Hypoglycemia and the origin of hypoxia-induced reduction in human fetal growth

Abstract

Background: The most well known reproductive consequence of residence at high altitude (HA >2700 m) is reduction in fetal growth. Reduced fetoplacental oxygenation is an underlying cause of pregnancy pathologies, including intrauterine growth restriction and preeclampsia, which are more common at HA. Therefore, altitude is a natural experimental model to study the etiology of pregnancy pathophysiologies. We have shown that the proximate cause of decreased fetal growth is not reduced oxygen availability, delivery, or consumption. We therefore asked whether glucose, the primary substrate for fetal growth, might be decreased and/or whether altered fetoplacental glucose metabolism might account for reduced fetal growth at HA.

Methods: Doppler and ultrasound were used to measure maternal uterine and fetal umbilical blood flows in 69 and 58 residents of 400 vs 3600 m. Arterial and venous blood samples from mother and fetus were collected at elective cesarean delivery and analyzed for glucose, lactate and insulin. Maternal delivery and fetal uptakes for oxygen and glucose were calculated.

Principal findings: The maternal arterial - venous glucose concentration difference was greater at HA. However, umbilical venous and arterial glucose concentrations were markedly decreased, resulting in lower glucose delivery at 3600 m. Fetal glucose consumption was reduced by >28%, but strongly correlated with glucose delivery, highlighting the relevance of glucose concentration to fetal uptake. At altitude, fetal lactate levels were increased, insulin concentrations decreased, and the expression of GLUT1 glucose transporter protein in the placental basal membrane was reduced.

Conclusion/significance: Our results support that preferential anaerobic consumption of glucose by the placenta at high altitude spares oxygen for fetal use, but limits glucose availability for fetal growth. Thus reduced fetal growth at high altitude is associated with fetal hypoglycemia, hypoinsulinemia and a trend towards lactacidemia. Our data support that placentally-mediated reduction in glucose transport is an initiating factor for reduced fetal growth under conditions of chronic hypoxemia.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.

Figures

Figure 1. Maternal glucose concentrations and uterine glucose delivery.

(A) Maternal arterial and venous plasma glucose concentrations in normal, term pregnancies at low vs. high altitude. Arterial glucose concentrations were similar at low (4.3±0.1 mM) and high altitude (4.5±0.1 mM, p = 0.06), but venous concentrations were reduced at high altitude (3.5±0.1 at 400 m and 3.2±0.1 at 3600 m, p−1.kg−1 uterine contents) and 3600 m (0.63±0.04 mmol.min−1.kg−1 uterine contents, p = 0.06).

Figure 2. Fetal glucose concentrations and fetal glucose consumption.

(A) Umbilical venous and arterial plasma glucose concentrations in normal, term pregnancies at 400 and 3600 m. Glucose concentrations were greater in the umbilical vein at 400 m (3.5±0.1 mM) than at 3600 m (2.9±0.1 mM, p

Figure 3. Indicators of fetal glucose metabolism.

(A) The fetal glucose/oxygen quotient (6 x fetal glucose consumption/fetal oxygen consumption) was greater at 400 m (1.64±0.13) than at 3600 m (1.18±0.13 p2 = 0.18, p<0.001 at 400 m; y = 0.23+0.35x, r2 = 0.43, p<0.0001 at 3600 m).

Figure 4. Fetal lactate and insulin response to altered glucose delivery and consumption.

(A) Umbilical venous plasma lactate concentrations in normal, term pregnancies at 400 vs 3600 m did not differ (3.36±0.09 mM at 400 m and 3.67±0.13 mM at 3600 m, p = 0.06), whilst arterial concentrations were lower at 400 m (3.22±0.11 mM) than at 3600 m (3.80±0.17, p−1 [54.1–170.1]; high altitude, median  = 53.5 pmol.ml−1 [38.2–79.2]). (C) Fetal glucose and insulin concentrations were positively correlated at both altitudes (400 m y  = −65.2+52.4x r2 = 0.34, p<0.0001; 3600 m y  = −49.4+40.5x, r2 = 0.33, p<0.0001) but neither the slopes (p = 0.42) nor intercept (p = 0.11) differed.