Pregnancy, Birth and Lactation

Maternal cardiac output rises about 40 percent in healthy human pregnancy, and mean arterial pressure changes only slightly. Two large changes run at once in opposite directions, and a pressure measured at the arm hides both. This lesson takes the pregnant body, the placenta and the fetus as one circulatory and endocrine problem, then follows the minutes in which the lungs replace the placenta. The ovarian cycle, fertilization and implantation come first, in ovarian cycle and fertility and animal reproduction.

Two circulations side by side, normally unmixed

In the human placenta, fetal villi, branched trees of fetal capillaries covered by a continuous layer called the syncytiotrophoblast, hang in an intervillous space filled with maternal blood. Maternal blood bathes the outside of the villus; fetal blood runs inside it. The two supplies lie close together and normally do not mix, and everything that moves between mother and fetus crosses the syncytiotrophoblast and the fetal capillary wall.

What crosses, and how, depends on the solute. Oxygen crosses by diffusion down its partial pressure gradient, with fetal hemoglobin on the far side holding oxygen more tightly at a given PO2 than adult hemoglobin does (oxygen transport). Glucose moves down a maternal-to-fetal concentration gradient by facilitated diffusion, mainly through the transporter GLUT1. Amino acids are different: fetal plasma concentrations of several are higher than maternal, and transfer depends on transport proteins in both syncytiotrophoblast membranes, the one facing maternal blood and the one facing the fetal capillary. Two membranes in series means two transport steps; transfer also depends on blood flow on both sides and on the amino acid concentrations in each compartment.

The static drawing was revised recently. Magnetic resonance imaging of human pregnancies in utero (Dellschaft and colleagues, 2020) found net blood speeds in some placental regions of around 0.04 cm per second in the healthy group (between faster jets of inflowing blood), and dynamic scans showed local contractions of the placental bed expelling maternal blood from the intervillous space. Note the limits of the finding: a net velocity from imaging is not volumetric perfusion, and no benefit to the fetus was shown.

A placenta that also works as an endocrine organ

Early in pregnancy, human chorionic gonadotropin (hCG) from the trophoblast supports the maternal corpus luteum, so progesterone production continues past the point where the luteum would otherwise regress (animal reproduction covers that handover). The placenta then becomes the main source, making progesterone from cholesterol precursors. Progesterone acts through its nuclear receptor (nuclear receptors) and promotes quiescence of the myometrium (smooth muscle).

Estrogens come from a division of labour. The placenta lacks part of the steroid pathway and converts DHEAS, a weak androgen precursor from maternal and fetal adrenal glands, to estradiol, and fetal 16-hydroxylated DHEAS to estriol, as gestation advances. The arrangement is called the maternal-fetal-placental unit because no one tissue holds the whole route (adrenal glands).

Human placental lactogen (hPL) is secreted by the syncytiotrophoblast and is structurally close to growth hormone and prolactin. It antagonizes insulin action: in an early study in adult volunteers, not a pregnancy cohort, twelve-hour infusions of hPL impaired glucose tolerance in seven of eight people tested. hPL is not the only candidate: placental growth hormone, also made by the syncytiotrophoblast, replaces pituitary growth hormone in maternal blood from mid-gestation and is thought to be a major driver of maternal insulin resistance, while hPL also promotes expansion of maternal beta cells. In mice, prolactin and placental lactogen act through the prolactin receptor on maternal beta cells, and deleting that receptor from beta cells left beta-cell mass late in pregnancy (gestational day 16.5) about half that of controls, though it still rose (insulin, glucagon and glucose).

The four-hormone diagram is an abridgement. Single-cell RNA sequencing of first- and second-trimester human placental cells (Liu and colleagues, 2018) distinguished fourteen cell subtypes and detected expression of 102 polypeptide-hormone genes. A transcript is not a secreted hormone, so read that number as a map of what the organ might make.

