Starling Matters

John Fuller Beckwith4/27/2026

In lymphedema education the Starling equation is central to understanding not only normal physiology of fluid exchange but also to understanding the pathophysiology of lymphedema. As one former faculty colleague stated, the Starling lecture is the “Mt. Everest” of the lectures we gave to students. Without getting lost in the details, the essence of understanding is the four Starling forces: hydrostatic pressure (or BCP - blood capillary pressure), oncotic pressure (or COPp - colloid osmotic pressure of the plasma), colloid osmotic pressure of the interstitium (COPi) and tissue pressure (or IP - interstitial pressure); and also that across the breadth of the blood capillary there is an even exchange between filtrated and reabsorbed fluid. The essence of the lymphedema pathophysiology lecture is that lymphatic absorption of filtrated fluid is the final factor that accounts for balance in fluid exchange.

Along comes Levick with a substantial challenge to this understanding. Over the past decade or so there has been within the lymphedema community a bit of a hyperbolic reaction to what is really nothing more than a refinement of Starling’s equation. From over-the-top exclamations such as “Starling is debunked” shouted out at conferences, to many presenters emphasizing we now have a new understanding of fluid exchange that replaces the old understanding, this is a case of throwing the baby out with the bath water.

The traditional Starling principle underlies much of the physiology and pathophysiology of lymphedema and as such is central to understanding the disease. A change in the Starling principle would be fundamental to our understanding of the causes and even treatment of lymphedema. In this paper I would like to examine the insights offered by Levick, put them in the context of commentary in the Foeldi textbook (2003) and thereby gain a useful and realistic perspective of Levick’s contributions.

I hope to demonstrate that the factors underpinning the Starling Principle remain the foundation of understanding edema and lymphedema, while the revised Starling Principle improves our understanding of how the Starling forces interact in relation to each other, the location where reabsorption occurs and pathophysiology of edemas. Further, it will be seen that instruction in the basic Starling Principle is still the key building block in understanding and treating the disease of lymphedema.

Two points to be clarified regarding Starling and Levick: first, the Starling principle was not meant to be a rigid model, never to be questioned. Rather it is a conceptual framework, or as Foeldi states, “…more a didactic model than a reflection of reality.” At the other end of the debate, Levick does not actually state there is no reabsorption in the venous end of the blood capillary, but rather that so-called steady state reabsorption is, “unlikely,” and there is no, “sustained,” reabsorption, and further that transient state reabsorption can and does occur in normal tissues. So neither Starling nor Levick is absolute and therefore not completely in opposition to one another. It should not be a question of one vs the other.

Levick laments that what Starling meant as a model became accepted as fact. Levick does support the basic principle that fluid homeostasis is based on the relationship between BCP and plasma proteins. However, he challenges traditional understanding of interstitial pressures, particularly tissue pressure (IP) and interstitial protein concentration pressures (COPi). His premise is that since there is little reabsorption downstream (venous end) in the blood capillary the lymph system must be responsible for absorption of most of the ultrafltrate. Levick examines the dynamic, variable nature of filtration volume and ultimately the new insights about the impact of the glycocalyx on the protein concentration gradient.

Levick takes three lines of evidence in suggesting a revision of the traditional Starling Principle: sum-of-forces, direct observation when BCP is low, and theoretical considerations. In the sum-of-forces argument Levick highlights two points: first, that inferior to the heart blood capillary pressure increases such that it exceeds plasma protein pressures and therefore there is only filtration (and hence no reabsorption) across the entire capillary structure; second, that in experiments using inserted wicks, Bates showed that tissue pressures were subatmospheric and interstitial proteins were of such a concentration that there was no net absorptive force. Foeldi counters this argument with the observation that the wick method of measuring tissue pressure causes immediate inflammation so that any measurement of pressure and protein concentration derived by this method of observation would not be reflective of normal tissue. Therefore the values Levick depends on for his conclusions are not accurate. This would undercut Levick’s basic argument. Also, Foeldi notes that other authors have pegged the interstitial pressure in the skin of the hand at +6.3 mmHg, clearly not subatmospheric, and makes the logical argument that if interstitial pressures were effectively negative pressures, then there would not be a force for lymph formation since the interstitial pressure could not overcome the basal lymph capillary pressure and therefore fluid could not flow to the lymph capillary because it would be against this pressure gradient. Again, raises the question if Levick is accurate in his formulations of interstitial pressure. Foeldi addresses the high capillary pressures argument on two fronts - first, that of course there is increase in capillary pressures inferior to the heart, but in the recumbent position capillary pressure will decrease such that at the venous end of the blood capillary it is less than the plasma protein pressures and therefore reabsorption occurs. Second is that capillary pressure varies continually with the systole and diastole of vaso-motion, so it is continually changing even in the recumbent position.

