Effects of warming up on reliability of anthropometric techniques in ankylosing spondylitis.код для вставкиСкачать
549 BRIEF REPORT EFFECTS OF WARMING UP ON RELIABILITY OF ANTHROPOMETRIC TECHNIQUES IN ANKYLOSING SPONDYLITIS W. NEAL ROBERTS, MATTHEW H. LIANG, LORNA M. PALLOZZI, and LAWREN H. DALTROY To determine the effect of warm-up on anthropometric measures, 10 patients with ankylosing spondylitis and 10 normal control subjects were measured over a l-hour period by one therapist. A series of 4 measures-a modified Schober’s test, finger-to-floor with the subject standing on a 23-em stool, chest expansion, and cervical rotation using an arthrodial protractorwere repeated completely 4 times. Subjects exercised between each series (except between the third and fourth series). Intra-rater reliability was determined by comparing measurement 3 with measurement 4, and the warm-up effect was determined by comparing measurement 1 with measurement 3. The Schober’s test and chest expansion are stable measures and are not affected by warm-up. However, the finger-to-floor and cervical rotation exercises require warm-up to ensure stable values for clinical or research followup of spondylitis. Anthropometric techniques have long been used to diagnose and monitor ankylosing spondylitis (AS). The measurement of spinal motion was first From the Departments of Medicine, Rheumatology and Immunology, Brigham and Women’s Hospital, Harvard Medical School, Robert B. Brigham Multipurpose Arthritis Center, Boston, Massachusetts. Supported by the World Rehabilitation Fund and NIH grants AM-36308 and RR-05669. W. Neal Roberts, MD: Instructor in Medicine, Harvard Medical School, and Arthritis Foundation Fellow (current address: Medical College of Virginia, Richmond); Matthew H. Liang, MD, MPH: Associate Professor of Medicine, Harvard Medical School; Lorna M Pallozzi, RPT; Lawren H. Daltroy, DrPH: Instructor, Harvard Medical School and Harvard School of Public Health. Address reprint requests to Matthew H. Liang, MD, MPH, Robert B. Brigham Multipurpose Arthritis Center, 16B, Brigham and Women’s Hospital, 75 Francis Street, Boston, MA 02115. Submitted for publication February 14, 1986; accepted in revised form October 13, 1987. Arthritis and Rheumatism, Vol. 31, No. 4 (April 1988) quantitated by Schober, by Moll and Wright, and by Sturrock and coworkers (1-3). Numerous quantitative methods have been proposed for assessing patients with AS, including radiologic techniques ( 4 3 , measurement devices such as the spondylometer (6), the goniometer (7), the inclinometer, and the cervical protractor @),other physical measurements such as chest expansion (9), occiput-to-wall, intermalleolar straddle, finger-to-floor, and neck rotation (lo), and various skin distraction techniques (1,11,12). No one technique is completely satisfactory, and the correlation of measurement techniques to one another ( 5 ) or to evaluation of function in patient’s terms is arguable (13). In one study of patients with rheumatoid arthritis or osteoarthritis, range of motion accounted for about one-half the variation in function (14). Range of motion incorporates many factors. For example, spinal range of motion decreases with age, shows wide variation within each decade, varies by sex, and is influenced by the presence of hip or knee disease (2). Nevertheless, the majority of published measurements incorporate, as a minimum, the Schober’s test, the modified Schober’s test used by Moll and Wright, and chest expansion. In evaluating measures for use in a clinical trial, we studied the effect of warm-up on measurement stability and reliability. Our findings are relevant for the choice of measurements for clinical trials and for clinical practice. PATIENTS AND METHODS Ten normal subjects and 10 AS patients from a registry at our center were studied after informed consent was obtained. The patients were evaluated by one physician to determine disease and medical char- BRIEF REPORTS 550 acteristics. Patients were entered into the study if their rheumatologist believed that they had a spondylarthropathy and radiographs showed an abnormality in at least one sacroiliac joint. Current radiographs were not obtained. Existing radiographs were reviewed independently in an unblinded manner by 2 rheumatologists not involved in care of the patients. Discrepancies between written evaluations by the clinical radiologist, clinical rheumatologist, or either of the study rheumatologists were resolved in favor of the lower New York criteria grade (0-IV) (15). In 3 patients whose films were not available for review, verbal or written reports of involved sacroiliac joints were accepted as an indication of sacroiliitis. One patient had bilateral grade IV sacroiliitis and 1 had unilateral grade I changes; the remaining patient had sacroiliac changes on both sides. Back pain was relieved by antiinflammatory medication in 9 patients and by exercise in 7. Clinical evaluation showed that only 1 patient had advanced disease, with severe impairment of spinal mobility and a high score on the Health Assessment Questionnaire (16). The Health Assessment Questionnaire detected some functional impairment for the group as a