The Journals of Gerontology Series A: Biological Sciences and Medical Sciences 58:M446-M452 (2003)
© 2003 The Gerontological Society of America
Acceleration Patterns of the Head and Pelvis When Walking Are Associated With Risk of Falling in Community-Dwelling Older People
Hylton B. Menz,
Stephen R. Lord and
Richard C. Fitzpatrick
Prince of Wales Medical Research Institute, University of New South Wales, Australia.
 |
Abstract
|
|---|
Background. A large proportion of falls in older people occur when walking, however the mechanisms underlying impaired balance during gait are poorly understood. This study evaluated acceleration patterns of the head and pelvis when walking on a level and an unpredictably irregular surface to determine whether older people at risk of falling demonstrate an impaired ability to stabilize the body under challenging conditions.
Methods. One hundred community-dwelling older people aged between 75 and 93 years were evaluated for their risk of falling using a range of physiological tests previously found to be accurate predictors of falling in prospective studies. Temporo-spatial gait parameters and acceleration patterns at the head and pelvis were then measured in three orthogonal planes while subjects walked on a flat corridor and an unpredictably irregular walkway. Harmonic ratios of head and pelvis accelerations in each plane were calculated to provide an indicator of stability.
Results. Subjects with a high risk of falling exhibited reduced temporo-spatial gait parameters and increased step timing variability. Harmonic ratios of acceleration patterns were reduced at the head and pelvis in the vertical and antero-posterior directions. These differences were particularly evident when walking on the irregular surface.
Conclusion. Older people at risk of falling adopt a more conservative basic walking pattern, but this does not ensure that the movements of the head and pelvis are stable. The irregular pelvis and head accelerations evident in the high risk group suggests that these subjects may have difficulty controlling trunk motion and maintaining a stable visual field when walking, particularly on irregular terrain.
UP to 70% of falls in older people occur when they are walking (1,2). In response to this observation, a number of investigations have been undertaken to determine how gait patterns in older people who fall differ from those who do not fall. These investigations have consistently shown that fallers walk more slowly (3,4) and exhibit greater variability in cadence (5,6) than nonfallers. However, these temporospatial parameters provide only indirect measures of stability, and therefore provide limited insights into the underlying mechanisms responsible for loss of balance when walking.
Sophisticated gait analysis techniques have been used to provide further clarification of this issue in a number of recent investigations. Winter and colleagues (7) have proposed an inverted pendulum model of gait in which the control of the forward trunk displacement is dependent on the generation of a counteracting hip joint force. Using this model, older subjects were found to have a smaller "index of dynamic balance" than younger subjects, representing an impaired ability to control the trunk. Similarly, Yack and Berger (8) assessed acceleration patterns at the level of the thoracic spine, and they reported that older subjects with self-reported balance problems exhibited less smooth acceleration patterns in the vertical and anteroposterior direction than those without balance problems. Contrary to these findings, Simoneau and Krebs (9) reported that whole body momentum during gait (calculated by use of an 11-segment kinematic model) did not differ between fallers and nonfallers; however, because of its small sample size (5 subjects), this study may not have been sufficiently powered to detect significant differences.
There are a number of limitations of the available literature pertaining to gait stability as an indicator of falls risk in older people. First, most sophisticated gait analysis techniques require the use of complex instrumentation, and therefore they are often undertaken on small numbers of subjects. This presents a major problem when older people are assessed, as there is considerable between-subject variability in walking patterns with advancing age. Second, the categorization of subjects into "faller" and "nonfaller" groups is often based on subjective information such as falls history or self-reporting of balance problems. Cummings and colleagues (10) have previously shown that up to 30% of older people with a confirmed history of falls did not recall falling, and subjective reports of balance ability in older people are only moderately associated with objective measures of balance performance (11). Finally, most gait studies evaluate subjects under controlled laboratory conditions that may not be representative of the circumstances likely to impair balance in real life.
In response to these limitations, we have developed a new technique for assessing stability when subjects are walking. This technique uses accelerometers placed on the pelvis and head to detect body accelerations when subjects are walking on a specially designed unpredictably irregular walkway (see Figure 1). We previously reported that, in young healthy subjects, walking on the irregular surface compared with a flat corridor led to larger accelerations at the pelvis; however, head accelerations remained relatively stable (12). This finding suggests that control of head movements is an important role of the body's postural control system when walking, in order to maintain a stable visual field and optimize conditions for the vestibular apparatus. Consistent with this suggestion, previous studies have indicated that older people and patients with vestibular impairment have difficulty controlling head movements during gait (13,14) and often suffer from oscillopsia (excessive movement of the visual field) when walking (15,16).
