- Email: [email protected]
Joint torques and dynamic joint stiffness in elderly and young men during stepping down
Clinical Biomechanics 18 (2003) 848–855 www.elsevier.com/locate/clinbiomech
Joint torques and dynamic joint stiffness in elderly and young men during stepping down Sally D. Lark
a,*
, John G. Buckley a, Simon Bennett a, David Jones b, Anthony J. Sargeant a,c
a
Active Life Span Research, Manchester Metropolitan University, Alsager ST7 2HL, UK School of Sport and Exercise Science, Birmingham University, Birmingham B15 2TT, UK Institute for Fundamental and Clinical Human Movement Sciences, Virje University, Amsterdam, The Netherlands b
c
Received 24 July 2002; accepted 27 June 2003
Abstract Objective. To compare the joint torque pattern and dynamic joint stiffness at the knee and ankle in elderly and young men during stepping down. Background. Adequate joint stiffness is critical during the single support phase to control forward and downward body momentum. Design. Six active elderly men (mean 67.7) and six young men (mean 23.6) of similar body mass and height, were filmed stepping down from one force platform to another. Repeated trials were undertaken at three different step heights (200, 250, and 300 mm). Method. Joint torques were determined for the ankle and knee of the support limb throughout the single support phase. The gradient of the joint torque–angle graph was calculated to define dynamic joint stiffness of the ankle and knee in two phases; (I) from initiation of movement until heel-off of the supporting limb, and (II) from heel-off of the supporting limb to contra-limb touch down. Results. Maximum ankle torque values were lower in the elderly and occurred at a larger dorsiflexion angle ðP < 0:05Þ. Knee torque patterns were similar in both groups. Phase I ankle stiffness was significantly less in the elderly (4.0–5.2 Nm/) at all step heights compared to the young (7.6 – 8.7 Nm/). In both groups ankle stiffness in Phase II increased with step height, while knee joint stiffness decreased. Conclusions. The different torque pattern and lower dynamic ankle stiffness in the elderly, particularly for Phase I, suggested an altered control strategy. These findings highlight the importance of dynamic ankle joint stiffness in stepping down. Relevance Understanding how the elderly step down may be important in developing strategies to prevent falls. 2003 Elsevier Ltd. All rights reserved. Keywords: Aging; Joint torques; Joint stiffness; Stepping down
1. Introduction The single support phase in stair descent requires sufficient lower limb strength to control and support the entire body mass on a single limb while moving for-
* Corresponding author. Address: School of Applied Sciences, University of Glamorgan, Rhondda Cynon Taff, Pontypridd CF37 1DL, UK. E-mail address: [email protected] (S.D. Lark).
0268-0033/$ - see front matter 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0268-0033(03)00150-5
wards and down. Previous research has shown that the elderly cannot develop ankle torque as rapidly as the young (Thelen et al., 1996), and muscle weakness and reduced flexibility at the ankle joint has been suggested to increase the risk of falling in the elderly (Whipple et al., 1987; Gehlesen and Whaley, 1990). It is not surprising therefore that stair descent by the elderly has previously been described as a Ôcontrolled fallÕ (Townsend et al., 1978). Earlier studies have investigated young to middle-aged subjects for kinematic and muscle power in stair walking (Andriacchi et al., 1980; Kowalk
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
et al., 1996). Yet none of these are comparison studies to determine altered strategies for stability. Consequently, in the review of stair negotiation in the elderly, Cavanagh et al. (1997) highlighted the need for a more comprehensive analysis of the joint kinetics required for safe stair descent. The present study was conducted in an attempt to address this need. During stepping down the control exerted by the support limb depends on the joint torque produced for any given joint angle. This relationship between joint torque and joint angle can be evaluated by modelling each joint as a torsional spring. As torque is developed to cause or control rotation of a joint, the joint stiffness also increases. Therefore dynamic joint stiffness can be defined as a measure of resistance to change in joint angle (Hunter and Kearney, 1982). The joint stiffness can be determined as the (positive) gradient of the torque–angle graph. Such an approach has been used to determine dynamic angular stiffness for the ankle during single limb support in walking (Siegler and Moskowitz, 1984; Davis and DeLuca, 1996; Lamontagne et al., 2000), running (Stefanyshyn and Nigg, 1998), and has also been used to determine the separate active and passive components of dynamic stiffness (Siegler and Moskowitz, 1984; Lamontagne et al., 2000). In this paper we follow Davis and DeLuca (1996), and define dynamic joint stiffness as the torque required to rotate the joint through a specified angle of rotation i.e., the resistance offered by the joint to a given rotation. Dynamic joint stiffness should not be confused with lower limb stiffness or muscle stiffness, which were investigated for the landing limb during stepping down by Hortobagyi and DeVita (1999, 2000). These authors suggested the stiffness of the landing limb was greater in the elderly as a result of a more upright lower extremity and a higher anticipatory muscle contraction, and they highlighted that this indicated a change in the postural control strategy. Once the swing limb lands on the lower step the individual is in double support, thus the trailing (support) limb must have an influence on the stiffness determined for the landing limb––this influence was assumed by Hortobagyi and DeVita (1999, 2000) to be minimal. A potentially more demanding phase is during single support when the individual rises up, and balances on the ball of their foot as they prepare to step down. Adequate joint stiffness is critical during this phase to control the forward and downward body momentum as such a decrease in the base of support has been cited as a risk factor in falls (Woolley et al., 1997). A lack of joint stiffness could result in instability and the movement becoming a Ôfalling forwardÕ rather than a controlled descent. Thus, the purpose of the present study was to determine whether the stance limb ankle and knee joint torques and dynamic joint stiffnesses in a group of elderly men during stepping down were comparable to those determined for a group of young men.
849
2. Methods Two matched groups of male subjects; six young (mean, 23.6 SEM 3.1 yr), and six elderly (mean, 67.7 SEM 3.5 yr) were recruited from the student population, and local community respectively. The study was performed in accordance with the Declaration of Helsinki and ethical approval for procedures was obtained from the Manchester Metropolitan University Ethical Committee. Each subject provided informed written consent prior to participation. Only men were tested to avoid any gender differences. Individuals who had chronic arthritis or rheumatism in the lower limbs, diabetes or a hospitalised fall in the previous 10 years were excluded from the study. A self-assessment questionnaire completed by all subjects indicated subjects in both groups were physically active; i.e. all elderly subjects engaged in light-to-moderate activities as defined by the Allied Dunbar National Survey (Allied Dunbar National Fitness Survey, 1992) for 30–60 min per day, except for two who exercised for approximately 80 min. All young subjects engaged in a minimum of 30 min of moderate exercise per day. The range of passive knee flexion, and ankle dorsiflexion were measured with a clinical inclinometer (Kuntovaline Oy, Helsinki, Finland; accuracy ±2) following the American Academy of Othopaedic Surgeons guidelines (American Academy of Orthopaedic Surgeons, 1965). Percentage body fat was estimated from the sum of four skin folds (Durnin and Womersley, 1973). A single optoelectric camera (E L I T E System, BTS, Milan, Italy) was used to record subjects stepping down from one Kistler force platform (Kistler, Winterthur, Switzerland) on to another. Reflective markers were attached to the right side of the subject, either directly to the skin or lycra clothing (such as a swimsuit). The markers (iliac crest, greater trochanter, lateral tibial condyle, lateral malleolus, and fifth metatarsal) were used to define a 2-D linked-segment model of the subject in the sagittal plane. Each subject was familiarised with stepping from each of three step heights (200, 250, and 300 mm) with their arms crossed in front of them (Fig. 1), before they were filmed performing three trials at each step height. Subjects started from a stationary standing position with feet comfortably apart and 25 mm from the front of the step. They were asked to step down at their own comfortable speed using a single step, and when finished to once again stand stationary. Arms were crossed over in front of the body to avoid using other balance strategies. The height order, which was completed in an ascending sequence, was not randomised for safety reasons. To eliminate potential fatigue effects subjects rested between trials. The step heights represented a United Kingdom regulation domestic step height of 200 mm (800 ), while 300 mm was representative of a train to platform or bus to curb height.
