Journal of Gerontology:

MEDICAL SCIENCES Copyright 2001 by The Gerontological Society of America

2001, Vol. 56A, No. 4, M248–M252

M248

Validity of the Multi-Directional Reach Test: A Practical Measure for Limits of Stability in

Older Adults

Roberta A. Newton

Department of Physical Therapy, Temple University, Philadelphia, Pennsylvania.

Background.

Falls occur not only in the forward direction, but also to the side and backward. The purpose of thisstudy was to develop a portable and valid tool to measure limits of stability in the anterior-posterior and medial-lateraldirections.

Methods.

Two hundred fifty-four community-dwelling older persons were administered the Berg Balance Test(BBT), the Timed Up & Go Test (TUG), and the Multi-Directional Reach Test (MDRT). For the MDRT, subjects per-formed maximal reaches with the outstretched arm forward (FR), to the right (RR), to the left (LR), and leaning back-ward (BR), with feet flat on the floor. Reach was measured by the subject’s total hand excursion along a yardstickaffixed to a telescoping tripod.

Results.

Mean scores on the MDRT were FR

5

8.89

6

3.4 in., BR

5

4.64

6

3.07 in., RR

5

6.15

6

2.99 in., and LR

5

6.61

6

2.88 in. Interclass Correlation (ICC2,1) for the reaches were greater than .92. Reliability analysis (Cronbach’sAlpha, .842) demonstrated that directional reaches measure similar but unique aspects of the MDRT. The MDRT dem-onstrated significant correlation with the BBT sum and significant inverse relationship with the scores on the TUG. Re-gression analysis revealed that activity level contributed to scores in the forward, right, and left direction and that fear offalling contributed to scores in the backward direction.

Conclusion.

The Multi-Directional Reach Test is an inexpensive, reliable, and valid tool for measuring the limits ofstability as derived by reach in four directions. Values obtained on relatively healthy community-dwelling older adultsserve as norms for screening patient populations.

S the population of older adults increases, there is anacute awareness of the impact of falls and fall-related

injuries because of morbidity, rising health care costs, andreduction in the quality of life for the elderly. Although re-search (1) and public health agendas (2) target the reductionof falls and related-injuries among older adults, the numberof hip fractures has increased as demonstrated in the mid-decade report of

Healthy People 2000

. Unintentional inju-ries are the sixth leading cause of death in individuals overage 65 and are the leading cause of death in people over age85. The majority of these deaths are related to falls.

The fear of falling is common in older adults who have orhave not sustained a fall.

Estimates range as high as 73% ofolder adults have a fear of falling, which may lead to a de-crease in activity levels and an increased risk of falls (3–5).Not only is fear of falling a recognized concern of olderadults, but older adults also fear not being able to get up oncea fall has occurred (6). Approximately half of the older adultswho have fallen once or more than once required help to getup after one of the fall episodes (6,7). These facts are ofconcern particularly in light of the numbers of older adultswho live alone.

The ability to confidently and safely perform routinedaily activities requires the older adult to generate appropriatemotor strategies to navigate in an environment containingboth stationary and moving objects within varying contexts.Balance is also needed while the individual manipulates the

environment and the feet are stationary. Activities such asbending or reaching up or to the side require shifting thecenter of gravity (COG) within the base of support (BOS).Once the COG moves outside the BOS, the limits of stabil-ity (LOS) for the currently executed balance strategy are ex-ceeded. An automatic movement strategy is executed tomaintain balance by either realigning the COG within theBOS or by evoking a step strategy and establishing a newBOS. If the appropriate movement strategy is not executed,the individual may stumble or fall in an attempt to regainbalance.

The Functional Reach Test (FRT) is an inexpensive andeasy to use tool to assess LOS in the forward direction (8).The FRT measures LOS only in the forward direction.Older adults with balance instability tend to overestimatetheir forward reach ability (9). An overestimation of forwardreach coupled with reduced postural limits (10,11) may causeloss of balance in older adults performing activities associ-ated with moving the COG toward the forward LOS.