whole: 4 patients had a disability index of 1 .O, 5 patients scored between 1.O and 1.5, and 1 patient with longstanding AS had a disability index of 2.65. We selected 4 measures for their convenience and because normative information was available. Each physical measurement was taken 4 times independently, by one observer (LMP). Warming up or stretching was done between measurements 1 and 2 and between measurements 2 and 3. In general, these consisted of 3 repetitions of the motion necessary for the measurement. Intra-rater reliability was studied between measurements 3 and 4.The precise protocols for each measure were as follows. Smythe test (10-cm segment method). The 10-cm segment method of the Smythe test requires that the subject stand with feet at shoulder width. A tape is used to mark the width of the stance. A line is drawn between the posterosuperior iliac spines (PSIS). The subject bends forward with knees straight, and 3 marks, at 10-cm intervals along the spine, are made in a cephalad direction from the PSIS line. The subject extends the spine in a prone position with arms straight, allowing the abdomen to sag toward the examining table. The 10-cm segments are then remeasured to give the resultant total measure. The Smythe test has been used for 15 years in Canada and has been described in a recent publication (12); its results are highly correlated with results of the modified Schober's test used by Moll and Wright, the Schober's test itself, and goniometric assessment. The 2 additional evaluation items on the Smythe test are spinal extension and mobility of the lower thoracic and upper lumbar regions. Finger-to-floor. Distance between the fingers and the floor was measured with the subject standing on a 23-cm stool (with stance marked to keep it constant), bent forward as far as possible with fingers reaching toward the floor. If the subject reached beyond his toes, the values were recorded as negative numbers. Chest expansion. Chest expansion was measured with the subject standing, hands on head and the tape placed circumferentially over the xiphoid. The subject inhaled maximally, then exhaled, and a value was recorded; the subject then inhaled maximally and a second value was recorded. Total change was recorded as the value at maximum inspiration minus the value at maximum expiration. Cervical rotation. Cervical rotation was measured with a protractor, with the subject seated (8). The 0" center point of the protractor was placed over the episternal notch and the 90" lines on the protractor were aligned with the antra of the ears. The cervical range of motion was recorded at the extremes of cervical rotation in both directions, and the total change was obtained by adding the measurements for the left and right rotations. Statistical analysis included paired-subject ttests to determine the significance of differences between the first and third measurements (the warm-up Table 1. Characteristics of subjects Men Age range (mean) Taking nonsteroidal antihlammatory drugs Radiographic involvement Sacroiliac Hip Lumbar Thoracic Cervical Current low back symptoms History of inflammatory low back pain Ankylosing spondylitis patients Normal subjects (n = 10) ~.. .., (n = 10) --, 9 9 6 31-54 (35) 0 10 - 2 - 20-47 (35) 1 0 1 9 9 \- 6 0 55 1 BRIEF REPORTS effect) and simple correlation between the third and fourth measurements to determine reproducibility. The paired t-test was used for change scores, to allow each subject to serve as his own control, taking into account the wide variability between subjects on baseline measurements. Pearson’s product correlation was chosen as the simplest representation of reproducibility, although this statistic is difficult to interpret when between-subject variance is small. RESULTS The characteristics of the AS patients and controls are presented in Table 1. Figure 1 shows the medians and ranges of the measures at each determination for patients and normal subjects. Table 2 summarizes the observations and indications that the intra-rater reliability was high, and shows the effect of warming up for each measurement on normal controls and AS subjects. Patients with AS all showed the same trend toward improvement with warm-up, except in an instance in which a patient with hamstring spasm during the finger-to-floor measurements regressed on the third measurement. As seen in Table 2, the AS patients’ improvement after warm-up was highly statistically significant (P < 0.002) for the “r OVERALL Reliability and susceptibility to warm-up of 4 measurement techniques in ankylosing spondylitis (AS) patients and normal subjects Table 2. Measurement, patient group Overall Smythe AS Normal Lower (lumbar) Smythe AS Normal Middle Smythe AS Normal Upper Smythe AS Normal Chest expansion AS Normal Finger-to-floor AS Normal Cervical rotation AS Normal 6-13 A 20-75 40-00 2 5 - 8 5 I RANGES A - 1 3 - + 4 4-135 3 3s-80 30-85 3 5 - 0 5 0.08 0.15 0.98 0.47 0.28 -0.13 0.34 >0.5 0.90 0.83 0.28 0.30 0.03 0.22 0.82 0.44 0.43 0.90 0.02 0.01 0.95 0.86 0.15 0.25 0.98 0.97 - 4.70 -4.85 <0.002 <0.01 0.96 0.83 18.1 16.2 <0.002 <O.OOI ------- 0-t43 160[ 2 3 4 -19-+2 -3-+395 -23-+25 -23-+45 -45-+365 -50-+36 CERVICAL ROTATION 4 