Spinal Flexibility and Individual Factors That Influence It


We conducted an investigation to examine the spinal flexibility of a large, adult population and to study the effects of other individual physical characteristics on spinal range of motion. The study group consisted of 3,020 blue collar employees (2,350 men and 670 women) who underwent a physical examination that included assessments of standing and sitting height, weight, shoulder flexibility, and spinal flexibility in the sagittal and frontal planes. Flexibility measures were correlated positively to one another; however, lumbosacral flexion measurements assessed by the modified Schober method correlated to the other flexibility measurements to a much lesser degree. Age, sex, and height affected ROM, as did obesity and the ratio of standing height to sitting height. The study findings indicate that spinal ROM covers a wide spectrum of values and is affected by many individual factors. Any attempts to determine what is normal, excessive, or diminished must take into account variations caused by age, sex, and other physical attributes.

Key Words: Lumbar region, Physical therapy, Spine.

Much attention has been given to flexibility of the lumbar spine and its relationship to back health. Exercises to stretch the back often are advocated in routine daily exercise programs for ath- letes and employees.1 Furthermore, spinal flexibility has been assumed to be such an important part of good back health that flexibility measures have been proposed for use in preplacement physical examinations and preemploy- ment screening in an effort to identify individuals thought to be at greatest risk for developing back problems.

Spinal range-of-motion measure- ments are a standard part of the evalu- ation of patients with back pain, with the most commonly assessed movement being flexion.2 Furthermore, such meas- urements are used extensively in the evaluation of permanent impairment in

individuals with longstanding back problems.3 The findings of such evalu- ations then are used as the basis for decisions regarding disability and com- pensation. In addition, many exercises and manual therapy techniques are used by patients with back pain to increase back flexibility with the hope of decreas- ing symptoms and speeding recovery.

Numerous techniques have been de- veloped to assess spinal flexibility. Fingertip-to-floor distance and sit-and- reach measurements have been ob- tained.4,5 Skin distraction tests such as the standard or modified Schober method also have been used.6 Even sub- jective reports obtained through ques- tionnaires have been used to compile flexibility information.7 In addition to the use of standard goniometers, devices such as inclinometers and spondylome- ters have been developed for spinal flex- ibility measurements.8-11 Two- and three-dimensional roentgenogram anal- yses have been used in both research and clinical settings.12-14 Flexibility even has been studied by various in vitro means.15

For spinal flexibility measurements to be meaningful to clinicians or research- ers, they must have normative infor- mation and an understanding of how different variables affect ROM. The pur- pose of our investigation was to examine

the spinal flexibility of a large, adult population to establish normative val- ues for the measures used. In addition, we wanted to determine whether any individual characteristics affect flexibil- ity. Thus, the flexibility measurements were compared with other individual physical characteristics to gain a better understanding of how such factors affect spinal ROM.



All blue collar employees receiving hourly wages at a Northwest aircraft manufacturing plant were given the op- portunity to participate in our study. The employees who volunteered to par- ticipate in the study received no finan- cial remuneration, but the study exam- inations were conducted at various times during their working hours. The 3,020 subjects who participated in the study represented 75% of all employees solicited. The study group included 2,350 men and 670 women between the ages of 21 and 67 years (Tab. 1). No individuals who volunteered to partici- pate were excluded from the study for any reason. After informed consent was obtained from the volunteers, each sub- ject was given a physical examination by one of three physical therapists.

Ms. Batti’e is a research scientist, Department of Orthopaedics, RK-10, University of Washington, Seattle, WA 98195 (USA).

Dr. Bigos is Assistant Professor, Department of Orthopaedics, University of Washington.

Ms. Sheehy is a predoctoral research associate, Department of Statistics, University of Washington.

Mr. Wortley is Southern Division Rehabilitation Coordinator, Care Enterprises, Orange, CA 92668.

This study was supported in part by National Institute for Occupational Safety Health, Grant 5 R010H00982-03, and the Boeing Company.

