This chapter answers Section G6(i) of the 2017 CICM Primary Syllabus, which expects the exam candidate to "explain the response of the circulation to changes in posture". In the normal ambulant person this response is mainly coordinated by the baroreceptor reflex, which is explained in detail elsewhere. Bizarrely, this has never come up in the exam, though it is absolutely ideal for viva and written questions. Of course, there are numerous possible changes in posture, and it would be impractical to deal with every possible variant, nor would they make satisfactory questions in entry-level exams.
What would be the minimum expected knowledge here, one wonders? With nothing official to guide us, we can only assume that CICM examiners would be disappointed if you did not at least understand the haemodynamic effects of transitioning from supine to upright position. However, there are many fascinating elements in the physiology of posture, and the effect of gravity on the circulation in a broader sense. Predictably, the end half of this chapter suffers from a significantly degraded focus on learner priorities. As the exam candidate may not be inclined to have their time wasted, memorable facts have been distilled and concentrated into this dense crystalline form:
- Hydrostatic contribution to blood pressure
- Gravity forces the redistribution of blood within the circulatory system
- The weight of the column of blood contributes 73.5 mmHg per 100cm of height
- Arterial effects:
- Decreased perfusion pressure in superior regions
- Increased perfusion pressure pressure in dependent regions
- Venous effects:
- Redistribution of venous blood volume (70% of total) into dependent regions
- Thus, decreased venous return
- Hydrostatic indifference point:
- The point inside the static circulatory system where pressure and therefore wall stress remains stable irrespective of the change in position.
- This point describes the reference point from which the hydrostatic pressure at the baroreceptors is determined
- When the baroreceptors are positioned above this point, the hydrostatic contribution to arterial pressure at the level of the baroreceptor is decreased.
- Effects of standing from supine position to upright position
- Venous return decreases, thus RV and LV stroke volume is decreased
- Carotid sinus baroreceptor pressure decreases because of:
- Elevation above hydrostatic indifference point
- Decreased cardiac output
- Baroreceptor reflex is activated, which:
- Increases heart rate by decreasing vagal tone (immediately)
- Increases peripheral vascular resistance (with a slight delay)
- Cerebral perfusion pressure also decreases;
- Cerebral vessels vasodilate to preserves blood flow
- The net effects are:
- Increased heart rate
- Increased blood pressure
- Decreased stroke volume
- Stable or slightly decreased cardiac output
- Stable cerebral perfusion
- In an awake patient standing up voluntarily, lower limb muscle pump activity ameliorates the decrease in venous return
Coonan & Hope (1983) have an excellent article to answer this exact question, including relevant details such as "what happens to the circulation when the anaesthetised patient is made to sit upright at 45 degrees?". Borst et al (1984) offers some experimental data of non-anaesthetised subjects, and is also free. Additionally, The Physiology Viva by Kerry Brandis has a great succinct one-page description, and one can be reasonably confident that all of the examiners also have a copy. Lastly, for thirty-nine pages of dense data and intense maths one may be referred to "Cardiovascular adjustments to gravitational stress" by Blomqvist et al (2011), which was the main source for what follows, and which should not be read by any of the trainees for any reason.
There is some discussion of this in the local chapter on the systolic mean and diastolic blood pressure, but the tone of that discussion is highly unprofessional, the author taking considerable liberties with the subject matter and at one point digressing on the subject of giraffes. To see what a more professional discussion of this might look like, the reader is directed to Hinghofer-Szalkay (2011) and Tansey et al (2019).
In summary, gravity (or any acceleration force for that matter) affects the distribution of blood within the circulatory system, because it gives weight to the mass of blood. To understand how this works, one must have the courage to see the human body for what it truly is - a fluid-filled column. Given the acceleration due to gravity (usually 9.8m/s2) and the density of blood (about 1055 g/l), that gives a pressure of 73.5 mmHg per every 100 cm of fluid height. Thus, in the upright man-column, there is a substantial difference in pressure between the head and the toes. To borrow values from a famous source:
The arterial circulation is at least protected from the worst of this by its high pressure and the muscularity of its walls, i.e. it behaves a bit like the rigid glassy pipe depicted above. Moreover there's not much blood in there, only perhaps 15-20% of the total circulatory volume. In contrast, the venous circulation has 70% of the blood, and weak thin walls, which make it a useless flaccid sack that deforms passively according to however the blood decides to slosh around inside it. There's probably some more professional way to word that. Venous intravascular pressure increases in the same direction as the vector of gravity or acceleration, which increases intravascular pressure, and because veins are more compliant than arteries, the dependent venous structures distended under that pressure and accommodate more volume. Basically as posture changes, the total venous blood volume redistributes:
How much volume redistributes? Turns out, a substantial amount. Sjöstrand (1953) reported that, when transitioning from standing to supine position, their five healthy male subjects had about 640ml of blood redistribute back out of the lower extremities. That was about 11% of their total blood volume.