The mother’s heart, lungs and kidneys reset

In healthy human pregnancy, cardiac output rises by around 40 percent and systemic vascular resistance falls by approximately 25 to 30 percent. Plasma volume expands proportionally more than red-cell mass, so hemoglobin concentration falls by dilution rather than because red cells were lost (blood cells and hemostasis). Minute ventilation can be up to 48 percent higher in the first trimester, through larger tidal volumes at an essentially unchanged respiratory rate; progesterone raises the sensitivity of the respiratory centre to carbon dioxide (control of breathing), arterial PCO2 settles at a plateau around 32 mmHg, and functional residual capacity falls as the diaphragm is displaced upward. Glomerular filtration rate is already up to 40 to 50 percent higher in the first trimester, and in pooled data from uncomplicated pregnancies serum creatinine was lowest at 15 to 21 weeks, 23.2 percent below nonpregnant values (glomerular filtration). Each is a group value from healthy pregnancies that varies with gestational age and method.

Now the arithmetic, with made-up round numbers for one healthy adult woman at rest. Take right atrial pressure as 0 mmHg, so mean arterial pressure equals cardiac output times systemic vascular resistance (blood pressure regulation).

mean arterial pressure = cardiac output x systemic vascular resistance

Pressure. Before pregnancy, cardiac output 5.0 L per min and mean arterial pressure 90 mmHg give resistance = 90 / 5.0 = 18.0 mmHg min per L. In pregnancy, output rises 40 percent to 7.0 L per min and resistance falls 28 percent, inside the reported range, to 18.0 x 0.72 = 12.96 mmHg min per L. New pressure = 7.0 x 12.96 = 90.7, about 91 mmHg. At the edges of the range, a 25 percent fall gives 94.5 mmHg and a 30 percent fall gives 88.2 mmHg. The two changes nearly cancel. Measured pressure does dip a little in the first and second trimesters and returns to nonpregnant levels in the third, so this is a first approximation.

Hematocrit and oxygen delivery. Before pregnancy, plasma 2.5 L and red cells 1.5 L give blood volume 4.0 L and hematocrit 1.5 / 4.0 = 37.5 percent. Suppose plasma rises 44 percent to 3.6 L and red cells 20 percent to 1.8 L. Blood volume is 5.4 L and hematocrit 1.8 / 5.4 = 33.3 percent. Assume hemoglobin per litre of red cells and arterial oxygen saturation are unchanged and ignore dissolved oxygen, so arterial oxygen content scales with hematocrit: 33.3 / 37.5 = 0.889. Oxygen delivery is relative output times relative content: 1.40 x 0.889 = 1.24, about 24 percent higher although content is about 11 percent lower.

Carbon dioxide. In the steady state, arterial PCO2 is proportional to CO2 production divided by alveolar ventilation. Start at 40 mmHg and suppose CO2 production rises 20 percent. For PCO2 to settle at 32 mmHg, alveolar ventilation must rise by the factor 1.20 x 40 / 32 = 1.50, that is 50 percent. Check: 40 x 1.20 / 1.50 = 32 mmHg. Alveolar ventilation is not minute ventilation, since part of each breath ventilates dead space, and the 48 percent above is minute ventilation.

Cardiac output change +40 percent (from 5.0 L per min)

Systemic resistance change −28 percent (from 18.0 mmHg min per L)

Plasma volume change +44 percent (from 2.5 L)

Red-cell volume change +20 percent (from 1.5 L)

Marks: lesson’s worked example (+40, −28, +44, +20) no change (0, 0, 0, 0)

Pressure: output × resistance

Blood volume: plasma + red cells

Hematocrit: red cells ÷ blood volume

Relative oxygen delivery

In this sketch (all of this is made up for the sketch unless the lesson gives it): the baselines are the lesson’s made-up round numbers for one healthy adult woman at rest, with cardiac output 5.0 L per min, mean arterial pressure 90 mmHg, resistance 18.0 mmHg min per L (90 ÷ 5.0), plasma 2.5 L and red cells 1.5 L (hematocrit 37.5 percent); right atrial pressure is taken as 0 mmHg; hemoglobin per litre of red cells is unchanged and dissolved oxygen is ignored, so oxygen content scales with hematocrit; the slider ranges, and any setting other than the lesson’s worked example, are choices made for the sketch. The lesson reports a cardiac output rise of around 40 percent and a fall in resistance of approximately 25 to 30 percent, and only supposes the plasma and red-cell rises. Real values are group values that vary with gestational age and method, and measured pressure dips a little in the first and second trimesters and returns to nonpregnant levels in the third, so this is a first approximation.

