Synonyms

Neuromuscular system; motor processes; physiological declines; loss of complexity with aging

Definition

The declines observed across numerous motor functions that develop as a consequence of the natural process of aging can be broadly viewed within the context of a general slowing of various physiological processes.

The processes of aging tend to progressively degrade the human motor system and reduce the ability of even healthy elderly individuals to move and perform skillfully in the tasks of everyday life (Spirduso 1985). At the behavioral level of analysis, these detrimental effects of aging are most typically manifest in the decrement of performance outcome, a change in indices of movement variability and/or a loss in the efficiency of movement. There is also an age-related change in the complexity of output of the human physiological and behavioral systems (Lipsitz 2002; Lipsitz and Goldberger 1992; Vaillancourt and Newell 2002).

A pervasive phenomenological change with aging that has been the subject of considerable experimental investigation is the progressive slowing of the neuromotor system (Cousins et al. 1998; Moehle and Long 1989; Deary et al. 2010). The slowing of movement-related properties is found across all levels of analysis of the system (behavioral, neural, chemical, thermal) and from dynamic analyses of the organization of subsystems in the control of movement behavior, such as the muscle, spinal cord, and brain. The slowing of the aging motor system is also apparent in a range of behavioral movement-related properties. Indices of slowing in aging have been reported across the spectrum of movement actions, from eye movements to fine motor skills involving a small number of joints/muscles such as in laboratory reaction time-movement time and finger-tapping tasks and more complex gross motor activities such as walking that involves the coordination and control of multiple elements of the skeletal-muscular system. Figure 1 provides a schematic overview of the predominant qualitative changes in movement tasks and properties of motor control with aging.

Aging and Slowing of the Neuromotor System, Fig. 1
figure 113figure 113

Schematic illustration of the various changes in motor outputs with aging. The general characteristics of the changes within each movement are also outlined

Full size image

The slowing of the motor system in aging leads to a loss of functional capacity, adaptability, and, in the ultimate expression, death (Birren and Fisher 1991). The changes in the timescales of motor output and the multiple processes that support it provide an interdisciplinary window into the motor control of aging. There has been an increasing interest in system’s frameworks of analyses to the timescales of change in aging given that it is difficult to ascribe a causal relation (as in the reductionist agenda) of one particular process to the performance decrement with aging.

Discrete Movement Tasks

Reaction Time. Assessment of reaction time (RT) in discrete movement tasks has been a common approach to determining the impact aging has on cognitive and neuromotor processes (Bunce et al. 2004; Graveson et al. 2015; Hultsch et al. 2002; Spirduso 1980; Spirduso et al. 1988; Williams et al. 2005; Welford 1988). Reaction time measures the latency from the presentation of a “go” signal to the onset of the appropriate movement response (Spirduso 1980, 1985). Typically, investigators assess the response latency of the subject under either simple reaction time (SRT) or choice reaction time (CRT) conditions. SRT consists of a single stimulus to begin the action that is paired to a single possible response. For CRT conditions, more than one stimuli is available to be presented, with each stimuli requiring a different response (e.g., the varying stimuli may relate to performing the action with a different effector or, when performing the action, moving to different targets). The time taken to perform the movement component of the task is referred to as movement time (MT). To minimize any anticipatory actions, the foreperiod (i.e., time from a “ready” signal to the presentation of the “go” signal) is usually varied so the individual cannot predict when the stimulus to start the movement response is provided (Welford 1971, 1988).

Typically, reaction time tends to progressively increase (i.e., individuals get slower in the latency of their responses) from their mid-20 years until the individual passes 70 years of age (Welford 1988). Naturally, the changes in age-related RT are more marked when the task response is more difficult, as where individuals are required to respond under CRT situations (Bunce et al. 2004; Spirduso 1980; Williams et al. 2005). The use of EMG to fractionate RT has shown that the majority of the simple and choice RT age effect is in the pre-motor phase – that is, the time from the onset of the stimulus to begin the action to the initiation of muscle activity (Clarkson 1978). Interestingly, the effects of age on both RT and the time it takes to perform the desired movement (MT) are not the same across genders, with males generally exhibiting faster (shorter latency) responses compared to females. However, both males and females are similarly influenced by condition effects such as complexity of stimulus context (Der and Deary 2006). An example of the typical pattern of change in simple RT with age is shown in Fig. 2.

