-&ea& Cluarterly for Exercise and Sport a 81998 by the American Alliance for Health, Physical Education, Recreation and Dance

.&- Vol. 69, No. 4, pp. 338-343

~r 6% . ..-. .

The Reliability and Validity of a Chair Sit-and-Reach Test as a Measure of Hamstring Flexibility in Older Adults

C. Jessie Jones, Roberta E. Rikli, Julie Max, and Guillermo Noffal

The purpose of this study was to examine the test-retest reliability and the n'terimr validity ofa newly developed chair sit-and- reach (CSR) test as a measure of hamstring flexibility in older adults. CSR perfonance was also compared to sit-and-reach (SR) and back-saver sit-and-reach (BSR) measures of hamstringflnnbtlity. To estimate ~el.iability, 76 ma and wmna age = 70.5 years) p e r f m d the CSR on 2 different days, 2-5 days apart. In the validity phase ofthe study, scores o f80 men and w o r n age = 74.2 years) were obtained on threefield test measures of hamstringflexibility (CSR, SR, and BSR) and on a cracraimion test (piometer measurement of a passive straight-leg raise). Results indicate that the CSR has good intrclass test-retest reliability (R = .92 for ma; r = .96 for w o r n ) , and has a moderate-to-good relationship with the criterion measure (r = .76 f m men; r = .81 for w m ) . The crilaon validity ofthe CSR for the male and female participants is comparable to that ofthe SR (r = .74 and r =. 71, respectively) and BSR (r =. 70 and r =. 71, respectively). Results indicate that the CSR test produces reasonably accurate and stable measures of hamstringflexibility. In addition, it appears t h k the CSR is a safe and socially acceptabk alterndiue to traditional floor sit-and-mh tests as a measure of hamstringjhzbility in older aduki.

Key wards: aging, field test, assessment, mobility

L ack of hamstring flexibility has been associated with low back pain, postural deviations, gait limitations,

risk of falling, and susceptibility to musculoskeletal in- juries (American College of Sports Medicine, 1995; Grabiner, Koh, Lundin, & Jahnigen, 1993; Kendall, Mc- Creary & Proyance, 1993; Liemohn, Snodgrass, & Sharpe, 1988). In older adults, tight .hamstrings especially can lead to reduced stride length and walking speed, which in turn can cause problems with dynamic balance (Brown, 1993). Due to the importance of hamstring flex- ibility, its measurement is included in most current fit- ness test programs including the AAHPERD Physical Best (American Alliance for Health, Physical Education, Recreation and Dance, 1988), the Prudential FITNESS GRAM (Cooper Institute for Aerobics Research, 1994),

Submitted: January 27, 1998 Accepted: May 4, 1998

C. Jessie Jones, Roberta E. Rikli, Julie Max, and Guillermo Noffal are with the Division of Kinesiology and Health Promotion at California State University-Fullerton.

the Ys Way to Physical Fitness (Gelding, Myers, & Sin- ning, 3989), the AAHPERD Functional Fitness Test for Adults Qver 60 (Osness et al., 1996), and the President's Challenge Fimess Test (President's Council on Physical Fitness and Sports, 1990).

- The most common method of assessing hamstring flexibility in the field setting has been the floor sit-and- reach (SR) test, originally reported by Wells and Dillon (1952). Recently, a modified one-leg version of the SR, the back-saver sit-and-reach (BSR), has been recom- mended as an altemative to the two-leg SR (Cooper In- stitute for Aerobics Research, 1994). The rationale for the BSR is based on the work of Cailliet (1988) who sug- gested that stretching one hamsuing at a time, instead of both at once, results in less stress and risk of injury for the low back and spine. Studies have shown that both the SR and BSR tests are highly reliable, with R values con- sistently above .90, and they have at least moderate cri- terion validity relative to goniometer-measured hamstring flexibility (rtalues range from .51 to 39; Jackson & Baker, 1986; Jackson & Langford, 1989; Patterson, Wiksten, Ray, Flanders, & Sanphy, 1996).

