Georg Ivanovas From Autism to Humanism - systems theory in medicine
A system is defined as robust when it continues to function in the face of outer perturbations (Wagner 2007: 1), or in the case of inner problems, like a temporal failure or permanent loss of some components (Macia/Solé 2008). The first part of this definition refers to the biological sight, whereas the second part is a more technical perspective. Due to its machine model current medical science is more concerned with the second part of the definition, and most of medical practice tends to improve defective components., at least in theory.
How can robustness be assessed?
This is not an easy question. Some maintain that robustness can be tested on the level of “a part (protein), a trait (wing shape), or a capability (amino acid biosynthesis). The less the feature changes in the face of perturbation, the more robust it is” (Wagner 2007: 3). This represents a typical reductionist point of view. In this view it seems to be logical to test, for example, genomic robustness to come to conclusions about the robustness of an organism (Lenski et al 2006). This is, however, a fallacy. What is true for a member of a class is no longer be true for the class itself. Such a violation of the logical types leads necessarily to a paradox (chap. 3.2). The paradox of robustness research shows up, when the notion of evolvability is introduced. Evolvability is either understood as the ability to adapt to changing conditions or to reproduce itself. Both meanings are an expression of autopoiesis and we can refer to what has been said before (chap. 4.8): In order to adapt to perturbations the autopoietic unit has to change, otherwise it will cess to exist. Or shorter: In order to be robust, an autopoietic unit has to be evolvable. Reductionism, however, suggested that the more robust a system is, the less evolvable it becomes (Wagner 2008). This paradox vanishes when two preconditions are given:
- The logical level has to be clear: For example, “a highly robust RNA genotype has low evolvability. In contrast, a highly robust phenotype has high evolvability” (Wagner 2008).
- Clear definitions have to be used. I suggest: A behaviour is robust when a system reveals a high evolvability. A system with a low evolvability should be called rigid.
This former definition could be refined by saying that a system is rigid if it shows characteristics of learning 0 or I. A system might be robust if it is shows characteristics of higher orders of learning. But such a hypothesis has to be tested further.
How is a system robust?
In mechanical systems robustness is achieved by redundancy. That is, multiple copies of a given component are available to the system. But this is not the case with the living systems. Actually redundancy would not make too much sense in the living. Under changing conditions (that is in the case of perturbations) multiple copies of the same component would not improve the adaptability and the evolvability.
Living systems attain their robustness through a mechanism called ‘degeneracy’ or ‘distributed robustness’, which is the ability of structurally different elements to perform the same function (Macia/Solé 2008). This has been metaphorically explained by a ‘neutral space’ containing a collection of equivalent solutions (Wagner 2007: 6). Meant is nothing else than the equifinal ability of a system to achieve a goal. This is realized by the interaction of the parts in a network. From brain research it is known that the nerves create new patterns and a new structure confronted with new tasks (Rae-Dupree 2008). Something similar is found in bacteria, where the genetical network improves its evolvability just by adding new links in the gene (Isalan et al 2008).
Distributed robustness has two main characteristics. The first is obvious and generally accepted: Robust living systems are nonlinear. The other characteristic is rather surprising: Nonlinear living systems are operating far from the equilibrium (Goldberg et al 2002). It is surprising as it is against the usual expectation that physiological control in healthy systems aims to reduce variability and to maintain physiological constancy (Goldberger et al 2002). But the opposite seems to be true (Buchanan 1998). Maintaining constancy is not the goal of physiological control (Goldberger et al 2002).
Fluctuations are essential for a living process. They are necessary to compensate perturbations. Famous is the picture of a tightrope walker making chaotic and sometimes strong movements in order to keep balance.
The opposite, rigidity is a characteristic of pathological states. In terms of emergence (chap. 4.10) it is a condition where an environmental perturbation does not lead to a series of reactional patterns. In the terms of synergetics (chap. 4.11) it is a ball in a valley, not able to leave it anymore, even if other valleys would be energetically more appropriate
Robustness and rigidity in the living
These theoretical considerations are supported by many physiological findings. It can be shown that robust systems reveal a nontrivial behaviour far from an equilibrium, whereas rigid systems are found in states of disease.
States of reduced reactions correspond often to severe developments. Babies are known to have large fluctuations in their physiological processes. Taking deep breaths the regular sighing - is essential for them (Baldwin et al 2004a). The reduction of fluctuations might indicate a breakdown of the whole system. Sudden infant death syndrome, a typical network pathology (Bajanowski/Poets 2004), is often preceded by monotonous heart rhythms (Casti 1997).
In an attempt to establish an algorithm describing the normal heart function it has been found that the pattern is chaotic, but not random. The heart ‘remembers’ the last 200 beats and steadily compensates recent rhythms (Buchanan 1998). That is, there is a structure in the fluctuations far from an equilibrium with an inner logic which can be described by mathematical models. In disease this ‘memory’ is erased. It is the loss of rhythms already seen in chronomedicine (chap. 5.3.b). This fact remains normally unnoticed as “traditional algorithms indicate higher complexity for certain pathologic processes associated with random outputs than for healthy dynamics exhibiting long-range correlations. This paradox may be due to the fact that conventional algorithms fail to account for the multiple time scales inherent in healthy physiologic dynamics” (Costa et al 2002). That is, chaotic fluctuations embedded in the general reactional pattern will look like random and superfluous from a linear point of view.
These characteristics become more clear in the following practical examples.
The first illustration (PhysioNet) shows four different types of heart rate. “A and C are from patients in sinus rhythm with severe congestive heart failure. D is from a subject with a cardiac arrhythmia, atrial fibrillation, which produces an erratic heart rate. The healthy record, B, far from a homeostatic constant state, is notable for its visually apparent nonstationarity and "patchiness." These features are related to fractal and nonlinear properties. Their breakdown in disease may be associated with the emergence of excessive regularity (A) and (C), or uncorrelated randomness (D). Of note in C is the presence of strongly periodic oscillations…, which are associated with Cheyne-Stokes breathing, a pathologic type of cyclic respiratory pattern. Quantifying and modelling the complexity of healthy variability, and detecting more subtle alterations with disease and aging, present major challenges in contemporary biomedicine.” (Goldberg et al 2002).
The second illustration (1) of the calcium concentration shows, again, that the pathological state is associated with rigidity whereas the healthy process is characterized by prominent fluctuations (Gerok 1990: 30). It shows the course of Calcium concentration in blood (above), of parathormone (middle) and its metabolite amino acid 44-68 (below).
a) Intense chaotic fluctuation of parathormone in the healthy
b) loss of nearly all oscillation in severe osteoporosis
This is in line with the finding that parathormone given continuously provokes bone loss, its intermittent administration fosters bone formation (Marx 2004b).
The third illustration (due to copyright reasons only in the original publication Ivanovas et al. 2007) shows the behaviour of a healthy family (probably no measurement but a guess). It demonstrates that ‘family balance’ is mainly a sort of a disequilibrium. It only has to remain in a manageable range (Minuchin/Fishman 1981: 22). This contradicts the myth that functioning marriages are free of difficult problems (Roberto 1991: 446). Contrarily, families in equilibrium are regarded as endangered or ill as they are more inclined to become symptomatic in phases of transition (chap. 5.2).
(1) Illustration by courtesy of Wolfgang Gerok