Aging and Slowing of the Neuromotor System, Fig. 2
figure 114figure 114

Age-related differences in simple reaction time responses. Data were attained from 75 healthy adults ranging in age from 30 to 80 years of age. The reaction time task involved depressing a computer button with the index finger in response to a light stimulus. Three individual trial responses from each person are shown in this figure

Full size image

The resultant age-related slowing of motor responses under the various RT situations is not simply reflected by changes in the average latency of the person’s response. Increases in intraindividual variability (i.e., trial-to-trial variation in RT performance) are also a function of healthy aging (Graveson et al. 2015; Hultsch et al. 2002; Light and Spirduso 1990). Together, the age-related slowing of RT and increased trial-to-trial RT variability has been linked with a general decline in cognitive functioning, including attentional and/or executive control mechanisms (Bunce et al. 2004; Deary et al. 2010).

Interestingly, the reports of increased intertrial variability with aging are consistent with the general view that the process of aging or the emergence of age-related diseases is reflected by changes in the variability and/or complexity of a given physiological process (Lipsitz 2002; Lipsitz and Goldberger 1992; Vaillancourt and Newell 2002). Indeed, from this perspective, a diverse range of studies have shown how the complexity/variability of such diverse physiological time series such as brain activity (i.e., EEG), neuromuscular function, respiratory and cardiovascular responses, balance, walking ability, physiological/pathological tremor, and hormone secretion is systematically affected by increasing age in adulthood (Hausdorff et al. 2005; Newell et al. 2006; Peng et al. 1995; Pincus 1994).

Movement Time. Within the context of the reaction time discrete movement paradigm, MT captures the time from initiation of the selected response to the termination of the movement (Schmidt and Lee 2011; Spirduso 1985). Similar to the results reported for RT changes with aging, older adults tend to exhibit a slowing of MT (Sleimen-Malkoun et al. 2013a; Temprado et al. 2013; Birren and Fisher 1991; Heitz and Schall 2012; Ketcham and Stelmach 2004). Based upon an understanding of the age-related changes in various physiological processes, a number of different explanations such as neural noise theory and the general slowing hypothesis were developed to explain the slowing of MT with aging (Schmidt and Lee 2011; Spirduso 1985). Aging individuals tend to follow the speed-accuracy relation described by Fitts’ (1954) law in aiming tasks, but the effect of task difficulty (amplitude increase and/or target size decrease) tends to slow the movement more than in young adults (Fitts 1954; Forstmann et al. 2011; Smith and Brewer 1995; Ketcham and Stelmach 2004; Sleimen-Malkoun et al. 2013b).

Continuous Movement Tasks

Aging-related effects of the slowing of the neuromotor system have also been studied in sequential and continuous movement tasks. The aging-related slowing of the motor system is observed in both the preferred rhythm and the maximal frequency (or minimal duration) of motor output for a given task.

Finger-tapping. The pattern and frequency of finger-tapping has been widely used to assess how aging or neurological disease impacts on central nervous system function (Aoki and Fukuoka 2010; Arunachalam et al. 2005; Cousins et al. 1998; Moehle and Long 1989). Consistent with the general trend of the observed slowing of movement responses, several studies have reported a decline in finger-tapping speed (i.e., declines in overall rate and longer inter-tap intervals) and increased variability of tapping responses in both healthy older adults and persons with neurological disorders such as Parkinson’s disease and Alzheimer’s disease and where damage to the cerebrum leads to declines in cognitive function (Cousins et al. 1998; Shimoyama et al. 1990). The basis for this decline appears to be embedded within neuromotor changes rather than being attributed to deficits in peripheral sensory function or force-producing capacity of the muscles involved in the task (Aoki and Fukuoka 2010).

Isometric Force Control. When grasping an object with the hand, there is a requirement to produce a certain level of (isometric) force in order to hold and manipulate the object (Flanagan et al. 1999). In performing these tasks, the resultant force profile is characterized by a series of small fluctuations or oscillations referred to as force steadiness or isometric force tremor (Enoka 1997; Christou and Carlton 2001). Consistent with the pattern of findings for other activities, older adults often exhibit reduced control in force production, as quantified by an increase in the amplitude of these fluctuations (Kinoshita and Francis 1996; Lazarus and Haynes 1997). Interestingly, this decrement in force-producing capacity has been interpreted to reflect changes in motor unit (MU) control and sensorimotor function and not simply in terms of muscle strength. The consequence of these changes is that elderly adults exhibit greater targeting error and isometric force variability.

