Why is human uterine artery blood flow during pregnancy so high?

Abstract

In healthy near-term women, blood flow to the uteroplacental circulation is estimated as 841 mL/min, which is greater than in other mammalian species. We argue that as uterine venous Po₂ sets the upper limit for O₂ diffusion to the fetus, high uterine artery blood flow serves to narrow the maternal arterial-to-uterine venous Po₂ gradient and thereby raise uterine vein Po₂. In support, we show that the reported levels for uterine artery blood flow agree with what is required to maintain normal fetal growth. Although residence at high altitudes (>2,500 m) depresses fetal growth, not all populations are equally affected; Tibetans and Andeans have higher levels of uterine artery blood flow than newcomers and exhibit normal fetal growth. Estimates of uterine venous Po₂ from the umbilical blood-gas data available from healthy Andean pregnancies indicate that their high levels of uterine artery blood flow are consistent with their reported, normal birth weights. Unknown, however, are the effects on placental gas exchange of the lower levels of uterine artery blood flow seen in high-altitude newcomers or hypoxia-associated pregnancy complications. We speculate that, by widening the maternal artery to uterine vein Po₂ gradient, lower levels of uterine artery blood flow prompt metabolic changes that slow fetal growth to match O₂ supply.

HUMAN UTEROPLACENTAL BLOOD FLOW

Pregnancy provokes enormous changes in the maternal cardiovascular system. Systemic vascular resistance and blood pressures fall, prompting a ∼30% rise in total blood volume and resting cardiac output. The reduction in uteroplacental vascular resistance is particularly marked, causing the major portion of the increased cardiac output to be directed to that circulation by term. Several reviews have cataloged the numerous systemic and uteroplacental factors involved in these cardiovascular changes (1, 2). Here, we address why human uterine artery blood flow is so high and the consequences of when such a rise in blood flow fails to occur.

Surprisingly, given the magnitude of rise in uterine artery blood flow, its measurement in unanesthetized women has only become available in the past few decades. Table 1 summarizes studies in which blood flow through the main uterine artery has been measured noninvasively using transabdominal Doppler ultrasound near term in healthy low-altitude women. Such studies rely on being able to use the convenient anatomical landmark where the uterine artery crosses over the external iliac artery to measure vessel diameter and blood flow velocity in the same anatomical location. Volumetric flow is then calculated as the product of uterine artery cross-sectional area (π·r², where r is the uterine artery radius) and the average blood flow velocity across several cardiac cycles. Necessary for making such measurements is a skilled ultrasound operator, who can accurately measure vessel diameter in longitudinal view and then carefully rotate the Doppler probe to measure flow velocity at a sufficiently low angle of insonation (ideally <30°).

Table 1.

Blood flow through the uterine arteries and to the uteroplacental circulation during normal, low-altitude human pregnancy

Reference	Week	Ut a diam, cm	Ut a flow velocity, cm/s	Unilateral Ut a Flow, mL/min	Bilateral Ut a Flow, mL/min	Total Uteroplac Flow/Fetal Weight*, mL/min/kg
Palmer (3)	36	0.34	61	382	(764)	263
n = 5
Konje (4)	38	0.48			970	292
n = 57
Jeffreys (5)	35-8	0.42	32	410	(820)
n = 57
Zamudio (6)	37.2	0.42	36		638	188
n = 44
Julian (7)	36	0.50	45		613	177
n = 18
Julian (8)	36	0.50	31	369	(738)	220
n = 28
Michelsen (9)	40	0.28		228	488	138
n = 179
Average	37.1	0.42	36	347	719	214
Including ∼17% from uterine branches of the two ovarian arteries					841	235

Bilateral Ut a flow values in parentheses are calculated as twice the unilateral value for studies in which only unilateral values were reported. Estimates for flow per kilogram fetal weight used values reported for birth weights in each study. *Birth weights reported for each study were used to calculate flow per kilogram fetal weight as fetal weight at the time of study was not reported. A, artery; diam, diameter; Ut a, uterine artery; uteroplac, uteroplacental.

