| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
|
|
|
|||||||
|
|||||||||
a Hebrew Rehabilitation Center for Aged, Beth Israel Deaconess Medical Center, and Harvard Medical School Division on Aging, Boston, Massachusetts
Lewis A. Lipsitz, Hebrew Rehabilitation Center for Aged, 1200 Centre Street, Boston, MA 02131 E-mail: Lipsitz{at}mail.hrca.harvard.edu.
Decision Editor: John A. Faulkner, PhD
 |
Abstract |
|---|
 Top Abstract Dynamics of Rest and...Alterations in Aging and...Restoring Healthy DynamicsReferences |
|---|
NORMAL physiologic function requires the integration of complex networks of control systems, feedback loops, and other regulatory mechanisms to enable an organism to perform a variety of activities necessary for survival. The control systems of the human body exist at molecular, subcellular, cellular, organ, and systemic levels of organization. Continuous interplay between the electrical, chemical, and mechanical components of these systems ensures that information is constantly exchanged, even as the organism rests. These dynamic processes give rise to a highly adaptive, resilient organism that is primed and ready to respond to internal and external perturbations.
Recognition of the dynamic nature of regulatory processes challenges the prevailing view of homeostasis, which asserts that all healthy cells, tissues, and organs maintain static or steady-state conditions in their internal environment. However, with the development of instruments that can acquire continuous data from physiologic processes such as heart rate, blood pressure, balance control (center-of-pressure displacements), nerve activity, or hormonal secretion, it has become apparent that these systems are in constant flux, even under so-called steady-state conditions. Yates introduced the term homeodynamics (1), which conveys the notion that the high level of bodily control required to survive depends on a dynamic interplay of multiple regulatory mechanisms rather than constancy in the internal environment.
Recently, there has been growing interest in the development of new measures to describe the dynamics of physiologic systems, and use of these measures to distinguish healthy function from disease, or to predict the onset of adverse health-related events (2). A variety of measures have been derived from chaos theory and the fields of nonlinear dynamics and statistical physics. Many of these are based on the concept of fractals.
The classic definition of a fractal, first described by Mandelbrot (3), is a geometric object with "self-similarity" over multiple measurement scales. For example, the ragged, irregular appearance of a coastline or cloud looks similar whether it is measured in inches, feet, or miles. In fact, the smaller the measuring device, the larger the length of a fractal object. This is a property known as power-law scaling. The outputs of dynamic physiologic processes such as heart rate, which are measured over time rather than space, also have fractal properties (4). Their oscillations appear self-similar when observed over seconds, minutes, hours, or days. Furthermore, they demonstrate power-law scaling in the sense that the smaller the frequency of oscillation of these signals, the larger their amplitude (amplitude squared is power). This power-law relation can also be expressed as A 
1/f ß, where A is the amplitude, f is the frequency, and ß is the scaling exponent (generally, 0 
ß 2).
Taking the log of each side of this equation yields a linear relationship between amplitude and frequency, the slope of which is the scaling exponent ß. Any process with 1/f or power-law scaling is fractal-like. The scaling exponent (or slope of the log–log relation) has been used to describe the complexity of a process; one with long-range correlations (self-similarity) across many decades in time or space is considered maximally complex and has a scaling exponent of 1. White noise or uncorrelated randomness lacks the long-range correlations characterizing complexity and has an exponent of 0. Brownian noise (a random walk process), which has only local, short-term correlations between neighboring points, has an exponent of 2. When applied to the frequency spectra of continuous biological signals from healthy subjects, such as heart rate (4)(5), blood pressure (5), electroencephalographic potentials (6), stride interval (7), and center-of-pressure displacements (8) (Fig. 1), this approach demonstrates complex fractal-like behavior.