So, now we can relate this thing to something vaguely familiar. From whatever slippery grasp one still might have on the concept of venous return, one might recall that the change in circulating volume affects venous return and cardiac output by shifting the vascular function curve left along the x-axis, effectively changing the mean systemic filling pressure. The MSFP is of course unchanged (the volume is not really lost, it's just spending some time in your legs), but this model does not permit that sort of sophistication. Thus:
In short, as the hydrostatic pressure gradient drags blood into the lower body, the upper body (where the heart happens to be) has to work with what is effectively a decreased circulating volume, and the cardiac output decreases as the result. This is also the mechanism by which the passive leg raise manoeuvre tests fluid responsiveness: by lifting the patient's legs over their haemostatic indifference point, you return about 600ml of blood back into the circulation, increasing the effective circulatory volume and boosting the cardiac output.
Before describing the reflex changes in the circulatory system due to posture, it seems important to discuss the concept of the hydrostatic indifference point. From the point of view of structuring the discussion, there is nowhere else to put this concept. It certainly does not belong at the beginning of the chapter, because it is entirely possible to explain all of the important points without it. On the other hand, the literature keeps referring to it, so it should probably be mentioned somewhere. It is a concept which comes from the assumption that, as the posture changes, some body regions will become uppermost and will decrease in pressure, others will become dependent and increase in pressure, and therefore somewhere there will be a middle point where the pressure will remain unchanged:
So that sounds logical, but - what is the point of knowing this? Helmut Hinghofer-Szalkay (2011) tries to explain. In his own words:
"At this specific location, there is probably little change in vessel volume, wall tension, and the balance of Starling forces after a positional manoeuvre. In terms of cardiac function, this is important because ...Baroreceptors pick up pressure signals that depend on their respective distance to hydrostatic indifference locations with any change of body position."
In other words, if you had all of your cardiac reflex sensors concentrated at a point where pressure and vessel wall tension did not change with changes in posture, those stretch-sensitive mechanosensors would have nothing to sense, and your cardiac reflexes would be oblivious to the major redistribution of volume which might have occurred.
Fortunately, all those sensors are generally distributed away from the arterial and venous hydrostatic indifference point. Which brings the question: where exactly the hell is it? Of course, there are macabre experiments from the early 20th century for us to enjoy. Clark et al (1934), for example, rotated the bodies of dead animals and recorded pressure values at different points, arriving at a spot in the upper IVC approximately 6cm below the heart which did not seem to change in pressure irrespective of how the carcass was twisted. Guyton & Greganti (1956) performed the same experiment with non-dead anaesthetised animals and ended up with a slightly different location, probably because the intact circulation had sympathetic tone. They came up with a spot in the right ventricle, just below the tricuspid valve:
The investigators did not stop with mere posture changes; they induced haemorrhagic shock in their animals, filled them with adrenaline and destroyed their spinal cords - the venous indifference point remained stationary. From this, the investigators drew considerable significance. "It is to be expected that venous pressures referred to the hydrostatic level of the right ventricle where the end-diastolic pressure is measured would not vary greatly as the animal’s body is rotated", they wrote, because the right ventricular diastolic pressure (which is essentially right atrial pressure) is a major determinant of venous return, in the clockwork universe of Guyton.
So, that's where the hydrostatic indifference point is. Or rather, that's where it is in the anaesthetised dog strapped to a tilt-table. In humans, it is probably extracardiac. Most resources (including modern ones like Petersen et al, 2014) report that it is a few centimetres below the diaphragm. Their references for this are always Gauer & Thron (1965), who published an influential chapter in the ancient and inaccessible Handbook of Physiology, and for some reason Ernst Wagner's undecipherable "Fortgesetzte Untersuchungen über den Einfluss der Schwere auf den Kreislauf" (1896). In short, the exact experiment is difficult to track down, but it appears that people familiar with the topic are convinced, and later experiments (again, Petersen et al, 2014) have confirmed this position. For them, it was 7 ± 4 cm below the 4th intercostal space.