Table 1 presents the values for uterine artery blood flow reported in seven studies in healthy, near-term sea-level pregnancies. Across all studies, bilateral uterine artery blood flow averaged 719 mL/min or 214 mL/min/kg fetal weight. This is an underestimate, however, since current methods only measure blood flow through the two uterine arteries and not the ∼17% supplied by the uterine branches of the two ovarian arteries (10). Including this, 17% raises total uteroplacental blood flow near term to 841 mL/min or 235 mL/min/kg fetal weight.

A COMPARATIVE PERSPECTIVE ON THE MAGNITUDE OF THE RISE IN UTERINE ARTERY BLOOD FLOW

The 235 mL/min/kg fetal weight value is higher in rodents whose uteroplacental blood flow measured using microspheres was 145 mL/min/kg total fetal weight (11). What accounts for such high levels of uteroplacental blood flow in human pregnancy? One factor is that not all the blood reaches the placenta as ∼25% is directed to the myometrium and cervix (12). Another factor is the existence of uterine arterial-venous anastomoses, evident as early as the first trimester and persisting throughout pregnancy, which can return ∼10% of the blood directly to the venous circulation before reaching terminal vessels (13). We concur with others (14) that a key factor stems from the venous equilibrator nature of human placental exchange and its consequences for fetal O₂ supply.

Unlike rodents, rabbits, and other smaller mammals with highly efficient countercurrent placental exchange systems, humans and other haplorrhine primates (i.e., Old and New World monkeys, apes) and sheep, often used as an experimental animal model, have a venous equilibrator or mixed pool system. The opposing directions of maternal and fetal blood flow in a countercurrent system (Fig. 1A) enable twice as an efficient transplacental exchange compared with a concurrent system in which maternal and fetal blood travels in the same direction (22) and three times as efficient as a venous equilibrator system (23) (Fig. 1B). In a venous equilibrator system, the terminal uterine (spiral) arteries inject blood at high velocity into the intervillous space, which becomes a mixed pool stirred by the fetal villi and has an intervillous Po₂ approximating that of the uterine vein. Given the need for an O₂ gradient of at least 10 mmHg for diffusion to the umbilical vein (i.e., the vessel carrying oxygenated blood to the fetus), the uterine vein Po₂ sets the upper limit for umbilical venous Po₂ (14).

Figure 1.

A: counter-current placental exchange allows the relatively high Po₂ in the maternal uterine artery (Ut a) to equilibrate with umbilical venous (Umb v) blood flowing from the placenta to the fetus, achieving an Umb v Po₂ that can exceed the uterine venous (Ut v) value, and operate at relatively low levels of maternal Ut a blood flow (15). B: a venous-equilibrator exchange system is much less efficient as Umb v blood equilibrates with the Po₂ in the intervillous space, which approximates that of the Ut v, and requires high levels of uterine artery blood flow (14) to maintain an estimated 10 mmHg Ut v to Umb v Po₂ gradient. Sea level umbilical blood gas values are from Refs. 14, 16, and 17) and Ut v Po₂ from Ref. 18. The high-altitude value for maternal artery Po₂ is from 304 healthy Andean pregnant women at 4,300 m (19) and maternal Ut v Po₂ of 34 mmHg was estimated by assuming that it was 10-mmHg greater than the reported Umb v value for babies born to 18 healthy Andean women delivering appropriate for gestational age birth weight infants at 3,600 m (20, 21).