|
Under basal resting conditions, most healthy physiologic systems demonstrate highly irregular, complex dynamics that represent multiple interacting influences operating over multiple time scales. Unlike classic fractal processes in physics in which the same mechanism is engaged over different scales, physiologic processes operate with different mechanisms interacting over a variety of time scales. This gives functionality to physiologic systems, enabling them to respond on any scale necessary to adapt to internal or external stresses. When the organism is suddenly perturbed or deviates from a given set of boundary conditions, many physiologic systems evoke single closed-loop mechanisms that operate over relatively short time periods to restore the organism to a new steady state that is appropriate for the new set of conditions. The response itself may be periodic with a single-frequency spectrum, rather than fractal with a broadband spectrum and power-law scaling. However, when a new steady state is achieved, it again demonstrates complex dynamics. The process by which physiologic systems evoke single-frequency responses to return to a dynamic steady state (or dynamic equilibrium) can be called reactive tuning. Therefore, studies attempting to define healthy physiologic function by the presence of long-range fractal-like correlations have to distinguish between transient response modes and long-term steady-state behavior. Furthermore, the dynamics observed under steady-state conditions might be used to gauge the capacity of an organism to mount an appropriate adaptive response when it is perturbed.
In this paper I suggest that dynamic, interacting networks of regulatory systems are the structural and functional scaffolds for healthy physiologic function. I will provide evidence for the degradation of these systems with age and disease and suggest that the loss of adaptive capacity that ensues may characterize the onset of frailty. Finally, I will conclude with recent preliminary evidence that age- or disease-related functional impairments may be reversible with a number of interventions that restore the underlying dynamics.
|
Dynamics of Rest and Response |
|---|
|
TopAbstractDynamics of Rest and...Alterations in Aging and...Restoring Healthy DynamicsReferences |
|---|
In general, the more complex a system is (i.e., the more interacting parts it has operating on different scales in time or space), the greater its functionality (although too complex a system may freeze up and lose functionality). The home heating example above is functional only during the winter months, and it can do nothing more than heat the house. However, adding an air conditioning unit and appropriate thermostat control will add year-round functionality. The greater complexity of the automobile enables it to respond to multiple road conditions from a parking lot to a highway. The more complex automobiles of today add functions such as cruise control, antilock brakes, 4-wheel drive, and traction control to enable them to adapt more effectively to a variety of conditions.
Although the mechanical analogy is an oversimplification, physiology can be similarly understood. Underlying the relatively uniform homeostatic responses of living organisms to a variety of stimuli lies a complex dynamic that creates functionality. Unlike mechanical systems that tend to fail when they become too complex (e.g., computers when too much software is installed), physiologic systems are built in such a way that larger-scale components (e.g., skeletal, nervous, and circulatory systems) protect the finer-scale parts (e.g., cells and organelles). In physiology, the greater the complexity, the greater the range of adaptive responses. In support of this notion, I will briefly review the origin of complex dynamics during resting steady-state conditions in a variety of physiologic processes, and the difference between resting dynamics and the response to a perturbation (reactive tuning).
Origin of Complex Dynamics in Physiology
Because of the ready availability of electrocardiographic and blood pressure monitoring devices in medical practice, cardiovascular dynamics are probably the most intensively studied in humans. The sympathetic and parasympathetic limbs of the autonomic nervous system probably account for most of the short-term (second-to-second) beat-to-beat variability in heart rate and blood pressure. On longer time scales of minutes to hours, hormonal and temperature influences may dominate, whereas over 24-hour periods, circadian rhythms exert their control. The importance of the autonomic nervous system in beat-to-beat cardiovascular variability is evident from pharmacologic blocking studies. During blockade of the parasympathetic and sympathetic nervous systems with atropine and propanolol, respectively, beat-to-beat heart rate fluctuations disappear. Yamamoto and colleagues have shown that vagal blockade with atropine decreases the fractal nature of heart rate variability (9), whereas beta-adrenergic blockade with propanolol does not (10). However, these authors comment that during beta blockade, activity of the parasympathetic nervous system alone is not sufficient to produce fractal-like 1/f scaling of heart rate variability (9).
The maturation of sympathetic innervation to the heart during the first month of life in neonatal swine provides another example of the influence of the autonomic nervous system on heart rate variability, and it supports the contribution of sympathetic innervation to heart rate complexity (11). As sympathetic neurons from the right stellate ganglion sprout connections to the heart during this time period, baby pigs develop increasing complexity in their heart rate time series. When the stellate ganglion is denervated at birth, the animals fail to develop the heart rate irregularity that is characteristic of mature animals.