Rule to keep: pressure follows the product of output and resistance, so a 40 percent rise in output against a 25 to 30 percent fall in resistance nearly cancel, while hematocrit follows the share of blood volume that is red cells, so oxygen delivery can rise even though oxygen content per litre of blood falls.

Three shunts route fetal blood past the liver and the lungs

The fetal lungs are fluid-filled, pulmonary vascular resistance is high, and gas exchange happens at the placenta. Oxygenated blood returns in the umbilical vein; a variable share passes through the ductus venosus, which bypasses the liver toward the inferior vena cava, while the rest perfuses the liver. From the right atrium a stream crosses the foramen ovale, an opening in the atrial septum, into the left atrium, so relatively well-oxygenated blood reaches the left ventricle and the vessels of the head. Blood entering the right ventricle meets the high resistance of the lungs and mostly leaves the pulmonary artery through the ductus arteriosus into the aorta, held open predominantly by prostaglandin E2 acting on EP4 receptors (the EP4 evidence is from mouse studies), with nitric oxide also contributing. Umbilical arteries carry blood back to the placenta, a low-resistance bed in parallel with the fetal body.

The classic measurements of that layout were made in fetal lambs. Rudolph and Heymann, working in the 1960s, catheterized fetal lambs inside the uterus and injected radioactively labelled microspheres sized to lodge in the first capillary bed they reached, then counted radioactivity organ by organ; the count in an organ is proportional to the share of cardiac output it received at that moment. If the shunts were minor anatomical curiosities, the lungs should have taken a normal share of output. Instead about half of combined ventricular output went to the placenta in young fetal lambs and about 40 percent near term, only a small share went to the lungs, a share that rose with gestation, and substantial streams crossed the foramen ovale and the ductus arteriosus. Those are sheep numbers from an instrumented preparation. Human fetuses differ: Doppler measurements put the lung share at about 13 percent of combined output at 20 weeks, rising to about 25 percent by 30 weeks (Rasanen and colleagues, 1996). Sources: Rudolph and Heymann, Circulation Research 21 (1967) 163-184; Rudolph and Heymann, Circulation Research 26 (1970) 289-299.

At birth the lungs replace the placenta

Aeration of the lungs lowers pulmonary vascular resistance, so pulmonary blood flow rises. Pulmonary venous return raises left atrial pressure, and once it exceeds right atrial pressure the flap of septum primum is pressed against the septum and the foramen ovale stops carrying a right-to-left stream. Losing the placental bed removes a low-resistance circuit, so systemic vascular resistance and ventricular afterload rise. Oxygen constricts the ductus arteriosus, shown directly on fresh human ductal rings studied ex vivo (Michelakis and colleagues, 2002). The 26 ducts came from infants about nine days old, and the experiment shows how isolated tissue responds to oxygen rather than the order of events in a newborn.

These endpoints do not arrive together. In fifty healthy term neonates examined repeatedly by ultrasound, the ductus arteriosus had closed in 94 percent before day 3, the ductus venosus in only 12 percent by then and all of them before day 18; Doppler detected flow across the foramen ovale in every infant. Functional closure, no detectable flow, and anatomical fusion are three different statements.

The modern experiment put a clock on the sequence. In instrumented preterm lambs delivered by caesarean section at 126 plus or minus 1 days of gestation, with flow probes on the pulmonary artery, the carotid artery and the ductus arteriosus, two orders of events were compared in six lambs each: clamping the cord before ventilation, or starting ventilation before clamping. If the cord were only a supply line, the order should have mattered little. Clamping first dropped mean right ventricular output from 114.6 plus or minus 14.4 to 38.8 plus or minus 9.7 mL per min per kg (group means plus or minus SEM) before ventilation began, a fall of about 66 percent, with heart rate also down by about 40 percent, so the fall in output is not purely a stroke-volume effect; ventilating first gave stable flows and pressures across the transition. The reading is that before aeration, umbilical venous return supplies much of the filling of the left ventricle, and aeration lets pulmonary venous return take that role over, so the order decides whether preload falls as afterload rises. The study shows no human outcome and cannot separate a transfusion of blood volume from preload replacement.