In this study, we expand on our previous work to determine whether gait alterations are associated with physiological falls risk in a large sample of community-dwelling older people. We hypothesized that older people with an elevated risk of falling would exhibit more irregular acceleration patterns than low-risk older people, despite the adoption of a reduced walking speed.
|
Methods
|
|---|
Subjects
One hundred older people (32 men and 68 women) aged 75 to 93 years (mean 79.9; SD = 4.0) were recruited from the community as part of a larger randomized controlled trial of tailored fall-prevention strategies. Names and addresses of people over the age of 75 years were randomly drawn from a membership database of a private health insurance company. These subjects were initially contacted by letter and asked to participate in the study. Subjects were then contacted by telephone and invited to the Falls and Balance Laboratory at the Royal North Shore teaching hospital in Sydney, Australia. Subjects were excluded from the study if they had Parkinson's disease or a Short Portable Mental Status Questionnaire score less than 7 (17). Transport was provided for those who could not make their own way to the hospital in order to maximize the participation rate of older people with mobility limitations. The prevalence of major medical conditions, medication use, physical activity, mobility, and activity of daily living (ADL) limitations is shown in Table 1. The Human Studies Ethics Committee at the University of New South Wales gave approval for this study, and informed consent was obtained from all subjects prior to their participation.
View this table:
[in this window]
[in a new window]
|
Table 1. Prevalence of Major Medical Conditions, Medication Use, Participation in Physical Activity, and Mobility and ADL Limitations in the Sample Group.
|
|
Assessment of Risk of Falling
All subjects underwent tests of vision, peripheral sensation, strength, reaction time, and balance. Details of the test procedures, and their test–retest reliability, are reported elsewhere (18,19). Performances on these tests were entered into a software program that generates an overall falls risk score, which is a single index derived from a discriminant function analysis from large prospective studies. This combination of tests correctly classifies subjects as fallers or nonfallers with an accuracy of 75–80% (20).
Gait Analysis
Linear accelerations of the body were measured along three orthogonal axes (vertical, anteroposterior, and mediolateral) by use of two triaxial piezoresistant accelerometers, one enclosed in a helmet placed on the head and the other firmly strapped onto the subject with a belt at the level of the sacrum. All subjects were provided with appropriately sized Oxford-style lace-up shoes. Subjects were then instructed to walk along a 20-m long by 1.5-m wide walkway consisting of 5-mm-pile artificial grass underlain with two layers of 20-mm-thick soft foam rubber and 20-mm-thick wooden blocks of varying sizes and shapes in an arbitrary manner. Two markings were made on the walkway 15 m apart to designate the trial distance. Subjects were also tested on a level corridor. Two trials of each condition were performed in a randomized order.
Full descriptions of the data-processing protocol, the derivation of acceleration variables, and test–retest reliability have been described previously (12). The following variables were calculated from the acceleration signals: (i) walking velocity (meters per second), (ii) cadence (steps per minute), (iii) average step length (centimeters), (iv) step timing variability, and (v) harmonic ratio of acceleration signals.
The harmonic ratio provides an indicator of the degree of rhythm of the acceleration signal. The basic underlying premise of this technique is that the unit of measurement from a continuous walking trial is a stride (two steps). A stable, rhythmic gait pattern should therefore consist of acceleration patterns that repeat in multiples of two within any given stride, as these patterns are therefore "completed" prior to taking subsequent strides. Accelerations patterns that do not repeat in multiples of two are problematic, as they produce out-of-phase accelerations that are not completed within each stride, and therefore manifest as irregular accelerations during a walking trial. Briefly, the technique involves decomposing the acceleration signal into individual harmonics by use of a finite Fourier transform. The summed amplitudes of the even-numbered harmonics are then divided by the summed amplitudes of the odd-numbered harmonics, providing a harmonic ratio. Higher ratios represent a more stable walking pattern, as a greater proportion of the acceleration signal is "in phase" with the subject's stride frequency (12).