Gait. Walking performance is another movement activity where declines are observed with increasing age in adulthood. The preferred walking speed of healthy older adults (i.e., over 60 years) tends to be significantly slower than healthy adults in their 20s, and walking speed continues to decline as the person ages further (Murray et al. 1969). One reason for this decrease in speed appears to be that older persons take a shorter step length in preference to altering (i.e., decreasing) step time (Himann et al. 1988; Winter et al. 1990; Owings and Grabiner 2004). Further adaptions utilized by older adults include increasing the proportion of time spent in double stance (i.e., both feet in contact with the surface of support), taking wider steps, and reducing the proportion of time spent in the swing phase during locomotion (Murray et al. 1969; Winter et al. 1990). The goal of these adaptations would appear to ensure an optimal level of dynamic balance during locomotion and prevent falls (Maki 1997). Figure 3 illustrates the general pattern of change in gait speed as a function of the normal process of aging.

Aging and Slowing of the Neuromotor System, Fig. 3
figure 115figure 115

Pattern of changes in individual walking speed as a function of increasing age. Gait speed data were attained while healthy individuals (n = 75) walked at their preferred speed on a 20 f. GAITRite pressure-sensitive walking surface. Three individual trial responses from each person are shown in this figure

Full size image

However, these decrements in walking performance are not the singular product of chronological aging per se. Several studies have found no gait differences between healthy young and elderly adults when the older individuals have been screened for potential physical impairments (Grabiner et al. 2001; Owings and Grabiner 2004). These findings indicate that factors other than chronological age such as fear of falling, decline in cognitive processing speed, decreased leg strength, and/or reduced leg range of motion are also likely to contribute to the slower walking speeds observed in the average elderly individual (Maki 1997; Kang and Dingwell 2008). These associations of potential causal factors provide indirect evidence that there are multiple contributions to the slowing of neuromotor system with aging.

Generalization of Intraindividual Movement Slowing Across Tasks. The findings on the behavioral slowing of the aging movement system in different motor tasks have typically been reported in isolation. This experimental design does not afford an examination of the generalization of intraindividual movement slowing that is assumed to hold in theories of aging. The limited studies on intraindividual generalization have reported modest correlations of movement slowing over tasks with the effects stronger as aging advances (Bielak et al. 2010; Dykiert et al. 2012).

Physiological Basis for Slowing

Physiological Function and Structure. While there seems to be little dispute regarding the general slowing of behavioral responses with aging and age-related diseases, the basis for such changes cannot be linked to any single defining factor. For example, the slowing of RT, increased trial-to-trial RT variability, and decreases in tapping speed have primarily been tied to a generalized decline in cognitive function. In contrast, changes in tremor and force production have been attributed to decline in neuromuscular function, particularly with respect to changes in MU capabilities. Many of the reported age-related declines have been linked to structural changes within the CNS itself that can include decreases in overall conduction velocity, age-related losses of white matter and gray matter, and degeneration of neurotransmitter systems (Zimmerman et al. 2006; Soares et al. 2014; Wang and Young 2014; Seidler et al. 2010). A consequence of these structural changes within the CNS is that, when performing the same motor task, older adults demonstrate increased activity across a wider network of motor areas within the brain (including the regions of the prefrontal cortex and basal ganglia) compared to younger adults (Seidler et al. 2010; Riecker et al. 2006; Ward 2006).

In addition to the central changes in function, there are a number of peripheral physiological changes that may impact on the ability of the older adult to respond quickly and appropriately. Central to these changes is the general decline in skeletal muscle function that leads to an overall decrease in muscle cross-sectional area, a reduction in muscle mass, and a decline in strength. Specific structural and functional neuromuscular changes that can arise with aging include increases in the variability of MU firing, atrophy of fast-twitch MUs, remodeling of MUs, and a decline in the number of alpha motor neurons (Erim et al. 1999). These peripheral changes have been linked to the slowing of gait responses, declines in isometric force control, and the altered dynamical structure of physiological tremor (Himann et al. 1988; Kang and Dingwell 2007; Morrison and Sosnoff 2009; Enoka 1997).

As an example, changes (slowing) in physiological tremor dynamics have been widely reported in healthy older individuals. The primary mechanism for this change has been some compromise in the neural output – the result of a general decline in the functional capacity of the aging system (Elble 1998; Morrison et al. 2006; Raethjen et al. 2000). Physiological tremor is an intrinsic property of a normal functioning nervous system which reflects the aggregated contribution from the mechanical resonant properties of the limb segment, cardiac mechanics, central neural mechanisms, and more peripheral neural contributions from stretch reflexes (Elble and Koller 1990; McAuley et al. 1997). The oscillations of the central neural component of physiological tremor are typically within the 8–12 Hz range and represent output from neural oscillatory structures including the basal ganglia, thalamus, inferior olive, and alpha motor neurons within the spinal cord (Elble 2000; McAuley et al. 1997). For older adults, changes in this intrinsic, involuntary motor output are reflected by increases in overall tremor amplitude and/or a decrease in frequency, with the tremor responses being observed at the lower range of the 8–12 Hz bandwidth (Elble 1998; Morrison and Sosnoff 2009; Raethjen et al. 2000). Thus, the general slowing of motor responses is also reflected by a loss of the fast timescale processes inherent in physiological tremor of postural control.