The classic Fick equation, where tissue oxygen consumption (V̇o₂) is the product of blood flow (Q) and the arteriovenous O₂ content difference (${\text{Ca}}_{{\text{O}}_2}$ − ${\text{Cv}}_{{\text{O}}_2}$),

$$\text{V}\dot{}{\text{o}}_2=\text{Q}\times ({\text{Ca}}_{{\text{O}}_2}-{\text{Cv}}_{{\text{O}}_2})$$

reveals the importance of maintaining a high blood flow for raising venous O₂ content and pressure. Applying the Fick equation to the uteroplacental circulation and rearranging it in terms of uteroplacental blood flow (Q_Utp) shows that Q_Utp equals fetoplacental O₂ consumption (V̇o_2FP) divided by the maternal arterial – uterine venous O₂ content difference (${\text{C}}_{\text{Uta}{\text{O}}_2}$ − ${\text{Cv}}_{\text{Uta}{\text{O}}_2}$)¹ or

$${\text{Q}}_{\text{Utp}}=\text{V}\dot{}{\text{o}}_{2\text{FP}}/({\text{C}}_{\text{Uta}{\text{O}}_2}-{\text{C}}_{\text{Utv}{\text{O}}_2})$$

It follows then that at a constant level of maternal ${\text{Ca}}_{{\text{O}}_2}$, hemoglobin-O₂ dissociation curve position and uteroplacental V̇o₂, a high Q_Utp narrows the uterine arterial-venous O₂ difference, raises intervillous and uterine vein Po₂, and facilitates O₂ diffusion to umbilical vein and fetal circulation.

TRANSPLACENTAL GAS EXCHANGE IN NORMAL AND FETAL GROWTH-RESTRICTED PREGNANCY AT LOW AND HIGH ALTITUDE

Few studies have evaluated transplacental nutrient exchange in vivo in human pregnancy (24). Only two have measured maternal arterial and uterine venous blood gases at low altitude, finding uterine venous Po₂ to range from 46 to 49 mmHg with an O₂ saturation of 72%–79% (18, 25). Using these values together with calculated arterial O₂ content and assuming that 60% of V̇o_{2 FP} is consumed by the fetus and 40% by the placenta (26), we used the Fick equation to calculate that a 796 mL/min total uteroplacental blood flow is required to supply the 1,575 micromoles O₂/min needed for a 36 wk, normal 3 kg fetus (Fig. 1B, left). This 796-mL/min value is consistent with the 719-mL/min reported at low altitude for bilateral uterine artery or the estimated total uteroplacental blood flow of 814 mL/min when the fraction from the uterine branches of the two ovarian arteries is included (Table 1). There is only one report in which maternal and umbilical respiratory gases, but not uterine artery blood flow, were measured in fetal growth-restricted pregnancies at low altitude (25). A higher, not lower, uterine vein Po₂ was found in the growth-restricted than in normal pregnancies, which was due to impaired placental diffusing capacity resulting from small placental size. However, most fetal growth-restricted pregnancies are not attributable to small placental size or other placental pathologies (27). In addition, an important factor contributing to fetal growth restriction is a lesser third trimester rise in uterine artery blood flow (28). Hence, further study is required for assessing the effect of variation in uterine artery blood flow on transplacental gas exchange in normal and fetal growth-restricted pregnancies at low altitude.

Residence at high altitude (>2,500 m) exerts one of the strongest depressant effects on fetal growth (29) and therefore provides a natural laboratory for investigating the relationship between uterine artery blood flow and transplacental gas exchange. Not all populations, however, are equally affected; genetically adapted groups, Tibetans and Andeans, are relatively protected from the altitude-associated fetal growth restriction seen in newcomers at the same elevation (30–32), even after several generations of high-altitude residence (33). No maternal arterial and uterine venous Po₂ and content together with umbilical venous and arterial respiratory gases are available from healthy high-altitude pregnancies, but maternal hemoglobin and arterial Po₂, umbilical vein, and arterial Po₂ values have been reported (19–21) (Fig. 1B, right). Applying the Fick equation and making the same assumptions as at sea level, we calculated that a total uteroplacental blood flow of 700 mL/min is required to maintain a uterine vein Po₂ of 34 mmHg (i.e., 10 mmHg greater than the umbilical venous value) and supply the 1,575 μmol O₂/min required for a 36-wk, normal 3-kg fetus. This 700-mL/min value agrees with the 711 mL/min value for bilateral uterine artery blood flow that we have reported for healthy Andean women at week 36 of pregnancy (8) but is somewhat lower than the 832 mL/min estimated for total uteroplacental blood flow when the contribution from the uterine branches of the two ovarian arteries is included. Unknown, however, are the effects on placental gas exchange of the lower levels of uterine artery blood flow seen in high-altitude newcomers (7, 34) or hypoxia-associated pregnancy complications at any altitude. Valuable, therefore, would be studies in high-altitude residents with and without fetal growth restriction to determine whether lower levels of uterine artery blood flow widen the maternal arterial – uterine vein O₂ content gradient, lower uterine and umbilical vein Po₂, and prompt metabolic changes responsible for slowing fetal growth to match O₂ supply with demand. Such studies should also include direct measurements of the hemoglobin-O₂ dissociation curve given the importance of the act of operating at a lower point on the curve.