Statistical Analysis
All analyses were performed with SPSS Release 10 for Windows (SPSS Inc., Chicago, IL). To determine mean differences in subject characteristics according to falls risk group, a series of one-way analyses of variance (ANOVAs) were used for age, height, and weight. A series of two-way repeated measures ANOVAs were performed on the gait variables with falls risk group (low, moderate, or high) as the between-subjects factor and walking surface (level or irregular) as the within-subjects factor. Following the determination of a significant main effect of falls risk group, a series of one-way repeated measures ANOVAs were used to assess for differences between gait parameters according to the falls risk group for each surface. Post hoc comparisons among the three falls risk groups were performed with the Hochberg post hoc test to adjust for the multiple comparisons.
|
Results
|
|---|
Falls Risk
The falls risk score ranged from -0.66 to 5.56 (mean 1.25; SD = 1.04). As the falls risk score was normally distributed, subjects were divided into three falls risk groups (based on tertiles) for subsequent analysis—low, moderate, and high. The mean and standard deviations for the falls risk score in each of these groups was as follows: low, n = 34, mean = 0.24, and SD = 0.37; moderate, n = 33, mean = 1.10, and SD = 0.32; and high, n = 33, mean = 2.45, and SD = 0.72. There was no difference in the proportion of women and men in each of the falls risk groups (
2 = 0.79, df = 2, and p =.637). The mean (SD) age for each of the three falls risk groups was 78.94 (3.44), 79.15 (3.83), and 81.48 (4.45) for the low-, moderate-, and high-risk groups, respectively. Subjects in the high-risk group were older than those in the moderate-risk group, with F(1,2) = 4.30 and p =.016. There were no significant differences in height, with F(1,2) = 0.07 and p =.931, or weight, with F(1,2) = 0.209 and p =.812, among the three falls risk groups. In the 12 months prior to the study, 38 subjects had at least one fall, and 10 had multiple falls. Multiple fallers had a significantly higher falls risk score than nonfallers or once-only fallers (t98 = -2.11; p =.037).
Differences in Temporospatial Gait Parameters Between Falls Risk Groups
High-risk fallers walked significantly more slowly and took shorter steps than low-risk fallers on the level surface, and both low- and moderate-risk fallers on the irregular surface. High-risk fallers also walked with a significantly slower cadence and greater step-timing variability than both moderate- and low-risk fallers on the irregular surface (see Table 2).
Differences in Harmonic Ratios Between Falls Risk Groups
At the level of the pelvis, high-risk fallers exhibited significantly smaller vertical harmonic ratios compared with low-risk fallers on both surfaces, and smaller anteroposterior and mediolateral harmonic ratios compared with low-risk fallers on the level surface (see Table 3). At the level of the head, high-risk fallers exhibited significantly smaller vertical and anteroposterior harmonic ratios compared with low risk fallers on both surfaces. Harmonic ratios at the head in the mediolateral direction were not influenced by falls risk group on either walking surface (see Table 4).
Figure 2 shows an accelerometry recording at the head from a low-risk subject and a high-risk subject. As can be noted from the figure, the high-risk subject exhibits smaller magnitude accelerations (as the subject is walking more slowly); however, the acceleration signal is less rhythmic. Figure 3 shows a harmonic analysis of acceleration signals at the head in the vertical and anteroposterior directions from a low-risk subject and a high-risk subject, highlighting the larger amplitude high-frequency components of the signal in the high-risk subject.
View larger version (24K):
[in this window]
[in a new window]
|
Figure 2. Accelerometry recording at the head from a low risk subject (left) and a high risk subject (right). Each recording represents 10 seconds of each trial. Scale in units of gravity (g)
|
|
View larger version (19K):
[in this window]
[in a new window]
|
Figure 3. Harmonic analysis of accelerometry signals at the head in the vertical and anteroposterior directions from a low-risk subject (left) and a high-risk subject (right), normalized to the second harmonic
|
|
|
Discussion
|
|---|
Falls that occur during ambulation account for up to 70% of all falls (1,2). Clearly, the task of maintaining balance during locomotion is a challenging one for the ageing postural control system. Although a number of investigations have reported that a range of temporospatial, kinematic, and kinetic variables differ between fallers and nonfallers (3–6,8,21,22), determining how these variables contribute to an overall pattern of movement has been difficult. In an attempt to simplify the myriad of movement patterns that may predispose a subject to falls, we have assumed that any abnormalities in the rhythmic movements of the legs or stability of the trunk will manifest as aberrant acceleration patterns at the head, pelvis, or both. Furthermore, the model we have used incorporates an unpredictably irregular surface obscured from the subject's view, as opposed to the discrete obstacles used in previous studies (23,24). This approach was taken to accurately represent many environmental conditions likely to impair stability, and to prevent subjects from developing anticipatory motor control strategies to avoid individual obstacles.