Older adults also fall laterally and backward (12). Hipfractures are more likely to occur with a sideways fall (13–15), and among women wrist fractures are more likely tooccur from falling on the outstretched hand in the backwarddirection (13). Induced sway in lateral direction duringblindfolded conditions has been found to be an excellentpredictor of fall risk (16,17). An assessment tool was devel-oped to measure medio-lateral reach, but as with the For-

A

MULTI-DIRECTIONAL REACH TEST

M249

ward Reach Test it is limited in the direction of reach (18).Knowledge of reach in one direction has not been shown topredict reach in other directions.

A primary aim of this study was to test the reliability andvalidity of the Multi-Directional Reach Test (MDRT) as aninexpensive screening tool to measure the limits of stabilityin four directions. That is, how far an individual can volun-tarily reach, thereby shifting the COG to the limits of theBOS with the feet stationary. A secondary aim was to deter-mine whether reach in one direction predicts reach in otherdirections. A preliminary introduction to this test is foundelsewhere (19).

M

ETHODS

Subjects were 254 community-dwelling older adults (meanage

5

74.1

6

7.9 years) from senior centers or residentialhousing centers located in North Philadelphia, a federallydesignated medically underserved area. Older adults wereexcluded if they were wheelchair bound or could not liftboth outstretched arms to 90

8

in the forward direction.Eighty-six percent of the population represented minoritybackgrounds. The majority of the participants were womenwho lived alone.

After signing an informed consent form, subjects com-pleted a questionnaire that included demographics, health,activity status, and fall history. Translators were used forSpanish-speaking subjects. The questionnaire included itemsrelated to health status (excellent, good, fair, poor), the num-ber of falls in the past six months, the frequency of perform-ing social and physical activities, and a fear of falling index(6,20). The Berg Balance Test (BBT) (21), the Timed Up &Go Test (TUG) (22), and the MDRT were administered.The customary assistive device was permitted only duringthe performance of the TUG.

The BBT assesses performance on 14 routinely performedactivities. Tasks include standing, turning 360

8

, picking upan object from the floor, and alternate foot placement (19). Azero to four point rating system is used for each task for amaximum score of 56. The TUG is a timed measure of anindividual’s ability to stand, walk 10 ft, and return to the ini-tial seated position. Subjects completed the TUG at theirtypical pace.

The following procedures were used for the MDRT (Fig-ure 1). A yardstick affixed to a telescoping tripod wasplaced at the level of the subject’s acromion process. Priorto the reach, the yardstick was leveled so that it was hori-zontal to the floor. The older adult lifted an outstretched armto shoulder height, paused for an initial reading, then reachedas far forward (FR) as possible. Instructions given to olderadults were “without moving your feet or taking a step, reachas far (direction given) as you can, and try to keep your handalong the yardstick.” For the backward direction the subjectwas asked to “lean as far back as you can.” Individuals usedtheir typical strategy to accomplish the task. Participantsused their arm of choice for the forward and backward tasks,and used the respective arm for the right and left reaches.The start and end positions of the index finger of the out-stretched hand were recorded, and the difference representedthe total reach for that direction. Feet were maintained flaton the floor; if the feet were moved, the trial was discarded.

Two trials were recorded for each direction. The procedurewas repeated with the backward (BR), right (RR), and left(LR) directions.

Appropriate descriptive statistics were used to analyze datafrom the questionnaire. Test-retest reliability was obtained forthe MDRT. Pearson correlation statistics were used to com-pare scores on the MDRT to scores on the BBT and TUG.Multiple regression analysis was used to determine how fac-tors such as health status, activity status, fall history, and fearof falling as measured by a Fear of Falling Index contributedto scores on the MDRT. ANOVA with post-hoc Bonferronianalysis was used to examine the mean for each direction ofthe MDRT in relation to the use of assistive devices and healthstatus (i.e., excellent, good, fair, poor).

R

ESULTS

Table 1 gives the socio-demographic and health statusdata for the sample. The group, by self-report, consideredtheir health status as good to excellent (67.1%), did not havea fear of falling (58.6%), and had not fallen in the past 6months (78.6%). Approximately 19% of the older adults didnot take medications, and 10% took more than four medica-tions. The majority of older adults did not use an ambulatorydevice (69.6%), and those using a mobility aid used a cane.