Similarly, the transplanted human heart has been shown by Kresh and Izrailtyan to develop increasing complexity of heart rate variability (measured by pointwise-correlation dimension analysis) following implantation (12). At implantation the donor heart manifests a metronome-like heart rate that evolves in a nonmonotonic fashion, increasing by 100 days, decreasing transiently by 20–30 months, and then rising to its peak by 7–10 years after implantation. This "dynamic reorganization in the allograft rhythm-generating system" probably represents the emergence of local intracardiac control mechanisms, followed by systemic hormonal and neural inputs that evolve over the course of graft– host adaptation.
The resting output of the brain's motor control system is an irregular, low-amplitude physiologic tremor. This probably arises from the rich interconnected network of neurons arising from the motor cortex, basal ganglia, cerebellum, and other structures. Using a network model of dopaminergic innervation in the nigrostriatal pathway of the brain to produce a time series characteristic of the normal aperiodic physiologic tremor, Edwards and colleagues have shown that parameter changes representing weakened synaptic connections result in the emergence of a periodic tremor resembling that of Parkinson's disease (13). This loss of complexity through "dynamic simplification," as they call it, not only suggests potential mechanisms underlying the complex output of the motor control system but also demonstrates how the loss of complex dynamics may be associated with the development of disease.
Mathematical models based on nonlinear dynamics and chaos theory can be used to illustrate how different levels of system inputs and different engagements of system components can generate a variety of dynamic behaviors. As shown in Fig. 2, the logistic equation will generate different dynamic behaviors, progressing from periodic to chaotic, depending on the initial value for parameter a. In similar fashion, the behavior of a given biologic system might differ, depending on the value of a particular input. For example, the output of the dopaminergic nigrostriatal system in the brain that controls movement differs according to the relative concentration of dopamine. Normal concentrations of dopamine produce smooth movements and an irregular, low-amplitude physiologic resting tremor, whereas low concentrations produce stiff movements with a periodic, high-amplitude Parkinsonian tremor. Excessive concentrations of dopamine produce dystonic and dyskinetic motor activity.
|
As shown in Fig. 3, the complex blood pressure dynamics of the supine resting state, represented by a broad power spectrum, give rise to low-frequency "Mayer waves" with a dominant frequency at approximately 0.1 Hz when a healthy young subject is tilted upright to mimic standing. This low-frequency blood pressure oscillation is thought to result from baroreflex-mediated vasomotor activity that counters the reduction in venous return to the heart through sympathetic activation and vasoconstriction, and prevents blood pressure overshoots through sympathetic withdrawal and vasodilation. This feedback mechanism keeps blood pressure in the range necessary to ensure adequate organ perfusion. Many elderly subjects (Fig. 3) have less complexity of heart rate and blood pressure dynamics in the supine and tilted positions (5). Moreover, they may fail to tune in on a dominant frequency when tilted (15). In Fig. 3, the hypertensive elderly subject does not develop low-frequency (baroreflex-mediated) systolic blood pressure oscillations when tilted, and the subject also experiences orthostatic hypotension. Thus, an age- or disease-related loss of complexity in resting dynamics may hinder reactive tuning and thereby lead to maladaptive responses to physiologic stresses. Further research is needed to determine whether the absence of reactive tuning predisposes to blood pressure destabilization or other functional alterations in physiologic systems.
|
In support of the preceding concepts, Butler and colleagues (18) found that orthostatic challenges (lower body negative pressure and head-up tilt) increased the ß coefficient (decreased the fractal dimension or complexity) of the 1/f heart rate spectrum in healthy young subjects. Those subjects who experienced presyncope during orthostatic stress had a significantly greater increase in ß at the highest levels of lower body negative pressure, indicating a greater loss of complexity in heart rate variability. These data suggest that the healthy heart rate response to tilt is associated with reduced complexity (i.e., reactive tuning), but in some individuals with maximal responses the cardiovascular system is unable to achieve a new steady state and syncope results.