850
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
contra-lateral limb
Fig. 1. Experimental set-up. Subject is standing on a step positioned on force platform 1 (P1) and steps down to platform 2 (P2). To aid kinematic analysis reflective markers were placed over (a) ASIS, (b) greater trochanter, (c) lateral condyles, (d) malleolus, (e) fifth metatarsal. The contra-lateral limb (limb landing on the lower force platform) is indicated with the dotted line.
The step was constructed from solid sections of wood and covered the entire area of one of the Kistler force platforms (400 mm · 600 mm). No covering was used on the step and each subject performed the test in bare feet. The force platform was set to zero with the step in place, so the recorded ground reaction forces resulted from movement of the subject alone. Signals from the force platform were amplified via a charge amplifer (Kistler, 8 channel, model 9865) before linking to the A/D convertor of the E L I T E system, which synchronised force plate and kinematic data at 100 Hz. Kinematic and ground reaction force data were filtered using the E L I T E systemÕs automatic (Lambda technique) filtering software (DÕAmico and Ferrigno, 1990) with a cut off frequency of 4–8 Hz. Angular displacements and velocities were determined for the knee and ankle of the supporting limb using the E L I T E software. The change in angular displacement during the stepping movement was calculated as the difference between the angle determined when standing in the anatomical position until the maximal knee flexion, or ankle dorsiflexion. Individual net joint torques for the knee and ankle were determined for the entire movement using a standard inverse dynamics approach (Winter, 1983), and incorporated the segmental anthropometric data of Drillis and Continni (1966). The vertical landing force (N/kg) registered by
the contra-lateral limb touching the lower platform was also recorded. Dynamic joint stiffness ðkstiff Þ was determined using a torsional spring model (Siegler and Moskowitz, 1984; Davis and DeLuca, 1996; Lamontagne et al., 2000). This approach determines joint stiffness as the gradient of the torque–angle graph and was used to calculate stiffness for two distinct sub-phases in the supporting limb. Phase I, was from when the torque at the ankle began to increase (for five consecutive frames), following an initial decrease in joint torque when the ankle first started to dorsiflex, up to the point of maximum torque, which coincided with the instance of heel-off (Fig. 2). The instance of heel-off was confirmed by assessing when the recording of the marker on the malleolus rose above its starting position by a minimum of 5 mm. Phase II was from the instance of maximum torque (balanced on the ball of the foot), to the instance of contra-lateral limb touch down. Contra-lateral limb touch down was determined as the instance when the vertical landing force rose above 50 N. The torque–joint angle graph was then plotted using Microsoft Excel 2000, and a linear regression line, for each of these sub-phases, was inserted graphically to the joint torque–angle graph using the ÔAdd trendlineÕ function. The gradient of this line indicated joint stiffness for each of the sub-phases; R2 values for the regression analyses were greater than 0.67 and 0.92 for the knee and ankle respectively. This indicated the modelling approach taken had an acceptable reliability. To give an indication of the speed of descent the total movement time was determined. This was calculated as the instance when the ankle dorsiflexion angle of the supporting limb changed, from its stationary starting position, by greater than 5 until the instance of contralateral limb touchdown. Time spent with the supporting limb in a foot-flat position (i.e. up to the instance of heel-off, Phase I), and the time spent on the ball of the foot (Phase II), were also recorded. Joint stiffness values for each subject were averaged over the three trials at each step height. Group mean values were then compared using an age group · step height repeated measures analysis of variances (A N O V A ). If a significant main effect for step height was found, a simple effects post-hoc analysis was used to determine where the differences were. An alpha level of P < 0:05 was used to determine the level of significance. In an attempt to objectively quantify the pattern of torque at the ankle and knee joints, torque values at joint angles 0, 5, 10, 15 and maximum dorsiflexion, and 10, 20, 40, 60, 80 and maximum knee flexion were determined. A repeated measures A N O V A was used to determine differences between age groups in the torque produced at each of these joint angles and were evaluated for ‘‘meaningfulness’’ by means of magnitude of effects (Thomas et al., 1991). Only effect
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
size values of >0.8 were considered as truly meaningful differences. Range of motion and anthropometric differences between the two age groups were determined by a Students paired t-test. Group mean and standard errors of the mean (SEM) are presented in the tables and results section. Analysis of variance was also used to compare total movement time, percentage of time spent with the foot flat, and the time spent on the ball of the foot in the supporting limb. The same analysis was used for the vertical landing force of the swing leg.
β
(a) 80
60
torque (Nm)
Phase I 40 α
851
Phase II
20 contra-lateral limb touch down
3. Results
0 -5
0
5
10
15
Table 1 shows the mean (and SEM) height and body mass of each group along with the passive range of motion, and percentage body fat. The elderly had 15% significantly less passive range of motion for knee flexion, and 41% less ankle dorsiflexion compared to the young group ðP < 0:05Þ. Percentage body fat was significantly greater in the elderly (25.4%) compared with young (17.2%).
angle (degrees)
-20 (b) 200
contra-limb touch down 150
torque (Nm)
Phase I
3.1. Temporal characteristics
100 Phase II 50
0 0 -50
20
40
60
80
100 120
angle (degrees)
Fig. 2. (a) A typical ankle torque vs ankle angle plot. The slope of the best-fit line between points a and b represents Phase I ankle stiffness. Point b also indicates the point of heel lift off. The slope of the line between b and contra-limb touch down represents Phase II ankle stiffness. (b) A typical knee torque vs knee angle plot. Phase I corresponds to Phase I in the ankle; note the slope is negative indicating as absence of stiffness. Phase II was thought to coincide with the eccentric action of the knee extensors as the CM was lowered.
The time the supporting limb spent in foot-flat, i.e. Phase I, as a percentage of the total movement time, was greater in the elderly compared to the young at the step heights, 250 and 300 mm (Table 2). As the total movement time, i.e. Phase I + II, was similar between the two age groups, then the time spent in Phase II (on the ball of the foot) as a percentage of the total movement time was significantly shorter in the elderly group. 3.2. Ankle torques The torque values at the specific ankle angles of 0, 5, 10, 15 and maximum dorsiflexion, showed similar differences between the elderly and young groups at all three step heights, thus to save repetition only data for the high step height is presented graphically (Fig. 3a).
Table 1 Subject characteristics Age (years) Height (m) Body mass (kg) R four skinfold (mm) % Body fat Max. passive knee flexion () Max. passive ankle dorsiflexion ()
Young n ¼ 6
Elderly n ¼ 6
23.6 (3.5) 1.80 (0.2) 79.7 (6.7) 37.7 (3.0) 17.2 (0.9) 124.3 (4.4) 17.4 (2.3)
67.7 (3.1) 1.74 (0.4) 81.7 (5.5) 49.5 (8.4) 25.4 (2.2) 109.1 (6.6) 11.3 (1.3)
Group mean values (SEM) for age, height, body mass, the sum of four skinfold measurements, % body fat, maximum passive knee flexion, and maximum passive ankle dorsiflexion are given. Significant differences between groups ðP < 0:05Þ are denoted by .
852
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
Table 2 Temporal characteristics and vertical landing force Step height Total movement time (ms)
Elderly Young Elderly Young Elderly Young Elderly Young
Ball of foot time (Phase II) % Flat foot time (Phase I) Contra-lateral limb vertical touch down force (N/kg)
200 mm
250 mm
300 mm
1120 (63) 1006 (67) 258 (58) 297 (23) 75.3 (5.7) 70.3 (0.7) 1.5 (0.1) 1.5 (0.1)
1246 (52) 1163 (51) 238 (32) 322 (11) 80.0 (1.8) 69.4 (1.2) 1.7 (0.2) 1.8 (0.2)
1217 (65) 1100 (42) 276 (48) 384 (17) 79.7 (4.5) 66.1 (2.4) 1.9 (0.2) 2.0 (0.2)
Group mean values (SEM) are presented for total movement time in single support (ms), time spent on the ball of the foot (ms), and the time spent flat footed as a percentage of total movement time. The vertical touch down force of the contra-lateral limb on the lower platform corrected for body mass is also presented. Significant differences ðP 6 0:05Þ between the two age groups is denoted by .
young at all three step heights (200 mm: 14.7 vs 21; 250 mm: 17 vs 24.6 ; 300 mm: 16 vs 24). Differences between groups in the torque produced at the ankle dorsiflexion angles 0, 5, 15 were meaningful (effect sizes >0.8), as was the dorsiflexion angle at the instance of maximum torque.
