A comparison of gait characteristics in young and old subjects.

Many studies have compared the gait patterns of young and old people.[1-6] The gait characteristics investigated included temporal and distance factors along with specific joint range-of-motion (ROM) measurements during gait,[1-5,7,8] or temporal and distance factors alone.[6,9]

In general, most investigators have found little difference in the joint ROM of the hip and knee between younger and older people.[1-4,8,10-12] Differences in total ankle excursion between young and old people during gait have been noted.[4,5,8] The difference in total ankle excursion noted between young and old people ranged from 2 to 7 degrees, with a specific decrease in plantar flexion at the end of the stance phase noted for older men compared with younger men.

Differences between young and old individuals in distance characteristics of gait have been more consistently reported. Comparisons of stride length measures of younger and older people indicate stride length decreases with age.[1-6] For younger subjects, stride lengths averaged 151 to 170 cm, whereas the stride lengths of older subjects averaged 135 to 153 CM.[1,3,5,6]

In contrast, absolute temporal characteristics of gait do not appear to differ with age. No significant differences in stride time values for young and old adults have been reported, with a mean for adults approximately equal to 1 second.[1-4,9]

Distance per unit of time measures of gait do appear to differ with aging. Older adults exhibit decreased gait velocity compared with younger adults, with gait velocity ranging from 143 to 160 cm/s in younger adults compared with 118 to 145 cm/s in older adults.[2-5] Differences in walking speed between young and old persons may affect the comparison of temporal and distance factors and joint excursions.[2,6,13-15] For example, longer stride lengths and greater hip, knee, and ankle angles have been noted with faster speeds.[6]

The purpose of this investigation was to describe the differences in nine gait characteristics (Tab. 1) between younger and older people in good health. Peak flexion and extension at the hip, knee, and ankle in the sagittal plane during free speed ambulation in younger people was compared to the peak flexion and extension of these joints during free speed ambulation in older people. Stride length and time, and velocity were also compared between groups.

[TABULAR DATA 1 OMITTED]

Method

Older and younger adults participated in a multivariate comparison study of gait characteristics at self-selected walking speeds,

Subjects

Sixty volunteers in good health-30 (15 male, 15 female) between the ages of 20 and 40 years and 30 (15 male, 15 female) between the ages of 60 and 80 years--were accepted as subjects for the gait analysis. Young subjects were a mean 28.2 years of age, 172.1 cm in height, and 64.9 kg in weight. Older subjects were a mean 67.4 years of age, 170 cm in height, and 72.1 kg in weight. A description of the subjects is shown in Table 2. Many of the subjects were employees or volunteers at Harmarville Rehabilitation Center, Pittsburgh, Pa. Each subject provided informed consent to undergo the screening and videotape analysis of walking.

[TABULAR DATA 2 OMITTED]

The health status of each subject was determined by a screening evaluation (Tab. 3) performed by a licensed physical therapist. The purposes of the screening evaluation was to ensure that subjects were in good health and had sufficient mobility and strength for gait. Such screening was believed to be essential to enable a more valid comparison of the gait characteristics studied in two different age groups.

Table 3. Variables Included in
Screening by the Physical Therapist
Medical history
Height
Weight
Posture
Trunk flexibility
Leg flexibility
Lower-extremity flexibility
Lower-extremity strength

All subjects participating were without appreciable leg-length discrepancy ([greater than or equal to] 0.5 cm), demonstrated freedom of movement in all planes of trunk motion, and complained of no functional limitations in trunk movements. Subjects demonstrated the mean lower-extremity ROM used in gait at the hip, knee, and ankle as reported by Murray and colleagues.[1,4,16] All subjects exhibited Good to Normal muscle strength throughout both lower extremities based on standard manual muscle testing.[17]

Procedure

Each subject was scheduled for one 45-minute session, which included the screening evaluation and videotaping. All screening and videotaping were performed in the outpatient physical therapy area at Harmarville Rehabilitation Center by the principal investigator (KMO). All subjects were required to wear shorts, sleeveless or shortsleeve shirts, and flat-sole shoes. Following successful completion of the screening, subjects were prepared for data collection. Reflective markers were placed at six locations along the right side of each subject: (1) iliac crest, (2) greater trochanter, (3) middle of the knee joint, (4) approximate center of mass of the lower leg and foot, (5) heel, and (6) fifth metatarsophalangeal joint (Fig. 1). All subjects were standing when the markers were placed.