Or at least that's the venous hydrostatic indifference point. Obviously there is only one true hydrostatic indifference point within a simple uninterrupted fluid compartment, like the gall bladder. The circulatory system is much more complex than that. We generally tend to assume that, as the arterial circulation is rather separate from the venous, it has its own indifference point, which is in a slightly different position. After observing a large number of rotating dogs, Smith & Guyton (1963) came to the conclusion that "in most normal dogs such an axis lay in the neck a few centimetres cephalad to the sternum", somewhat below their carotid baroreceptors. In the human, the location of the arterial hydrostatic indifference point is probably slightly different, and apparently roughly at the level of the heart. Again this comes from Gauer & Thron (1965), and so the original experiment is lost in the depths of time, but most authors who reference them seem to put this point around the aortic root.
At this stage the pragmatic intensivist will belch obscenities, demanding answers to the question: what is the point of all this, and why should we care? Well. The theory is that the location of this point is a determinant of the sensed change in pressure by all the mechanoreceptors. Specifically, it determines the hydrostatic contribution to blood pressure at those receptors, which is an important contributor to the total pressure. For example, for the carotid baroreceptors, hydrostatic pressure at the carotid sinus changes with posture according to the following equation:
ρgh × sin(α)
- ρ is the density of blood,
- g is the acceleration due to gravity,
- h is the distance between the baroreceptor and arterial hydrostatic indifference point, and
- α is the angle of tilt of the carotid artery measured from the horizontal.
This equation comes from Heldt et al (2002) who needed such mathematical trickery to create a computational model of the cardiovascular responses to posture, but in fact equations are not required to understand this intuitively apparent truth. A childish diagram is probably enough:
So, in summary, the pressure gradient in the upright arterial circulation decreases the perfusion pressure of the brain (which is now far above the heart). Taking a normal ICP as 5-10 mmHg, the cerebral perfusion pressure is now 29-34 mmHg, which is far below what you might typically expect. On top of that, as mentioned above, a change in posture from supine to upright decreases venous return and cardiac output. So, the cardiac output has fallen, and the perfusion pressure of the brain has dropped. Needless to say, an organism couldn't remain erect and bipedal for long unless it had developed some mechanisms to maintain its blood pressure in this position. These mechanisms are the subject of the rest of this chapter.
Finally after 2,000 words the reader encounters something with clinical relevance. In summary, when a supine person abruptly transitions to an upright position, the following events take place:
To illustrate what happens graphically, here are some data from Katkov & Chesthukhin (1980). The article itself, commissioned by the USSR Ministry for Health, can't be found anywhere, but fortunately the graphs were reproduced by Blomqvist et al (2011). The originals have been reproduced here after some considerable modification, which (the author flatters himself) have enhanced their clarity.
Yes, the Soviet scientists who collected these data measured the arterial systolic and diastolic pressure from the foot, which is why that measurement went up. The difference between the supine and erect measurements in pressure corresponded very closely to what might be predicted purely from hydrostatic estimates alone. From this, we can infer that carotid sinus pressure would have dropped proportionally, though we have no direct human evidence for it.
What does all this reflex activity look like in vivo, plotted over time? Here's an excellent recording from Critchley et al (1997) which documents the effects of a sudden 55 head-up tilt in a healthy young male:
The precise timing of sympathetic and parasympathetic activity has been estimated rather than measured. In trying to model the changes in pressure and flow which occur during postural changes, Olufsen (2005) had assumed that:
"..parasympathetic withdrawal induces fast (within 1−2 cardiac cycles) increases in heart rate, whereas sympathetic activation yields a slower (within 6−8 cardiac cycles) increase in vascular resistance, vascular tone, and cardiac contractility and a further increase in heart rate"
From basic channel physiology, you might expect the parasympathetic response here to be somewhat faster, as acetylcholine acts directly to open potassium channels. Some seconds later, sympathetic effects catch up when enough cAMP is produced.