Previously, we reported that when healthy young subjects were walking on an unpredictably irregular surface, their gait patterns differed only slightly from those when walking on a level surface (12), and this suggested that the main motor control pattern adopted by healthy young subjects to stabilize the body was an increase in step length, in an attempt to minimize the number of foot contacts over a given distance. In contrast, the older subjects in the current study significantly reduced their velocity and step length when walking on the irregular surface. These results suggest that older people adopt a more conservative, cautious gait pattern than younger people, possibly in an attempt to minimize the displacement of the upper body. The acceleration data provide useful insights into the effect that these gait alterations have on the stability of the head and trunk. The vertical and anteroposterior harmonic ratios at the pelvis and head were significantly smaller in older subjects with higher physiological falls risk. This suggests that older people at risk of falling have difficulty controlling the rhythmic displacements of the trunk that occur during walking. Despite some differences in methodology, similar results were reported by Yack and Berger (8), who found that the harmonic ratio of vertical accelerations at the thoracic spine was smaller in 10 subjects with self-reported balance problems compared with 10 subjects without balance problems and 10 young control subjects. Our results confirm these preliminary findings in a much larger sample, using a more objective measure of physiological falls risk. Furthermore, in the current study, the differences in acceleration patterns according to falls risk were particularly evident when subjects were walking on the irregular surface—a more challenging test condition representative of the circumstances likely to predispose to falls in real life.
Upper-body instability when subjects are walking on an irregular surface has important ramifications for head control and gaze stabilization. It has previously been shown that the magnitude and frequency of head movements when subjects are walking are sufficient to activate vestibular reflexes, and that patients with bilateral vestibular deficits have difficulty controlling head and eye movements when they are walking (14,15). The possible relationship between poor head control and falls in older people has also been suggested by Wu (25), who showed that older people exhibit much larger head movements in response to sudden anteroposterior platform perturbations than young people, suggesting a slowing of the cervicocollic reflex response with advancing age. In the absence of adequate head control, the ability of the vestibular system to maintain a stable visual field is compromised. Impaired head control and gaze stabilization may play a role in increasing falls risk. Older people who fall have been shown to place greater reliance on visual input (26) to compensate for age- or disease-related impairments in the somatosensory or vestibular systems. It is likely that the excessive reliance on inaccurate visual information regarding the orientation of the body in relation to its surroundings will limit an older person's ability to respond adequately to unexpected perturbations during gait.
Conclusions
This investigation provides clear evidence of stability-related gait variables strongly associated with increased risk of falling in a large sample of older people. Despite walking more slowly, older people with a high risk of falling exhibited more variable step timing, and less rhythmic acceleration patterns of the pelvis and head. Furthermore, the associations between these variables and risk of falling were particularly evident when the subjects were walking on an unpredictably irregular surface. These findings suggest that older people with an elevated risk of falling may have difficulty controlling the movement of their trunk and maintaining a stable visual field when they are walking on irregular terrain.
|
Acknowledgments
|
|---|
This project was funded by the National Health and Medical Research Council Health Research Partnerships Grant, Prevention of Injuries in Older People.
We acknowledge Susan Murray, Kirsten Chapman, Bridget Munro, and Anne Tiedemann for conducting the physiological tests, and Dr. Susan Ogle, Head of the Aged Care and Rehabilitation Department, for the use of the testing facilities at the Royal North Shore Hospital.
Address correspondence to Hylton B. Menz, PhD, Prince of Wales Medical Research Institute, University of New South Wales, High Street, Randwick, New South Wales 2031, Australia. E-mail: h.menz{at}unsw.ed.au
Received August 24, 2002
Accepted September 16, 2002
|
References
|
|---|
- Cali CM, Kiel DP. An epidemiologic study of fall-related fractures among institutionalized older people. J Am Geriatr Soc.. 1995;43:1336-1340.[Medline]
- Norton R, Campbell AJ, Lee-Joe T, Robinson E, Butler M. Circumstances of falls resulting in hip fractures among older people. J Am Geriatr Soc.. 1997;45:1108-1112.[Medline]
- Gehlsen GM, Whaley MH. Falls in the elderly: Part I, gait. Arch Phys Med Rehabil.. 1990;71:735-738.[Medline]
- Wolfson L, Whipple R, Amerman P, Tobin JN. Gait assessment in the elderly: a gait abnormality rating scale and its relation to falls. J Gerontol Med Sci.. 1990;45:M12-M19.