100 80 torque (Nm)
* 60
*
40 *
20
Phase I
3.3. Knee torques
0 0
5
10
(a)
15
20
25
angle (degrees) Phase II
Phase I 250
*
torque (Nm)
200 150 * 100 50 0 0
10
20
30
40
50
60
70
80
90
(b)
Knee torque values at specific knee angles (again just for the high step height) are shown in Fig. 3b. The initial increase in torque was less marked in the elderly except at the 250 mm step height. The torque at the start of Phase I (20 of knee flexion) was 29%, 33% and 40% less in the elderly at the step heights 200, 250 and 300 mm respectively. At 40 and 60 of knee flexion, there was little difference in torques between the age groups. The knee flexion angle of the support limb, at the instance of touch down of the contra-lateral limb, was less in the elderly than in the young, and the maximum torque, which occurred at this point, was significantly less (14%) for the elderly at all step heights.
Fig. 3. Group mean ankle (a) and knee (b) torques (Nm) for elderly ð Þ and young ðNÞ subjects at specific angles for 300 mm step height. Only Phase I is shown as the ankle begins to plantarflex with heel lift off. The start of Phase I is indicated by the broken line. ( ) Indicates significant difference ðP < 0:05Þ.
3.4. Ankle dynamic stiffness
The younger subjects tended to have a wider range of joint torques. The initial torque value for quiet standing (i.e. 0 dorsiflexion) was significantly higher in the young (53–55 Nm) than in the old (38–42 Nm) at all step heights. However, by 5 dorsiflexion the torque values were lower in the young (20–30 Nm) compared to the old (37–40 Nm). Ankle torques for the young then steadily increased until they peaked at the instance of heel-off (85–94 Nm). In contrast ankle torques for the elderly remained fairly constant (37–44 Nm) from inital standing until approximately 15 dorsiflexion, before increasing to a maximum at the instance of heel-off (76– 83 Nm). Maximum ankle torque in the elderly occurred at a significantly greater dorsiflexion angle than in the
The R2 values of the linear regression analyses were similar (0.92) in both age groups for all step heights. Due to less torque being developed over a larger range of ankle dorsiflexion, the elderly had significantly ðP < 0:05Þ lower dynamic stiffness at the ankle joint for Phase I at all step heights (Fig. 4a). These differences ranged from 31% at 200 mm step height, 53% at 250 mm, and 37% at 300 mm. The elderly subjects also had a lower (20%) dynamic ankle joint stiffness during Phase II than the young group although this did not reach conventional levels of significance (Fig. 4b). There are no Phase II ankle stiffness values for the 200 mm step height as a flat-foot was maintained for practically the entire single support phase, that is, until contra-lateral limb touch
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
stiffness (Nm/degree)
25 20 15 10
*
*
*
250
300
5 0
(a)
200
stiffness (Nm/degree)
20 15 10 5 0
stiffness (Nm/degree) (c)
3.6. Contra-lateral limb touch down
4. Discussion 200
250
300
5 4 3 * 2 1 0
the gradient of the plotted line for this phase was negative and hence indicated a lack of stiffness (compliance) at the knee. Therefore only knee stiffness for Phase II are presented here (Fig. 4c). The elderly group consistently had less dynamic stiffness at the knee than the young, but this did not reach conventional levels of significance except at the highest step height. There was a step height effect, that is, knee stiffness decreased with increasing step height in both age groups.
Despite the observed differences in the joint torques and dynamic joint stiffness at the ankle and knee, there was no statistical difference between age groups in the peak vertical landing force for the contra-lateral limb at touch down (Table 2). It did however systematically increase with step height in both age groups, from 1.5 to 2.0 N/kg.
25
(b)
853
200
250 step height (mm)
300
Fig. 4. Comparison of group mean joint stiffness between elderly ðÞ and young ðjÞ subjects for (a) the ankle joint during Phase I, (b) the ankle joint and (c) the knee joint during Phase II. ( ) Indicates significant difference ðP < 0:05Þ.
down. Ankle stiffness values for Phase II for both age groups were significantly greater ðP < 0:05Þ than those for Phase I. These increases were between 200% and 300% across all step heights. 3.5. Knee dynamic stiffness The R2 values of the linear regression analyses were smaller for the young (0.67–0.68) than for the elderly (0.80–0.86). Initially there was a decrease in torque during Phase I (Fig. 2b), which presumably corresponded with the ÔunlockingÕ of the knee. Consequently
It has previously been shown that compared to the elderly, the young can develop torque more rapidly at the ankle joint, and this is thought to be essential for quick recovery following balance perturbations (Thelen et al., 1996). In the present study there were consistent differences, in Phase I, in the ankle torque pattern between the two groups across the three step heights; which implies an age-dependent strategy for stepping. The elderly developed a relatively constant torque over a large range of ankle motion, whereas in the young, torque initially reduced then increased sharply to a maximum, which was significantly higher than that in the elderly. In addition the maximum ankle torque in the elderly was produced at a dorsiflexion angle that was approximately 10 greater than in the young. Interestingly, the maximum dorsiflexion angle attained in Phase I by the young subjects was only slightly less than their maximal passive dorsiflexion (17.4, SEM 2.3), whereas in the elderly the maximum dorsiflexion angle (23–24) was approximately 11 beyond their maximum passive dorsiflexion (11.3, SEM 1.3). This extreme angle of dorsiflexion for the elderly subjects allowed them to maintain a foot flat position for longer, and was likely a strategy for maintaining a large base of support for longer, and thus stability. At such an angle the ability to produce adequate net torque at the ankle has been shown to diminish (Marsh et al., 1981). This suggests the elderly may be unable to respond to balance perturbation at these extreme ankle angles during stepping down. Furthermore the angle 15–17 may be a critical point for the elderly during stepping in terms of stability. Future work investigating falls in the elderly could easily test these hypotheses.
