This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation

Effects of Aging on Automatic and Effortful Processes in Bimanual Coordination

^a McMaster University, Ontario, Canada.

Laurie R. Wishart, School of Rehabilitation Science, McMaster University, 1280 Main St. West, Hamilton, Ontario, L8S 4K1, Canada E-mail: wishartl{at}fhs.mcmaster.ca.

Decision Editor: Toni C. Antonucci, PhD

	Abstract

TOP Abstract Experiment 1 Experiment 2 General Discussion References

 
Two experiments are reported that compared younger and older adults on their performance of two bimanual temporal coordination tasks at varying movement speeds. In many cases, older adults performed as well as younger adults at all speeds of an in-phase coordination pattern and at slow speeds of an anti-phase pattern for both coordination accuracy and stability. Age differences tended to emerge most prominently at high speeds for the anti-phase pattern. These findings are consistent with the aging literature regarding automatic and effortful processing distinctions, suggesting that relative age differences become magnified when effortful resources are required for motor performance.

PROFICIENT motor functioning becomes an increasingly important lifestyle factor as human beings age because mobility, dexterity, and coordination affect one's capacity for independent living. One goal of research on aging effects in motor skills is to provide some understanding of the normal expectations about performing well-learned skills (motor control) and acquiring new skills (motor learning). The majority of the work to date in these areas has focussed on motor control issues, and a large portion of this work has examined tasks for which speed was the main criterion (e.g., responding to a stimulus or aiming at a target). Older adults are generally slower to respond to a stimulus than are younger adults, and this age difference is magnified if choices between stimuli and/or responses must be made (e.g., Fozard, Vercruyssen, Reynolds, Hancock, and Quilter 1994). Age differences also occur in how rapid actions are prepared prior to movement initiation (for a review, see Seidler and Stelmach 1995) and in the visual control of single-limb aiming movements (Chua, Pollock, Elliott, Swanson and Carnahan 1995; Darling, Cooke, and Brown 1989; Liao, Jagacinski, and Greenberg 1997; Lyons, Elliott, Swanson, and Chua 1996). An excellent overview of these and related findings is provided in Spirduso 1995.

A motor control issue that has received much less attention concerns the effects of aging in tasks that require movements that are not performed as fast as possible but instead emphasize movement coordination. Bimanual temporal coordination, in particular, serves as an excellent task for the study of aging and motor control for several reasons: (a) The task requires the control of multiple degrees of freedom, making it a relatively complex skill that shares a common basis with many activities of daily life; (b) the primary measure of temporal coordination, relative phase, provides insight regarding how inherent (natural) coordination patterns change with age; and (c) new coordination patterns can be learned, and the effects of aging on learning new motor tasks is an important lifestyle issue that has received little attention.

There have been few studies completed on bimanual coordination in the elderly population. In two studies involving spatial coordination (Contreras-Vidal, Teulings, and Stelmach 1998; Stelmach, Amrhein, and Goggin 1988), older adults showed considerable performance decrements in relation to younger adults. In contrast, using a temporal coordination task, Greene and Williams 1996 found aging effects that were quite specific to the nature of the performed coordination pattern—patterns that research with young adults has shown to have very distinct coordination characteristics. An in-phase temporal coordination task represents a pattern of activation in which the mirror image actions of the upper limbs are performed simultaneously. An anti-phase pattern represents an alternating activation of the mirror image movements, resulting in same-direction actions. Previous research has shown that young adults can perform these patterns accurately and consistently at low movement frequencies. However, as speed is increased, the anti-phase pattern loses stability and spontaneously switches to an in-phase pattern (see Kelso 1995; Schmidt and Lee 1999, for a review of this evidence for young adults). In young adults, the in-phase and anti-phase bimanual coordination patterns are inherent in the performer's repertoire—at comfortable speeds these patterns require no practice to be performed at a level that is very efficient, and further practice does not enhance performance. As movement speed increases, the in-phase pattern remains accurate and stable, whereas the anti-phase pattern loses stability, resulting in one of two options: (a) the anti-phase pattern switches to an in-phase pattern or (b) the individual consciously intervenes and, with considerable effort, maintains the anti-phase pattern. One interpretation of these results is that in-phase coordination represents a temporal patterning of muscle activations that is relatively automatic at all movement frequencies. In contrast, anti-phase coordination may be considered an automatic process at slow speeds but requires effortful processing at high movement frequencies (Baldissera, Cavallari, Marini, and Tassone 1991).

The importance of bimanual temporal coordination as a task to study aging effects is revealed when comparing the putative roles of automaticity and controlled processing in this motor task with the role of automatic and effortful processing as a theoretical construct in aging (Craik and Jacoby 1996; Hasher and Zacks 1979, Hasher and Zacks 1988). This perspective of aging suggests that behaviors which are "automatic" (or which involve relatively nonconscious or nonattentive processing) are retained and performed well by older individuals. In contrast, behaviors that require effortful processing (consciously controlled decision making) typically show rather dramatic declines as a function of aging. To our knowledge this perspective on aging processes has not been extended to the study of human motor behavior, although it does provide an excellent basis on which to make predictions about the performance of inherent coordination patterns as a function of age. For example, Greene and Williams 1996 found not only a similar tendency in both younger and older adults to maintain a stable in-phase pattern at high movement frequencies, but also that older adults tended to switch from the anti-phase to the in-phase pattern at a much lower frequency compared with younger adults. One interpretation of this evidence is that the performance of the relatively automatic (in-phase) pattern was maintained very well by participants of all ages, whereas the relatively more controlled (anti-phase) pattern showed rather dramatic age-related differences as movement frequency increased (i.e., as effortful demands on movement control were amplified).