The iliac crest marker was placed so as to reflect the angle of the pelvis on the femur (pelvic angle). The iliac crest marker was placed along a line cortgecting the anterior superior iliac spine and the posterior superior iliac spine at a point at which a perpendicular line passing through the greater trochanter intercepted. If an anterior pelvic tilt was present, the iliac crest marker placement was anterior to the greater trochanter, and if a posterior pelvic tilt was present, the iliac crest marker placement was posterior to the greater trochanter (Fig. 2). With the knee in extension, the knee marker was placed at the midpoint of a line connecting the mid-popliteal fold and the middle of the body of the patella.

The center of mass of the lower leg and foot was estimated to be 60% of the distance between the knee marker and the lateral malleolus, measured from the knee. The center of mass of the lower leg and foot marker was used instead of a marker at the lateral malleolus because paths created by the ankle and foot markers crossed over one another making it difficult to discern the individual positions of the markers during analysis. The marker for the center of mass of the lower leg was placed in line with the knee marker and the lateral malleolus.

The heel and metatarsal markers were placed directly on the outside of the subject's shoe, The heel marker was placed at the posterolateral base of the heel, and the metatarsophalangeal marker was placed at the site of the fifth metatarsophalangeal joint. The latter site was determined when the subject was asked to lift the heel, flexing the foot at the metatarsophalangeal joints.

Subjects walked along a 6-m walkway at a self-selected speed. Videotaping was done from the right side only. Subjects received initial coaching and practice to determine which foot to begin walking with in order for one complete gait cycle to be videotaped. The area videotaped included approximately 2.7 m of the walkway, allowing for only one complete gait cycle for most subjects. At least three walks were filmed for each subject to ensure one walk with adequate marker visibility. The walk with the best visibility of markers was chosen.

Instrumentation

Gait analysis was performed using the ExpertVision[TM] Motion Analysis[TM] system.(*) This integrated computervideo system is tailored for motion analysis and consists of (1) an NEC TI-23A CCD video camera[dagger] with a 12.5-mm television lens on a tripod approximately 81 cm above the floor, (2) a high-intensity light[double dagger] mounted parallel to the video camera, (3) a Panasonic Model TR-124MA video monitor,[subsections] (4) a Panasonic Model AG-6300 VHS player,[sections] (5) a VP-110 video processor,(*) and (6) an IBM AT computer[parallel] with display terminal and mouse.

A tripod with the video camera and high-intensity light were positioned perpendicular to the 6-m walkway at a distance of 6.7 m. Behind the subject, a curtain served as background. Fluorescent lighting was used in the walkway area. Subjects wore dark clothing when possible to avoid reflections that would be digitized by the video processor, making analysis difficult.

The video camera recorded the motion image and the reflective images from the 2-cm diameter reflective markers placed at six positions along the subject's right side. The video processor produced digital outlines of the reflective markers at 60 frames per second, which were fed into the computer. "Centroids" (representing the geometric center of the marker) were calculated by the computer from the digitized images and connected to form "paths" for each marker.

A distance scale factor was entered into the computer for each subject. The investigator drew a line joining the hip and knee centroids on the computer monitor, and the actual distance (in centimeters) between the hip and knee markers measured on the subject was equated with the number of pixels (the computer unit of distance) between these two points. The computer recorded the events of right heel-strike numerically and graphically. The points of heel-strike were used in calculating stride length, stride time, and walking speed.