This article was submitted October 7, 1985; was with the authors for revision 16 weeks; and was accepted July 10, 1986. Potential Conflict of Inter- est: 4.

Volume 67 / Number 5, May 1987 653

D ow

nloaded from https://academ

ic.oup.com /ptj/article/67/5/653/2728198 by APTA M

em ber Access user on 14 O

ctober 2020

Fig. 1. Markings and initial position for the modified Schober method are on the left. The final position from which the measurement is obtained is on the right.


The examination of each subject in- cluded assessments of standing height, sitting height, weight, shoulder flexibil- ity, and spinal flexibility in the sagittal and frontal planes. Because of time con- straints and the simplicity of the tests, each measurement was obtained only once. A modified Schober test was used to measure lumbar flexion.6 The test was chosen for its simplicity and because of its high correlation with forward flexion measurements of the lumbosacral spine obtained through radiographs (r = .97) as reported by Macrae and Wright.6 The modified Schober method requires only a plastic tape measure and pen to make three markings on the skin overlaying the lumbosacral spine. With the subject standing erect, the first mark is placed at the lumbosacral junction, as indicated by the dimples of Venus. A second mark is placed 5 cm below the lumbosacral junction, and a third 10 cm above the junction. The subject then is asked to bend foward as far as possible, as though to touch the toes, and the new distance between marks two and three is meas- ured. Lumbar flexion is expressed as the difference between this measurement and the initial distance of 15 cm (Fig. 1). Errors in identifying the lumbosacral junction of up to 2 cm result in an error of only a few degrees, as compared with flexion measurements obtained radio- graphically.6

TABLE 1 Study Group Profile

Age Group (yr)

20-29 30-39 40-49 50-59 60+

Men (n)

763 694 480 326


Women (n)

168 192 158 125 27

Other flexibility tests were chosen for their ease of application and popularity in various clinical settings. Sit-and-reach measurements were obtained to assess overall flexibility in forward flexion, with measurements recorded as the dis- tance (in centimeters) between the fin- gertips and the toes. The subject’s knees were straight throughout the test, and the ankles were maintained at 90 de- grees by having the soles of the feet pressed against a board placed perpen- dicular to the sitting surface. During the testing, lateral flexion was examined by recording the difference between the po- sition of the fingertips in erect standing and in maximal side bending to the right and left sides. The subject was instructed to keep the knees straight and both heels on the floor when bending to the side. Shoulder flexibility was assessed by hav- ing the subject lie prone with the arms stretched overhead while holding onto a wand. The subject then was asked to raise the wand as high as possible off the mat without bending the elbows or lift-

ing the chin off the mat. The distance (in centimeters) from the mat to the highest point that the wand was raised then was measured.

Data Analysis Data analysis included graphing the

data in the form of histograms, scatter plots, and box plots to depict more clearly the general distribution. We then tabulated the data, sorting the different ROM data by age and sex and using the means and standard deviations as sum- mary statistics. These tabulations were used to assist in the selection of variables to be included in the analyses of covar- iance (ANCOVAs) for each of the flex- ibility measures.16 In each ANCOVA, a flexibility measure was the dependent variable, and variables such as age, sex, height, and weight were used as inde- pendent variables.

Our goal when analyzing the data was to develop a regression model for each of the flexibility measures that would permit determination of the effects of variables such as age, sex, weight, or height on flexibility. When studying the effects of age, individuals were catego- rized by age group, for example, 20 to 29 years or 30 to 39 years. The effect of age then was modeled as a trend effect, such that subjects in the 20- to 29-year- old age group were given the value 1 for age, and those in the combined 50+ age group were given the value 4. This new variable was included as a dependent


D ow

nloaded from https://academ

ic.oup.com /ptj/article/67/5/653/2728198 by APTA M

em ber Access user on 14 O

ctober 2020

RESEARCH TABLE 2 Means and Standard Deviationsa of Flexibility Measurements by General Age Group