These complementary automatic/effortful processing views of aging and bimanual temporal coordination, together with the initial findings of Greene and Williams 1996, form the basis for the two experiments that follow. Our purpose was to replicate and extend the findings of Greene and Williams, using a different coordination task and methodological variations to test the following predictions. The performance of the in-phase bimanual coordination pattern, representing an automatic process at all speeds, should show relatively few differences in performance accuracy or stability when younger and older adults are compared. The anti-phase pattern, which may be considered to be automatic at relatively slow speeds, should be relatively unaffected by age until intrusions on the capability to maintain the pattern begin (at higher speeds). At that point, the influence of conscious, effortful processing behaviors will determine the nature of the coordination pattern, and age-related differences are expected to occur.

	Experiment 1

TOP Abstract Experiment 1 Experiment 2 General Discussion References

 
In the first experiment, a group of younger adults was compared with two groups of older adults on the performance of the in-phase and anti-phase patterns at four speeds that were paced by an auditory metronome. A fifth speed was also examined at which participants were asked to perform the bimanual coordination patterns at a preferred movement frequency (i.e., the speed at which they felt most comfortable when performing the patterns).

Method
Participants
We examined groups of younger adults (), younger-old adults (), and older-old adults (). The younger adults were recruited from McMaster University. Older adults were recruited from a list of independent, community living volunteers comprised of respondents to local media advertisements. The older adults were reimbursed for parking or public transportation expenses. All participants read and signed consent forms prior to testing.

General measures of motor speed and cognitive functioning were collected to ensure that participants within the older age groups were representative of their age group populations. The following tests were performed: The Mini Mental State Exam (Folstein, Folstein, and McHugh 1975) was used as a screening tool for mental status (minimum acceptable score of 23), the Digit Span Test (WAIS-R; Wechsler 1981) was used as a measure of working memory, and the Functional Reach task (Duncan, Weiner, Chandler, and Studenski 1990; Weiner, Duncan, Chandler, and Studenski 1992) provided a general measure of motor function status. All participants scored within the age-expected norms on these tests.

Apparatus.
The task was to move two slide carriages with both upper limbs simultaneously and continuously for 15 s. The slide carriages were mounted on ball bearing casings and slid with minimal friction on top of two steel rods such that movement could occur parallel to the participant's chest. The entire apparatus was mounted securely on a desk. To move the apparatus, participants grasped wooden dowel handles that were attached upright in the middle of the sliding carriages. A movement of the slide carriage away from the center of the apparatus required a shoulder abduction/external rotation. To move the slide toward the center of the apparatus required a shoulder adduction/inward rotation. A height-adjustable chair was used so that the forearms could be positioned parallel to, but not resting upon, the table top with the elbows flexed at a 90-degree angle, while holding the dowels. The chair was positioned so that the body midline was centered between the two carriages.

Two 16 cm (6.3 in) regions were clearly marked on the base of the apparatus, one for each slide carriage, which defined the amplitude boundaries (the points at which the continuous, oscillatory movements were to be reversed). Each carriage could be moved a maximum horizontal distance of 23 cm (9 in; i.e, 7 cm [2.8 in] further than the goal amplitude). Linear potentiometers (Duncan Electronics 612R12KL.08) were attached to each of the slide carriages to encode displacements.

A customized configuration of the LabWindows software program (Version 2.2.1, National Instruments Corporation, Austin, TX) was run on an 80486 PC. This program controlled the temporal events within each trial (the initiation and termination of a trial, and the frequency of an auditory pacing signal) and recorded displacement data at 200 Hz. The auditory pacing signal was amplified by a tone generator (Lafayette Instrument Co. 58025).

Procedures.
All participants performed two bimanual timing patterns (in-phase and anti-phase) under five different movement frequency conditions (at metronome frequencies of 0.5, 1.0, 1.5, and 2.0 Hz and also at a preferred speed that was not paced by a metronome). The length of each trial was 15 s and the intertrial interval was about 7–10 s. The total duration required for participation in the experiment, including instructions and collection of demographic information, was approximately 60 min for the younger adults and 90 min for the older adults.

Both coordination patterns required simultaneous, continuous movement of the two slide carriages by the upper limbs. On each beat of the metronome both slide carriages were to have completed one complete cycle (e.g., from its position nearest the body midline to maximum lateral displacement and return). Performance of the in-phase coordination pattern required that the left and right arms move as mirror images—when the left slide carriage was at its closest point relative to the body midline so too was the right carriage, both limbs moving away from and toward the midline simultaneously. Performance of the anti-phase coordination pattern was like the actions of a car's front windshield wipers—when the left carriage was at the point nearest the midline, the right carriage was at its maximum outward point, with both limbs moving side to side, simultaneously, always in the same direction. These patterns were described both verbally and using written instructions (with illustrations).