854
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
The ankle dynamic stiffness values calculated in this study for Phase I (3–8 Nm/ or 0.05–0.11 Nm/kg/) are comparable to those determined by Davis and DeLuca (1996) for walking in paediatric subjects (0.0598 Nm/ kg/), and by Stefanyshyn and Nigg (1998) for running (5.7 Nm/) and sprinting (7.4 Nm/). However, the values for Phase II (13–18 Nm/) are much higher than those reported previously. This likely reflects that during this phase the body is being lowered to a height that is lower than the level of the support limb––which was not the case for Phase I, or indeed for walking or running. During Phase I the elderly had 31–53% less ankle joint stiffness compared to the young. During this phase, footflat was maintained while the ankle and knee joint flexed under the eccentric control of the plantarflexor and knee extensors, and ankle stiffness did not change irrespective of step height. The eccentric plantarflexor torque required is dependent upon body mass, where a larger body mass requires a greater internal joint torque. As the subjects in this study were height and weight matched it was not a consideration here. Phase I coincided with the forward movement of the body mass in preparation to being lowered in the subsequent phase (Phase II). During Phase II, when subjects had to balance on the ball of the foot while the contra-lateral limb (and body mass) were being lowered, ankle dynamic stiffness increased, from that determined in Phase I, by 200–300%. Differences in the stiffness for Phase II, between the two age groups were moderate (20%) and non-significant, and this may be why the touch down force of the contra-lateral limb was similar for both groups (Table 2). Clearly the elderly have the potential to develop an equivalent degree of stiffness as the young. Therefore the significantly lower stiffness values observed in Phase I for this group, was likely due to an altered control strategy rather than impaired function, or a change in the active vs passive stiffness components of the muscle-tendon complex (Siegler and Moskowitz, 1984; Lamontagne et al., 2000). The total stepping down movement time was similar between the two groups, however the elderly maintained, across all step heights, a foot-flat position for a longer relative time than the young. The altered control strategy would appear to be one that maintains the horizontal position of the centre of mass (CM) closer to the ankle joint. Such a CM position would result in a reduction in the required ankle torque, and thus a reduction in the dynamic stiffness at the ankle (which was the case on both counts). Apart from a significantly lower knee torque in the elderly at the start of Phase I, both groups exhibited a similar torque pattern with increasing knee flexion, and this pattern was independent of step height. Although differences between groups were only significant at the highest step height, knee stiffness (Phase II only) was consistently lower in the elderly than in the young. In addition knee stiffness was significantly less than that
produced at the ankle for Phase II ðP < 0:05Þ, and was seen to decrease with increasing step height (Fig. 4c), unlike ankle stiffness, which increased with increasing step height (Fig. 4b). The knee stiffness values, ranging from 1.35 to 3.55 Nm/, are similar to those reported for descending stairs (1.7–2.6 Nm/, Kowalk et al., 1996; McFayden and Winter, 1988; Andriacchi et al., 1980). The findings presented here for the knee tend to support the notion, that an altered control strategy for stepping down was achieved by the elderly by modulating the ankle kinetics, with no significant change to the knee kinetics. As such, these findings suggest that exercise intervention and/or rehabilitation strategies in the elderly should focus on activities that develop and/or maintain ankle and foot function. In the present study dynamic joint stiffness was determined as the gradient of the joint torque–angle graph. However, the inverse dynamics approach used to determine the joint torques does not account for the individual contributions of agonist and antagonist muscles. Consequently the stiffness values were based solely on the net torque determined. Several authors have measured muscle co-activity and have found greater co-activation with increased joint stability (Marsden et al., 1983; Baratta et al., 1988; Hortobagyi and DeVita, 2000). Unfortunately there is little research available on muscle co-contraction in the elderly, and this is an area that needs further investigation, particularly for functional activities. In addition, the model used in the present study to calculate joint stiffness (Davis and DeLuca, 1996) does not separate viscous effects from elastic effects, thus the values reported may have been over-estimated (Zhang et al., 1998). Further limitations noted are the lack of maximal torque data for knee extension and plantarflexion. Such data would indicate whether the subjects were operating at, or near, their maximum during stepping down, particularly as in this study the elderly typically had a reduced lean body mass, which may have contributed towards a reduced maximal torque. In addition, although the torque pattern, and dynamic joint stiffness for the ankle were found to be significantly different between groups, the tests were carried out on groups of a small sample size. Thus future work, using groups with larger sample size, are needed to substantiate the findings presented here. Finally, to make the findings more generalizable, future work also needs to include female subjects.
5. Conclusions The study compared the joint torque pattern and dynamic joint stiffness at the knee and ankle in elderly and young men during stepping down. Possibly to maintain a larger base of support for as long as possible, the elderly remained in a flat-foot position for a longer
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
period of time before rising up onto the ball of their foot. Maximum ankle torque values, recorded during this phase of stepping down, were significantly lower in the elderly and were achieved at significantly greater dorsiflexion angles across all step heights. As a consequence, ankle stiffness was significantly lower in the elderly than in the young. Knee torque patterns were similar in both groups, and except for the highest step, knee stiffness values for the two groups were also similar. By modulating the ankle kinetics with no significant change to the knee kinetics, these findings indicate that an altered control strategy during stepping down was adopted by the elderly subjects. These findings suggest that exercise intervention and/or rehabilitation strategies in the elderly should focus on activities that develop and/or maintain ankle and foot function. References Allied Dunbar National Fitness Survey: main findings. 1992. Commissioned by the Sports Council and Health Education Authority, UK. American Academy of Orthopaedic Surgeons. 1965. Joint motion: method of measuring and recording, USA. Andriacchi, T.P., Andersson, G.B.J., Fermier, R.W., Stern, D., Galante, J.O., 1980. A study of lower-limb mechanics during stair climbing. J. Bone Joint Surg. 62A, 749–757. Baratta, R., Solomonow, M., Zhou, B.H., Letson, E.D., Chuinard, R., DÕAmbrosia, D., 1988. Muscular coactivation: the role of the antagonist musculature in maintaining knee stability. Am. J. Sports Med. 16, 113–122. Cavanagh, P.R., Mulfinger, L.M., Owens, D.A., 1997. How do the elderly negotiate stairs? Muscle Nerve Suppl. 5, S52–S55. DÕAmico, M., Ferrigno, G., 1990. Technique for the evaluation of derivatives from noisy biomechanical displacement data using a model-based bandwidth selection procedure. Med. Biol. Eng. Comput. 25, 407–415. Davis, R.B., DeLuca, P.A., 1996. Gait characterization via dynamic joint stiffness. Gait Posture 4, 224–231. Drillis, R., Continni, R. 1966. Body segment parameters. Technical Report 1166.03, NY University, New York. Durnin, J.V.G.A., Womersley, J., 1973. Body fat assessed from total body density and its estimation from skin-fold thickness: measurements on 481 men and women aged from 16 to 72 years. Br. J. Nutr. 2, 77–97. Gehlesen, G.M., Whaley, M.H., 1990. Falls in the elderly: part II balance, strength, and flexibility. Arch. Phys. Med. Rehab. 71, 739– 741.
855
Hortobagyi, T., DeVita, P., 1999. Altered movement strategy increases lower extremity stiffness during stepping down in the aged. J. Gerontol.: Biol. Sci. 54A, B63–B70. Hortobagyi, T., DeVita, P., 2000. Muscle pre- and coactivity during downward stepping are associated with leg stiffness in aging. J. Electromyogr. Kinesiol. 10, 117–126. Hunter, A.F., Kearney, R.E., 1982. Dynamics of human ankle stiffness: variation with mean ankle torque. J. Biomech. 15, 747– 752. Kowalk, D.L., Duncan, J.A., Vaughan, C.L., 1996. Abduction– adduction moments at the knee during stair ascent and descent. J. Biomech. 29, 383–388. Lamontagne, A., Malouin, F., Richards, C.L., 2000. Contribution of passive stiffness to ankle plantarflexion moment during gait after stroke. Arch. Phys. Med. Rehab. 81, 351–358. Marsden, C.D., Obeso, J.A., Rothwell, J.C., 1983. The function of the antagonist muscle during fast limb movements in man. J. Physiol. 335, 1–13. Marsh, E., Sale, D., McComas, A.J., Quinlan, J., 1981. Influence of joint position on ankle dorsiflexion in humans. J. Appl. Physiol. Respirat. Environ. Exercise Physiol. 51, 160–167. McFayden, B.J., Winter, D.A., 1988. An integrated biomechanical analysis of normal stair ascent and descent. J. Biomech. 21, 733– 744. Siegler, S., Moskowitz, G.D., 1984. Passive and active components of the internal moment developed about the ankle joint during human ambulation. J. Biomech. 17, 647–652. Stefanyshyn, D.J., Nigg, B.M., 1998. Dynamic angular stiffness of the ankle joint during running and sprinting. J. Appl. Biomech. 14, 292–299. Thelen, D.G., Schultz, A.B., Alexander, N.B., Ashton-Miller, J.A., 1996. Effects of age on rapid ankle torque development. J. Gerontol. 51A, M226–M232. Thomas, J.R., Salazar, W., Landers, D.M., 1991. What is missing in p less than 0.5? Effect size. Res. Quart. Exercise Sport 62, 344– 348. Townsend, M.A., Lainhart, S.P., Shiavi, R., 1978. Variability and biomechanics of synergy patterns of some lower-limb muscles during ascending and descending stairs and level walking. Med. Biol. Eng. Comput. 16, 681–688. Whipple, R.H., Wolfson, L.I., Amerman, P.M., 1987. The relationship of knee and ankle weakness to falls in nursing home residents: an isokinetic study. J. Am. Geriatr. Soc. 35, 13–20. Winter, D.A., 1983. Moments of force and mechanical power in jogging. J. Biomech. 16, 91–97. Woolley, S.M., Czaja, S.J., Drurry, C.G., 1997. An assessment of falls in elderly men and women. J. Gerontol.: Biol. Sci. Med. Sci. 52A, M80–M87. Zhang, L-Q., Nuber, G., Butler, J., Bowen, M., Rymer, W.Z., 1998. In vivo human knee joint dynamic properties as functions of muscle contraction and joint position. J. Biomech. 31, 71– 76.