Age Group











763 168 694 192 480 158 326 125 87 27

Modified Schober

7.26 6.47 7.29 6.26 6.90 6.03 6.36 6.19 6.22 6.05


1.58 1.28 1.19 1.25 1.6 2.8 2.17 1.3 1.32 1.2

Sit and Reach

2.10 2.89

-0.48 2.40

-3.49 1.45

-7.43 -0.39 -8.98 -3.52


8.8 8.1 9.0 7.9 9.2 8.7 9.3 8.2

10.3 7.87

Side Bending (Right)

23.39 22.79 21.89 21.66 19.82 19.11 17.75 18.74 17.82 16.95


3.89 3.84 3.6 3.3 3.6 3.4 3.4 3.14 3.78 4.42

Side Bending (Left)

23.55 23.18 22.06 21.67 20.07 19.32 17.95 18.69 17.74 17.52


4.0 3.8 3.79 3.3 3.79 3.48 3.56 3.24 3.55 4.52

Shoulder Lift

32.61 30.24 29.77 27.72 24.90 23.81 20.71 19.96 18.25 16.36


11.3 10.2 11.04 11.9 10.4 8.7

10.3 9.1 9.6


variable in the ANCOVA. The AN- COVA allows for the control of other variables that may be associated with the specific variable under study. Thus, the true effect of the variable of interest on ROM is determined more precisely, and confounding factors are minimized.


The means and standard deviations for the various flexibility measures by age group and sex are shown in Table 2. The trends and distributions of the four flexibility measures, with medians and ranges, are displayed in Figures 2 through 5. All flexibility measures were correlated positively to one another; the degree of flexibility as assessed by one measure usually was similar to those assessed by the other measures. Corre- lation coefficients ranged from .38 to .44 when comparing sit-and-reach, lat- eral flexion, and shoulder flexibility measurements to one another. Because right and left side-bending scores were similar within individuals (r = .88), an average of the two measurements was used for further analyses. The modified Schober method was unique in that it showed little relationship to the other flexibility measures with correlation coefficients of .15 to .24, indicating that it may be measuring a different com- ponent of flexibility. A principal com- ponents analysis further demonstrated that the modified Schober method was explaining a different component of the variability in the flexibility measure- ments than the other measures.16

The ANCOVA demonstrated that the additional physical attributes also were related to the flexibility measures. With age and sex controlled, height was re- lated significantly to side-bending, sit-

Fig. 2. Range of shoulder flexibility measurements by age and sex.

Fig, 3. Range of sit-and-reach measurements by age and sex.

a All measurements are in centimeters.

Volume 67 / Number 5, May 1987 655

D ow

nloaded from https://academ

ic.oup.com /ptj/article/67/5/653/2728198 by APTA M

em ber Access user on 14 O

ctober 2020

Fig. 4. Range of modified Schober measurements by age and sex.

Fig. 5. Range of average side-bending measurements by age and sex.

and-reach, and modified Schober meas- urements. The regression model shows that for every increase in height by one standard deviation, the average side- bending measurement increased by 1.8 cm, the average sit-and-reach measure- ment decreased by 2.4 cm, and the av- erage modified Schober measurement increased by 0.6 cm. The ratio of stand- ing height to sitting height also was ex- amined. For every increase of one stand- ard deviation in the ratio of sitting height to standing height, decreases of 0.44 and 1.35 cm were estimated in the average side-bending and sit-and-reach measurements, respectively.

Obesity (defined as weight/height3) also was related significantly to the flex-

ibility measurements. For every increase in obesity by one standard deviation (calculated by using logarithms), in- creases of 0.4 and 1.0 cm were estimated in the modified Schober and sit-and- reach measurements, respectively.

After controlling for all of the addi- tional physical variables examined, we studied the relationships between ROM and the variables of age and sex. An interesting finding with regard to flexi- bility and aging was that a difference in the effects of aging appeared to exist between the sexes; the rate of decrease in spinal flexion, as measured through the modified Schober (Fig. 6) and the sit-and-reach techniques, was signifi- cantly less for women than for men.