Figure 1. Positions of the Multi-Directional Reach Test. A, Posi-tion for the individual to reach forward or lean backward. B, Positionfor the individual to reach to the right. The task was repeated usingthe left arm.

Table 1. Sociodemographic and Health Status

Mean age 74.1

6

7.9 yGender Self-report health status

Men 21.5

†

Excellent 13.5Women 78.5 Good 53.6

Race Fair 27.8African American 69.0 Poor 5.2Hispanic 17.1 Assistive deviceCaucasian 11.9 No 69.6Other 2.0 Yes 30.4

Living status Fear of fallingAlone 51.6 Not afraid 58.6With others 48.4 Afraid, same activity level 24.3

Medications Afraid, decr activity level 17.1None 19.3 Tripped/Fallen in the past 6 mo1–4 70.5 No 78.6

.

4 10.2 Yes 21.4

†

Values are percentages.

M250

NEWTON

Table 2 shows the scores for the BBT, TUG, and MDRT.Distribution for the average reaches in the four directions islocated in Figure 2. Interclass correlation, ICCs (1,2), for theMDRT were FR

5

0.942, BR

5

.0929, RR

5

.0926, and LR

5

.0947.

t-

tests for paired samples demonstrated no significant dif-ference between the two trials for reach in the forward direction.Significant differences were noted between trials for the back-ward, right, and left directions, indicating a practice effectacross the two trials; however, effect sizes were small: FR

5

0.09, BR

5

0.14, RR

5

0.18, and LR

5

0.21. Table 3 presentsa correlation matrix for the four directions of reach. A strongcorrelation for the reaches is noted (

p

z

.0004). Cronbach’s al-pha, a measure of the internal consistency of the test instrument,was 0.842 (Table 3). Deletion of any one of the test items didnot significantly improve the alpha value.

The Pearson correlation statistic (Bonferroni adjust-ment,

p

z

.0004) revealed that scores on the MDRT demon-strated significant correlation with the BBT sum: FR (

r

5

0.476), BR (

r

5

0.356), RR (

r

5

0.389), and LR (

r

5

0.390). Italso revealed that scores on the MDRT demonstrated signif-icant inverse relationships with the scores on the TUG: FR(

r

5

2

0.442), BR (

r

5

2

0.333), RR (

r

5

2

0.260), and LR(

r

5

2

0.310).

Table 2. Scores on the Berg Balance Test, Timed Up & Go, and Multi-Directional Reach Test

Test Mean

SD

Min Max

BBT (max.

5

56) 48.6 6.0 23.0 56.0TUG (s) 15.6 7.0 6.0 47.8MDRT

FR (in) 8.89 3.4 0.5 16.8BR (in) 4.64 3.1 0.4 14.0RR (in) 6.86 3.0 0.7 18.2LR (in) 6.61 2.9 0.0 14.4

Note

: BBT

5

Berg Balance Test; TUG

5

Timed Up & Go Test; MDRT

5

Multi-Directional Reach Test; FR

5

arm stretched forward; BR

5

leaning back-ward; RR

5

arm stretched to the right; LR

5

arm stretched to the left.

Figure 2. Distribution of reach scores for the Multi-Directional Reach Test (average of two trials). A, Histogram of forward reach scores. B,Histogram of backward reach scores. C, Histogram of right reach scores. D, Histogram of left reach scores.

MULTI-DIRECTIONAL REACH TEST

M251

Table 4 shows the mean scores for each direction of reachin relation to various health factors. Multiple regressionanalysis examining activity level, fear of falling, health sta-tus, and fall history revealed that the activity level contrib-uted significantly (

p

,

.0004) to the scores in the forward,right, and left directions. Fear of falling, as measured by theFear of Falling Index, contributed significantly to scores inthe backward direction. Subjects were subdivided relative tohealth status, use of an assistive device, and history of falls.Mean scores on the MDRT for each subcategory were ob-tained (Table 4). No significance associations (

p

.