Joint torques and dynamic joint stiffness in elderly and young men during stepping down Sally D. Lark
a,*
, John G. Buckley a, Simon Bennett a, David Jones b, Anthony J. Sargeant a,c
a
Active Life Span Research, Manchester Metropolitan University, Alsager ST7 2HL, UK School of Sport and Exercise Science, Birmingham University, Birmingham B15 2TT, UK Institute for Fundamental and Clinical Human Movement Sciences, Virje University, Amsterdam, The Netherlands b
c
Received 24 July 2002; accepted 27 June 2003
Abstract Objective. To compare the joint torque pattern and dynamic joint stiffness at the knee and ankle in elderly and young men during stepping down. Background. Adequate joint stiffness is critical during the single support phase to control forward and downward body momentum. Design. Six active elderly men (mean 67.7) and six young men (mean 23.6) of similar body mass and height, were filmed stepping down from one force platform to another. Repeated trials were undertaken at three different step heights (200, 250, and 300 mm). Method. Joint torques were determined for the ankle and knee of the support limb throughout the single support phase. The gradient of the joint torque–angle graph was calculated to define dynamic joint stiffness of the ankle and knee in two phases; (I) from initiation of movement until heel-off of the supporting limb, and (II) from heel-off of the supporting limb to contra-limb touch down. Results. Maximum ankle torque values were lower in the elderly and occurred at a larger dorsiflexion angle ðP < 0:05Þ. Knee torque patterns were similar in both groups. Phase I ankle stiffness was significantly less in the elderly (4.0–5.2 Nm/) at all step heights compared to the young (7.6 – 8.7 Nm/). In both groups ankle stiffness in Phase II increased with step height, while knee joint stiffness decreased. Conclusions. The different torque pattern and lower dynamic ankle stiffness in the elderly, particularly for Phase I, suggested an altered control strategy. These findings highlight the importance of dynamic ankle joint stiffness in stepping down. Relevance Understanding how the elderly step down may be important in developing strategies to prevent falls. 2003 Elsevier Ltd. All rights reserved. Keywords: Aging; Joint torques; Joint stiffness; Stepping down
1. Introduction The single support phase in stair descent requires sufficient lower limb strength to control and support the entire body mass on a single limb while moving for-
* Corresponding author. Address: School of Applied Sciences, University of Glamorgan, Rhondda Cynon Taff, Pontypridd CF37 1DL, UK. E-mail address: [email protected] (S.D. Lark).
0268-0033/$ - see front matter 2003 Elsevier Ltd. All rights reserved. doi:10.1016/S0268-0033(03)00150-5
wards and down. Previous research has shown that the elderly cannot develop ankle torque as rapidly as the young (Thelen et al., 1996), and muscle weakness and reduced flexibility at the ankle joint has been suggested to increase the risk of falling in the elderly (Whipple et al., 1987; Gehlesen and Whaley, 1990). It is not surprising therefore that stair descent by the elderly has previously been described as a Ôcontrolled fallÕ (Townsend et al., 1978). Earlier studies have investigated young to middle-aged subjects for kinematic and muscle power in stair walking (Andriacchi et al., 1980; Kowalk
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
et al., 1996). Yet none of these are comparison studies to determine altered strategies for stability. Consequently, in the review of stair negotiation in the elderly, Cavanagh et al. (1997) highlighted the need for a more comprehensive analysis of the joint kinetics required for safe stair descent. The present study was conducted in an attempt to address this need. During stepping down the control exerted by the support limb depends on the joint torque produced for any given joint angle. This relationship between joint torque and joint angle can be evaluated by modelling each joint as a torsional spring. As torque is developed to cause or control rotation of a joint, the joint stiffness also increases. Therefore dynamic joint stiffness can be defined as a measure of resistance to change in joint angle (Hunter and Kearney, 1982). The joint stiffness can be determined as the (positive) gradient of the torque–angle graph. Such an approach has been used to determine dynamic angular stiffness for the ankle during single limb support in walking (Siegler and Moskowitz, 1984; Davis and DeLuca, 1996; Lamontagne et al., 2000), running (Stefanyshyn and Nigg, 1998), and has also been used to determine the separate active and passive components of dynamic stiffness (Siegler and Moskowitz, 1984; Lamontagne et al., 2000). In this paper we follow Davis and DeLuca (1996), and define dynamic joint stiffness as the torque required to rotate the joint through a specified angle of rotation i.e., the resistance offered by the joint to a given rotation. Dynamic joint stiffness should not be confused with lower limb stiffness or muscle stiffness, which were investigated for the landing limb during stepping down by Hortobagyi and DeVita (1999, 2000). These authors suggested the stiffness of the landing limb was greater in the elderly as a result of a more upright lower extremity and a higher anticipatory muscle contraction, and they highlighted that this indicated a change in the postural control strategy. Once the swing limb lands on the lower step the individual is in double support, thus the trailing (support) limb must have an influence on the stiffness determined for the landing limb––this influence was assumed by Hortobagyi and DeVita (1999, 2000) to be minimal. A potentially more demanding phase is during single support when the individual rises up, and balances on the ball of their foot as they prepare to step down. Adequate joint stiffness is critical during this phase to control the forward and downward body momentum as such a decrease in the base of support has been cited as a risk factor in falls (Woolley et al., 1997). A lack of joint stiffness could result in instability and the movement becoming a Ôfalling forwardÕ rather than a controlled descent. Thus, the purpose of the present study was to determine whether the stance limb ankle and knee joint torques and dynamic joint stiffnesses in a group of elderly men during stepping down were comparable to those determined for a group of young men.
849
2. Methods Two matched groups of male subjects; six young (mean, 23.6 SEM 3.1 yr), and six elderly (mean, 67.7 SEM 3.5 yr) were recruited from the student population, and local community respectively. The study was performed in accordance with the Declaration of Helsinki and ethical approval for procedures was obtained from the Manchester Metropolitan University Ethical Committee. Each subject provided informed written consent prior to participation. Only men were tested to avoid any gender differences. Individuals who had chronic arthritis or rheumatism in the lower limbs, diabetes or a hospitalised fall in the previous 10 years were excluded from the study. A self-assessment questionnaire completed by all subjects indicated subjects in both groups were physically active; i.e. all elderly subjects engaged in light-to-moderate activities as defined by the Allied Dunbar National Survey (Allied Dunbar National Fitness Survey, 1992) for 30–60 min per day, except for two who exercised for approximately 80 min. All young subjects engaged in a minimum of 30 min of moderate exercise per day. The range of passive knee flexion, and ankle dorsiflexion were measured with a clinical inclinometer (Kuntovaline Oy, Helsinki, Finland; accuracy ±2) following the American Academy of Othopaedic Surgeons guidelines (American Academy of Orthopaedic Surgeons, 1965). Percentage body fat was estimated from the sum of four skin folds (Durnin and Womersley, 1973). A single optoelectric camera (E L I T E System, BTS, Milan, Italy) was used to record subjects stepping down from one Kistler force platform (Kistler, Winterthur, Switzerland) on to another. Reflective markers were attached to the right side of the subject, either directly to the skin or lycra clothing (such as a swimsuit). The markers (iliac crest, greater trochanter, lateral tibial condyle, lateral malleolus, and fifth metatarsal) were used to define a 2-D linked-segment model of the subject in the sagittal plane. Each subject was familiarised with stepping from each of three step heights (200, 250, and 300 mm) with their arms crossed in front of them (Fig. 1), before they were filmed performing three trials at each step height. Subjects started from a stationary standing position with feet comfortably apart and 25 mm from the front of the step. They were asked to step down at their own comfortable speed using a single step, and when finished to once again stand stationary. Arms were crossed over in front of the body to avoid using other balance strategies. The height order, which was completed in an ascending sequence, was not randomised for safety reasons. To eliminate potential fatigue effects subjects rested between trials. The step heights represented a United Kingdom regulation domestic step height of 200 mm (800 ), while 300 mm was representative of a train to platform or bus to curb height.