Although the SR and BSR are generally considered acceptable field test measures of hamstring flexibility for

nnrlr. n ---- r-- rnnn

most age groups, many older people, due to their medi- cal conditions or functional limitations (e.g., obesity, lower back pain, lower body weakness, hip and knee re- placements, and severely reduced flexibility), find it dif- ficult or impossible to get down and up from the floor position for these tests. Also, we have found that some older adults, possibly due to weak abdominal musclq as well as tight hamstrings, cannot hold a sitting posi&a.f on a flat surface, especially with both legs extended, and will start to fall backward during testing. Therefore, to increase the rate of participation for our older adult cli- ents and decrease the risk of injury, we experimented with a chair sit-and-reach (CSR) test as an alternative to the SR and BSR The CSR test might best be described as a modified version of the BSR test, in that only one leg at a time is involved in the testings thus, reducing the stress on the lower back and spine. The CSR requires participants to sit near the front edge of a chair, extend- ing one leg straight out in front ofthe hip, with the other leg bent and slightly off to the side. The two-fold purpose of this study was to (1) determine the test-retest reliabil- ity of the CSR, and (2) to evaluate the validity of the CSR by comparing CSR scores to a criterion (goniometer) measure of hamstring flexibility in older adults. CSR pe?formance also was compared to other common field test measures of hamstring flexibilivhe SR and BSR.

Methods m

Participants

Seventy-six older adults (34 men and 42 women, M age = 70.5 years) were solicited from a university-based exercise program to participate in the reliability phase of the study. A different group of 80 volunteers (32 men and 48 women, Mage = 74.2 years) were recruited from nutrition and exercise classes at a nearby retirement community. The criteria for inclusion in the study were that the participants be over the age of 60 years, have no musculoskeletal limitations which would prohibit their performance on the tests, and agree to sign astate- ment of informed consent.

Procedures

Prior to all testing, participants performed an 8-min warm-up and static stretch routine emphasizing the lower body. Participants in the reliability study, conducted a p proximately 4 weeks prior to the validity study, performed the CSR on 2 different days, 2-5 days apart- Testing pru tocols during reliability testing, including technician training procedures, were the same as those described below in the validity phase of the study. So that interrater reliability would be reflected in the reliability analysis,

&~ca, w, Max; -.

-A- -: . - .

different technicians were used to c& Day 1 and Day 2. Day 2 technicians w the scores obtained on Day 1.

Participants in the validity study were asskssed on the CSR, SR, and BSR in a counterbalanced (rotating) or- der determined prior to the test day and indicated on the score card. The CSR, SR, and BSR tests were con- hc&d by a team of six graduate students and six older adult volunteer technicians, all of whom had participated in a group training session led by the study coordinators. During the h i n h g , technicians practiced on each other until they demonstrated proper procedures to the study coordinators. The goniometer assessment of hamstring flexibility, administered after completion of the other three tests, was administered by three experienced clini- cians (two physical therapy aides and one athletic trainer). A pilot study, utilizing a subsample of 19 par- ticipants, indicated that the interrater reliability of the 3 clinicians was .92. The clinicians administering the g u niometer tests were unaware of scores received on the CSR, SR, and BSR tests.

Only the p-4med leg score (the leg yielding the best score) was used for the CSR, BSR and goniometer tests. Once the preferred leg was determined, that score was held constant throughout all three of the single-leg mea- sures. Following a demonstration of each test, 2 practice trials and two test trials were given for each of the mea- sures. Participants were reminded to exhale as they bent forward, avoid bouncing or rapid, forceful movement, and never stretch to the point of pain. On all sit-and- reach measures, if the knee(s) started to bend, the par- ticipants were asked to slowly sit back until the knee(s) were straight before scoring. The best of the two test tri- als' (scored to the nearest 1/2 in.) was used for subse- quent analysis on the SR, BSR, and CSR. The average of the two test trials (scored to the nearest degree) was used k r the goniorneter test. All measures were administered on the same day, and all tests were conducted with the participants' shoes on. With older adults, -the time re- quired for removing shoes can be extensive and is often prohibitive when testing groups within the field setting.