.5) werenoted between mean scores on the MDRT and these catego-ries of health status.

D

ISCUSSION

Limitations in the ability to shift the center of gravity to-ward the limits of stability may result in reductions in theactivity level of older adults. The use of force plates to ex-amine postural responses to unexpected perturbing condi-tions or to measure forward reach may not be practical forscreening, particularly in community settings (23–25). Inaddition, force-plate studies using destabilizing forces evokea compensatory motor response compared with an anticipa-tory balance strategy evoked during voluntary movement ofthe COG when the person performs activities of daily living.

This study examined the reliability, validity, and practi-cality of the MDRT to measure limits of stability in settingswhere computerized force-plate systems are too expensiveor where a more mobile clinical assessment is required. Whenvoluntarily moving the center of gravity toward the limits ofstability (with feet not moving), a greater excursion is ex-pected in the forward compared with the backward direction,and medio-lateral excursions should be symmetrical. Thisexcursion pattern was observed with the MDRT (Figure 2).Reaches to the right and left were symmetrical and thereforewere not influenced by either arm or foot dominance. Be-cause of the biomechanical arrangement of the ankle and

foot, there is a greater capability for forward excursion ofthe body over the foot than for backward excursion. In addi-tion, fear may be contributing to this decreased amount ofBR as noted by a significant affect of fear, as measured bythe Fear of Falling Index, on the BR scores. Although nottested, muscle strength of the anterior tibialis and hip abduc-tor muscles may contribute to reach scores in the respectivedirections.

In this group of community-dwelling older adults, no as-sociation was noted between assistive device use, health sta-tus, or self-report of falls. However, the results indicate atendency for lower mean reaches for each of these catego-ries. To determine if the MDRT predicts falls, a sample ofcommunity-dwelling older adults with medically documentedfalls is currently under study. The group of older adults repre-sents primarily a culturally diverse sample (86.1% AfricanAmerican and Hispanic) from a relatively low socio-economicregion (North Philadelphia). The predictability of the MDRTfor assistive device usage and falls needs to be examined inother socio-economic and ethnic groups, in homebound com-munity-dwelling older adults, and in those residing in longterm care facilities.

In this study, older adults voluntarily performed the MDRTin a freestanding environment and not close to a wall as notedin the FRT. Therefore, the MDRT may be more challengingbecause there is no wall to provide perceived security. Thisfactor may account for the lower mean FR observed in theMDRT (8.89 in) as compared with published mean values ofthe FRT (10.9 in.) (26). A lower reach score also occurredbecause the subjects were instructed to maintain the feet flaton the floor and not to take a step. Such instructions providea more accurate assessment of the limits of stability. Overbal-ancing and taking a step indicates that the postural strategyused to maintain COG over the BOS was exceeded and that anew movement strategy was evoked to maintain balance.

When administering the MDRT, it is recommended thattwo trials be given for each direction. Although there was anotable difference between the first and second trials in thebackward, right, and left directions, the effect size was small.That is, the relative magnitude of the differences between thetwo trials for a particular reach direction was small. Variationin the movements used to perform the reach may account forthe differences noted between the two trials per task. A cur-rent study will examine the method of reach in each directionthrough component analysis of movement patterns used inthe upper extremities, trunk, and lower extremities.

Cronbach’s alpha was not improved with the deletion ofany one of the reach directions, indicating that each direc-tion measures similar but unique aspects. This finding indi-

Table 3. Reliability Analysis (Cronbach’s Alpha) for theMulti-Directional Reach Test Correlation Matrix

Direction FR BR RR LR

FR 1.0000BR 0.4979 1.0000RR 0.6176 0.4957 1.0000LR 0.6104 0.4926 0.7493 1.0000Alpha

5

0.842

Note

: Abbreviations as in Table 2.