Two user programs assisted in analyzing the heel path to note the events of right heel-strike that determine the limits of the gait cycle. A user program was written to view the horizontal position of the heel marker versus time. First, a numerical listing of heel-position points, time, and frame numbers was analyzed. Plateaus in this listing indicated stance phase, when the heel was not moving. The position of the heel at the beginning of each plateau was determined and entered into a second user program, which created a graph of the heel position versus time (Fig. 3). The two heel-strike events were marked on the graph for visual verification that the events occurred at the beginning of the plateaus. The points of heel-strike were important in calculating the temporal and distance factors of gait, such as stride distance, stride time, and walking speed.

The maximum and minimum joint angles at the hip, knee, and ankle during one gait cycle were derived from the data input to the computer Fig. 4). The hip joint excursion is equal to 180 degrees minus the angle formed by the lines connecting the greater trochanter to the iliac crest and the greater trochanter to the knee marker (reading clockwise). Positive values represent hip flexion, and negative values represent hip extension. Knee joint excursion is equal to the angle formed by the lines connecting the greater trochanter to the knee joint and the greater trochanter to the center of mass of the lower leg (reading clockwise) minus a constant of 180 degrees. Positive values represent knee flexion, and negative values represent knee hyperextension, Ankle joint excursion is represented by the angle between a line connecting the heel and fifth metatarsophalangeal joint and the neutral position of the foot (determined by drawing a line perpendicular to the line connecting knee and leg markers). A positive angle represents plantar flexion, and a negative angle represents dorsiflexion.

Reliability

One walk of six subjects in each group was reanalyzed to determine test-retest reliability of the method of data recording. The intraclass correlation coefficients (type 2,1) for repeat analysis of the nine gait variables measured for the same walk were r=.92 or greater (Tab. 4).[18] The computer analysis of the gait characteristics recorded was consistent with the repeat analysis. The reliability testing of marker placement and videotaping was not done as part of this study, which could account for some errors in data collection.

Table 4. Interclass Correlation
Coefficients ([ICC]2,1)(a) for Reliability
Study
Gait Characteristic              ICC
Hip extension                    .996
Hip flexion                      .990
Knee extension                   .994
Knee flexion                     .995
Ankle dorsiflexion               .958
Ankle plantar flexion            .995
Stride length                    .956
Stride time                      .923
Velocity                         .969
(*) The ICC (2,1) determines the correlation
between initial and repeat analyses of the
same walk. Reliability study based on reanalysis
of 6 subjects in each group (total of 12
subjects).

Data Analysis

The means, standard deviations, and 95% confidence intervals were calculated for the nine gait characteristics in each group (Tabs. 5, 6). These values were used for comparison between the two groups using a multivariate analysis of variance (MANOVA). The specific MANOVA, the Hotelling's [T.sup.2] multivariate test of significance between groups, was calculated. Initially, the Levene test was run to determine whether the group data met the assumptions of the statistical test. Univariate F tests were then calculated for the nine gait variables. The acceptable level of significance was determined by dividing the alpha level .05 by the number of variables to obtain an adjusted probability value (ie, P=.05/9 variables=.006).

[TABULAR DATA 5 OMITTED]

Table 6. 95% Confidence Intervals
Gait Cheracteristic                 Young People           Older People
Hip extension (degree])                 13-7                   11-6
Hip flexion(degree])                    23-29                  26-32
Knee extension (degree])                 1-4                    5-8
Knee flexion (degree])                  65-68                  67-71
Ankle dorsiflexion (degree])            14-10                  15-13
Ankle plantar flexion (degree])         25-31                  21-26
Stride length (cm)                      147.56-156.48          136.72-146.89
Stride time (s)                         1.08-1.13              1.09-1.17
Velocity (cm/s)                         132.81-143.16          121.04-133.01

Results

The results of this multivariate comparison of gait in young and old people indicated the groups differed in their gait performance based on the nine gait variables studied (F=4.44; df=9,50; P<.001). Two of the gait characteristics--peak knee extension ROM and stride length--were independently different between the young and old groups (F= 15.23; df= 1,58; P<.006 and F=9.55; df= 1,58; P<.006, respectively). Peak knee extension and stride length were significantly less in the older persons compared with the younger persons. The difference in walking speed between groups approached the .006 level of significance. Older persons exhibited a slower walking speed than did younger persons.