850
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
contra-lateral limb
Fig. 1. Experimental set-up. Subject is standing on a step positioned on force platform 1 (P1) and steps down to platform 2 (P2). To aid kinematic analysis reflective markers were placed over (a) ASIS, (b) greater trochanter, (c) lateral condyles, (d) malleolus, (e) fifth metatarsal. The contra-lateral limb (limb landing on the lower force platform) is indicated with the dotted line.
The step was constructed from solid sections of wood and covered the entire area of one of the Kistler force platforms (400 mm · 600 mm). No covering was used on the step and each subject performed the test in bare feet. The force platform was set to zero with the step in place, so the recorded ground reaction forces resulted from movement of the subject alone. Signals from the force platform were amplified via a charge amplifer (Kistler, 8 channel, model 9865) before linking to the A/D convertor of the E L I T E system, which synchronised force plate and kinematic data at 100 Hz. Kinematic and ground reaction force data were filtered using the E L I T E systemÕs automatic (Lambda technique) filtering software (DÕAmico and Ferrigno, 1990) with a cut off frequency of 4–8 Hz. Angular displacements and velocities were determined for the knee and ankle of the supporting limb using the E L I T E software. The change in angular displacement during the stepping movement was calculated as the difference between the angle determined when standing in the anatomical position until the maximal knee flexion, or ankle dorsiflexion. Individual net joint torques for the knee and ankle were determined for the entire movement using a standard inverse dynamics approach (Winter, 1983), and incorporated the segmental anthropometric data of Drillis and Continni (1966). The vertical landing force (N/kg) registered by
the contra-lateral limb touching the lower platform was also recorded. Dynamic joint stiffness ðkstiff Þ was determined using a torsional spring model (Siegler and Moskowitz, 1984; Davis and DeLuca, 1996; Lamontagne et al., 2000). This approach determines joint stiffness as the gradient of the torque–angle graph and was used to calculate stiffness for two distinct sub-phases in the supporting limb. Phase I, was from when the torque at the ankle began to increase (for five consecutive frames), following an initial decrease in joint torque when the ankle first started to dorsiflex, up to the point of maximum torque, which coincided with the instance of heel-off (Fig. 2). The instance of heel-off was confirmed by assessing when the recording of the marker on the malleolus rose above its starting position by a minimum of 5 mm. Phase II was from the instance of maximum torque (balanced on the ball of the foot), to the instance of contra-lateral limb touch down. Contra-lateral limb touch down was determined as the instance when the vertical landing force rose above 50 N. The torque–joint angle graph was then plotted using Microsoft Excel 2000, and a linear regression line, for each of these sub-phases, was inserted graphically to the joint torque–angle graph using the ÔAdd trendlineÕ function. The gradient of this line indicated joint stiffness for each of the sub-phases; R2 values for the regression analyses were greater than 0.67 and 0.92 for the knee and ankle respectively. This indicated the modelling approach taken had an acceptable reliability. To give an indication of the speed of descent the total movement time was determined. This was calculated as the instance when the ankle dorsiflexion angle of the supporting limb changed, from its stationary starting position, by greater than 5 until the instance of contralateral limb touchdown. Time spent with the supporting limb in a foot-flat position (i.e. up to the instance of heel-off, Phase I), and the time spent on the ball of the foot (Phase II), were also recorded. Joint stiffness values for each subject were averaged over the three trials at each step height. Group mean values were then compared using an age group · step height repeated measures analysis of variances (A N O V A ). If a significant main effect for step height was found, a simple effects post-hoc analysis was used to determine where the differences were. An alpha level of P < 0:05 was used to determine the level of significance. In an attempt to objectively quantify the pattern of torque at the ankle and knee joints, torque values at joint angles 0, 5, 10, 15 and maximum dorsiflexion, and 10, 20, 40, 60, 80 and maximum knee flexion were determined. A repeated measures A N O V A was used to determine differences between age groups in the torque produced at each of these joint angles and were evaluated for ‘‘meaningfulness’’ by means of magnitude of effects (Thomas et al., 1991). Only effect
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
size values of >0.8 were considered as truly meaningful differences. Range of motion and anthropometric differences between the two age groups were determined by a Students paired t-test. Group mean and standard errors of the mean (SEM) are presented in the tables and results section. Analysis of variance was also used to compare total movement time, percentage of time spent with the foot flat, and the time spent on the ball of the foot in the supporting limb. The same analysis was used for the vertical landing force of the swing leg.
β
(a) 80
60
torque (Nm)
Phase I 40 α
851
Phase II
20 contra-lateral limb touch down
3. Results
0 -5
0
5
10
15
Table 1 shows the mean (and SEM) height and body mass of each group along with the passive range of motion, and percentage body fat. The elderly had 15% significantly less passive range of motion for knee flexion, and 41% less ankle dorsiflexion compared to the young group ðP < 0:05Þ. Percentage body fat was significantly greater in the elderly (25.4%) compared with young (17.2%).
angle (degrees)
-20 (b) 200
contra-limb touch down 150
torque (Nm)
Phase I
3.1. Temporal characteristics
100 Phase II 50
0 0 -50
20
40
60
80
100 120
angle (degrees)
Fig. 2. (a) A typical ankle torque vs ankle angle plot. The slope of the best-fit line between points a and b represents Phase I ankle stiffness. Point b also indicates the point of heel lift off. The slope of the line between b and contra-limb touch down represents Phase II ankle stiffness. (b) A typical knee torque vs knee angle plot. Phase I corresponds to Phase I in the ankle; note the slope is negative indicating as absence of stiffness. Phase II was thought to coincide with the eccentric action of the knee extensors as the CM was lowered.
The time the supporting limb spent in foot-flat, i.e. Phase I, as a percentage of the total movement time, was greater in the elderly compared to the young at the step heights, 250 and 300 mm (Table 2). As the total movement time, i.e. Phase I + II, was similar between the two age groups, then the time spent in Phase II (on the ball of the foot) as a percentage of the total movement time was significantly shorter in the elderly group. 3.2. Ankle torques The torque values at the specific ankle angles of 0, 5, 10, 15 and maximum dorsiflexion, showed similar differences between the elderly and young groups at all three step heights, thus to save repetition only data for the high step height is presented graphically (Fig. 3a).
Table 1 Subject characteristics Age (years) Height (m) Body mass (kg) R four skinfold (mm) % Body fat Max. passive knee flexion () Max. passive ankle dorsiflexion ()
Young n ¼ 6
Elderly n ¼ 6
23.6 (3.5) 1.80 (0.2) 79.7 (6.7) 37.7 (3.0) 17.2 (0.9) 124.3 (4.4) 17.4 (2.3)
67.7 (3.1) 1.74 (0.4) 81.7 (5.5) 49.5 (8.4) 25.4 (2.2) 109.1 (6.6) 11.3 (1.3)
Group mean values (SEM) for age, height, body mass, the sum of four skinfold measurements, % body fat, maximum passive knee flexion, and maximum passive ankle dorsiflexion are given. Significant differences between groups ðP < 0:05Þ are denoted by .
852
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
Table 2 Temporal characteristics and vertical landing force Step height Total movement time (ms)
Elderly Young Elderly Young Elderly Young Elderly Young
Ball of foot time (Phase II) % Flat foot time (Phase I) Contra-lateral limb vertical touch down force (N/kg)
200 mm
250 mm
300 mm
1120 (63) 1006 (67) 258 (58) 297 (23) 75.3 (5.7) 70.3 (0.7) 1.5 (0.1) 1.5 (0.1)
1246 (52) 1163 (51) 238 (32) 322 (11) 80.0 (1.8) 69.4 (1.2) 1.7 (0.2) 1.8 (0.2)
1217 (65) 1100 (42) 276 (48) 384 (17) 79.7 (4.5) 66.1 (2.4) 1.9 (0.2) 2.0 (0.2)
Group mean values (SEM) are presented for total movement time in single support (ms), time spent on the ball of the foot (ms), and the time spent flat footed as a percentage of total movement time. The vertical touch down force of the contra-lateral limb on the lower platform corrected for body mass is also presented. Significant differences ðP 6 0:05Þ between the two age groups is denoted by .
young at all three step heights (200 mm: 14.7 vs 21; 250 mm: 17 vs 24.6 ; 300 mm: 16 vs 24). Differences between groups in the torque produced at the ankle dorsiflexion angles 0, 5, 15 were meaningful (effect sizes >0.8), as was the dorsiflexion angle at the instance of maximum torque.