Measures

Chair Sit-and-Reach. Following a demonstration, par- ticipants sat on a folding chair (17-in. high seat) and moved forward until they were sit- near the front edge. (The chair was placed against a wall and checked to see that it would remain stable throughout the testing). Par- ticipants were asked to extend their preferred leg in front of their hip, with the heel on the floor and foot dorsiflexed (at approximately a 90" angle), and bend the other leg so that the sole of the foot was flat on the floor about 6-12 in. to the side of the body's midline. With the extended leg as straight as possible and hands on top of each other with palms down (tips of the middle fingers

Jones~iddi, Max, and Noffal

em@, participants were to "slowly bend forward at the hip joint, keeping the spine as straight as possible and the head in normal alignment with the spine (not tucked) ." Participants were instructed to reach down the extended leg in an attempt to touch the toes. The par- ticipant heM a brief static pwktion (for 2 s), while the adminitrator recorded the "reached score" using an 18- in. ruler positioned p a e l to the lower leg (shin; see Figure 1). The mki& of the toe at the end of the shoe represented a "aeso* WQre. Reaches short of the toes were recorded as minus scores* and reaches beyond the toes were recorded as plus scores. Test-retest reliability esti- mates for the CSR are reported in the results section of this paper,

Sitand-Reuch. The SR test was administered using the procedures outlined in the Osness et al. manual (1995). A yardstick was placed on the floor, with a 12-in. strip of masking,tape positioned at the 20-in. mark (6 in. on each - side of the yardstick). Following a demonstration, partici- pants sat on the floor with their shoes on, legs M y ex- tended, a yardstick between their legs ("On mark of yardstick toward the hips), and their heels 12 in. apart. Throughout testing, the administrator checked to ensure that the heels remained at the 2&n. mark. With the ex- tended leg as stiaight as possible, hands on top of e ~ h other (tips of the middle fingers even), and palms do.4ln, the participant slowly reached forward sliding the hack along the yardstick as far as possible. In this study, die,. 20-in. mark represented a zero scare, with reaches sh& '

of the mark recorded as mfnb scores and reaches b yond 20 in. as plus scores. Scores were recorded to tb nearest 1/2 inch. Previous studies indicate that reliabiil- ity estimates for the SR are consistently high (.% 1: R< .99; Bravo et al., 1994; Jackson & Baker, 1986; ~acGoa& Langford, 1989; Shaulis, Golding, & Tandy, 1994).

Back Saver Sit-and-Reach. The procedures for the BSR were similar to those described in the Prudential FIT-

NESSGRAM (Coo- In.sti~~~prh&pbPics Research, 1994), with the major e x c e ~ ; w .$at the foot was not positioned agaihst a sit-and-reach bx.'~nscea$, foot placement was similar to that used for the SR (i.s,, heel positioned even with the 20-in, mark on a yardstick and 6 in. to the side). Participants were asked to extend their preferred leg only and bend the other leg so that the sole of the foot was flat on the floor, 6712 in. to the side of the yardstick. The 2Wi. mark represented a zero score, with reaches short of the mark recorded as minus scores and reaches beyond,the mark as plus scores. BSR mea- sures have been found to be highly reliable for both male and female participants (R= 99) (Patterson e t al., 1996).

Gmkmetm-Mau-. The goniometer assess- ment of hamstring flexibility was administered after completion of the other three tqts by experienced ex- aminers who were unaware of ,&e %ores participants received on the earlier tests, Follqwj~g pracedures out- lined by the American ~cade&~.of.~rtho~edic Surgeons (1966), a goniometer was used to measure hamstring flex- ibilig during a passive straightlleg raise, This test was selected because of its prevalent acceptance as a crite-

. .rion mea,sure . f ~ r hanns;tring.flexibility and its high reli- a W $ $95 < R < .99) (Jackson & Baker, 1986; Jackson & bngford, 1989; Patterson, et al., 1996). As indicated e&ikr;$ke irrtemrer reliability for the examiners in this 4m4ii3#z$, .92,&sed on a subsample of 19 participants.

&& $r6tdc~~J~involved ahgmng the axis of the goni- 'a&ter with the axisof the hip joint The stationary arm - - .- - was placed in hel&ith the trunk, with the movable arm positioned in line with the femur. With the knee held

- straig>t, the participant's preferred leg was passively ' moved into hip flexion until tightness was felt. A techni- cian assisted *th"moving the leg through flexion and keeping the participant in the correct position, while the clinician recorded the scores to the nearest degree.

Data Analysis

Test-retest reliability was estimated by calculating the intraclass coefticiknt (R) using one-way analysis of vari- ance (ANOVA) procedures appropriate for a single trial (Baumgartner &Jackson, 1995). Pearson correlation analysis was used to determine the relationships between the (2% SR, BSR, andhe criterion goniometer measure- ment. Niiety-five percent confidence intervals were com- puted for all correlation coefficients using Fisher's "Z transformation" procedures (Glass & Hopkins, 1984; Morrow &Jackson, 1993).