Table 4. Mean Multi-Directional Reach Test Scores in Relation to Selected Measures of Falls Status

Assistive Device Health Status Falls

DirectionNo

Mean (

SD

)Yes

Mean (

SD

)Excellent

Mean (

SD

)Good

Mean (

SD

)Fair

Mean (

SD

)Poor

Mean (

SD

)No

Mean (

SD

)Yes

Mean (

SD

)

Forward (in) 9.72 (3.19) 7.10 (3.11) 9.29 (3.37) 9.07 (3.46) 8.81 (3.33) 6.43 (2.94) 9.07 (3.19) 8.38 (4.07)Backward (in) 5.39 (3.19) 3.25 (2.20) 5.16 (3.18) 5.09 (3.19) 3.86 (2.69) 3.33 (2.90) 4.80 (3.1) 4.06 (2.94)Right (in) 7.57 (2.96) 5.62 (2.55) 7.54 (4.10) 7.0 (2.89) 6.53 (2.57) 5.25 (2.46) 7.08 (3.04) 6.12 (2.76)Left (in) 7.33 (2.80) 5.33 (2.45) 6.94 (3.03) 6.56 (2.83) 6.60 (2.99) 5.72 (1.57) 6.86 (2.80) 5.67 (3.06)

M252

NEWTON

cates that measurement in a single direction (e.g., forwardreach) does not necessarily predict reach values in the otherdirections. Because falls occur in all directions, reach as ameasure of the limits of stability needs to be assessed in alldirections.

Although the effect of test order was not examined, it isrecommended that a standard order of forward, backward,right, and left directions be used to compare the scores withpublished values. Because the backward direction is associ-ated with the Fear of Falling Index, this direction should notbe the first direction tested.

Scores on the MDRT demonstrated an appropriate posi-tive relationship with scores on the BBT; that is, the higherthe scores on the BBT (max

5

56), the greater the distancereached in all directions on the MDRT. Conversely, scoreson the MDRT demonstrated an appropriate inverse relation-ship with scores on the TUG; that is, the faster the individ-ual performed the TUG (lower timed scores), the greater thedistance reached in all directions on the MDRT. The signif-icant but low correlation between scores in the MDRT andscores on the BBT and TUG indicate that these screeningtests measure similar but unique aspects of balance abilities.Therefore, it may be justified to perform these tests in con-junction with one another to obtain a more comprehensiveassessment of balance abilities. The MDRT measures thelimits of stability with the feet stationary; the TUG mea-sures balance abilities during various phases of standing,walking and turning, and the BBT measures balance abili-ties while the individual performs tasks that mimic tasksroutinely performed in every day life.

In summary, the MDRT is a valid and reliable clinicalmeasure for limits of stability. The screening tool can beused with other tools such as the BBT and TUG to obtain acomprehensive clinical measure of balance abilities in olderadults. On the basis of such assessments, effective preventa-tive and rehabilitative measures can be developed.

Acknowledgments

This work was partially supported by HSRA/PHS 1 D27 AH 00720.

I thank DeWitt Kay, PhD, for his statistical support.

Address correspondence to Roberta A. Newton, PhD, Department ofPhysical Therapy, Temple University, 3307 North Broad St, Philadelphia,PA 19140. E-mail: rnewton@astro.temple.edu

References

1. Tinetti ME. Prevention of falls and fall injuries in elderly persons: aresearch agenda.

Preventive Medicine

. 1994;23:756–762.2. U.S. Department of Health and Human Services.

Healthy People 2000:National Health Promotion and Disease Prevention Objectives

. Washing-ton, DC: U.S. Department of Health and Human Services; 1990. DHHSPublication No. (PHS) 91-50213.

3. Tinetti ME, Speechley M, Ginter SF. Risk factors for falls among el-derly persons living in the community.

N Engl J Med

. 1988;319:1701–1707.

4. Maki BE, Holliday PJ, Topper AK. Fear of falling and postural perfor-mance in the elderly.

J Gerontol

Med Sci.

1991;46:M123–M131.

5. Howland J, Lachman ME, Peterson E, Cote J, Kasten L, Jette A. Covari-ates of fear of falling and associated activity curtailment.

Gerontologist

.1998;38:549–555.