The auditory tone was described to the participants as an aid to help pace the desired movement frequency. For the in-phase task, participants were told to attempt to have both slide carriages at the inner-most positions (near the body midline) coincident with each beat of the metronome. For the anti-phase task, participants were told to try to coincide the position of the slide carriages such that one limb was positioned near their body midline, while the other was at the maximum lateral point of the apparatus on the metronome beat. For the preferred trial speeds, participants were told that no metronome would be heard and that they should perform the movement patterns at a speed that was comfortable and which would allow for their best performance. In addition, participants were asked to try to move the slide carriages in a rhythmic, fluid manner without stopping, to maintain their pace with the beating of the metronome, and to stay with the coordination pattern as best as possible throughout the trial. None of the participants had difficulty in understanding and following these instructions.

The experimenter performed one demonstration of each coordination task. The participants were then given one practice trial at each of the 10 experimental combinations (2 Patterns x 5 Speeds). Following practice, the experimental trials were run in a standard order. The order was five in-phase trials (the first trial at the preferred speed, then one trial each at metronome frequencies 0.5, 1.0, 1.5, and 2.0 Hz). The same order was then followed for five anti-phase trials. This set of 10 trials was repeated two additional times, resulting in 30 experimental trials in total.

Data analyses.
The measure of coordination commonly used in this temporal coordination task is relative phase (). This measure captures the relative time at which one limb advances through its movement cycle in relation to the advancement of the other limb through its cycle. When both limbs' advancement coincide temporally, they are moving in-phase; when in asynchrony, the limbs are in anti-phase. To compute relative phase, the velocities and amplitudes of the right and left limbs were rescaled to the interval (-1, 1) for each cycle of oscillation. Relative phase angles were then computed using the following formula:

where is the relative phase between limbs at each sample, X is the position of the limb within a cycle rescaled to the interval (-1, 1) (dX/dt) refers to the normalized instantaneous velocity, and R and L, are the right and left limbs (Scholz and Kelso 1989, p. 129).

Discrete measures of relative phase were determined at the peak extension and peak flexion positions of the right limb (the reference limb) for each cycle. The mean of these relative phase angles over cycles provided a measure of average relative phase for a trial. In-phase coordination is characterized as a relative phase of 0°; anti-phase is a 180° relative phase. To compare the performance of these two patterns directly, we calculated the absolute difference between the observed mean and the goal relative phase (0° or 180°) for each trial (the absolute difference maintains within-participant bias but prevents between-participant differences, in opposite directions of bias, from nullifying each other). The resulting error score (absolute mean error) provided a measure of temporal coordination accuracy. The standard deviation of the individual measures of relative phase about the scores that comprised a trial mean provided a measure of consistency (or coordination stability). The observed movement frequency was also analysed.

Statistical analyses were performed using analyses of variance (ANOVAs). All ANOVAs were mixed (split-plot) designs, with age (younger, younger-old, older-old) being the between-group factor. Coordination pattern (in-phase and anti-phase) and movement speed (0.5, 1.0, 1.5, 2.0 Hz, or preferred) were the within-participant factors. Post hoc comparisons of means were performed using the Tukey HSD procedure. The probability level for achievement of statistical significance was .05. Although all significant effects are reported, our primary interest here is in the effects that involve age groups.

Results
Performance of participants under the different movement speed conditions for the in-phase and anti-phase patterns is illustrated in Fig. 1 and Fig. 2. Note that in these figures, trials that were paced by the metronome are represented by unfilled symbols and plotted according to the metronome condition. The preferred movement frequency trials are represented by filled symbols and plotted along the abscissa according to the observed frequency.

View larger version (19K): [in this window] [in a new window]   Figure 1. Absolute mean error performance in relative phase for the older-old, younger-old, and younger groups of participants as a function of metronome frequency and coordination pattern in Experiment 1. Metronome-paced trials are represented by unfilled symbols; preferred frequency trials are represented by filled symbols and plotted along the abscissa in correspondence with the observed frequency.

View larger version (19K): [in this window] [in a new window]   Figure 2. Within-participant standard deviation in relative phase for the older-old, younger-old, and younger groups of participants as a function of metronome frequency and coordination pattern in Experiment 1. Metronome-paced trials are represented by unfilled symbols; preferred frequency trials are represented by filled symbols and plotted along the abscissa in correspondence with the observed frequency.

Accuracy (absolute mean error).
Fig. 1 illustrates the performance accuracy data. In general, the two older groups were relatively close to the younger group for both movement patterns at the preferred and slow metronome frequencies. However, at higher speeds (1.5 and 2.0 Hz), the older groups' performance became less accurate. These observations were substantiated by a significant group main effect, , a Group X Speed interaction, . Subsequent post hoc tests on the interaction revealed no age differences at 0.5 and 1.0 Hz or at the preferred speed. At 1.5 Hz the younger adults were more accurate than the younger-old adults, who, in turn, were more accurate than the older-old adults. At 2.0 Hz, the younger adults were more accurate than both older groups, who were not different from each other.

In addition, the ANOVA revealed main effects for speed, , p < .001, and pattern, p < .001, as well as a Speed x Pattern interaction, , p < .001. These results replicate earlier findings with young adults.