If reactive tuning is operative, one would expect the resting, equilibrium relationship between two interacting control systems to be altered during a given perturbation in order to restore the system to a new steady state. In this case, the response may not be predicable from the resting dynamics of the system. The process of cerebral autoregulation provides a good example of a physiologic process in which the resting relationship between blood pressure and cerebral blood flow is altered during acute hypotension, so that abrupt changes in blood pressure do not threaten cerebral blood flow. Using the transfer function between blood pressure and cerebral blood flow velocity oscillations while subjects rested in a sitting position, we were unable to predict changes in cerebral blood flow from changes in blood pressure that occurred when subjects suddenly stood up (19). This is probably due to autoregulatory vasodilation of the peripheral cerebral vasculature, which restores blood flow in response to reduced perfusion pressure. The activation of a focused adaptive mechanism to preserve cerebral blood flow (cerebral autoregulation) in a healthy young subject is evident in the dynamics of the blood flow velocity time series shown in Fig. 4. Note how the dynamic response is greatly damped in the elderly subjects with and without hypertension.
|
The adaptive advantage of complex behavior is also evident in recent work on social networks. Several studies have shown strong relationships between social integration and mortality following myocardial infarction, functional recovery from stroke, and the development of dementia (21)(22). These data lend further support to the notion that the underlying complexity of a system—in this case, an individual's social integration and connectedness—improves one's ability to overcome severe illness.
Thus, complex network-like infrastructures that permit dynamic interactions between individual components during steady-state conditions appear to be associated with a focused adaptive response at times of stress. This concept of reactive tuning may explain a wide range of behaviors, such as when a chaotic stock market makes a unidirectional correction in response to a change in interest rate or when birds on differing flight paths flock together to migrate over long distances (23). These observations highlight the importance of distinguishing the complex nonlinear dynamics of the basal state from the less complex, linear dynamics that arise in response to a given stimulus. Furthermore, they suggest the need for further research to determine whether complex fractal-like structures and processes have evolved in biologic systems to permit an efficient adaptive response that can rapidly restore the organism to equilibrium during the exigencies of everyday life.
|
Alterations in Aging and Disease: The Pathway to Frailty |
|---|
|
TopAbstractDynamics of Rest and...Alterations in Aging and...Restoring Healthy DynamicsReferences |
|---|
For this discussion, it is important to recognize the difference between variability and complexity. As shown in Fig. 5, a highly variable signal, such as a high-amplitude sine wave, is minimally complex and can be described mathematically by a simple periodic function. In contrast, a highly complex signal may have marked irregularity, but less variability (lower amplitude), requiring an aperiodic, nonlinear function to describe it. Therefore, loss of amplitude (variability) of a periodic process such as melatonin (25) or thyroid-stimulating hormone secretion (26) that occurs with aging should not be confused with a loss of complexity. Furthermore, aging can be associated with a decrease in complexity, but greater variability in physiologic systems. The concept of reactive tuning suggests that a loss of complexity during resting conditions impedes an individual's ability to mount a focused response during stress. This would result in greater variability of the response. In addition, complexity loss within elderly individuals could result in increased variability of a given response between individuals. Thus, compared with healthy young individuals, healthy elderly people may look more alike under steady-state conditions, but less alike when challenged by internal or external perturbations.
|
|
|
|
|
Restoring Healthy Dynamics |
|---|
|
TopAbstractDynamics of Rest and...Alterations in Aging and...Restoring Healthy DynamicsReferences |
|---|
Single Interventions
One of the most widely studied single interventions to improve functional ability is exercise. Aerobic exercise training for 6–9 months has been shown to restore heart rate variability in older adults (38)(39)(40), while a 6-month home-based program of strength and balance training reduced the fluctuations in stride time dynamics associated with gait instability (41). These studies reported measures of variability, rather than complexity, leaving some uncertainty as to whether exercise interventions indeed restore the complex fractal-like dynamics that characterize healthy physiologic functions. During acute treadmill exercise, the short-term correlation properties of heart rate dynamics are increased (42), indicating they become less complex, as would be expected if reactive tuning were operative. It is not yet known whether chronic exercise training has an effect on resting heart rate dynamics.