Using a regression model, for example, the sit-and-reach measurement in the 20- to 29-year-old age group was 1.72 cm less for women than for men. In the combined 50+ age group, however, the measurement was 6.21 cm greater for women than for men (Fig. 7). The mean side-bending measurement decreased by 2.0 cm for every 10-year increase in age for both sexes.


Past studies have agreed that a general decrease in spinal ROM occurs in aging adults.5,13,17,18 Many conflicting reports exist, however, regarding the impor- tance of other factors in relation to flexibility. Many authors have found differences between the sexes, with most reporting higher spinal flexibility values for men than for women, particularly in the sagittal plane.5,13,17,18 Others report no significant sex differences.10,13 Our study demonstrated a decided effect of sex on ROM, even when controlling for other factors affecting ROM. Further- more, a difference in the effect of aging was demonstrated between the sexes. Troup et al first intimated this difference when they found lumbar ROM de- creased significantly in men but not in women in a group of 230 subjects.18 Macrae and Wright similarly reported that a decrease in sagittal mobility of the spine tended to be less in women than in men.6

The additional variables of standing height, ratio of standing height to sitting height, and obesity previously had not been studied extensively and were shown to have a significant effect on flexibility in the sagittal plane. These relationships cannot be explained easily, with the exception of the standing-to- sitting height ratio and the sit-and-reach test. Regarding the ratio of standing height to sitting height, individuals with longer rather than shorter than average trunks for their heights may have less difficulty reaching their toes, thus achieving a greater sit-and-reach meas- urement.

Another finding of clinical and re- search importance was that lumbar ROM, as measured through the modi- fied Schober method, bore little relation to the outcome of the sit-and-reach test. This finding adds further support to the idea that sit-and-reach, fingertip-to- floor, and other general measures of for- ward flexion are not adequate expres- sions of spinal flexibility. Considering


D ow

nloaded from https://academ

ic.oup.com /ptj/article/67/5/653/2728198 by APTA M

em ber Access user on 14 O

ctober 2020


Fig. 6. Trends in mean modified Schober measurements with age.

Fig. 7. Trends in mean sit-and-reach measurements with age.

that these measures are used currently at least as often as more specific meth- ods to examine overall lumbar mobility, such measures likely reflect mobility at the hips rather than at the spine.5

Similar to earlier studies,13,15 our re- sults indicate that ROM in the various age and sex groups covers a wide spec- trum of values and that a substantial overlap exists between the groups. We also found that spinal flexibility is af-

fected by many individual factors. Thus, the use of flexibility measurements as a screening tool for the identification of individuals who exhibit ROM values outside the accepted norms would ap- pear to be no easy task. To be meaning- ful, normative information must take into account such variables as age, sex, height, sitting-to-standing height ratio, and obesity. A more important issue in the use of flexibility measures, however,

might be how they relate to back health, or more specifically how spinal flexibil- ity relates to the risk for back problems or the recovery from them.

When reviewing the scientific litera- ture, little evidence exists to support the use of exercises to maintain or increase spinal ROM as a protective measure against back problems, despite their popularity. Biering-Sorensen examined spinal flexibility in a large group of in- dividuals, following them for one year for subsequent back injuries. He re- ported that men who experienced an episode of back pain had significantly greater flexibility than those who did not experience back pain.5 Similarly, How- ell found the incidence of low back pain was higher in elite women rowers who stretched their backs regularly than in those rowers who did not stretch their backs. A significant correlation also ex- ists between hyperflexion and an in- creased incidence of back pain.4 Reports on studies examining the relationship between spinal mobility and a history of back pain have been varied. Some re- searchers have reported a general de- crease in spinal mobility associated with a history of back problems,5,8 whereas others have reported the opposite of in- creased spinal mobility.19 Others report that a history of incapacitating back pain had no effect on spinal ROM.13 Longitudinal studies assessing premor- bid spinal flexibility are needed to define clearly the relationship between spinal flexibility and back problems. Only then will we know how to use fully the infor- mation provided by our study.