Results

Figure 1: Chair sit-and-reach. Descriptive statistics of participants in the reliaviity

phase of the study are presented in Table 1. Test-retest

nnrlr. n ---- L-- rnna

means and standard deviations, inuaclass R dues ; &d ~ I I B S S I ~ ~ pmwd-diffem somewbat from the SR and reliability confidence intervals are prese~nkd5n ~akde "2 . BSR. @i4BedCSS scores-repret the distance reached The high intraclass cortehrim t R 2292 f ' r men and R relative t;cr the tip oftk toes. On the SRand MR, scores =.96 for women), t o g e ~ e ' r ~ : a ' ~ ~ ~ ~ t c h q e r ep ren t the dis-e reached rehtive to a k e on the in scores from '&if 1 testing to Iday 2 { p > .05),'irrdicate floor even with the heel@) of the foot, that irlmurements are highly stable. Mthough data Table 4 contains the correlation r values and 95% i n i d l y werk recorded hi in. ( f o ~ ease in interpreting c d d e n c e intervals indicating the relationship between results to ofder adults), scores weie transformed into cm the field sit-and-reach tests and the criterion measure. for data an* and repoi-ti%. As indi&ed m the table, the correlations between the '

Descriptive chzmkteiisti'cs and mean flexibility scores CSR-and the caerion (goniometer-measured flexibility) of the validity study partiC:lpmts are presented in Table for both male and female participants ( r = .76 and 31, ?.'independent t test analyses 'of the data indicate that respectivdxy) Ws comparable to and, in fact, slightly fie women were more flexibk than irhe men on all four greater than-the correlations of the SR (r = .74 and .71, ha;nscriclg measures ( p < .~obl').'&o, M W A analysis respectively) and BSR with the criterion measures ( r = inqcates significant differknkes in Sedbility scores p r e ; f ,

duced by the CSR, SR, a ) I .

98.2, p < .0001. Post hoc co ibilii scores are better with or B!3R (p c .OOT). Gores not significantly'different &;'a, iii Sc?t, .& within participant groups all flexibility measures). ever, should be i n t q r e

74.53 (5.69) 74.02 (6.67) 175.1 8 (1 1.67) 159.38 (7.96) 79.36 (10.71) 62.07 (9.29)

Table 1. Means and standard deviations of descriptive characteristics for reliability study participants

Men (n = 34) Women ( n = 42) M SD M SD

Age (years) 72.62 (6.57) 69.11 (5.12) Height (cm) 177.01 (7.37) 163.14 (5.79) Weight (kg) 83.14 (16.61) 71.19 (14.33)

Note. M = mean; SD = standard deviation.

Table 2 Test-retest means, standard deviations, intraclass reliability estimates, and 95% confidence intervals for the chair sit-and-reach reliability study

Test 1 Test 2" R GI M SD M SD

Chair sit-and-reach (cm) Total (n = 76) -4.83 (14.48) -5.33 (14.22) .95 (.92-.97)

Men (n = 34) -13.46 (13.72) -12.45 (14.22) -92 (.85--96)

Women (n = 42) .23 (12.20) .25 (1 1.68) .96 (.93-.98)

Note. M = mean; SD = standard deviation; R = intraclass reliability estimate; CI = 95% confidence interval. "ANOVA analysis revealed no significant differences between Test 1 and Test 2 scores for male or female participants ( p > .05).

and-reach (cm) -9.69 (13.43)' 3.18 (12.27) 4.41" Sit-and-reach (cm) -20.84 (12.81)? -5.73 (11.57) 5.48" Back saver (cm) -20.90 (14.4412 -5.37 (12.04) 5.21" Goniometer (") 74.72 (14.24) 91.29 (12.04) 5.50"

Note. M = mean; SD = standard deviation. Chair sit-and-reach and backsaver scores represent only the preferred leg (defined as the leg which results in the better score). 'Scores represent distance reached relative to tip of toes 'Scores represent distance reached relative to line on the floor even with heel(s) of the feet * t ratios comparing flexibility scores of men and women are statistically significant ( p < .0001, df= 78).