Using the slope of the log-transformed cardiac interval power spectrum (1/f slope) as well as a technique called detrended fluctuation analysis (32) that quantifies the short-term (4–11 beats) correlation properties of the cardiac interbeat interval data, Lin and colleagues (43) showed that the beta-adrenergic blocker atenolol increased fractal-like heart rate complexity in 10 patients with advanced congestive heart failure over a 1- and 3-month treatment period. Unfortunately, the study was not placebo controlled. Additional research is needed to confirm this result and determine whether an increase in complexity is associated with physiologic and functional improvements in such patients.
Estrogen replacement therapy is another single intervention with wide-ranging effects on urogenital health, bone turnover, cardiovascular physiology, and neurocognitive function. In addition to its well-known effects on improving the structural integrity of osteoporotic bone, and thus reducing fractures, we have recently shown it to improve integrated baroreflex regulation of vascular sympathetic outflow when given to postmenopausal women for 6 months (44). The functional implications of this result are not yet known.
Multifactorial Interventions
The traditional approach to human biology and medicine is reductionistic, organizing bodies of knowledge into individual organ systems (e.g., cardiology, nephrology, pulmonology, etc.) and explaining health and disease as the absence or presence of specific abnormalities in these organs. However, because the human body is a complex system, its function in health and disease cannot be fully explained by an understanding of its component parts. In recent years it has become evident that most pathologic conditions are systemic illnesses, affecting a multitude of organ systems and functional processes. Treatments that have an impact on multiple components of an illness have been found to be more effective than standard unidimensional approaches. For example, multifactorial interventions have been shown to be effective treatments for common geriatric syndromes such as falls (45) or delirium (46), which are often due to interacting abnormalities in multiple organ systems. The recognition that congestive heart failure is a systemic disease characterized not only by a reduction in cardiac inotropy but also by increased sympathetic nervous system activity, activation of the renin–angiotensin system, and skeletal muscle dysfunction has led to more effective treatments with combinations of beta-blockers, angiotensin-converting enzyme inhibitors, and exercise training. Similarly, refractory depression is now being managed with combinations of antidepressants that affect multiple neurotransmitters, and diabetes is being effectively treated with pharmacologic agents to improve insulin action as well as its release. Although to my knowledge the effect of multifactorial interventions on the dynamics of specific physiologic systems has yet to be studied, these interventions apparently have favorable effects on the overall integrity and integration of multiple organs and processes that are necessary for healthy physical or cognitive function. As is characteristic of nonlinear systems, the whole is greater than the sum of its parts.
It is clear that future genomic therapies for disease and disability will also require nonreductionistic, multifactorial approaches. Although the correction of a single genetic defect will be effective for certain conditions, the majority of chronic diseases will require polygenic interventions that correct not only individual gene products but also the spatial and temporal characteristics that enable these products to promote effective function of an entire organism.
External Dynamic Control Techniques
Insight into the role of nonlinear dynamics in healthy physiology is leading to a variety of new approaches for the treatment of disease that directly influence system dynamics. Examples include the following: ameliorating deconditioning and muscle weakness by administering hormones in pulsatile fashion, improving insulin action and glucose utilization through oscillating insulin infusions (27), and increasing blood oxygenation during artificial ventilation by varying end-expiratory pressure with the addition of noise (47).
One area in which dynamic control processes are likely to make important therapeutic contributions is in the prevention and termination of cardiac arrhythmias. Garfinkel and colleagues used nonlinear dynamics to stabilize drug-induced irregular cardiac activity in a section of tissue from the interventricular septum of the rabbit heart (48). By plotting one interbeat interval versus the previous interbeat interval of the normal cardiac rhythm, one can create a "first-return map" that displays recurrent patterns of points representing an underlying periodic fixed point. Using pacing procedures to alter the interbeat intervals, or pharmacologic approaches to change atrioventricular nodal conduction, one can perturb the aperiodic activity of a cardiac arrhythmia and force it into a stable rhythm (49). In the future, algorithms could be built into external pacing devices to terminate arrhythmias by exploiting their underlying nonlinear dynamics.