Discussion

The findings of this multivariate comparison of gait in young and old people indicated peak knee extension ROM and stride length differed between the groups. Stride time values were very similar between younger and older people in our study younger group: X = 1. 1 1, SD = 0.07 seconds; older group: X = 1. 13, SD = 0.11 seconds). Stride time values in our study were consistent with those reported for other studies.[1,3,4,9] The results of previous investigations and of our study support the notion that older people take smaller and slower steps compared with younger people. The result is an almost equal stride time for each group, but a slower speed for the elderly.

In our study, peak hip flexion was slightly increased in the older group compared with the younger group, although this finding was not statistically significant. Murray et all and Finley et al[8] also found peak hip flexion to be slightly increased in older people compared with younger people. Slight increases in hip flexion during gait in older people may be an unconscious precaution against tripping or an attempt to maintain stride length.

Hip extension values in our study were similar to those reported by Murray and colleagues[1,4] and Finley et al.[8] Finley et al reported hip extension in younger and older people to be 9 degrees, whereas in our study hip extension was 10 degrees (SD=9[degrees]) in the younger people and 8 degrees (SD=6[degrees]) in the older people. Murray and colleagues[1,4] report hip extension in healthy men and women to be approximately 10 degrees.

Knee extension in our study was less in the older group (A = 7[degrees], SD = 4[degrees]) than in the younger group (X=30, SD=4[degrees]). Murray and colleagues[1,3] found knee extension to be almost identical in various age groups. During the initial screening for our study, several older people exhibited slight knee flexion with relaxed standing, whereas younger people were more likely to exhibit complete knee extension or hyperextension. Young people may exhibit greater ligament laxity at the knee, whereas older people may have an increased stiffness of soft tissue structures at the knee. Another possibility is that older people may exhibit more knee flexion throughout the gait cycle to unconsciously lower their center of gravity.

Knee extension in our study was less in both age groups than in other studies[1,4,8,11] Differences in knee motion during gait among studies could also be due to measurement methods, including marker placement and calculation of the knee angle. For example, Murray and colleagues[1,3] manually measured knee motion from the image of reflective markers on a photograph.

Knee flexion in our study was similar to that found by Winter et al.[11] Winter et al found knee flexion to be 67 degrees (SD=8[degrees]), whereas in our study knee flexion was 66 degrees (SD=4[degrees]) in the younger group and 69 degrees (SD=5[degrees]) in the older group. Murray et al3 found knee flexion to be greater in younger men than in older men. Knee flexion values were greater in our study compared with other studies.[4,11,10,12] This finding may be an indication of a healthier, more active sample in our study, or it may reflect differences in measurement of knee motion during gait among studies,

Ankle dorsiflexion was increased slightly in older people and ankle plantar flexion was decreased in older people compared with younger people in our study. Similarly, Murray and colleagues[1,3] reported decreased ankle plantar flexion and increased ankle dorsiflexion in older men compared with younger men.

Dorsiflexion and plantar-flexion ROMs were larger in our study for both younger and older people than in at least three other studies reviewed.[1,5,8] In the study by Finley et al,[8] subjects walked very slowly encumbered by measuring equipment, which could account for decreased ankle motion during gait, In the two other studies,[1,5] different ankle excursions reported may be due to the angle used to represent ankle motion. Many subjects in our study, however, also exhibited considerable ankle ROM when measured goniometrically during screening and indicated being active in regular exercise including walking, running, and golfing.

Some of the differences between our findings of gait changes in older people and the findings reported in previous investigations may be due to the differences in the health status of the older people studied. That is, some gait changes found in older people in previous studies may have been the result of pathology, not just aging. Extensive prescreening (Tab. 3) resulted in the subjects accepted in our study being free from pathology such as stroke, head injury, spinal cord injury, poliomyelitis, or arthritis. All subjects had normal flexibility and strength in the lower extremities and normal posture and leg length, Many subjects reported being involved in regular physical activity or exercise, although this was not a criterion for acceptance into the study.