6. Tinetti ME, Liu WL, Claus EB. Predictors and prognosis of inabilityto get up after falls among elderly persons.

JAMA

. 1993;269:65–70.7. Nevitt MC, Cummings SR, Hudes ES. Risk factors for injurious falls:

a prospective study

.

J Gerontol

Med Sci

. 1991;46:M164–M170.8. Duncan PW, Weiner DK, Chandler J, Studenski S. Functional reach: a

new clinical measure of balance.

J Gerontol

Med Sci

. 1990;45:M192–M197.

9. Robinovitch S, Cronin T. Perception of postural limits in elderly nurs-ing home and day care participants.

J Gerontol

Biol Sci

. 1999;54A:B124–B130.

10. Duncan PW, Studenski S, Chandler J, Prescott B. Functional reach:predictive validity in a sample of elderly male veterans.

J GerontolMed Sci.

1992;47:M93–M98.11. King MB, Judge JO, Wolfson L. Functional base of support decreases

with age

.

J Gerontol Med Sci.

1994;49:M258–M263.12. Cummings SR, Nevitt MC. Non-skeletal determinants of fractures: the

potential importance of the mechanics of falls.

Osteoporosis Int

. 1994;4(suppl 1):S67–S70.

13. Nevitt MC, Cummings SR. Type of fall and risk of hip and wrist frac-tures: the study of osteoporotic fractures. The Study of OsteoporoticFractures Research Group.

J Am Geriatr Soc

. 1993;41:1226–1234.14. Slemenda C. Prevention of hip fractures: risk factor modification.

AmJ Med

. 1997;103(2A):65S–71S.15. Greenspan SL, Myers ER, Kiel DP, Parker RA, Hayes WC, Resnick

NM. Fall direction, bone mineral density, and function: risk factors

for hip fracture in frail nursing home elderly.

Am J Med

. 1998;104:539–545.

16. Maki BE, Holliday PJ, Topper AK. A prospective study of posturalbalance and risk of falling in an ambulatory and independent elderlypopulation.

J Gerontol

Med Sci

. 1994;49:M72–M84.17. Topper AK, Maki BE, Holliday PJ. Are activity-based assessments of

balance and gait in the elderly predictive of risk of falling and/or typeof fall?

J Am Geriatr Soc

. 1993;41:479–483.18. Brauer S, Burns Y, Galley P. Lateral reach: a clinical measure of me-

dio-lateral postural stability.

Physiother Res Int

. 1999;4:81–88.19. Newton RA. Balance screening of an inner city older adult population

.Arch Phys Med Rehabil

. 1997;78:587–591.20. Tinetti ME, Mendes de Leon CF, Doucette JT, Baker DL. Fear of fall-

ing and fall-related efficacy in relationship to functioning amongcommunity-living elders.

J Gerontol

Med Sci. 1994;49:M140–M147.21. Berg KO, Wood-Dauphinee SL, Williams JI, Maki B. Measuring bal-

ance in the elderly: validation of an instrument. Can J Pub Health.1992;83(suppl)12:S7–S11.

22. Podsiadlo D, Richardson S. The timed “Up & Go”: a test of basicfunctional mobility for frail elderly persons. J Am Geriatr Soc. 1991;39:142–147.

23. Thelen DG, Wojcik LA, Schultz AB, Ashton-Miller JA, AlexanderNB. Age differences in using a rapid step to regain balance during aforward fall. J Gerontol Med Sci. 1997;52A:M8–M13.

24. Luchies CW, Wallace D, Pazdur R, Young S, DeYoung AJ. Effects ofbalance assessment using voluntary and involuntary step tasks. J Ger-ontol Med Sci. 1999;54A:M140–M144.

25. Wernick-Robinson M, Krebs DE, Giorgetti MM. Functional reach:does it really measure dynamic balance? Arch Phys Med Rehabil.1999;80:262–269.

26. Weiner DK, Duncan PW, Chandler J, Studenski SA. Functional reach:a marker of physical frailty. J Am Geriatr Soc. 1992;40:203–207.

Received August 13, 1999Accepted March 31, 2000Decision Editor: William B. Ershler, MD