Can Physical Resilience and Frailty in Older Adults be Linked by the Study of Dynamical Systems?

Introduction

Practitioners of Geriatric Medicine often help older patients decide whether proposed medical treatments or procedures are worth the risk of potentially devastating adverse outcomes. Although many studies have helped to identify risk factors that predispose to adverse health outcomes in older adults following medical treatments or procedures, few have allowed investigators to more deeply understand the complex biology that makes up a resilient response to medical treatments or procedures and their incumbent physical stress on the older adult. Indeed, why some older adults rebound when they experience such stressors—or are robust against insults in the first place, whereas others decline precipitously upon exposure to even mild challenges, remains largely unknown. Defining the basis of physical resilience remains one of the most complex challenges facing clinical investigators working to improve the health and safety of older adults. It is also potentially one of the most rewarding as it holds great promise for improving the health and well-being of older adults.

Building on this need, and on our long-standing work to understand the biological basis of physical frailty, we have developed a conceptual framework for resiliency that hypothesizes a specific physiological underpinning, rooted in the dynamical interplay of stress-response systems. While we entertain that resiliency and frailty may entail dimensions that do not overlap, we have long hypothesized that physical or syndromic frailty is a disordered state of the same dynamical systems that produce resiliency when these same physiological systems are functioning in good order. Indeed, in our construct, frailty emerges as individual stress response systems or their networked connections erode past a functional threshold, while resiliency remains intact when the stress response systems remain functionally robust. Hence, in our view, robustness, resilience, and frailty are distinct but related and possibly overlapping concepts that tie back to homeostasis.

Resiliency is a dynamic construct that can be best ascertained through “dynamic stimulation tests”. Indeed, this approach is also likely useful for identifying non-resilient persons before they encounter a major stressor as framed within a classical dynamical systems approach(1), as evidenced by a series of hallmark studies of physiological stimulation tests that were deployed in participants in the Women’s Health and Aging Studies (WHAS)(2–4). Building on this approach, we maintain that the capture of stimulated (dynamic) and unstimulated baseline measures from multiple physiological systems can be highly informative for ascertaining resiliency (5–8). Systems we have in mind include those that regulate glucose tolerance, hypothalamic pituitary adrenal (HPA) axis, autonomic nervous system, inflammatory response, and catecholamine production. This choice reflects our hypothesis that the stress response systems, represented on a measurable physiological level, lie at the center of physiological systems that either promote resilience when functioning optimally or drive frailty when functioning poorly.

The dynamical systems framework by which we conceptualize the etiology of resiliency is in part informed by Csete and Doyle, who wrote that it “is in the nature of their robustness and complexity that biology and advanced engineering are most alike (9).” In engineering, robustness and resiliency are tied to a machine’s complexity, which confers flexibility by which to maintain functioning when faced with stressors. Hallmarks, and their analogy to human physiology, include:

• Modules - “sub-machines” or components which implement a part of the machine’s functioning. In humans, these could be cells, or organelles, or organs, or individual physiological systems. Robust health of each subcomponent tends to confer robust health of the machine as a whole.

• Protocols - rules by which processes governing functioning of modules and relationships between modules are managed. Even if all modules are intact, impaired protocols can lead to system failure. In the present context, protocols are the mechanisms by which physiological systems are internally regulated and co-regulated.

• Redundancy - well-designed complex systems have backup modules and protocols which can take over if primary modules or protocols become impaired.

To implement a dynamical systems framework, we have designed research to collect measurements of physiological “modules” we hypothesize as central to resiliency and will use dynamic paradigms to assess the protocols by which they function and interact. Such a framework has clinical implications, as it might be able to distinguish whether resilience relies on the protection of key “modules” or instead on key “protocols” connecting systems to one other, or rather whether there are a multitude of ways to get to the endgame where resiliency is lost. Analysis will follow the approach proposed in Varadhan et al. (2008), where a stimulus-response modeling paradigm was detailed for deriving summary “parameters” of system functioning (1).

Differentiating Physical Resilience from Homeostasis, Robustness and Physical Frailty

As we developed our conceptual framework of physical resiliency, we carefully considered prior definitions of these constructs. Resilience has been defined in the recent gerontological literature as “a characteristic which determines one’s ability to resist or recover from functional decline following health stressor(s)” (10) and “the ability to resist or recover from adverse effects of a stressor” (11). Ukraintseva (2016) pointed out an important distinction between the ability to resist deviation from original state (which they called “robustness”) and the ability to recover after such deviation (which they called “resilience per se”) (12). We and others have suggested that dynamical systems biology influences both frailty and resiliency (1). Here we further clarify the distinction between homeostasis, the closely related notions of robustness and resilience, and physical frailty.

Homeostasis is the maintenance of a physiological state within certain (narrow) “control” limits. Consider the regulation of electrolyte-water balance in cells: When this is perturbed, then the body responds to reestablish its steady-state level. Note that the control limits for the physiological state variable remain the same after the perturbation. Homeostasis may also be called ‘stability’ in the terminology of dynamical systems.

