and FiO2 0.21 during dark phase), with ad libitum access to HFD. Group 4 was exposed to IH and TRE. Cages were changed each morning to minimize food hoarding. Glucose tolerance tests (GTT) were performed before and after the exposures, and dual energy x-ray absorptiometry (DEXA) was performed on a subset. Serum insulin was measured. Results Fasting glucose was reduced in IH-TRE (-25.4±5.6 mg/dL, p< 0.001) but not in other groups. GTT area under the curve was increased in RA-ad lib (3155±924 mg*min/dL, p=0.002) and IH-ad lib groups (5267±1161 mg*min/dL, p< 0.001), but not in TRE. Insulin was increased in RA-ad lib relative to other groups (p< 0.010 for all), implying insufficient insulin secretion, particularly in IH-ad lib. Body fat was reduced with both IH (p=0.028) and TRE (p< 0.001). Serum 1,5-anhydroglucitol, inversely related to pancreatic β-cell dysfunction, was increased in TRE (p=0.011). Conclusion In IH modeling OSA, TRE improves glucose homeostasis, via improved pancreatic insulin secretion. Intermittent fasting may be a viable therapeutic option in OSA. Support (if any) This project was supported by NIH K08HL143140 and UCSD RS295R.