- Lord SR, Lloyd DG, Li SK. Sensori-motor function, gait patterns and falls in community-dwelling women. Age Ageing.. 1996;25:292-299.[Abstract/Free Full Text]
- Hausdorff J, Rios D, Edelberg H. Gait variability and fall risk in community-living older adults: a 1-year prospective study. Arch Phys Med Rehabil.. 2001;82:1050-1056.[Medline]
- Winter DA, Patla AE, Frank JS, Walt SE. Biomechanical walking pattern changes in the fit and healthy elderly. Phys Ther.. 1990;70:340-347.[Abstract/Free Full Text]
- Yack HJ, Berger RC. Dynamic stability in the elderly: identifying a possible measure. J Gerontol Med Sci.. 1993;48:M225-M230.
- Simoneau GG, Krebs DE. Whole-body momentum during gait: a preliminary study of non-fallers and frequent fallers. J Appl Biomech.. 2000;16:1-13.
- Cummings SR, Nevitt MC, Kidd S. Forgetting falls—the limited accuracy of recall of falls in the elderly. J Am Geriatr Soc.. 1988;36:613-616.[Medline]
- Myers AM, Powell LE, Maki BE, Holliday PJ, Brawley LR, Sherk W. Psychological indicators of balance confidence: relationship to actual and perceived abilities. J Gerontol Med Sci.. 1996;51A:M37-M43.[Abstract]
- Menz HB, Lord SR, Fitzpatrick RC. Acceleration patterns of the head and pelvis when walking on level and irregular surfaces. Gait Posture. In press.
- Hirasaki E, Kubo T, Nozawa S, Matano S, Matsunuga T. Analysis of head and body movements of elderly people during locomotion. Acta Otolaryngol.. 1993;501:25-30.
- Takahashi M, Hoshikawa H, Tsujita N, Akiyama I. Effects of labyrinthine dysfunction upon head oscillation and gaze during stepping and running. Acta Otolaryngol.. 1988;106:348-353.[Medline]
- Hillman EJ, Bloomberg JJ, McDonald PV, Cohen HS. Dynamic visual acuity while walking in normals and labyrinthine-deficient patients. J Vestib Res.. 1996;6:295-313.[Medline]
- Schubert MC, Herdman SJ, Tusa RJ. Vertical dynamic visual acuity in normal subjects and patients with vestibular hypofunction. Otol Neurotol.. 2002;23:372-377.[Medline]
- Pfieffer E. A short portable mental status questionnaire for the assessment of organic brain deficit in elderly patients. J Am Geriatr Soc.. 1975;23:433-441.[Medline]
- Lord SR, Clark RD, Webster IW. Postural stability and associated physiological factors in a population of aged persons. J Gerontol Med Sci.. 1991;46:M69-M76.
- Lord SR, Ward JA, Williams P. Exercise effect on dynamic stability in older women: a randomised controlled trial. Arch Phys Med Rehabil.. 1996;77:232-236.[Medline]
- Lord SR, Ward JA, Williams P, Anstey K. Physiological factors associated with falls in older community-dwelling women. J Am Geriatr Soc.. 1994;42:1110-1117.[Medline]
- Kerrigan DC, Lee LW, Nieto TJ, Markman JD, Collins JJ, Riley PO. Kinetic alterations independent of walking speed in elderly fallers. Arch Phys Med Rehabil.. 2000;81:730-735.[Medline]
- Kerrigan DC, Lee LW, Collins JJ, Riley PO, Lipsitz LA. Reduced hip extension during walking: healthy elderly and fallers versus young adults. Arch Phys Med Rehabil.. 2001;82:26-30.[Medline]
- Chen H-C, Ashton-Miller JA, Alexander NB, Schultz AB. Stepping over obstacles: gait patterns of healthy young and old adults. J Gerontol Med Sci.. 1991;46:M196-M203.
- Begg R, Sparrow W. Gait characteristics of young and older individuals negotiating a raised surface: implications for the prevention of falls. J Gerontol Med Sci.. 2000;55A:M147-M154.[Abstract/Free Full Text]
- Wu G. Head movement during sudden base translations as a measure of risks for falls in the elderly. Clin Biomech.. 2001;16:199-206.
- Lord SR, Webster IW. Visual field dependence in elderly fallers and non-fallers. Int J Aging Hum Devel.. 1990;31:269-279.
This article has been cited by other articles:
|
|
|
|
 
D. S. Marigold and A. E. Patla
Adapting Locomotion to Different Surface Compliances: Neuromuscular Responses and Changes in Movement Dynamics
J Neurophysiol,
September 1, 2005;
94(3):
1733 - 1750.
[Abstract]
[Full Text]
[PDF]
|
|
|
Top
Abstract
Methods