Stability (standard deviation).
Fig. 2 (performance stability) illustrates that, compared with the anti-phase pattern, all age groups performed the in-phase pattern under all movement speeds with relatively equal stability. The performance of the anti-phase pattern showed marked differences amongst the age groups. The young participants were able to maintain stability across all movement speeds. Both groups of older adults lost stability for the anti-phase pattern as the metronome speed was increased from 1.0 to 1.5 Hz and from 1.5 to 2.0 Hz. Moreover, the older-old group's anti-phase performance became much more unstable than the younger-old group's performance at the faster metronome speeds. These observations were statistically verified by a group main effect, , p < .001, two-way interactions between group and speed, , p < .001, and between group and pattern, , p < .001, and by a three-way interaction between group, speed, and pattern, , p < .001. Post hoc tests on the three-way interaction revealed no age group differences at any of the movement speeds for the in-phase pattern. There were also no age group differences for the anti-phase pattern at the 0.5 Hz, 1.0 Hz, and preferred movement speed conditions. At the 1.5 Hz speed, the younger group was significantly more consistent than the younger-old group, who were more consistent than the older-old group. At the 2.0 Hz speed, the younger and younger-old groups were not different, but both were more consistent than the older-old group.

In addition to the above effects, there were also main effects for speed, , p < .001, and pattern, , p < .001, and a Speed x Pattern interaction, , p < .001. Again, these findings replicate previous findings for young adults.

Discussion
On the basis of an automatic/effortful processing distinction of aging effects (Craik and Jacoby 1996; Hasher and Zacks 1979, Hasher and Zacks 1988), we predicted that an in-phase (automatic) coordination pattern would be performed equally well by both older and younger individuals. Indeed, our results showed that the in-phase pattern was performed relatively well by both groups of older adults when compared with the younger group. Performance speed and coordination stability were equally well maintained across all metronome frequency levels for the in-phase task in all age groups. In contrast, the absolute mean error data revealed that the older adult groups sacrificed some performance accuracy for both the in-phase and anti-phase tasks at the higher movement frequencies. However, although the three-way interaction was not significant, it is clearly illustrated in Fig. 1 that this decreasing accuracy was more dramatic for the anti-phase pattern than for the in-phase pattern. Combined with the comparatively undiminished performance for stability and speed measures for the in-phase pattern across all age groups, the automatic/effortful processing distinction for aging effects in bimanual coordination fit the obtained results reasonably well.

Previous research suggests that the anti-phase task is also performed as an automatic process at slow speeds (Baldissera et al. 1991). In the present study, performance accuracy, stability, and movement frequency were maintained equally well for all age groups at the preferred frequency, and at 0.5 and 1.0 Hz. When the performance of the anti-phase task was stressed by higher metronome frequencies, however, there was evidence for some differences in the way that younger and older adults coordinated their actions. The younger adults were remarkably good at maintaining movement frequency, accuracy, and stability. The older adults tended to slow down (but not significantly), became biased away from the anti-phase pattern, and were more variable in their control of this coordination pattern. Moreover, this set of results was magnified more so for the older-old group of participants than for the younger-old group, thus amplifying the age-related differences with which effortful control was implemented to deal with the demands of the anti-phase pattern at high frequencies. These findings are also consistent with predictions of the automatic/effortful distinction (Craik and Jacoby 1996; Hasher and Zacks 1979, Hasher and Zacks 1988).

The automatic/effortful processing distinction also accounts well for the findings of Greene and Williams 1996. In their study, participants tended to switch from the anti-phase pattern to the in-phase pattern as movement frequency increased. Following the phase transition however, the older adults performed similarly to the younger adults, because both were now performing the (automatic) in-phase pattern. Pattern switching did not occur in the present study however, and differences between the younger and older adults were magnified because both age groups were (effortfully) attempting to remain in the anti-phase pattern. This difference in switching behavior exhibited in the two studies was very likely due to a difference in the instructional sets used. In the present study, participants were told to try to perform the same movement pattern throughout an entire trial. In contrast, participants in the Greene and Williams 1996 study were told to not intervene if they felt a pattern switch occurring—if a switch from anti-phase to in-phase began to occur, participants were told not to resist the changing coordination pattern (cf. Kelso 1995).

The influence of instructions on the dynamics of movement coordination is not a trivial issue and has been found to have profound effects on performance in young adults. Previous research has shown that, when told to remain in the instructed pattern at all times, young adults show increased instabilities at higher metronome frequencies but no evidence of switching from an anti-phase to an in-phase pattern (Lee, Blandin, and Proteau 1996). Thus, the often-observed, non-linear switch in coordination patterns may, at least in part, be due to the tendency to succumb to, or to resist, effortfully, the attraction of the in-phase pattern.

In terms of the aging influences in this automatic/effortful processing distinction, we offer the following interpretation. As the maintenance of an anti-phase pattern becomes stressed by an increasing movement frequency, participants in these studies are faced with two options. One option (Greene and Williams 1996) is to relax the effortful processing that must be engaged to maintain the anti-phase pattern. In this case, a more automated performance tendency or "habit" emerges as both younger and older individuals revert to an equivalent level of performance, now in an in-phase (automatic) pattern. The other option (as found in the present study) is for the participants to try to maintain the anti-phase coordination pattern. This occurs at the cost of a higher involvement in effortful processing and, thus, a diminished level of performance in the older adults.

The above interpretation, however, is contingent on the assumption that the differences between the findings of Greene and Williams 1996 and the present study were due to differences in instructional sets used in the two studies. A more convincing interpretation would be supported if both the Greene and Williams pattern of results and those of the present study could be replicated within the same experiment—a goal that represented the purpose of the following experiment.