MEASUREMENT MEASUREMENT RANGES 1 .oo MEASUREMENT 55-135 5 2 0.72 \ MEASUREMENT 85-13 5-145 0.97 0.84 FINGER-TO-FLOOR SMYTHE 19-13 55-140 3 5 - 0 5 40-105 RANGES A 110-160 25-145 120-165 3 5 - I50 120-175 3 5 - 160 110-170 41 160 - Figure 1. For each of 4 anthropometric measures, median (not mean) values before and after warm-up stretching are displayed for 10 normal controls (A) and 10 ankylosing spondylitis patients (0).In each panel, the origin is non-zero for the ordinate, and the ranges of the y values are listed below the x-axis. (4) P$ >0.5 0.4 sion, and finger-to-floor values are in cm, cervical rotation values are in degrees.. $ Paired t-test. oi 4-125 Warm-up effectt * Between measurements 3 and 4 (Pearson product correlation). t Difference between measurements 1 and 3. Smythe, chest expan- -lo RANGESA Intra-rater reliability* BRIEF REPORTS 552 finger-to-floor and cervical rotation measures. Low between-subject variance was noted for normal subjects in the Smythe test and may explain the apparent poor reproducibility of results in this group. up selectively affects some measures is a subject for future investigation. DISCUSSION REFERENCES A variety of factors influence anthropometric techniques, but the individual contribution of these factors to measurement error has not been studied in detail. That warm-up influences patient performance is not surprising, but the published literature in this field makes no mention of this factor, and our observation that some measures are not affected seems somewhat paradoxical. How does one explain our findings? Improved performance with repetition may be influenced by the patient learning what is expected, or by the examiner providing encouragement. We judge this explanation to be unlikely, however, since our subjects neither had difficulty comprehending the task nor were given any unusual encouragement. If improved motivation or understanding had been a factor, we would have expected it to have been operative for all measures. Variation between measurements might have been due to the timing of medication taken by AS patients, but this cannot explain the similar results in controls. A more probable explanation may be that the techniques are measuring different joints or muscles, some of which are inherently more susceptible to warm-up than are others, o r that some movements, such as chest expansion, are used constantly and thus, are maximally warmed up at the start. Parenthetically, we also observed that some subjects with severe AS improved their performances after the initial warm-up but then had decreased mobility, presumably due to spasm and pain. In addition, cervical rotation takes as many as 6 repetitions before maximum warm-up is achieved and measurements are stabilized. Warm-up improves performance for patients and normal subjects in the measures of finger-to-floor distance and cervical rotation, but not in the overall Smythe test or chest expansion measures. Other measurement techniques may also be affected by warmup. Measurement error for finger-to-floor distance and cervical rotation may be reduced by a warm-up protocol to achieve stability of performance. Why warm- 1. Schober P: Ledenwirbel saoule und kreuzachmerchen. Munchener Medizinische Wochenschrift 84:336, 1937 2. Moll JMH, Wright V: Normal range of spinal mobility: an objective clinical study. Ann Rheum Dis 30:381, 1971 3. Sturrock RD, WojtulewskiJA, Hart FD: Spondylometry in a normal population and in ankylosing spondylitis. Rheumatol Rehabil 12:135-142, 1973 4. Tanz SS: Motion of the lumbar spine: a roentgenological study. AJR 69:399-412, 1953 5. Portek I, Pearcy MJ, Reader GP, Mowat AG: Correlation between radiographic and clinical measurement of lumbar spine movement. Br J Rheumatol 22: 197-205, 1983 6. Dunham WF: Ankylosing spondylitis: measurement of hip and spine movement. Br J Phys Med 12:126-129, 1949 7. Boone DC, Azen SP, Chun-Mei L, Spence C, Baron C, Lee L: Reliability of goniometric measurements. Phys Ther 11:1355-1360, 1978 8. Cheshire MW: New apparatus: a device for measuring rotation of the neck. Arch Phys Med 38592, 1957 9. Moll JMH, Wright V: An objective clinical study of chest expansion. Ann Rheum Dis 31:l-8, 1972 10. Defibaugh JJ: Measurement of head motion. Amer J Phys Ther 44:157-162, 1964 11. Macrae IF, Wright V: Measurement of back movement. Ann Rheum Dis 28584-589, 1969 12. Miller MH, Lee P, Smythe HA, Goldsmith CH: Measurement of spinal mobility in the sagittal plane: new skin contraction technique compared with established methods. J Rheumatol 11507-51 1, 1984 13. Badley EM, Wood PHN: The why and the wherefore of measuring joint movement. Clin Rheum Dis 8533-544, 1982 14. Badley EM, Wagstaff S, Wood PHN: Measures of functional ability (disability) in arthritis in relation to impairment of range of joint movement. Ann Rheum Dis 43:563-564, 1984 15. Moll JMH, Wright V: New York criteria for ankylosing spondylitis: a statistical evaluation. Ann Rheum Dis 32:354-358, 1973 16. Fries JF, Spitz P, Kraines RG, Holman HR: Measurement of patient outcome in arthritis. Arthritis Rheum 23:137-145, 1980 Acknowledgment. The helpful comments and assistance of Dr. Steven H. Stern, MD, are appreciated.