Basic gait parameters: Reference data for normal subjects ...

1 'V"ZeDepartment ofVeterans AffairsJournal of Rehabilitation Researchand Development Vol. 30 No. 2 1993 Pages 210 223A Technical NoteBasic gait parameters : Reference data for normal subjects ,10-79 years of ageTommy Oberg, MD, PhD; Alek Karsznia, PT, PhD; Kurt Oberg, PhDDepartment of Biomechanics and Orthopaedic Technology, University College of Health Sciences, S-551 11 Jonkoping, Sweden; Department of Agricultural Engineering, Swedish University of Agricultural Sciences,S-750 07 Uppsala, SwedenAbstract Basicgait parameters were extracted from 233healthy subjects 116 men and 117 women, 10 to 79 yearsof age.

2 The measurements were made in a gait laboratoryon a m walkway. The results are presented in a seriesof Reference tables for slow, normal , and fast gait . Mean,standard deviation, coefficient of variation, 95010confi-dence intervals, and95%prediction intervals were calcu-lated. Significant sex differences exist in all gait parame-ters. In a two-way analysis of variance (ANOVA) model,there was a statistically significant age-variability for gaitspeed and step length at normal and fast gait , but not forstep frequency. In the step length parameter there was asignificant interaction effect of age and sex at normal andfast gait .

3 The Reference data are considered valid in anindoor laboratory words: gait , gait parameters , stride Basic gait parameters most frequently usedare velocity, step length, and step frequency. Manyreports are concerned with pathological gait , butsuch data must be compared with valid normalAddress all correspondence and requests for reprints to: Tommy Oberg,MD, PhD, Director, Associate Professor, Department of Biomechanicsand Orthopaedic Technology, University College of Health Sciences,S-551 11 Jonkoping, data to be interpretable. Published dataare generally limited to specific groups, for example, normal men (1), normal women (2), and elderlywomen (3,4).

4 These reports are often based onrelatively small numbers of subjects , spread overmany age groups. Different investigators use varyingunits of measurement, further hampering compari-sons. Reference data for outdoor walking, based onmeasurements from 260 subjects , men and womendivided into four age groups, have been published(5). The aim of this study was to present referencedata on Basic gait parameters for normal AND METHODSS ubjectsTwo hundred and forty healthy subjects wereexamined. However, seven subjects were excludedfrom the original material, 3 girls and 3 boys aged0-9 years, and one subject aged 80, because theywere too few to represent an age group.

5 Age and sexcharacteristics are shown inTable AnalysisWe have used the gait analysis method that wasdeveloped at the Biomechanics Laboratory, Univer-sity of California, Berkeley, California, and the210211 OBERGet al. Basic gait ParametersTable 1. 10 m long, including acceleration and decelerationAge and sex characteristics of the subjects . distances. Two photocells with m intervals,NumberAge group, yearsMenWomenTotal10-1927275420-29151530 30-3915153040-4915153050-5915153060-6915 153070-79141529 Total116117233self-aligning electrogoniometers, a computer, and aplotter constitute the equipment used(Figure 1).

6 Themeasurement was performed between the two pho-tocells. Goniometry was not used in the presentstudy. Heel strike was indicated by means of amanual temporal gait parameters ( gait speed, steplength, and step frequency) were collected duringslow, normal , and fast gait . The subject had to walkbetween the photocells 13 times 10 times withoutgoniometers, and 3 times with goniometers. Thewas calculated for each gait of the 10 measurements without goniometersStatistical MethodsDepartment of Orthopaedic Surgery at the Univer- Analysis of variance (ANOVA) and regressionsity of Uppsala, Sweden (6). The method was later analysis were performed according to standardfurther developed at the Department of Biome- methods (7,8).

7 All computations were made with achanics and Orthopaedic Technology, Jonkoping, commercial statistics package for a personal com-Sweden. The gait laboratory has a walkway about puter, SYSTAT between lights DTime =TDi0 0 0 0 1 2 3 410 Numberof steps= NTime forNsteps =TNAverage velocity, V = D/7D cm/sAverage stepfrequency,F = N/TNsteps/sAverage steplength, LrV.*cmFigure top view of the walkway showing the measured variables desired for analysis (after ()berg and Lamoreaux, 1979).212 Journal of Rehabilitation Research andDevelopment Vol. 30 No. 21993p-valueMenWomenGait speedSlow gaitN .S .< gait <.

8 gait < < lengthSlow gaitN .S .< gaitN .S .< gaitN .S .< frequencySlow gaitN .S .N . gait < . gait < .S .N .S . = Not significant (p> ) normal gait < < gait < < length Slow gait N .S . < DISCUSSION normal gait N .S . < ConsiderationsFast gait < < The results of all measurements, including gaitStep frequency Slow gait N .S . N .S . analysis, are dependent on the test conditions. normal gait N .S . < Consequently, Reference data are only valid for aFast gait N .S . < test situation similar to the conditions of the originalreference test situation. gait analysis data should N .S . = Not significant (p> ) always be interpreted with regard to a thoroughlydefined test situation.

9 Dahlstedt (9) reviewed 15articles with respect to walking speed. He foundTable 3. values of m/s for men and m/s forGait parameters . Two-way analysis of variance. women for normal self-selected gait . The meanp-value speed in shopping centers was m/s for women,Age Age and m/s for men (10). However, laboratory*group Sex Sex gait data cannot be considered valid for suchGait speed Slow gait N .S . < . walking situations. We have a that gait velocity data reported fromNormal gait < < . long walkways tend to be higher than data obtained Fast gait < < . from a short walkway.*Step length Slow gait N.

10 S. < < These observations are supported by the results ofa study by Waters et al. (5), who examined 260 normal gait < < < subjects aged 6 to 80 years along a level Fast gait < < . outdoor track of m. In the present study, usingStep frequency Slow gait N .S . N .S .. a m walkway with a acceleration distance, normal gait N .S . < . a m retardation distance, we found self-selected gait speed to be 118-134 cm/s Fast gait N .S . < N ..S . = Not significant (p> ) *Healthy subjects tested in a long basement corridor, 65 m walkingdistance (Oberg U, (*erg T, unpublished observations).RESULTSR eference tables for slow, normal , and fast gaitfor different age groups and sexes are shown in theAppendix(see p.))