Conclusions

In this multivariate comparison of nine gait characteristics, the groups of young and old people studied differed in their gait performance. The old people had significantly less peak knee extension and shorter stride length than demonstrated by the younger group of people. Based on the results of this study, it may be important to closely evaluate knee extension in the elderly during gait and consider developing treatment strategies to normalize knee exiension during gait to improve an individual's gait pattern or performance. Stride length and speed also appear to be important to measure and address during treatment sessions. Improvement in speed of gait could serve to document the overall efficiency of gait.

In the future, a gait analysis of older people who are physically active versus older people who are sedentary might more clearly define the impressions resulting from our study. A longitudinal study in which the gait of several people is analyzed at various age milestones would also be a meaningful way of determining the effects of aging on gait characteristics. Lastly, intervention studies could be conducted to determine whether improving knee extension, stride length, or walking speed in older adults improves functional gait.

Acknowledgments

We would like to express our thanks to Mrs Pauline Egan for her assistance in locating subjects, Pam Wickemeyer, PT, for willingly posing for photographs, Anne McDermott, PT, for taking the photographs, Jim Ostrosky for his technical assistance with data collection and statistics, and the physical therapy staff at Harmarville Rehabilitation Center Inc for their support. We also give a special thanks to all of our subjects.

References

[1] Murray MP, Drought AB, Kory RC, Walking patterns of normal men. J Bone Joint Surg [Am]. 1964;46:335-360. [2] Murray MP, Kory RC, Clarkson BH, Sepic SB. Comparison of free- and fast-speed walking patterns of normal men. Am J Phys Med. 1966; 45:8-24. [3] Murray MP, Kory RC, Clarkson BH. Walking patterns in healthy old men. J Gerontol. 1969; 24:169-178. [4] Murray MP, Kory RC, Sepic SB. Walking patterns of normal women. Arch Phys Med Rehabil, 1970;51:637-650. [5] Hageman PA, Blanke DJ. Comparison of gait of young women and elderly women. Phys Ther. 1986;66:1382-1387. [6] Larish DD, Martin PE, Mungiole M. Characteristic patterns of gait in the healthy old. Ann NY Acad Sci. 1988;515:18-32. [7] Blanke DJ, Hageman PA. Comparison of gait of young men and elderly men. Phys Ther, 1989;69:144-148. [8] Finley FR, Cody KA, Finizie RV. Locomotion patterns in elderly women. Arch Phys Med Rehabil. 1969;50:140-146. [9] Gabell A, Nayak USL. The effect of age on variability of gait. J Gerontol 1984;39:662-666. [10] Sutherland DH, Hagy JL. Measurement of gait movements from motion, picture film. J Bone Joint Surg [Am]. 1972;54:787-797. [11] Winter DA, Quanbury AO, Hobson DA, et al. Kinematics of normal locomotion: a statistical study based on TV data. J Biomech. 1974;7:479-486. [12] Isacson J, Gransberg L, Knutsson E. Three-dimensional electrogoniometric gait recording. J Biomech. 1986;19:627-635. [13] Larsson LE, Odenrick P, Sandlund B, et al. The phases of the stride and their interaction in human gait, Scand J Rehabil Med 1980;12: 107-112. [14] Lamoreux LW. Kinematic measurements in the study of human walking. Bull Prosthet Res, 1971;3:3-84. [15] Frigo C, Eng D, Tesio L. Speed-dependent variations of lower-limb joint angles during walking. Am J Phys Med. 1986;65:51-62. [16] Murray MP. Gait as a total pattern of movement. Am J Phys Med. 1967;46:290-333. [17] Kendall HO, Kendall FP, Wadsworth GE. Muscles: Testing and Function. 2nd ed. Baltimore, Md: Williams & Wilkins; 1971. [18] Shrout PE, Fleiss JL. Intraclass correlations: uses in assessing rater reliability. Phychol Bull. 1979;86:420-428.

