Endocrine Abstracts

Introduction: Mountain dwellers show a lower prevalence of diabetes, which among countless other factors could be due to lower partial oxygen pressure at high altitude. The present study was to investigate in an experimental setting, if and how reduced oxygen availability affects blood glucose.

Methods: Male obese mice on high fat diet were continuously maintained under a normobaric hypoxic atmosphere (10% oxygen) for three months. To exclude indirect effects via blunted appetite and reduced weight gain, control groups kept on normal air were fed restrictedly, so to match the weight curves of their hypoxia-exposed counterparts. To track down involved mechanisms, wild type mice as well as mice, which expressed the erythropoetin receptor in the haematopoetic lineage only, were treated with erythropoetin (300 U/kg i.p., three injections per week for two months). The immediate response to infusion of donor erythrocytes was also examined. Treatment-induced effects on blood glucose and insulin sensitivity were studied.

Results: Life under hypoxia increased the haematocrit (44±1 vs. 55±1 %, P<0.001) and lowered blood glucose of obese mice (166±5 vs. 132±3 mg/dl, P<0.001), which went along with improved insulin sensitivity (HOMA: 1.4±0.2 vs. 0.60.±0.1, P<0.001; euglycaemic-hyperinsulinaemic clamp experiments, mg glucose/kg/min: rate of disappearance, 42±2 vs. 50±2,P=0.03; rate of appearance, 9±4 vs. -5±1,P=0.008). Parameters of lipid metabolism were unaffected by hypoxia. Similar effects were observed in wild type mice under regular injections of erythropoetin (haematocrit: 44±2 vs. 69±3 %, P<0.001; blood glucose: 173±7 vs. 108±9 mg/dl, P<0.001; HOMA: 1.8±0.4 vs. 0.3±0.1,P=0.008). Absence of the erythropoetin receptor in all tissues except bone marrow and spleen did not impair the glucose lowering action of erythropoetin injections (haematocrit: 46±2 vs. 63±1 %, P<0.001; blood glucose: 160±4 vs. 97±10 mg/dl, P<0.001; HOMA: 2.6±0.7 vs. 0.8±0.3,P=0.048). This excluded direct action of erythropoetin on non-haematopoetic organs as the cause of decreased blood glucose. In line with such evidence for haematopoesis-mediated lowering of blood glucose, an acute increase in the haematocrit from 46±1 % to 60±1 % (P<0.001) by infusion of donor erythrocytes induced a significant reduction in blood glucose (169±6 vs. 135±4 mg/dl, P<0.001).

Conclusion: Reduced blood glucose in mice living in a hypoxic environment is associated with improved insulin sensitivity and is obviously the direct consequence of an erythropoetin-mediated increase in the haematocrit. Elevated haematocrit thus likely explains lower blood glucose found in people living at high altitude as well as in patients under treatment with erythropoetin.