-- --

Table 4. Correlations and 95% confidence intervals of chair sit- and-reach, sit-and-reach, and back-saver sit-and-reach scores with goniometer-measured flexibility

Men ( n = 32) Women (n = 48) r CI r CI

CSR .76 (.57-.88) .81 (.69-.89)

SR .74 (.54-.86) .71 (.54-.83)

BSR .70 (.48-.84) .71 (.54-.83)

Note. r = correlation; CI = 95% confidence interval; CSR = chair sit-and-reach; SR = sit-and-reach; BSR = back-saver sit-and- reach.

.lone% R i a Max, and Noifal

.70 m d .71, respectively). However, none of the differ- 'ences between the correlation values reached statistical significance at the .05 level.

T i Discussion a

Although the SR and the BSR are the most com- monly used field measures of hamstring flexibility in current fitness test batteries, both tests have inherent limitations for older adults who may have difficulty get- ting down and up from the floor or have difficulty sit- ting on a level surface with legs extended. Therefore, a chair sit-and-reach was proposed as an alternative pro- cedure for psessiog hamstring flexibiliya procedure which would presumably enable more older adults to participate and minimize the possibility'of injury during testing. The purpose of this study was to examine the relationship of the CSR to other measures of hamstring flexibility in older adults, particularly with respect to the test-retest reliability and criterion validity of the CSR.

Results indicate that the CSR test has good stability reliability for both men (R =.92) and women (R = .96) and that its criterion validity correlations (r= .76 for men and .81 for women) are slightly greater (although not statistically different) than for SR or BSR procedures (see Table 4). The validity coefficients faund in this stu$ for CSR, SR, and BSR (.70 > R> .81) are similar to those found in other studies with other age groups. Jacksun and Baker (1986) andJackson and Langford (1989) re- ported validity coefficients for the SR test ranging from .64 to .88 in studies involving teenage and middle-age participants, respectively. Also, Patterson et al. (1996), in a study involving 11-15-year-olds, reported fairly com- parable BSR coefficients for male participants (left leg = .68; right leg = .72 ), but somewhat lower values for female participants (left leg = .51; right leg =.52). The high CSR reliability values for the older adults in this .

study were also similar to the SR and BSR values reported in other studies, with R coefficients in all cases consis- tently above .90.

Although the findings of this study and others indi- cate that the CSR, SR, and BSR all have comparable re- liability and validity coefficients for participants who can perform the tests, 8 (approximately 10%) of the origi- nal volunteer participants in the validity study had to be excluded because they either could not or would not get down on the floor for sit-and-reach testing. No partici- pants, on the other hand, were eliminated due to their inability to perform the chair sit-and-reach test. Also, in spite of our emphasis on proper spotting of the partici- pants in this study, one female participant did fall back- ward during the SR testing, hitting her head on the gymnasium floor. As indicated earlier, weakened uunk muscles or tight hamstrings, or both, make it difficult

for some older adults to-.maim& a straight-leg sitting position on the fioor. No injuries occurred during test- ing on the CSR. However, careful spotting is recom- mended when assessing frail participants or individuals with balance problems.

In concl~~sion, the measurement of hamstring flex- ibility is an important component of health-related fit- ness. Results of this study indicate that the CSR is highly reliable and has moderate validity as a measure of ham- string flexibility. Further, the CSR appears to be a safe and socially acceptable assessment procedure for older adults and can measure each leg separately to detect any bilateral differences in hamsuing flexibility. Early detec- tion of shortened hamstrings would be valuable for the practitioner in providing feedback for exercise prescrip tion to help correct or reduce imbalances that may lead to mobility problems and potential injuries. Furthermore, in an era of assessment and accountability for health care (Russek, Wooden, Ekedahl &Bush, 1997), the CSR could potentially provide practitioners an excellent outcome measure for assessing the benefits of therapeutic inter- vention with this population, although additional stud- ies are needed to test the ability of the CSR to detect cbange over time. Also, more studies are recommended to investigate the reliability and validity of the CSR with physically frail and disabled populations. The relation- ship of the CSA in this study to other hamstring flexibil- i q meisures must be delimited to the population studied, apparently healthy older adults with no major orthopedic limi -

w - . ~ d ~ R

~efreien ces

American Alliance for Health, Physical Education, Recreation and Dance. (1988). AAHPERD Physical Best. Reston, VA: Author.

American Academy of Orthopedic Surgeons. (1966). Joint motion: Method of mewring and recording. Edinburgh, UK: Livingstone.

American College of Sports Medicine. (1995). Guidelims for exercise testing and pmcription. Philadelphia: Lea & Febiger.