While the predominant view is that the slowing of movement responses is primarily driven by declines in physiological processes, an alternate (but related) consideration relates to the possibility that older adults select different strategies when performing movement tasks compared to younger adults. Under RT conditions, for example, there is evidence to indicate the older individual is often more careful and cautious in their selection of when to respond – in effect trading speed of movement for accuracy of performance (Bunce et al. 2004; Hultsch et al. 2002; Light and Spirduso 1990; Williams et al. 2005; Spirduso 1985). Thus, in comparison to younger adults, older persons may prioritize minimizing performance errors over moving faster, and so the observed slowing of responses may actually reflect that they occupy a different criterion position on the speed-accuracy continuum (Spirduso 1985; Welford 1988). Figure 4 illustrates this pattern whereby older adults may operate on a different point and/or curve with regard to the relation between speed of response and target accuracy (Salthouse 1979). However, this is not to say that this trade-off is voluntarily driven and occurs independent of any age-related changes in the underlying neurological structures. For example, studies have reported that the adoption of a more cautious selection strategy in older adults could also be reflective of alteration in the activation pattern and/or impaired neural connectivity between such regions as the supplementary motor areas and striatum (Bogacz et al. 2010; Forstmann et al. 2011; Heitz and Schall 2012). Supporting this view, van Dyck and colleagues (2008) reported that declines in dopaminergic function within the basal ganglia can be linked to the progressive slowing of RT in older adults (van Dyck et al. 2008).

Aging and Slowing of the Neuromotor System, Fig. 4
figure 116figure 116

Illustration of the potential differences in the speed-accuracy trade-off as a function of increasing age. The increased response time for older adults in comparison to young adults may reflect that they operate on a different point and/or curve. For example, to achieve a similar degree of accuracy as the young adults, older persons may operate on a different curve (3), or if they operate on the same curve (2), they would trade-off accuracy for speed (Adapted from Spirduso 1985)

Full size image

The consequences of the declines across various physiological systems are not simply restricted to performance within the context of a single task. One of the major health concerns for older adults is the likelihood of suffering a fall (Tideiksaar 1998; Tinetti et al. 1988). The trend of a slowing of responses, including decreased strength, slowing of reactions, walking slower, loss of physiological variability, impaired balance, changes in visual and/or sensory function, and declines in cognitive functioning all are factors that are linked with (and contribute to) increased falls risk for older individuals (Close et al. 2005). There is little doubt that the combined gradual slowing of responses across a range of physiological and behavioral outputs are driving factors underlying the increased occurrence of falls in older adults. However, the consequences are not simply limited to the immediate outcomes of suffering a fall (e.g., injury, death), as the long-term effects can be just as problematic. Indeed, a previous fall can be the precursor for a downward cascade of decline, as many people become less physically active, which can lead to further losses of muscle strength, adopting a slower, more cautious walking pattern, and exhibit increased tiredness following a fall. All these outcomes can ultimately lead to a further increased risk of falling and are viewed as markers for the descent into physical frailty (Fried et al. 2001).

Aging and the Adaptation of Multiple Timescales

The preceding sections show the pervasive and well-established examples of behavioral and physiological slowing of the neuromotor system as a function of aging. For the majority, these examples rest on the traditional distributional analyses of temporal components of behavioral responses and activities that are driven by the mean and standard deviation of the dependent variable in question (e.g., RT, MT, finger-tapping rates, average gait speed, and cadence). This standard approach, however, takes out the roles of time- and frequency-dependent structure in a time series of behavioral output in spite of being concerned about the role of time in aging and, more generally, developmental processes.

Since the early 1990s, there has been a concerted effort to introduce a new view to understanding the problems of aging that is formulated around the general umbrella theoretical constructs of self-organization and the emergent complexity (Lipsitz and Goldberger 1992; Vaillancourt and Newell 2002). The construct of self-organization in behavior is tied to the emergent dynamics, their change over the life span, and the contribution of different timescales to this process. The timescales provide a window into the role of different processes in a systems framework to movement behavior and its change over time. In this view, a timescale is not merely the duration of an event as in the typical psychological framework, but is an interval that arises from the intrinsic dynamics of the system (growth-decay and/or oscillatory processes).