Run the mechanism backwards as a test. If pulmonary vascular resistance fails to fall after birth, as in persistent pulmonary hypertension of the newborn, pulmonary blood flow stays low, so pulmonary venous return and left atrial pressure stay low; right atrial pressure can remain the higher of the two, so the foramen ovale flap is not held shut and flow continues right to left; pulmonary artery pressure can stay at or above aortic pressure, so the ductus arteriosus keeps carrying blood toward the aorta; and systemic arterial PO2 stays low, removing one stimulus that constricts the duct.

Labour, and then milk

Parturition is better understood in sheep than in humans. In sheep, a rise in fetal adrenal cortisol near term shifts the placenta’s steroid output away from progesterone and toward estrogens and prostaglandins, the ovine progesterone-withdrawal pathway. In humans circulating progesterone stays elevated until birth, so the withdrawal that releases the myometrium is thought to be functional or local rather than systemic, and no single human trigger is established. In labour, stretch of the cervix and vagina can drive a positive-feedback loop in which oxytocin release increases uterine contraction, which increases stretch (tissues and homeostasis sets out that vocabulary, and reports the mice lacking oxytocin that delivered without obvious deficits), and prostaglandins contract the myometrium and soften the cervical stroma.

Milk has two hormones with two jobs, and students reverse them. The postpartum fall in progesterone, with prolactin maintained, permits secretory activation. Prolactin, from the anterior pituitary, supports secretion of milk by the alveolar cells (endocrine axes and testing covers its dopamine control). Oxytocin, released from the posterior pituitary by magnocellular hypothalamic neurons (hypothalamus and autonomic control), contracts the myoepithelial cells wrapped around each alveolus and ejects milk already stored. Suckling is the main afferent limb for both, and infant cues such as crying can also trigger oxytocin release: prolactin fills the alveolus, oxytocin empties it.

Design problem

In the preterm-lamb experiment, aerating the lungs before clamping the cord largely preserved right ventricular output (a fall of about 22 percent against about 66 percent). Two mechanisms could explain that: pulmonary venous return supplies replacement preload, or the placenta transfuses blood volume into the lamb during the extra interval. Using instrumented lambs, design the measurements and groups that would separate them, and say what each explanation predicts.

One way to do it

Record continuously in every animal: pulmonary arterial, ductal and carotid flow, arterial and central venous pressure, and umbilical flow before clamping. Add a blood-volume measure across the transition, such as dilution of a tracer that stays in the vascular space, plus serial hematocrit.

Groups, gestation-matched and randomised to order: clamp then ventilate; ventilate then clamp; and a third group ventilated first with the umbilical vein occluded for the same interval, so no placental blood can return; if the arteries stay patent, fetal blood can drain into the placenta, so the volume measure has to be read with that in mind. Add a fourth group, clamped first and then given a measured transfusion of blood volume before ventilation, to test the transfusion arm directly. No single group settles it; the comparison across groups does.

Predictions. If replacement preload is doing the work, the ventilate-first and vein-occluded groups both show pulmonary blood flow rising early and output largely preserved, even though the vein-occluded group gains no placental blood and may lose some. Transfusion without aeration should preserve output only while the transfusion runs. If transfusion is doing the work, the vein-occluded group loses much of the benefit of ventilating first; preserved output in the intact ventilate-first group comes with a measurable rise in blood volume; and transfusion alone preserves output without aeration.

What would ruin it: unmatched gestational age, unequal aeration between groups, and hematocrit alone as a volume measure, since fluid shifts change it with no transfusion. The limit stands: preterm lambs, a mechanism, not a practice.