	Experiment 2

TOP Abstract Experiment 1 Experiment 2 General Discussion References

 
The purpose of this experiment was to compare the effects of aging, under two different instructional sets, on the performance of in-phase and anti-phase coordination patterns at various metronome frequencies. The design of this study represents a comparison of the instructional methods used by Greene and Williams 1996 and the methods used in Experiment 1. The hypothesis was as follows: Under instructions to not intervene in a pattern switch at higher speeds, older adults and younger adults will show more similarities in coordination behavior than they will when the instructions ask the individuals to maintain the pattern at all times. The hypothesis is based on the premise that succumbing to the pull towards the automatic, in-phase pattern will be similar in older and younger adults, resulting in similar performance tendencies. In contrast, maintaining the anti-phase pattern at increased movement speeds will require enhanced effortful processing resources, which will reduce the coordination capabilities of the anti-phase pattern at higher speeds for the older compared with the younger adults.

Only one older-age group was used in the present study because there were only small differences between the younger-old and older-old groups in Experiment 1. Instructional set was manipulated as a between-groups factor, such that half of the participants in each of the age groups received instructions to "not intervene," compared with the other half who received instructions to "stay" with the pattern (as in Experiment 1). Also, to stress the coordination capabilities of the younger and older adults, we did not use the preferred and 0.5-Hz movement frequencies in this experiment and instead included a 2.5-Hz frequency. We ran another study using a within-participants design prior to the present investigation. In that study the younger participants could follow the instructions to either stay with the intended pattern or to not intervene in a pattern switch without difficulty. In contrast, the older adults experienced considerable carry-over effects when changing from one instructional set to the other instructional set. Thus, the experiment was repeated here using a between-participants design.

Method
Participants.
Sixteen young adults () and sixteen older adults (), who had not participated in Experiment 1 were recruited for this study. All details regarding the recruitment and screening of participants were identical to Experiment 1 except that the Functional Reach task was replaced by a more general measure of movement agility, the Get-Up-And-Go task (Mathias, Nayak, and Isaacs 1986; Podsiadlo and Richardson 1991). All participants scored within age-expected norms. Prior to testing, the participants within each age group were randomly assigned to one of two instruction subgroups, with the restriction that for each subgroup.

Apparatus and procedure.
The apparatus and experimental setup were identical to Experiment 1. Many of the basic experimental methods were similar as well. Movements were paced by an auditory metronome at frequencies of 1.0, 1.5, 2.0 and 2.5 Hz (the self-paced and 0.5-Hz conditions were not used; the 2.5-Hz condition was added to this experiment). Four trials per experimental condition were tested in an order that was similar to Experiment 1. All instructions regarding the description of the two coordination patterns were the same as in Experiment 1.

The major difference in this experiment was the inclusion of two instructional sets, tested as a between-participant component of the experimental design. One subgroup within each of the age groups was given the identical instructional set as used in Experiment 1. These Stay subgroups were instructed to stay with the intended movement coordination pattern throughout the entire trial. They were further instructed that if they were to "lose" the coordination pattern at any point in the trial, they should try to reacquire the pattern. In contrast, the Do-not-intervene subgroups were told that if they were to lose the intended movement pattern during a trial, they should abandon the intended pattern and perform the more comfortable pattern. These latter instructions are similar to those used in the aging study by Greene and Williams 1996 and in a number of studies conducted with younger adults (for a summary, see Lee et al. 1996).

Data analyses.
The analysis of relative phase as the measure of coordination and the determination of accuracy and stability measures were identical to Experiment 1. The ANOVA model was a four-way, Age x Instructional Set x Pattern x Speed Design, with repeated measures on the last two factors. Similar to Experiment 1, we have reported all significant effects but focus on the main effects and interactions that involve age groups.

As found in a study with young adults (Lee et al. 1996), the Do-not-intervene strategy often results in a bimodal distribution of relative phase measures for the anti-phase pattern within a trial. Thus, similar to our earlier work, we have provided a more detailed analysis of the anti-phase data. Frequency counts of the discrete measures of relative phase across the four trials of each anti-phase condition were partitioned into categories that were either near the anti-phase pattern (i.e., prior to a pattern switch) or near the in-phase pattern (following a pattern switch). We used a bandwidth of ± 30° relative phase to operationally define each category, which is roughly equivalent to twice the standard deviation under most of the anti-phase conditions examined in Experiment 1. Thus, the frequency counts summarized the proportion of relative phase measures between 150° and 210° relative phase (surrounding anti-phase) and between -30° and 30° relative phase (surrounding in-phase—after a pattern switch).

Results
Observed movement frequency.
These frequency data replicate the findings from Experiment 1 and do not appear to change as a function of instructional set. Reliability of this interpretation is supported both by a three-way interaction between age, pattern, and speed, , p < .001, and by the absence of a significant four-way interaction (which would have included instructional set), , p > .2. Post hoc comparisons of the triple interaction revealed that the only significant difference between the age groups in slowing of observed frequency was the older group's performance of the anti-phase pattern at 2.5 Hz (the speed that had not been tested in Experiment 1).

In addition to this three-way interaction, there were main effects for age, , p < .01, pattern, , p < .001, and speed, , p < .001. There were also two-way interactions between age and pattern, , p < .001, age and speed, , p < .001, and pattern and speed, , p < .001. There was no significant main effect or interactions involving instructional set.