Baumgartner, T. A., &Jackson, A. S. (1995). Measuremat f w Eoaluation in physical educalion and exercise sciozce. Dubuque, IA: Brown & Benchmark.

Bravo, G., Gauthier, P., Roy, P., Tessier, D., Gaulin, P., Dubois, M., & Peloquin. L. (1994). The functional fitness assess- ment battery: Reliability and validity data for elderly women. Journal $Aging and Physical Activity, 2,67-79.

Brown, M. (1993). The well elderly. In A. Guccione (Ed.): Ge- riatricph~~calthmapy (pp. 391401). St. Louis, MO: Mosby.

Cailliet, R (1988). Low back pain syndrome. Philadelphia: F. A. Davis.

Cooper Institute for Aerobics Research. (1994). The A w l a t i d FiTWSSGRAM test &ministration manual. Dallas, TX: Au- thor.

Glass, G. V, & Hopkins, K. D. (1984). Statistical methods in edw carimc and PsychorOgY. Englewood Clifh, NJ: RenticeHall.

Golding, L. A, Myers, C. R, & Sinning, W. E. (1989). The Y's way to physicatfines. Champaign, IL: Human Kinetics Publishers.

Grabiner, M. IL, Koh, T. J., Lundin, T. M., & Jahnigen, P. W. (1993). Binematics of recovery from a stumble. Journal of Geronfdtgy, 48, M97-M102.

Jackson, A. W., & Baker, A. A. (1986). The relationship of the sit and reach test to criterion measures of hamsuing and back flexibility in young females. Research Quurterly forf i- &e and Spmt, 57,183-186.

Jackson, A, & Langford, N. J. (1989). The criterion-related validity of the sit-and-reach test: Replication and extension of previous findings. Research Quarterly farE& and 60, 384-387.

Kendall, F. P., McCreary, E. K, & Provance, P. G. (1993). Muck: Testing and function (4th ed.) . Baltimore, MD: Williams & W1lkins.

Liemohn, W., Snodgrass, L. B., & Sharp=, G. L. ( 1 9 8 8 1 r F ,solved controversies in back management-* rcv ie~Jw- nal of Orthqpaedic and Sports Physical Thwafi, 9,2354-244:

Morrow, J. R, &Jackson, A W. (1993). How "significant6 is ybw reliability? Reswch Qyrierly fwEawcise and Sport, 64,352- 355.

Osness, W. H., Adrian, M., Clark, B., Hoeger, W., Rabb, D., & Wiswell, R (1996). Fumtiodfiilness a,wmmnt fmadults wer 60 years. Dubuque, IA: Kendall-Hunt.

Patterson, P., Wiksten, D. L., Ray, L., Flanders, C, & Sanphy, D. (1996). The didity and reliability of the back saver sit- and-reach test in middle school girls and boys. Research Qtw~terly far Ewzise and Spuyt, 64,448-45 1 .

President's Council on Physical Fitness and Sports. (1990). PCPFS President's chnllages physical fitness program test manzral Washington, DC: Author.

Russek, L., Wooden, M., Ekedahl, S. & Bush, A (1997). Atti- tudes toward standardized data collection. Phymat T h h , 77, 714-729.

Shaulis, D., Mding, L. A., & Tandy, R D. (1994). Reliabiity of the AAHPERD functional fitness assessment across multiple practice sessions in older men and women. Jour- nal of Aging and Physical Activify, 2,273-279.

Wells, R F., & Dillon, E. K. (1952). The sit-and-reach t e s d , test of back and leg flexibility. Research Qua&$, 23, 115- 118.

Note

1. Test theory suggests that with multiple trials the aver- age score is normally a more reliable indication of per- formance than the best score. However, because SR, BSR, and CSR performance has been found to be quite reli- able using the "best score" protocol (above .90) , we be- lieve this method is justified and, in fact, recommended in gouptesting situations where efficiency in testing time L &tical.

This research was supported by a grant from Pacificare Health Systems: The authors thank the s-and all voI- unteer partkipants from the Leisure World and Morningside Retirement Communities and from the Lifespan Wellness Clinic f o ~ their assistance with this study. Please address all correspondence regarding this article to C. Jessie Jones, Division of Kinesiology and Health Promotion, California State University-Fullerton, Fullerton, CA 92834.

E-mail: jjones@fidlerton.edu