Invited Commentary

Normative studies can provide valuable information to the clinician about performance in the nondisabled population. Ostrosky and colleagues attempted to describe selected gait characteristics in healthy older and younger people. Although the findings of decreased stride length and decreased velocity in the older population when compared with the younger population are consistent with those of earlier reports,[1-3] the most remarkable finding of the study is the lack of difference between the two groups on most of the measured variables. I believe that methodological issues can explain the lack of differences found.

Ostrosky et al examined peak hip, knee, and ankle angles during gait. In this study, the finding of decreased peak knee extension during the gait cycle in the older group (60-80 years of age) is difficult to interpret. The authors did not describe the phase of gait in which each angle was measured. Without knowing the relationship between joint motion and phase of gait, the utility of this finding is difficult to ascertain. This information would be treated differently depending on where the finding of decreased peak knee extension was found within the gait cycle. If the decreased peak knee extension occurred during the swing phase in the older group when compared with the younger group, the information may not be clinically meaningful. The authors report only the finding of decreased peak knee extension during the gait cycle. They do not indicate where this difference in peak angles compared within or between groups occurred, Did the peak knee extension angle compared occur at the same point in the gait cycle for all individuals? The functional ramifications would be different depending on where in the gait cycle peak knee extension occurred.

There are two points in the stance phase where the knee approaches full extension: weight acceptance and terminal stance (push-off) (Figure). If some of the peak knee extension angles compared occurred at weight acceptance and some at push-off, no meaningful inferences can be drawn from the data. The finding of decreased peak knee extension in the older group could provide important information if we knew where in the gait cycle it occurred. If the finding of decreased knee extension in this study occurred at loading response in the older group when compared with the younger group, this is an interesting finding. It may mean that older individuals use a flexed knee gait to increase their ability to accommodate to changes in surface or to increase the stability of this joint. No conclusions can be drawn, however, from the information presented.

I am also concerned about drawing a conclusion when the average difference in peak knee extension was 4 degrees between the groups and the standard deviation was also 4 degrees. Does this statistically significant difference between the groups translate to a clinically meaningful difference?

In previous normative studies of gait,[1-3] the subjects were grouped in 5- to 10-year increments beginning at 20 years of age and ending at 70 years of age, comparing the findings of each decade with those of every other decade. The groupings used in this study may have masked real differences between the older and younger groups. The authors did not group the subjects by decade, but instead compared the means of the 20- to 40-year-old individuals against those of the 60- to 80-year-olds, and the differences were not great. Scatter plots of the data examining each of the nine gait variables by age may have revealed age-related differences.

Groups that spanned 20 years, in which the means of measured variables overlapped to a large extent, may have hidden clinically meaningful findings.

How generalizable are the results? From the description of the methods, the population of subjects appears to be a sample of convenience. The authors did not discuss the normal weekly activity of the subjects. Did this sample include older individuals who do not go out everyday or who lead a sedentary life? How representative are the older individuals in this study of the healthy older population? The septuagenarians and octogenarians who participated in this study may be a select group and may not represent the population as well as would a true random sample. As a clinician trying to interpret and apply these findings, this information is important.

If the findings of this study are biased to a select group of older individuals, then the results of the study may not provide the clinician with truly generalizable information.

Marty Eastlack, PT

Program in Physical Therapy

University of Delaware

303 McKinly Lab

Newark, DE 19716

References

[1] Murray MP, Drought AB, Kory RC. Walking patterns of normal men. J Bone Joint Surg [Am]. 1964;46:335-360, [2] Murray MP, Kory RC, Clarkson BH, Sepic SB. Comparison of free- and fast-speed walking patterns of normal men. Am J Phys Med 1966; 45:8-24. [3] Murray MP, Kory RC, Sepic SB, Walking patterns of normal women. Arch Phys Med Rehabil. 1970;51:637-650.

Author Response

We agree that the most remarkable finding of the study is the lack of difference between the two groups for most of the measured variables. We think it is clinically important and perhaps encouraging to recognize that gait can remain typical of the gait of young adults with advancing age in healthy older people.