Accuracy (absolute mean error).
Performance accuracy was affected by all of the factors in the experiment—although there were some unexpected results associated with age, resulting in a significant four-way interaction, , p < .001. Post hoc comparisons substantiated what is illustrated in Fig. 3—age effects occurred most noticeably at 2.0 and 2.5 Hz, for the anti-phase pattern, under the Do-not-intervene instructional set. However, this effect was due to the significantly less accurate performance of the younger group compared with the older participants. We will describe this effect in detail later.

View larger version (21K): [in this window] [in a new window]   Figure 3. Absolute mean error performance in relative phase for the older and younger groups of participants as a function of metronome frequency, coordination pattern, and instructional set in Experiment 2.

Stability (standard deviation).
The stability data are presented in Fig. 4. Similar to Experiment 1, the differences between age groups were relatively small. Although the older, Stay instructional group appeared to show less stability in the intermediate anti-phase speeds, these differences were not statistically reliable. Indeed, the ANOVA revealed no main effects or interactions involving age as a factor (all ps > .10). The Do-not-intervene set resulted in particularly unstable performances for the anti-phase pattern at the 2.0- and 2.5-Hz speeds, which was due to the increased incidents of pattern switches (see next section). This instructional set influence resulted in a significant main effect, , p < .01, a Set x Pattern interaction, , p < .001, a Set x Speed interaction, , p < .05, and a three-way interaction between set, pattern, and speed, , p < .01. The ANOVA also revealed a pattern main effect, , p < .001, a speed main effect, , p < .001, and a Pattern x Speed interaction, , p < .001.

View larger version (21K): [in this window] [in a new window]   Figure 4. Within-participant standard deviation in relative phase for the older and younger groups of participants as a function of metronome frequency, coordination pattern, and instructional set in Experiment 2.

The means of the data are summarized in Table 1 . Considering first only the data near 180°, the ANOVA revealed significant main effects for instructional set, , p < .001, and speed, , p < .001. A significant Age x Speed interaction, , p < .001, revealed that the older participants displayed significantly fewer relative phase measures near 180° at 1.5 Hz, but significantly more at 2.5 Hz than did the younger participants. Also, a significant Set x Speed interaction, , p < .001, revealed that the Stay instructional set produced more performances near 180° than did the Do-not-intervene set at all speeds except 1.0 Hz.

Discussion
Overall, the results from this study provided additional evidence for the role of automatic and effortful processing in aging effects on bimanual timing performance. Consistent with predictions, younger and older adults performed virtually identically on the in-phase pattern at all movement frequencies (including the 2.5-Hz speed, which had not been examined in Experiment #1). This finding replicates the variability and frequency data from Experiment 1 and also extends the finding to the accuracy measure as well. The relative absence of aging effects in bimanual coordination capabilities fits well within the characterization of in-phase coordination as an automatic skill that does not diminish in performance capability with age (Craik and Jacoby 1996; Hasher and Zacks 1979, Hasher and Zacks 1988).

In addition, the performance of the anti-phase pattern conformed to the expectations of both automatic and effortful processing. At slow speeds (1.0 Hz), both younger and older groups performed at similar skill levels. As the metronome frequency was increased (to 1.5 and 2.0 Hz), the older adults' performance became less accurate in both instructional sets and less consistent in the Stay group. These findings are also relatively consistent with expectations and the automatic/effortful processing framework.

The highest movement frequency conditions in this experiment produced interesting, though unexpected results. For the Stay groups, the younger participants became much more variable, and their stability equalled that of the older participants at this speed (Fig. 4). This finding did not conform to predictions, as we expected the increased effort required under the Stay set to increase, not decrease, the differences between age groups. However, recall that the increase in metronome frequency (especially at 2.5 Hz) for the older group's anti-phase performance was not matched well by their performance speed—the older adults were trading off speed for accuracy and stability. We argue that this trade-off was necessitated by the increased effort required by the anti-phase task under the more demanding speed conditions. Thus, considering all of these data together, the results do conform reasonably well with expectations.

At first glance, the results for the Do-not-intervene groups did not conform to predictions at the highest movement frequency. Somewhat surprisingly, performance accuracy and stability were not only similar for the younger and older adults at the 2.5-Hz speed, but were slightly better for the older adults (right sides of Fig. 3 and Fig. 4). However, this performance advantage was again likely due to the slower speed at which the older adults moved in response to the 2.5-Hz metronome frequency. The older adults seemed to prefer to remain in an effortful "fight" to maintain the anti-phase pattern, even though they were instructed to let the pattern switch to in-phase. It was predicted that there would not be significant differences between the age groups when performing with these instructions. However, the older adults appeared reluctant to "not intervene." This could have been a conscious decision, with the older adults possibly believing that they would be giving up on the intended task goal too easily (even though the instructions were made very clear to all participants). Alternatively, the reluctance to not intervene may have been a nonconscious decision, as research has previously shown that once older adults begin to use effortful processing there is a reduction in the cognitive flexibility that would allow them to switch intended goals (Craik and Jacoby 1996). Either way, this reluctance to adhere to the temporal constraints allowed the older participants to keep their performance more stable and avoid pattern transitions. It was not the case that the older participants could not move at these faster speeds, as they managed to achieve these speeds when producing the in-phase pattern.

These trade-offs observed for the older adults may further explain the lack of predicted switching behavior at the higher movement frequencies to the in-phase pattern. As may be seen in the bottom, right portion of Table 1 , the younger groups performed a much higher proportion of relative phase measures near the in-phase pattern than did the older adults at the 2.0- and 2.5-Hz speeds. Thus, compared with the older adults' coordination behavior, the Do-not-intervene instructional set seemed to affect the younger adults' switch from an effortful processing behavior to an automatic behavior with a much more complete and consistent influence than did the effect that the instructional set had on the older adults' motor behavior.