|
Acknowledgments |
|---|
Received October 1, 2001
Accepted November 12, 2001
|
References |
|---|
|
TopAbstractDynamics of Rest and...Alterations in Aging and...Restoring Healthy DynamicsReferences |
|---|
This article has been cited by other articles:
|
|
 |
|
  L. Ferrucci The Baltimore Longitudinal Study of Aging (BLSA): A 50-Year-Long Journey and Plans for the Future J. Gerontol. A Biol. Sci. Med. Sci., December 1, 2008; 63(12): 1416 - 1419. [Full Text] [PDF]
|
|
|
|
 |
|
 J. J. Sosnoff and K. M. Newell Age-Related Loss of Adaptability to Fast Time Scales in Motor Variability J. Gerontol. B. Psychol. Sci. Soc. Sci., November 1, 2008; 63(6): P344 - P352. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. S. Heffernan, J. J. Sosnoff, C. A. Fahs, K. K. Shinsako, S. Y. Jae, and B. Fernhall Fractal scaling properties of heart rate dynamics following resistance exercise training J Appl Physiol, July 1, 2008; 105(1): 109 - 113. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. Desquilbet, L. P. Jacobson, L. P. Fried, J. P. Phair, B. D. Jamieson, M. Holloway, J. B. Margolick, and for the Multicenter AIDS Cohort Study HIV-1 Infection Is Associated With an Earlier Occurrence of a Phenotype Related to Frailty J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2007; 62(11): 1279 - 1286. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Novak, K. Hu, M. Vyas, and L. A. Lipsitz Cardiolocomotor Coupling in Young and Elderly People J. Gerontol. A Biol. Sci. Med. Sci., January 1, 2007; 62(1): 86 - 92. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. P. Fried, E. C. Hadley, J. D. Walston, A. B. Newman, J. M. Guralnik, S. Studenski, T. B. Harris, W. B. Ershler, and L. Ferrucci From Bedside to Bench: Research Agenda for Frailty Sci. Aging Knowl. Environ., August 3, 2005; 2005(31): pe24 - pe24. [Abstract] [Full Text]
|
|
|
|
 |
|
 M. M. Sturman, D. E. Vaillancourt, and D. M. Corcos Effects of Aging on the Regularity of Physiological Tremor J Neurophysiol, June 1, 2005; 93(6): 3064 - 3074. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. A. Lipsitz Physiological Complexity, Aging, and the Path to Frailty Sci. Aging Knowl. Environ., April 21, 2004; 2004(16): pe16 - pe16. [Abstract] [Full Text]
|
|
|
|
|
|
J. Walston Frailty--The Search For Underlying Causes Sci. Aging Knowl. Environ., January 28, 2004; 2004(4): pe4 - 4. [Abstract] [Full Text]
|
|
|
|
|
|
J. J. Sosnoff, D. E. Vaillancourt, and K. M. Newell Aging and Rhythmical Force Output: Loss of Adaptive Control of Multiple Neural Oscillators J Neurophysiol, January 1, 2004; 91(1): 172 - 181. [Abstract] [Full Text]
|
|
|
|
|
|
W. A. Banks and J. E. Morley Memories Are Made of This: Recent Advances in Understanding Cognitive Impairments and Dementia J. Gerontol. A Biol. Sci. Med. Sci., April 1, 2003; 58(4): M314 - 321. [Full Text] [PDF]
|
|
|
|
|
|
S. H. Tariq MD KNOWLEDGE ABOUT LOW TESTOSTERONE IN OLDER MEN J. Gerontol. A Biol. Sci. Med. Sci., April 1, 2003; 58(4): M382 - 383. [Full Text] [PDF]
|
|
|
|
|
|
H. T. Blumenthal The Aging-Disease Dichotomy: True or False? J. Gerontol. A Biol. Sci. Med. Sci., February 1, 2003; 58(2): M138 - 145. [Full Text] [PDF]
|
|
|
|
|
|
J. E. Morley, H. M. Perry III, and D. K. Miller Editorial: Something About Frailty J. Gerontol. A Biol. Sci. Med. Sci., November 1, 2002; 57(11): M698 - 704. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||