Levick’s second line of evidence is experimental observations of fluid exchange when capillary pressure is changed abruptly (transient state) versus when capillary pressure is maintained steady for at least 2 minutes (steady state). The experimental finding was that in the transient state reabsorption does in fact occur in the venous end of the blood capillary as capillary pressure goes below COPp. However, in the steady state, as capillary pressure decreased across the capillary, there was a corresponding linear reduction in filtration rate, but as capillary pressure went below COPp the curve was no longer linear and there was no reabsorption, but rather it reverts to slight filtration. Hence Levick states there is no reabsorption in the blood capillary in the steady state, though there is reabsorption in the transient state.

The third line of evidence is the theoretical rationale that COPi is not steady and will change, inversely, as the filtration rate changes across the length of the capillary, increasing as filtration decreases and counteracting potential reabsorption. This is because raising the capillary filtration rate ‘dilutes’ the macromolecules of the interstitial fluid but a decreasing capillary filtration rate concentrates the macromolecules, hence increasing COPi. Foeldi does not disagree with this, acknowledging that the basic Starling equation is a didactic model and in reality the forces vary, however this dynamic fluctuation is not enough to completely negate reabsorption.

The commentary on the glycocalyx, which came after any possible input from Foeldi, appears to indicate a couple of things: first, that the glycocalyx is the semipermeable layer that determines COP pressure gradient, second that the subglycocalyx region, as opposed to the pericapillary interstitial region, is where the effective protein concentration resides, and lastly that the subglycocalyx COP can be only 10% of the COP in pericapillary interstitial tissue. Therefore, it is proposed that when determining Starling forces, COPi should be replaced with COPg. Even so, Levick does not say COPi has no bearing, just that is appears to have less bearing than assumed in the Starling Principle.

Interestingly, Levick does not comment on the effect of lymphedema on the Starling forces. With his focus on the dynamic (variable) concentration of interstitial proteins, some commentary on the increase in interstitial protein concentration due to lymphatic mechanical insufficiency would seem warranted. And it would be appropriate in the lymphedema world that we should scrutinize all of this in the context of lymphedema.

Summary

Levick and Foeldi make a number of the same points, meanwhile where Foeldi disagrees he has made some effective counterpoints to Levick’s views. The sum of the new view is that there appears to be less (or much less) reabsorption of filtrate downstream (at the venous end of the blood capillary) than represented in the classic Starling equation and so we must consider that the lymph system assumes a greater responsibility for fluid homeostasis. From close scrutiny of this debate, it does not seem reasonable to take the position there is no reabsorption in the venous capillary, but rather that reabsorption can and at times does occur in the venous end of the capillary, that a substantial amount of reabsorption occurs in the lymph nodes, and that, as Foeldi states, reabsorption can occur all along the path of lymphatic return if the pressure and osmotic gradients support it. The relevant point is that the pressures, capillary and interstitial, and the protein concentrations, plasma and interstitial, still matter. If they change, filtration changes, fluid homeostasis is challenged and edema/lymphedema may occur. This is not different and is keeping with the didactic conceptual model of the Starling Principle.

Foeldi’s statement of the continued value of the Starling Equation is the most important view, reflecting that the factors that increase ultrafiltration will lead ultimately to edema/lymphedema. Foeldi’s statement (with slight editing):

The “Starling Equation” is extremely valuable, not for calculations, but because it readily shows that, per unit of time, the volume of net ultrafiltrate will increase:

  1. if the capillary filtration coefficient increases (inflammation);
  2. if the pressure within the blood capillary increases (hyperemia);
  3. if interstitial pressure decreases (loss of tissue compliance);
  4. if the colloid osmotic reflection coefficient decreases; (injury/inflammation)
  5. if the concentration of plasma proteins decreases (hypoproteinemia); or
  6. if the concentration of interstitial proteins increases (lymphedema).

Therefore, Foeldi holds that the basic Starling Principle is key, while Levick proposes a reassessment of where reabsorption occurs. It would seem reasonable to be able to hold both of those views at the same time, since one does not contradict the other. And it all leads to a greater appreciation for the importance of lymphatic function.

The problem is many in the CDT community think this article is so revolutionary that it shakes apart the foundation of our understanding of lymph formation and fluid exchange. This interpretation is simply wrong. In my view, what this article does is parse the Starling equation to more refined detail and inter-relations. And in the conclusion helps us to understand that perhaps there is actually no or little reabsorption occurring at the venous end. Thus, the lymphatics are even more important. Hence it reinforces rather than undermines our conception of lymph formation and pathophysiology of lymphedema. That means we should teach the Starling equation just as we have been, that we must continue to pay proper attention to the four Starling forces, but maybe with the caveat that the lymph capillary may take way more than the 10% of ultrafiltrate. If anyone tells you differently, they are missing the point.