100 80 torque (Nm)
* 60
*
40 *
20
Phase I
3.3. Knee torques
0 0
5
10
(a)
15
20
25
angle (degrees) Phase II
Phase I 250
*
torque (Nm)
200 150 * 100 50 0 0
10
20
30
40
50
60
70
80
90
(b)
Knee torque values at specific knee angles (again just for the high step height) are shown in Fig. 3b. The initial increase in torque was less marked in the elderly except at the 250 mm step height. The torque at the start of Phase I (20 of knee flexion) was 29%, 33% and 40% less in the elderly at the step heights 200, 250 and 300 mm respectively. At 40 and 60 of knee flexion, there was little difference in torques between the age groups. The knee flexion angle of the support limb, at the instance of touch down of the contra-lateral limb, was less in the elderly than in the young, and the maximum torque, which occurred at this point, was significantly less (14%) for the elderly at all step heights.
Fig. 3. Group mean ankle (a) and knee (b) torques (Nm) for elderly ð Þ and young ðNÞ subjects at specific angles for 300 mm step height. Only Phase I is shown as the ankle begins to plantarflex with heel lift off. The start of Phase I is indicated by the broken line. ( ) Indicates significant difference ðP < 0:05Þ.
3.4. Ankle dynamic stiffness
The younger subjects tended to have a wider range of joint torques. The initial torque value for quiet standing (i.e. 0 dorsiflexion) was significantly higher in the young (53–55 Nm) than in the old (38–42 Nm) at all step heights. However, by 5 dorsiflexion the torque values were lower in the young (20–30 Nm) compared to the old (37–40 Nm). Ankle torques for the young then steadily increased until they peaked at the instance of heel-off (85–94 Nm). In contrast ankle torques for the elderly remained fairly constant (37–44 Nm) from inital standing until approximately 15 dorsiflexion, before increasing to a maximum at the instance of heel-off (76– 83 Nm). Maximum ankle torque in the elderly occurred at a significantly greater dorsiflexion angle than in the
The R2 values of the linear regression analyses were similar (0.92) in both age groups for all step heights. Due to less torque being developed over a larger range of ankle dorsiflexion, the elderly had significantly ðP < 0:05Þ lower dynamic stiffness at the ankle joint for Phase I at all step heights (Fig. 4a). These differences ranged from 31% at 200 mm step height, 53% at 250 mm, and 37% at 300 mm. The elderly subjects also had a lower (20%) dynamic ankle joint stiffness during Phase II than the young group although this did not reach conventional levels of significance (Fig. 4b). There are no Phase II ankle stiffness values for the 200 mm step height as a flat-foot was maintained for practically the entire single support phase, that is, until contra-lateral limb touch
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
stiffness (Nm/degree)
25 20 15 10
*
*
*
250
300
5 0
(a)
200
stiffness (Nm/degree)
20 15 10 5 0
stiffness (Nm/degree) (c)
3.6. Contra-lateral limb touch down
4. Discussion 200
250
300
5 4 3 * 2 1 0
the gradient of the plotted line for this phase was negative and hence indicated a lack of stiffness (compliance) at the knee. Therefore only knee stiffness for Phase II are presented here (Fig. 4c). The elderly group consistently had less dynamic stiffness at the knee than the young, but this did not reach conventional levels of significance except at the highest step height. There was a step height effect, that is, knee stiffness decreased with increasing step height in both age groups.
Despite the observed differences in the joint torques and dynamic joint stiffness at the ankle and knee, there was no statistical difference between age groups in the peak vertical landing force for the contra-lateral limb at touch down (Table 2). It did however systematically increase with step height in both age groups, from 1.5 to 2.0 N/kg.
25
(b)
853
200
250 step height (mm)
300
Fig. 4. Comparison of group mean joint stiffness between elderly ðÞ and young ðjÞ subjects for (a) the ankle joint during Phase I, (b) the ankle joint and (c) the knee joint during Phase II. ( ) Indicates significant difference ðP < 0:05Þ.
down. Ankle stiffness values for Phase II for both age groups were significantly greater ðP < 0:05Þ than those for Phase I. These increases were between 200% and 300% across all step heights. 3.5. Knee dynamic stiffness The R2 values of the linear regression analyses were smaller for the young (0.67–0.68) than for the elderly (0.80–0.86). Initially there was a decrease in torque during Phase I (Fig. 2b), which presumably corresponded with the ÔunlockingÕ of the knee. Consequently
It has previously been shown that compared to the elderly, the young can develop torque more rapidly at the ankle joint, and this is thought to be essential for quick recovery following balance perturbations (Thelen et al., 1996). In the present study there were consistent differences, in Phase I, in the ankle torque pattern between the two groups across the three step heights; which implies an age-dependent strategy for stepping. The elderly developed a relatively constant torque over a large range of ankle motion, whereas in the young, torque initially reduced then increased sharply to a maximum, which was significantly higher than that in the elderly. In addition the maximum ankle torque in the elderly was produced at a dorsiflexion angle that was approximately 10 greater than in the young. Interestingly, the maximum dorsiflexion angle attained in Phase I by the young subjects was only slightly less than their maximal passive dorsiflexion (17.4, SEM 2.3), whereas in the elderly the maximum dorsiflexion angle (23–24) was approximately 11 beyond their maximum passive dorsiflexion (11.3, SEM 1.3). This extreme angle of dorsiflexion for the elderly subjects allowed them to maintain a foot flat position for longer, and was likely a strategy for maintaining a large base of support for longer, and thus stability. At such an angle the ability to produce adequate net torque at the ankle has been shown to diminish (Marsh et al., 1981). This suggests the elderly may be unable to respond to balance perturbation at these extreme ankle angles during stepping down. Furthermore the angle 15–17 may be a critical point for the elderly during stepping in terms of stability. Future work investigating falls in the elderly could easily test these hypotheses.