	General Discussion

TOP Abstract Experiment 1 Experiment 2 General Discussion References

 
Perhaps the most interesting (and encouraging) message that can be taken from the present studies is that the decline in motor functioning that accompanies advancing age is selective, and not absolute. The present results suggest that coordination skills that can be performed relatively automatically (i.e., in-phase coordination at all speeds and anti-phase at slow or preferred speeds), are retained well in terms of performance stability (both Experiments 1 and 2) and accuracy (Experiment 2), and do not require a trade-off in speed to remain so. Regardless of whether the task performed in these experiments is an innate skill that is hard-wired or one that has been highly overlearned, the evidence in favor of characterizing it as automatic is strong. There was relatively little evidence that advancing age had any impact at all in the performance of either the in-phase coordination pattern or the anti-phase pattern at slow or preferred speeds. Differences due to the effects of age were found when performance stability and bias diminished (at higher speeds), thus requiring effortful intervention by the performer. These findings support a theoretical perspective of the influence of age within the automatic/effortful processing framework reasonably well.

In-phase and anti-phase coordination patterns represent two of the infinite number of temporal phase relations that can be performed bimanually. Considerable evidence suggests that in-phase and anti-phase are the only two patterns that have a natural coordination preference (Kelso 1995). Other patterns can be learned, however, whereby one limb is performed with a temporal lag, or phase offset, relative to the other limb (for a review, see Lee 1998). Research with young adults has shown that the difficulty in learning these new patterns is exacerbated by the fact that coordination is pulled toward the existing, natural patterns—performing a new pattern requires that the individual "break away" from the pull towards the in-phase and anti-phase patterns. Based on the automatic/effortful processing distinction, the effects of aging would be expected to increase to the degree that effortful processing must be engaged in order to perform a new, unlearned coordination pattern. Indeed, recent research has revealed that older adults have much more difficulty (compared with younger adults) with the performance of these new tasks because of the additional cognitive load that must be engaged for learning to occur (Cunningham, Wishart, and Lee 1999; Swinnen et al. 1998).

This study is the first evaluation of the extension to voluntary movement of the critical theoretical distinction developed in the cognitive aging research between automatic and consciously controlled processes. This extension is based on the premise that there are both automatic and effortful processes involved in movement control. Until now, the investigation of the influence of aging on these processes has used a research paradigm focused on language. In this research paradigm, the automatically driven processes are considered to be the result of a strong link or habit between the stimulus and the response. For example, word phrases such as "bread and" often are linked to commonly used or learned completions, such as "butter" in this example. In older adults these automatic links are maintained, whereas the ability to inhibit these links to learn new associations becomes more difficult. Similarly, for voluntary movement of limbs, there are movement patterns that are considered to be relatively automatic. For example, rhythmical movements of the upper extremities in which the homologous muscles are activated simultaneously (in-phase movements) are described as automatic in that they can be preformed consistently at many different speeds without conscious effort. Anti-phase movements in which the rhythmical movement results from synchronous activation of opposite muscles are also described as automatic but become less automatic at higher movement speeds. The stability of the in-phase pattern is further characterized by the tendency for movements to be attracted to this pattern. For example, at higher speeds there is a tendency for younger adults to make an unconscious switch from the less automatic anti-phase movements to the more automatic in-phase movement. The premise of this proposed direction of research is that a parallel can be drawn between habitual movement patterns and the habitual, or well-learned, word associations. The existence of these automatic movement patterns or habits provides the opportunity to expand previous cognitive work and to investigate the role of automatic and effortful processes in performing and learning new movements. However, future research will need to address the viability of this extension. There are many possible limitations to extending this construct to movement, including the differences in the brain structures that support speech and movement and the overall functional purpose of the two systems.

The experiments described in this article established the relative stability of the in-phase and anti-phase movements in older adults, just as older adults retain automatic word associations. Also, they tested the effect of the Do-not-intervene versus Stay strategies to determine if there were age-related changes that were similar to those found in the cognitive research. We hypothesized that compared with younger adults, older adults would have more difficulty inhibiting the in-phase movement pattern, just as they have more difficulty inhibiting well-learned words, and as a result would switch more quickly from the more effortful pattern to the more habitual movement pattern. Instead, older adults slowed down the frequency of the movement and remained in the more effortful anti-phase movement pattern. It seems that the older adults actively chose not to switch—a strategy that is hypothesized to require cognitive effort. This hypothesis is currently under investigation using a divided attention manipulation.

The application of the consciously controlled versus automatic distinction to movement provides a theoretical structure to explore the effects of aging on control of coordinated movement. Future research will need to determine the extent of this application, including the relationship between the attentional set on the automatic and the controlled mechanisms of movement as well as how these mechanisms affect the learning of new movements by older adults.

	Acknowledgments

 
We gratefully acknowledge the Canadian Fitness and Lifestyle Research Institute and the Ontario Ministry of Health for their support of the present research. We also thank Larry Jacoby for sharing his thoughts regarding some of the issues discussed in this article.

Received for publication January 26, 1998. Accepted for publication December 14, 1999.