A statistically significant difference was found with knee extension throughout the gait cycle in this multivariate analysis of gait characteristics between the two age groups. With reference to at what point in the gait cycle the difference in knee extension was noted, the Figure shows that at both weight acceptance and at push-off, the difference in knee extension between the older and younger groups is similar. The pattern of knee flexion and extension we found in the younger and older groups resembles that described by previous investigators[1] related to the gait cycle.

The age groups studied were based on the low variability for gait characteristics within groups of people (ie, 20-40 and 60-80 years of age) demonstrated by Murray and colleagues.[1,2] We also attempted to account for variability within the groups by using a larger number of subjects in each group than previously studied.[1-6]

As for previous studies of the gait of older persons,[1-6] a sample of convenience was used in our study. When using a sample of convenience, it is important to clearly describe the population studied. We do believe the group of older people studied is representative of a healthy population of older people. As mentioned in the article, they were screened for chronic disease and normal range of motion and strength in the lower extremities. Many of the subjects reported being active in walking and golf activities, and many were volunteers at a local rehabilitation center-indicating a level of confidence in their own health status and a desire to keep active through service to others.

We hope this response and the Figure help to clarify some of the questions posed by the commentator.

References

[1] Murray MP, Drought AB, Kory RC. Walking patterns of normal men. J Bone Joint Surg [Am], 1964;46:335-360 [2] Murray MP, Kory RC, Clarkson BH, Walking patterns 'in healthy old men, J Gerontol. 1969; 24:169-178, [3] Murray MP, Kory RC, Clarkson BH, Sepic SB. Comparison of free- and fast-speed walking patterns of normal men. Am J Phys Med. 1966; 45:8-24. [4] Finley FR, Cody KA, Finizie RV. Locomotion patterns in elderly women, Arch Phys Med Rehabil. 1969;50:140-146. [5] Hageman PA, Blanke DJ. Comparison of gait of young women and elderly women. Phys Ther. 1986;66:1382-1387. [6] Blanke DJ, Hageman PA. Comparison of gait of young men and elderly men, Phys Ther, 1989;69:144-148. (*) Motion Analysis Corp, 93 Stony Cir, Santa Rosa, CA 95401. [dagger] NEC Corp, NEC Bldg, 33-1, Shiba 5-chrome, Minato-ku, Tokyo 108, japan. [double dagger] Redlake Corp, 1711 Dell Ave, Campbell, CA 95008. [sections] Panasonic Industrial Co, Audio-visual Systems Division Executive Office, One Panasonic Wa Secaucus, NJ 07094. [parallel IBM Corp, PO Box 1328-C, Boca Raton, FL 33432.

KM Ostrosky, PT, is Geriatric Physical Therapist III, University of Pittsburgh Medical Center, and Clinical Faculty Member, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh, 101 Pennsylvania Hall, Pittsburgh, PA 15261 (USA). At the time this research was completed, she was Supervisor, Physical Therapy Department, Harmarville Rehabilita Center Inc,. Pittsburgh, PA 15238. Address all correspondence to Ms Ostrosky.

JM VanSwearingen, PhD, PT, is Assistant Professor and Coordinator of the Graduate Focus in Geriatr Physical Therapy, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh.

RG Burdett, Phd, PT, is Acting Chairman, Department of Physical Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh.

Z Gee, PT, GCS, is Director of Physical Therapy, Leader Nursing and Rehabilitation Center, North Hills, Pittsburgh, PA 15237.

This study was completed in partial fulfillment of the requirements for Ms Ostrosky's Master of Science degree in Geriatric Physical Therapy, University of Pittsburgh, This study was approved by the institutional review boards of the University of Pittsburgh and Hannarville Rehabilitation Center Inc.

The results of this study were presented in poster format at the Annual Conference of the American Physical Therapy Association; June 24-28, 1990; Anaheim, CA.

This article was submitted February 4, 1993, and was accepted January 6, 1994,