This approach incorporates the use of nonlinear dynamics, frequency analysis, and time series analysis to provide additional insight as to process of aging and/or disease. While distributional analysis of variables through a mean and SD is still useful, it does not directly address the time- and frequency-dependent properties of physiological and behavioral data that are often more sensitive to age-related change. Moreover, Gilden et al. (1995) showed that even the pattern of RT responses, rather than exhibiting a normal distribution, was more appropriately characterized by complex frequency and nonlinear tendencies (the pattern was referred to as an example of 1/f noise). Subsequently, there have been many developments around this dynamic theme to traditional human performance variables in different experimental paradigms.

Hausdorff and colleagues (1997) studied the increase in the degree of variability of the gait pattern in aging adults and disease states such as Huntington’s and Parkinson’s disease (Hausdorff et al. 1997). They used spectral analysis and detrended fluctuation analysis to reveal the structure in the gait cycle beyond mean and SD of stride length (see Fig. 5). Their central finding was that the variability of the gait cycle exhibits properties of a self-similar system. That is, fluctuations of the gait cycle exhibit long-range correlations such that the stride properties of any given cycle are dependent on a cycle previously at rather remote times, perhaps hundreds of cycles earlier in the locomotion sequence. Their analyses showed, as others have since, that the dependence of the stride interval decays as function of a power law, suggesting a fractal pattern to the structure of the variability of the gait cycle over time. An important consequence of this finding is that it shows that the variability of the gait cycle is not that of a signal plus noise process, as has been viewed traditionally in studies of gait variability and assumed more generally in age-related performance decrements. Rather, this result highlights that there is an inherent dynamic structure to the variability of movement patterns and that deviations from the typical pattern of complexity may reveal insights as to the impact aging and disease on the selected motor output.

Aging and Slowing of the Neuromotor System, Fig. 5
figure 117figure 117

Differences in the pattern of stride interval (top, a) during walking in healthy young and older adults. Plots of the resultant differences in signal complexity (bottom, b) using detrended fluctuation analysis are also shown (Figure adapted from Hausdorff et al. 1997)

Full size image

Sosnoff and Newell (2008) examined the age-related loss of adaptability to fast timescales in the motor variability of isometric force production (Sosnoff and Newell 2008). The sensorimotor outputs to differing time and frequency properties (1/f noise structures) of target-force waveforms were studied. By having force-tracking pathways that followed different fractal noise structures, the manipulation of timescales in the task demands could directly be accomplished (i.e., changing the relative contribution of long- and short-frequency processes). The results showed that, when compared to younger adults, the older persons were progressively less able to approximate the lighter-color noise force targets and utilize information in the higher frequencies of the target signal.

The findings of Sosnoff and Newell (2008) are consistent with aging and the loss of complexity hypothesis of Lipsitz and Goldberger (1992), given that there was a declining ability with aging to use the faster timescales of sensorimotor control in force output. However, several studies have now shown that the particular directional effect of the loss or gain of complexity of force is moderated by the differential impact of task demands (Vaillancourt and Newell 2002). This is consistent with the general view that behavior is an emergent property of the confluence of constraints of the individual, environment, and task and that complexity is an emergent feature of the interaction of the three classes of constraint and not a property that should be viewed as within the body.

Concluding Comments

An inevitable consequence of the aging process is that behavioral movement speed slows across all movement domains. Although the lifestyle of the older individual in terms of health status and exercise habits can slow this decline in movement speed to some degree (Spirduso 1980, 1985), the degradation across all levels of the central-peripheral nervous system means that slowing of movement responses in the aging adult is pervasive and has many specific manifestations, only the major ones of which are addressed here. It follows then that all theories of aging, whether psychological, physiological, biological, or more general systems accounts, have tended to address this important phenomenon (Spirduso 1985, 2005).

The challenge is that correlates of the slowing of movement speed with aging can be found at many theoretical levels of analysis thereby lending support to the veracity and relevance of all theories of aging although no single unified theory has adequately captured the full scope of the declines seen across the various movement domains. Spirduso (1985, 2005) has proposed that the most compelling hypotheses to explain the age-related behavioral slowing are to be found in the various manifestations of biological deterioration that induce the slowing of movement in action seen in the aging adult. The system’s approach to the loss of complexity (Lipsitz and Goldberger 1992) provides a complementary framework for investigating the array of age-related changes found for physiological and behavioral processes. Contemporary research on movement and aging is still focused on hypotheses of network signal and connectivity issues in the aging neuromotor system.

Cross-References