854
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
The ankle dynamic stiffness values calculated in this study for Phase I (3–8 Nm/ or 0.05–0.11 Nm/kg/) are comparable to those determined by Davis and DeLuca (1996) for walking in paediatric subjects (0.0598 Nm/ kg/), and by Stefanyshyn and Nigg (1998) for running (5.7 Nm/) and sprinting (7.4 Nm/). However, the values for Phase II (13–18 Nm/) are much higher than those reported previously. This likely reflects that during this phase the body is being lowered to a height that is lower than the level of the support limb––which was not the case for Phase I, or indeed for walking or running. During Phase I the elderly had 31–53% less ankle joint stiffness compared to the young. During this phase, footflat was maintained while the ankle and knee joint flexed under the eccentric control of the plantarflexor and knee extensors, and ankle stiffness did not change irrespective of step height. The eccentric plantarflexor torque required is dependent upon body mass, where a larger body mass requires a greater internal joint torque. As the subjects in this study were height and weight matched it was not a consideration here. Phase I coincided with the forward movement of the body mass in preparation to being lowered in the subsequent phase (Phase II). During Phase II, when subjects had to balance on the ball of the foot while the contra-lateral limb (and body mass) were being lowered, ankle dynamic stiffness increased, from that determined in Phase I, by 200–300%. Differences in the stiffness for Phase II, between the two age groups were moderate (20%) and non-significant, and this may be why the touch down force of the contra-lateral limb was similar for both groups (Table 2). Clearly the elderly have the potential to develop an equivalent degree of stiffness as the young. Therefore the significantly lower stiffness values observed in Phase I for this group, was likely due to an altered control strategy rather than impaired function, or a change in the active vs passive stiffness components of the muscle-tendon complex (Siegler and Moskowitz, 1984; Lamontagne et al., 2000). The total stepping down movement time was similar between the two groups, however the elderly maintained, across all step heights, a foot-flat position for a longer relative time than the young. The altered control strategy would appear to be one that maintains the horizontal position of the centre of mass (CM) closer to the ankle joint. Such a CM position would result in a reduction in the required ankle torque, and thus a reduction in the dynamic stiffness at the ankle (which was the case on both counts). Apart from a significantly lower knee torque in the elderly at the start of Phase I, both groups exhibited a similar torque pattern with increasing knee flexion, and this pattern was independent of step height. Although differences between groups were only significant at the highest step height, knee stiffness (Phase II only) was consistently lower in the elderly than in the young. In addition knee stiffness was significantly less than that
produced at the ankle for Phase II ðP < 0:05Þ, and was seen to decrease with increasing step height (Fig. 4c), unlike ankle stiffness, which increased with increasing step height (Fig. 4b). The knee stiffness values, ranging from 1.35 to 3.55 Nm/, are similar to those reported for descending stairs (1.7–2.6 Nm/, Kowalk et al., 1996; McFayden and Winter, 1988; Andriacchi et al., 1980). The findings presented here for the knee tend to support the notion, that an altered control strategy for stepping down was achieved by the elderly by modulating the ankle kinetics, with no significant change to the knee kinetics. As such, these findings suggest that exercise intervention and/or rehabilitation strategies in the elderly should focus on activities that develop and/or maintain ankle and foot function. In the present study dynamic joint stiffness was determined as the gradient of the joint torque–angle graph. However, the inverse dynamics approach used to determine the joint torques does not account for the individual contributions of agonist and antagonist muscles. Consequently the stiffness values were based solely on the net torque determined. Several authors have measured muscle co-activity and have found greater co-activation with increased joint stability (Marsden et al., 1983; Baratta et al., 1988; Hortobagyi and DeVita, 2000). Unfortunately there is little research available on muscle co-contraction in the elderly, and this is an area that needs further investigation, particularly for functional activities. In addition, the model used in the present study to calculate joint stiffness (Davis and DeLuca, 1996) does not separate viscous effects from elastic effects, thus the values reported may have been over-estimated (Zhang et al., 1998). Further limitations noted are the lack of maximal torque data for knee extension and plantarflexion. Such data would indicate whether the subjects were operating at, or near, their maximum during stepping down, particularly as in this study the elderly typically had a reduced lean body mass, which may have contributed towards a reduced maximal torque. In addition, although the torque pattern, and dynamic joint stiffness for the ankle were found to be significantly different between groups, the tests were carried out on groups of a small sample size. Thus future work, using groups with larger sample size, are needed to substantiate the findings presented here. Finally, to make the findings more generalizable, future work also needs to include female subjects.
5. Conclusions The study compared the joint torque pattern and dynamic joint stiffness at the knee and ankle in elderly and young men during stepping down. Possibly to maintain a larger base of support for as long as possible, the elderly remained in a flat-foot position for a longer
S.D. Lark et al. / Clinical Biomechanics 18 (2003) 848–855
period of time before rising up onto the ball of their foot. Maximum ankle torque values, recorded during this phase of stepping down, were significantly lower in the elderly and were achieved at significantly greater dorsiflexion angles across all step heights. As a consequence, ankle stiffness was significantly lower in the elderly than in the young. Knee torque patterns were similar in both groups, and except for the highest step, knee stiffness values for the two groups were also similar. By modulating the ankle kinetics with no significant change to the knee kinetics, these findings indicate that an altered control strategy during stepping down was adopted by the elderly subjects. These findings suggest that exercise intervention and/or rehabilitation strategies in the elderly should focus on activities that develop and/or maintain ankle and foot function. References Allied Dunbar National Fitness Survey: main findings. 1992. Commissioned by the Sports Council and Health Education Authority, UK. American Academy of Orthopaedic Surgeons. 1965. Joint motion: method of measuring and recording, USA. Andriacchi, T.P., Andersson, G.B.J., Fermier, R.W., Stern, D., Galante, J.O., 1980. A study of lower-limb mechanics during stair climbing. J. Bone Joint Surg. 62A, 749–757. Baratta, R., Solomonow, M., Zhou, B.H., Letson, E.D., Chuinard, R., DÕAmbrosia, D., 1988. Muscular coactivation: the role of the antagonist musculature in maintaining knee stability. Am. J. Sports Med. 16, 113–122. Cavanagh, P.R., Mulfinger, L.M., Owens, D.A., 1997. How do the elderly negotiate stairs? Muscle Nerve Suppl. 5, S52–S55. DÕAmico, M., Ferrigno, G., 1990. Technique for the evaluation of derivatives from noisy biomechanical displacement data using a model-based bandwidth selection procedure. Med. Biol. Eng. Comput. 25, 407–415. Davis, R.B., DeLuca, P.A., 1996. Gait characterization via dynamic joint stiffness. Gait Posture 4, 224–231. Drillis, R., Continni, R. 1966. Body segment parameters. Technical Report 1166.03, NY University, New York. Durnin, J.V.G.A., Womersley, J., 1973. Body fat assessed from total body density and its estimation from skin-fold thickness: measurements on 481 men and women aged from 16 to 72 years. Br. J. Nutr. 2, 77–97. Gehlesen, G.M., Whaley, M.H., 1990. Falls in the elderly: part II balance, strength, and flexibility. Arch. Phys. Med. Rehab. 71, 739– 741.
855
Hortobagyi, T., DeVita, P., 1999. Altered movement strategy increases lower extremity stiffness during stepping down in the aged. J. Gerontol.: Biol. Sci. 54A, B63–B70. Hortobagyi, T., DeVita, P., 2000. Muscle pre- and coactivity during downward stepping are associated with leg stiffness in aging. J. Electromyogr. Kinesiol. 10, 117–126. Hunter, A.F., Kearney, R.E., 1982. Dynamics of human ankle stiffness: variation with mean ankle torque. J. Biomech. 15, 747– 752. Kowalk, D.L., Duncan, J.A., Vaughan, C.L., 1996. Abduction– adduction moments at the knee during stair ascent and descent. J. Biomech. 29, 383–388. Lamontagne, A., Malouin, F., Richards, C.L., 2000. Contribution of passive stiffness to ankle plantarflexion moment during gait after stroke. Arch. Phys. Med. Rehab. 81, 351–358. Marsden, C.D., Obeso, J.A., Rothwell, J.C., 1983. The function of the antagonist muscle during fast limb movements in man. J. Physiol. 335, 1–13. Marsh, E., Sale, D., McComas, A.J., Quinlan, J., 1981. Influence of joint position on ankle dorsiflexion in humans. J. Appl. Physiol. Respirat. Environ. Exercise Physiol. 51, 160–167. McFayden, B.J., Winter, D.A., 1988. An integrated biomechanical analysis of normal stair ascent and descent. J. Biomech. 21, 733– 744. Siegler, S., Moskowitz, G.D., 1984. Passive and active components of the internal moment developed about the ankle joint during human ambulation. J. Biomech. 17, 647–652. Stefanyshyn, D.J., Nigg, B.M., 1998. Dynamic angular stiffness of the ankle joint during running and sprinting. J. Appl. Biomech. 14, 292–299. Thelen, D.G., Schultz, A.B., Alexander, N.B., Ashton-Miller, J.A., 1996. Effects of age on rapid ankle torque development. J. Gerontol. 51A, M226–M232. Thomas, J.R., Salazar, W., Landers, D.M., 1991. What is missing in p less than 0.5? Effect size. Res. Quart. Exercise Sport 62, 344– 348. Townsend, M.A., Lainhart, S.P., Shiavi, R., 1978. Variability and biomechanics of synergy patterns of some lower-limb muscles during ascending and descending stairs and level walking. Med. Biol. Eng. Comput. 16, 681–688. Whipple, R.H., Wolfson, L.I., Amerman, P.M., 1987. The relationship of knee and ankle weakness to falls in nursing home residents: an isokinetic study. J. Am. Geriatr. Soc. 35, 13–20. Winter, D.A., 1983. Moments of force and mechanical power in jogging. J. Biomech. 16, 91–97. Woolley, S.M., Czaja, S.J., Drurry, C.G., 1997. An assessment of falls in elderly men and women. J. Gerontol.: Biol. Sci. Med. Sci. 52A, M80–M87. Zhang, L-Q., Nuber, G., Butler, J., Bowen, M., Rymer, W.Z., 1998. In vivo human knee joint dynamic properties as functions of muscle contraction and joint position. J. Biomech. 31, 71– 76.