DOES THE FALL IN UTERINE ARTERY BLOOD FLOW CURTAIL FETAL GROWTH?

The altitude-associated reduction in fetal growth is not simply due to reduced uterine artery blood flow since even the lower blood flows observed at high altitude are sufficient to maintain uteroplacental O₂ delivery at near sea-level values (6, 35–37). Furthermore, the decrease in fetal growth has begun by pregnancy week 28 when, given its still relatively small size, O₂ supply would be expected to be sufficient for meeting fetal O₂ demands. Across many organs of the body, metabolic activity is closely linked with blood flow, with vasodilation supported by metabolic by-products such as K⁺, adenosine, lactate, H⁺, and CO₂, thereby integrating O₂ demand with supply (38). While it has long been known that fetal growth restriction develops when fetal O₂ supply drops below a “margin of safety” (39), unclear are the pathways linking perfusion and metabolism. Ovine models of heat-induced placental insufficiency result in hypoxemic and growth-restricted fetuses with altered nutrient utilization and insulin sensitivity (40–42) but the contribution of hypoxia to these fetal metabolic responses remains a critical knowledge gap. While a focus within the field has been on how tissue metabolism affects blood flow and nutrient delivery, we suggest that hypoxia itself, whether resulting from placental insufficiency or lower uterine artery blood flow, can serve to widen the uterine arterial-uterine venous O₂ content gradient, lower umbilical venous Po₂, and prompt metabolic changes to slow fetal growth to match O₂ supply with demand.

Studies in indigenous high-altitude populations support integration of blood flow with uteroplacental and/or fetal metabolism. Genetic selection in Andeans for a variant of PRKAA1, the gene encoding the α-1 catalytic subunit of adenosine monophosphate kinase (AMPK), is associated with a normal pregnancy rise in uterine artery diameter and normal fetal growth (43). AMPK has potent pleiotropic effects, acting both as a vasodilator in the uterine circulation (44) and a highly conserved cellular energy gauge with multiple downstream transcriptional effects on metabolism, including activation of mitochondrial biogenesis, glucose utilization, and glycolysis to increase ATP production (45). AMPK is activated in placental tissues from high- versus low-altitude pregnancies with appropriate for gestational age birth weight infants (46). Furthermore, pharmacological AMPK activation in mice selectively raised uterine artery blood flow and halved the altitude-associated reductions in fetal weight (47). Thus, activation of AMPK as well as likely other important metabolic regulators could serve as mechanisms by which uteroplacental blood flow is integrated with fetoplacental O₂ and other nutrient demands.