So far the discussion has been focused on the results derived from conscious young adults standing upright from a supine position, which, though it illustrates the physiological principles, is probably not the scenario of interest for the intensivist. Young conscious people rarely find themselves in intensive care units, and unconscious patients rarely find themselves going from supine to upright. The anaesthetised crowd is much more likely to experience a seated position, eg. the "beach chair" pose for the purpose of shoulder surgery. To describe the haemodynamic changes in this scenario, Coonan & Hope (1983) present the following table:
|Parameter||Change from supine to seated position|
|Blood pressure||Increases 0-40%|
|Cardiac output||Decreased 12-20%|
|Heart rate||Remains largely unchanged|
|Systemic vascular resistance||Increases 50-80%|
|Cerebral blood flow||Decreases 15%|
As you can see, this is essentially the same series of changes as is seen in alert patients on tilt-tables. These values can probably be also applied to patients in the ICU, who spend most of their time in Fowler's position with the head up 30-45°. Unfortunately, they are very vague and generic (serious, a 0-40% range for change in blood pressure). Thankfully, Kubota et al (2011) were able to measure the changes directly using ten healthy university students. Their tabulated data was so good, the author couldn't help but graph it. In short, when compared to what happens in the supine-to-standing situation, the changes due to Fowler's position are trivial:
Thus far, we have been discussing the cardiovascular events which take place when somebody takes your baroreceptors and hoists them high above the hydrostatic indifference point. That is not the only direction things could go. Critical care patients also occasionally end up with their head down and their feet up, often in a misguided application of the Trendelenburg position within a context of hypotension. The whole point of this manoeuvre, as originally devised, was to make the abdominal organs slosh into the chest, exposing more of the pelvis to surgical access; it was apparently borrowed from "gelding and spaying methods involved in animal breeding" and was already described in thirteenth-century surgical manuals (Belloni, 1949), well before 1985 when one of Friedrich Trendeleburg's students had described it.
For a period, it was believed that this position has some benefit in the management of shock. As a random sample of the contemporary literature, Segal & Aisner (1944) recommended that "for the treatment of primary or initial shock which occurs immediately following trauma the Trendelenburg position is indicated and of value". Subsequently, this has been thoroughly debunked, and most of the time these days in the ICU the Trendelenburg position is used only transiently, to protect patients from air embolism during the insertion of central lines. However, it still has haemodynamic effects, and somebody somewhere is right now preparing an exam question to challenge you about it.
Fortunately, there are some good data to describe this. Sibbald et al (1979), writing in an era when every patient had a PA catheter, described the following changes in a group of mixed ICU patients, not all of whom were shocked:
|Parameter||Change from supine to 15-20º head down position|
|Blood pressure||Increases ~5%|
|Cardiac output||Remains essentially the same|
|Heart rate||Remains largely unchanged|
|Systemic vascular resistance||Decreases ~ 5%|
|Pulmonary artery wedge pressure||Increases bu 3-4 mmHg|
In short, because the venous volume of the lower body ends up being redistributed to the upper, one might expect cardiac preload to increase, and therefore the cardiac output. In practice, the effect of this position on carotid baroreceptors results in a decrease in systemic vascular resistance and cardiac contractility, which means the blood pressure remains essentially the same. In short, no haemodynamic benefit occurs when hypotensive or hypovolemic patients are placed in this position. Moreover, the increased venous pressure in the veins of the head and neck act to decrease venous drainage from the brain, reducing cerebral perfusion. Shenkin et al (1949) measured a 17% decrease in cerebral blood flow as the consequence of Trendeleburg position, albeit over a short period of observation.
It is probably worth finishing a discussion of postural haemodynamic effects with a discussion of prone position. It is unlikely to ever become an exam question at the level of the CICM First Part because the answer is still debated at the highest levels. Observe this paper by Jozwiak et al (2013), the editorial by Sheldon Magder (2013), the response by Albert & Hubmayr (2014). The situation is made more complex by the fact that in the ICU, the patient who ends up prone is also the patient with the most severe ARDS, who is therefore subject to all sorts of ventilatory ultraviolence, which is not without its own effect on the cardiovascular system. Fortunately, Jon-Emile Kenny from PulmCCM does an excellent job of unravelling the haemodynamic effects of prone position from the haemodynamic effects of high-pressure ventilation.
First, let us consider the prone patient whose lungs aren't turning into concrete. With the baroreceptors remaining in the same plane as during supine position, there should be no hydrostatic effects here. Thus, the main difference in haemodynamics must come from the compression of abdominal structures. This might mobilise some extra blood from the splanchnic circulation (which you might expect to improve venous return), or it may increase the resistance to venous return via the compression of the inferior vena cava. Which will it be, in any given patient? To avoid this question, the abdomen is generally protected from direct pressure during prone positioning.