Robustness is defined as the ability to maintain performance (phenotypic stability) in the face of diverse internal and external perturbations(13). While robustness is related to homeostasis, it covers a larger class of phenomena. Homeostasis, for example, deals with the state of the system, whereas robustness deals with function or performance of the system. While homeostasis is limited to local perturbations around equilibrium, robustness applies to phenomena that are dynamic and far from equilibrium. When discussing robustness, we have to specify the phenotypic features which are essentially unchanged, as well as the nature and degree of perturbations for which the invariance holds. Our characterization of robustness, based essentially on Stelling(13) and Kitano(14), is different from that of Ukraintseva(12) who defined robustness as the ability to resist deviation from original state. In our view, a robust system can move far away from its original equilibrium into a new state of equilibrium, without any discernible change in its performance. For example, in a robust system built with redundancies, a stressor might completely knock out one of the homeostatic pathways, but another pathway can come into play and ensure that the performance is not affected.

Resilience is a more diversely defined concept than robustness, even in fields—like ecology and engineering (15)—with a relatively long usage of the concept. We anticipate that the research we are undertaking will inform key questions over its nature and clinical implications in older adults, such as whether it can be distinguished from robustness and, if so, how “bouncing back” may be operationalized. For the present, we define resilience as the ability of a system to recover from a perturbation of sufficiently large magnitude (a stressor) that the system is pushed into a state far from its original equilibrium state, ultimately retaining essential identity and function. When discussing resilience, we have to specify the phenotypic features whose identity is retained, as well as the nature and degree of perturbations.

Robustness and resilience are closely related, but with one subtle difference: Whereas a robust system maintains its phenotypic stability quantitatively, a resilient system retains its phenotypic identity qualitatively, meaning that there could be some decrease in the function or performance. A non-resilient system, in contrast, is unable to retain its phenotypic identity in the wake of a stressor. For example, a resilient person may recover all of his/her functionality, with some minor deficits (e.g., slightly slower gait), after experiencing a stroke, whereas a non-resilient person who was functionally intact becomes disabled after having a stroke (e.g., needing a walking assistive device). By becoming disabled the person loses his/her phenotypic identity in the sense that there is a major qualitative shift in the phenotypic manifestation. Figure 1(a, b, c and d, based on Oken (16)) depicts these three concepts. Figure 2 delineates these concepts in a different manner.

Figure 1.

A hypothetical and abstract representation of possible responses of a physiological system to a clinical stressor. The solid circle (ball) represents the state of a physiological system. Under small perturbations (1a), the system establishes homeostasis and is stable. When perturbed by a stressor of sufficiently large magnitude, the system is displaced from its original state and establishes another equilibrium state. The system is robust (1b) when it maintains its functionality intact under the new equilibrium. The system is resilient (1c) when it maintains its essential functionality under the new equilibrium. The system is non-resilient (1d) when it loses its essential functionality under the new equilibrium.

Figure 2.

The pre-stressor and post-stressor levels of function of a physiological system. A robust system (triangle) maintains its level of function; a resilient system (circle) may lose some degree of function, although it retains its essential function; a non-resilient system (square) loses its essential function.

Physical Frailty is often defined as “age-associated depletion in physiological reserves resulting in increased vulnerability to stressors”(17). Vulnerability in this statement is rather loosely defined. It may refer to any decline in phenotypic function after experiencing a stressor, in which case frailty would be synonymous with lack of robustness, or it may refer to a substantial and lasting decline in function resulting in loss of phenotypic identity, in which case frailty would be synonymous with lack of resilience. One major point of clarification is that both robustness and resilience have to be carefully indexed with reference to the stressor (S) and the relevant phenotype (P), whereas frailty definition does not typically involve such indexing. Thus, frailty may be viewed as a global construct. It can either be a predictor of different types of system-specific robustnesses and resiliencies, or be a result of a non-robust or non-resilient response to a stressor, or both. Thus, in our view robustness, resilience, and frailty are distinct but related concepts, and the study of specific systems that underlie them can be highly informative to their ultimate definitions.

Conclusion

In summary, the study of physical resiliencies through the lens of dynamical systems physiology highlights the potentially deep biological connections between resiliency, robustness, and frailty. As clinical studies progress in this area, biological data related to stressor-specific and global resiliency traits will be important in consolidating around specific definitions of each construct. That in turn will help to move the field forward towards etiologic discovery and potential preventive and treatment strategies. Potential important implications of such studies would be the identification of 1) specific age-related molecular changes that drive stress-response system decline, 2) key alterations in specific stress-response systems that drive downstream declines in other systems, 3) stressors specific to clinical intervention that drive non-resilient responses, and 4) key physical and cognitive measures that help to better define resilient responses across a number of stressors. Such discoveries, in turn, would make possible interventions to promote resiliency and identify frail individuals before they are exposed to harms. We are hopeful that by so doing, the identification of specific measures that can better characterize and define physical resiliency and frailty across the field of Geriatric Medicine stand to open a new chapter in prevention, treatment and management for older adults.