Support for venous Po₂ influencing tissue metabolism comes from observations made during the Operation Everest II study in which blood gases and maximal O₂ uptake were measured at simulated altitudes up to and including the summit of Mt. Everest (48). It has long been known that V̇o_2max falls acutely and after acclimatization to high altitude; however, identifying the responsible mechanisms has proven elusive. Quite possibly, the answer stems from the need to maintain sufficient venous Po₂ in the exercising muscle to permit adequate O₂ diffusion to the mitochondria for oxidative phosphorylation. Consistent with such an expectation, maximal exercise and O₂ uptake were maintained in direct proportion to mixed venous Po₂ (Fig. 2). When venous Po₂ is not maintained at high altitudes, muscle substrate preference switches away from fatty acid oxidation toward more O₂-efficient glucose utilization. With prolonged exposures to extreme altitude (>5,300 m), a loss of muscle mitochondrial capacity further limits muscle O₂ demand which, together with metabolic and microcirculatory impairments, prevents raising arterial Po₂ from fully restoring V̇o_2max (49, 50).

Figure 2.

Maximal exercise capacity or rate of whole body O₂ consumption (V̇o_2max) declines in direct proportion to the fall in estimated exercising-muscle Po₂ at progressively higher altitudes. Values shown are from mixed venous Po₂ samples obtained in eight healthy males during maximal exercise at the progressively higher altitudes present during a simulated, hypobaric chamber ascent of Mt. Everest (48).

If the pregnancy rise in uterine artery blood flow fails to occur, we hypothesize that a widening of the uterine arterial-venous PO₂ gradient would occur chronically and, in turn, decrease uterine vein and intervillous PO₂, umbilical vein PO₂, placental mitochondrial electron transport chain function, and fetal growth. Support for this comes from the stabilization of HIF-1α and increased downstream expression of the microRNA, miR-210 observed in newcomer high-altitude residents (51). Of note, miR210 was similarly found to suppress placental mitochondrial capacity in preeclampsia (52). Decreased mitochondrial oxidative phosphorylation at altitude could, in turn, increase placental reliance on glycolytic ATP production via AMPK activation and/or HIF-1α-mediated stimulation of genes encoding glycolytic enzymes (53, 54). Greater placental glucose consumption and a consequent reduction in fetal glucose availability have been suggested as the cause of the reduction in fetal growth seen at high altitudes (55). However, direct evidence for an increase in placental glucose consumption is lacking (56) and the lower venous glucose levels characteristic of high-altitude residents (57, 58) are due to increased skeletal muscle, not placental, glucose consumption (54). Therefore, the metabolic factors involved in suppressing fetal growth are likely to be more complex, potentially involving an increase in pyruvate flux from fetal to uteroplacental tissues for use as a substrate for energy production and changes in preference for pyruvate oxidation over fatty acids under conditions of chronic hypoxia (59). In addition, perhaps sustained placental hypoxia modifies the cross talk among AMPK, HIF-1α, and mTOR signaling pathways given our observation of a lack of evidence for mTOR inhibition despite activation of AMPK in placental tissues from women residing at >2,500 m compared with women residing at 100 m (46).

In summary, blood gas and nutrient data are clearly needed from both the maternal and fetal sides of the placenta to understand the factors responsible for the long-observed, hypoxia-associated reductions in fetal growth. Such information can be expected to advance our understanding not only of the mechanisms by which O₂ and other nutrients regulate fetal growth but also lead to therapeutic interventions for treating or ultimately preventing pregnancy complications that restrict fetal growth at any altitude.

GRANTS

Grant support for research cited here was provided by National Institutes of Health Grants HD088590 (to LGM and CGJ), HL138181 (to CGJ), HL 079647 (to LGM), TW007957 (to LGM and CGJ), DK108910 (to SRW); and; Action Medical Research SP4545, and the British Heart Foundation PS/02/002/14893 and RG/07/004/22659 (to AJM).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

L.G.M., A.J.M., and C.G.J. conceived and designed research; C.G.J. performed experiments; L.G.M. and S.R.W. analyzed data; L.G.M., S.R.W., R.A.L., A.J.M., and C.G.J. interpreted results of experiments; L.G.M. and R.A.L. prepared figures; L.G.M. drafted manuscript; L.G.M., R.A.L., A.J.M., and C.G.J. edited and revised manuscript; L.G.M., S.R.W., R.A.L., A.J.M., and C.G.J. approved final version of manuscript.