	References

TOP Abstract Experiment 1 Experiment 2 General Discussion References

 

• Baldissera F., Cavallari P., Marini G., Tassone G., 1991. Differential control of in-phase and anti-phase coupling of rhythmic movements of ipsilateral hand and foot. Experimental Brain Research 83:375-380. [Medline]
• Chua R., Pollock B. J., Elliott D., Swanson L. R., Carnahan H., 1995. The influence of age on manual asymmetries in movement preparation and execution. Developmental Neuropsychology 11:129-137.
• Contreras-Vidal J. L., Teulings H. L., Stelmach G. E., 1998. Elderly subjects are impaired in spatial coordination in fine motor control. Acta Psychologica 100:25-35. [Medline]
• Craik F. I. M., Jacoby L. L., 1996. Aging and memory: Implications for skilled performance. Rogers W. A., Fisk A. D., Walker N., , ed.Aging and skilled performance: Advances in theory and applications 113-137. Erlbaum, Mahwah, NJ.
• Cunningham, S., Wishart, L., & Lee, T. D. (1999). Effects of aging in the learning of a new coordination pattern. Manuscript submitted for publication.
• Darling W. G., Cooke J. D., Brown S. H., 1989. Kinematics of arm movement in elderly humans. Neurobiology of Aging 10:149-157. [Medline]
• Duncan P. W., Weiner D. K., Chandler J., Studenski S., 1990. Functional reach: A new clinical measure of balance. Journal of Gerontology: Medical Sciences 45:M192-M197.
• Folstein M. F., Folstein S. E., McHugh P. R., 1975. "Mini-mental state." A practical method for grading the cognitive state of patients for the clinician. Journal of Psychiatric Research 12:189-198. [Medline]
• Fozard J. L., Vercruyssen M., Reynolds S. L., Hancock P. A., Quilter R. E., 1994. Age differences and changes in reaction time: The Baltimore longitudinal study of aging. Journal of Gerontology: Psychological Sciences 49:P179-P189.
• Greene L. S., Williams H. G., 1996. Aging and coordination from the dynamic pattern perspective. Ferrandez A. M., Teasdale N., , ed.Changes in sensory motor behavior in aging 89-131. Elsevier, Amsterdam, the Netherlands.
• Hasher L., Zacks R. T., 1979. Automatic and effortful processes in memory. Journal of Experimental Psychology: General 108:356-388.
• Hasher L., Zacks R. T., 1988. Working memory, comprehension, and aging: A review and a new view. Bower G.H., , ed.The psychology of learning and motivation (Vol. 22) 193-225. Academic Press, New York.
• Kelso J. A. S., 1995. Dynamic patterns: The self-organization of brain and behavior MIT Press, Cambridge, MA.
• Lee T. D., 1998. On the dynamics of motor learning research. Research Quarterly for Exercise and Sport 69:334-337. [Medline]
• Lee T. D., Blandin Y., Proteau L., 1996. Effects of task instruction and oscillation frequency on bimanual coordination. Psychological Research 59:100-106. [Medline]
• Liao M., Jagacinski R. J., Greenberg N., 1997. Quantifying the performance limitations of older and younger adults in a target acquisition task. Journal of Experimental Psychology: Human Perception and Performance 23:1644-1664.
• Lyons J., Elliott D., Swanson L. R., Chua R., 1996. The use of vision in manual aiming by young and older adults. Journal of Aging & Physical Activity 4:165-178.
• Mathias S., Nayak U. S. L., Isaacs B., 1986. Balance in elderly patients: The "get-up and go" test. Archives of Physical Medicine and Rehabilitation 67:387-389. [Medline]
• Podsiadlo D., Richardson S., 1991. The timed "up & go": A test of basic functional mobility for frail elderly persons. Journal of the American Geriatric Society 39:142-148.
• Schmidt R. A., Lee T. D., 1999. Motor control and learning: A behavioral emphasis 3rd ed. Human Kinetics, Champaign, IL.
• Scholz J. P., Kelso J. A. S., 1989. A quantitative approach to understanding the formation and change of coordinated movements patterns. Journal of Motor Behavior 21:122-144. [Medline]
• Seidler R. D., Stelmach G. E., 1995. Reduction in sensorimotor control with age. Quest 47:386-394.
• Spirduso W.W., 1995. Physical dimensions of aging Human Kinetics, Champaign, IL.
• Stelmach G. E., Amrhein P. C., Goggin N. L., 1988. Age differences in bimanual coordination. Journal of Gerontology: Psychological Sciences 43:P18-P23.
• Swinnen S. P., Verschueren S. M. P., Bogaerts H., Dounskaia N., Lee T. D., Stelmach G. E., Serrien D. J., 1998. Age-related deficits in motor learning and differences in feedback processing during the production of a bimanual coordination pattern. Cognitive Neuropsychology 15:439-466.
• Wechsler D., 1981. WAIS-R manual Psychological Corporation, New York.
• Weiner D. K., Duncan P. W., Chandler J., Studenski S. A., 1992. Functional reach: A marker of physical frailty. Journal of the American Geriatric Society 40:203-207.

This article has been cited by other articles:

				E. A. Christou and L. G. Carlton Old Adults Exhibit Greater Motor Output Variability Than Young Adults Only During Rapid Discrete Isometric Contractions J. Gerontol. A Biol. Sci. Med. Sci., December 1, 2001; 56(12): B524 - 532. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Download to citation manager Citing Articles Citing Articles via HighWire PubMed PubMed Citation