Exploring the magnitude of these effects, Dharmavaram et al (2006) performed some measurements and cardiac echo studies of elective outpatients undergoing lumbar spine surgery. Their data is comprehensive, and includes a comparison of different surgical techniques, of which the Jackson table probably resembles "ICU prone" best, as in this scenario the abdomen is relatively free.
|Parameter||Change from supine to 15-20º head down position|
|Mean arterial pressure||Increased by 3.6%|
|Cardiac output||Decreased by 8.6%|
|Heart rate||Decreased by 9.4%|
|Stroke volume||Decreased by 8.%|
|Pulmonary artery wedge pressure||Increases bu 3-4 mmHg|
As you can see, the effects were relatively minor. The authors specifically ascribed this to the fact that the abdomen was totally unobstructed. In contrast, with the use of the Siemens positioning system ( a partially compressed abdomen), the stroke volume and cardiac index were substantially lower.
Now, the ARDS patient. Not flipped on a Jackson table, but positioned carefully (lovingly, delicately) over a cumulonimbus of stacked pillows. Also, ventilated with ungodly pressures. What effects might his have? Well. The reasonable person would deflect this question by saying that the effects would be extremely individual. The following (conflicting, contradictory) effects can be expected, most of them concerning the right side of the circulation:
So, in summary, the patient could get worse, or better, or stay the same, and it is difficult to predict which it would be, except for the situation where the patient is significantly hypovolaemic. In the presence of insufficient volume, abdominal compression effects would have minimal positive and maximal negative effect, producing haemodynamic unpleasantness irrespective of whatever improvements might take place in the pulmonary circulation. Fortunately, intensivists are generally guilty of being overgenerous with their fluids. As an example of what happens in the real world, one can look at the table of adverse events from the PROSEVA trial. 14.8% of the prone group patients recorded hypotensive events (SBP <60 mmHg for more than 5 minutes), a complication which was more common in the supine group (21%).
The attentive reader will by this stage have firmly grasped that the force of gravity plays something of a role in all of this. From that, it should follow that the absence of gravity would render this entire discussion pointless, as the concept of the hydrostatic indifference point would no longer have any meaning - no matter how you rotate the weightless human body, none of the fluid compartment would experience any acceleration in any direction and there would never be "superior" or "dependent" areas. Released from the surly bonds of Earth, the baroreceptors would only be reacting to cardiogenic and vascular blood pressure changes, as all hydrostatic phenomena would be eliminated. Or, rather, the hydrostatic pressure in all corners of the circulatory system would be similar, and so one might assume that it might be approximately the same as that of a supine person having a lie-down at normal Earth gravity.
Looking at direct measurements, that is not entirely what happens. Sure, water and blood redistribute out of the lower body (Moore & Thornton in 1987 determined that the total cephalad fluid shift is probably close to 2L, one litre from each leg). However, this does not achieve a predictable Guytonian result where the CVP and MSFP increases. Buckley et al (1996) measured the CVP of Spacelab astronauts (using a PICC-like pressure transducer catheter which was inserted before launch). Observe, their recordings, somewhat degraded by motion artifact as the subjects ascended into the heavens on a pillar of fire:
As you can see, the CVP actually decreased in orbit (from 5-8 mmHg to 2.5 mmHg). This is completely the opposite of what you might have expected. The fluid in the legs, you might reason - surely it should distribute cephalad, increasing the volume of the venous compartment? So why does the CVP drop?
There are multiple reasons. The main one is likely the effect of a reduced hydrostatic pressure throughout the body fluid compartments. With less pressure in these compartments, and less muscle contraction (i.e. all the muscles normally used to maintain posture are no longer required), the effect is similar to smooth muscle relaxation. Moreover, there is a reduction in intrathoracic pressure, including the loss of compression of venous structures by the weight of the mediastinum. Thus, the CVP drops probably as the consequence of a widespread systemic redistribution of venous blood. Moreover, though the CVP might decrease, the transmural pressure of the central venous compartment is actually increased as the intrapleural pressure is lower, and this is supported by the echocardiographic findings (increase right atrial distension and increased stroke volume).
Fritsche-Yelle et al (1996) reported the data reproduced above, demonstrating that in the absence of constantly changing pressures, the baroreceptor reflex grows lazy and complacent. The heart rate and blood pressure of eightless astronauts is chronically lower than pre-flight measurements, and their baroreflex gain is decreased (i.e. for a given change in pressure, there is a smaller change in heart rate). After the trip is over, the deconditioned cardivascular system struggles to maintain normal values. This is not helped by the fact that the volume of the myocardial muscle decreases, by up to 8% in the first week (Tanaka et al, 2017). In short, the adaptation to proper gravity after microgravity is much more difficult.