This study indicates that the spinal ROM of a large sample of adult blue collar employees is affected by many individual factors and covers a wide range of values. Determinations of what is normal, excessive, or diminished, therefore, must take age, sex, and other physical attributes into account. These considerations make the identification of individuals who lie outside acceptable ROM norms no simple task.

Volume 67 / Number 5, May 1987 657

D ow

nloaded from https://academ

ic.oup.com /ptj/article/67/5/653/2728198 by APTA M

em ber Access user on 14 O

ctober 2020

REFERENCES 1. Locke JC: Stretching away from back pain,

injury. Occup Health Saf 52:8-13, 1983 2. Rae PS, Waddell G, Venner RM: A simple

technique for measuring lumbar spinal flexion. J R Coll Surg Edinb 29:281-284, 1984

3. Guides to the Evaluation of Permanent Impair- ment, ed 2. Chicago, IL, American Medical Association, 1984, chap 1

4. Howell DW: Musculoskeletal profile and inci- dence of musculoskeletal injuries in lightweight women rowers. Am J Sports Med 12:278-282, 1984

5. Biering-Sorensen F: Physical measurements as risk indicators for low-back trouble over a one- year period. Spine 9:106-119, 1984

6. Macrae JF, Wright V: Measurement of back movement. Ann Rheum Dis 28:584-589, 1969

7. Bird HA, Eastmond CJ, Hudson A, et al: Is generalized joint laxity a factor in spondylolis- thesis? Scand J Rheumatol 9:203-205, 1980

8. Mayer TG, Tencer AF, Kristoferson S, et al: Use of noninvasive techniques for quantifica-

tion of spinal range-of-motion in normal sub- jects and chronic low-back dysfunction pa- tients. Spine 9:588-595, 1984

9. Dunham WF: Ankylosing spondylitis: Measure- ment of hip and spine movements. British J Phys Med 12:126-129, 1949

10. Loebl WY: Measurement of spinal posture and range of spinal movement. Annals of Physical Medicine 9:103-110, 1967

11. Fitzgerald GK, Wynveen KJ, Rheault W, et al: Objective assessment with establishment of normal values for lumbar spinal range of mo- tion. Phys Ther 63:1776-1781, 1983

12. Pearcy MJ, Portek I, Shepherd J: Three-dimen- sional x-ray analysis of normal movement in the lumbar spine. Spine 9:294-297, 1984

13. Tanz SS: Motion of the lumbar spine: A roent- genologic study. American Journal of Roent- genology, Radium Therapy and Nuclear Medi- cine 69:399-412, 1953

14. Stokes IA, Wilder DG, Frymoyer JW, et al: Assessment of patients with low back pain by biplanar radiographic measurement of interver- tebral motion. Spine 6:233-238, 1981

15. Taylor J, Twomey L: Sagittal and horizontal plane movement of the human lumbar vertebral column in cadavers and in the living. Rheumatol Rehabil 19:223-231, 1980

16. Johnson RA, Wichem DW: Applied Multivariate Statistical Analysis. Englewood Cliffs, NJ, Prentice-Hall Inc, 1982, chap 8

17. Moll JMH, Wright V: Normal range of spinal mobility: An objective clinical study. Ann Rheum Dis 30:381-386, 1971

18. Troup JDG, Hood CA, Chapman AE: Measure- ments of the sagittal mobility of the lumbar spine and hips. Annals of Physical Medicine 9:308-321,1968

19. Howes RG, Isdale IC: The loose back: An unrecognized syndrome. Rheumatology and Physical Medicine 11:72-77,1972


D ow

nloaded from https://academ

ic.oup.com /ptj/article/67/5/653/2728198 by APTA M

em ber Access user on 14 O

ctober 2020

Get Help for All Your Academic Writing Assignments. We are here to make your academic life more successful and less stressful. Over 10 years, we deliver the promised services. Think wise, buy custom essay! Just tell us “Quickly write my essay!” and provide us with the details of the order and instructions – we will make sure it will be completed on time by our professional writing team

Are you finding it difficult to write your essays, research papers or term papers. Hold on – expert help is on the way!