Unique Cardio‐metabolic Consequences of Superimposed Sustained and Intermittent Hypoxia

Abstract

Rationale: The overlap syndrome of chronic obstructive pulmonary disease (COPD) and sleep disordered breathing (SDB) is associated with higher morbidity and mortality relative to either disease alone, and poorer outcomes are associated with worsened hypoxia. People at high altitude are susceptible to SDB, and hypoxic burden in these individuals is associated with metabolic dysfunction and cardiovascular disease. We aimed to characterize the cardio‐metabolic consequences of superimposed sustained plus intermittent hypoxia, as a model of COPD/SDB overlap syndrome, or high‐altitude hypoxia with periodic breathing. Methods:C57BL/6J mice (n=8/group) were subjected to one of four oxygen profiles for six weeks. In room air (RA), mice were kept at FiO2=0.21; in sustained hypoxia (SH), they were kept at FiO2=0.10. In intermittent hypoxia (IH), mice were kept at FiO2=0.21 for 12 hours, and FiO2 fluctuated between 0.21 and 0.06, 60 times/hour, for 12 hours. In overlap hypoxia (OH), mice were kept at FiO2=0.13 for 12 hours, and FiO2 fluctuated between 0.13 and 0.06, 60 times/hour, for 12 hours. Blood pressure and heart rate variability were measured weekly, and right ventricular systolic pressure (RVSP) was measured prior to sacrifice. Intraperitoneal glucose tolerance test (GTT) was performed at baseline and before sacrifice. Results: Mice in RA gained weight (+1.3g), whereas mice in SH lost weight (‐0.5g, p=0.010 vs RA), and there were no changes in IH or OH (IH: ‐0.1g, p=0.056; OH: +0.9g, p=0.876). Blood pressure was highly variable and did not significantly differ by group. With respect to heart rate variability, the LF/HF ratio at 4 weeks increased in SH and IH, without any significant change in RA or OH mice. IH did not increase RVSP relative to RA (p=0.977), whereas SH and OH increased RVSP by 52% (p<0.001) and 20% (p=0.002), respectively. Fasting blood glucose did not change in RA (‐3 mg/dL, p=0.808), but decreased in both SH and OH (‐57 mg/dL, p=0.001, and ‐41 mg/dL, p=0.004), and increased in IH (+25 mg/dL, p=0.001). Hepatic glycogen content reflected fasting glucose levels. The difference in area under the GTT curve (final‐initial) was unchanged in OH versus RA (p=0.666), but increased in IH (p=0.015) and decreased in SH (p=0.004). Hepatic IL‐1β was increased 22% in IH (p=0.019) but not in other groups, and TNF‐α and IL‐6 were similar between groups. Hepatic malondialdehyde, an oxidative stress marker, was increased in each group relative to RA (SH: +14%, p=0.003, IH: +23%, p<0.001, OH: +35%, p<0.001). Conclusions:Overlap hypoxia may represent a unique and deleterious cardiometabolic phenotype, characterized by more severe pulmonary hypertension than chronic intermittent hypoxia, without the protective metabolic effects of sustained hypoxia. Hepatic oxidative stress was worse in OH than in other tested hypoxia profiles.