The Relation Between Walking Capacity and Clinical Correlates in Survivors of Chronic Spinal Poliomyelitis

Research Design 

A nonexperimental, exploratory research design was used. Subjects participated on 3 days (visits), with a minimum of 1 rest day between and a maximum of 3. The order of administration of the clinical measures is presented in appendix 1.

Participants 

A convenience sample was recruited from a university-based postpolio clinic. Recruitment was conducted through a publicity statement.

Two practitioners confirmed that the inclusion criteria for participating in the study were met. These included (1) a history of chronic spinal poliomyelitis (at least 15y postonset), (2) independent ambulation with or without a walking aid, (3) the ability to understand and read basic English, and (4) the capability to provide informed consent. People were excluded if they had (1) conditions that contraindicated performing a 6MWT; (2) insufficient hemodynamic stability (dysrhythmia, heart or lung disease) that would contraindicate exercise testing17; and (3) other neuromuscular and musculoskeletal conditions.

Baseline Assessment 

Before the initial assessment, subjects were asked to refrain from excessive exertion or exercise on the day before and on the day of the assessment. General information was collected as part of the clinical data—that is, name, sex, age, height, weight, and whether the diagnosis of poliomyelitis and PPS had been confirmed. Height and weight were measured with a calibrated medical scale by the same person across all participants. As part of routine clinical assessment, the following categories were assessed and recorded with a checklist: (1) current disease status including progression of late-onset symptoms, (2) a lifestyle overview, and (3) comorbidity. Types of walking aids were recorded including orthosis and braces (shoe/half/full), cane(s), crutch(es), or walker.

Primary Independent Variable: 6MWT Distance 

The primary independent measure of interest was distance walked in the 6MWT.18, 19, 20

Clinical Correlates 

The clinical correlates of interest in relation to walking capacity in the 6MWT included (1) rate-pressure product (RPP),21 (2) walk intensity—that is, Physiological Cost Index (PCI) ([peak heart rate – resting heart rate]/walking velocity),22, 23 (3) exertion (RPE),21 (4) pain (modified Borg scale),24 (5) fatigue (modified Borg scale),24 (6) lower-extremity muscle strength,25 (7) functional reach as an index of standing balance,26 (8) balance confidence (16-item Activities-Specific Balance Confidence [ABC] Scale),27 (9) corrected limb-length discrepancy,28 and (10) pulmonary function (FEV1, FVC, FEV1/FVC, percentage of maximum predicted).29

Procedures 

Data Analysis 

SPSS softwarea was used for data analysis. Descriptive statistics were used to summarize the subject descriptive data and the dependent variables. The intraclass correlation coefficient (ICC) was used to examine test-retest reliability of 6MWT distance, lower-extremity muscle strength, balance, and balance confidence across trials. Pearson product-moment correlation was applied to examine the relations between walking distance and clinical correlates (ie, PCI, RPP, RPE, complaints of pain and fatigue, lower-extremity muscle strength, balance, balance confidence, corrected limb-length discrepancy, pulmonary function) in our cohort. One-way analysis of variance was used to examine differences across repeated measures of the 6MWT distance, muscle strength, balance, and balance confidence over multiple trials. Multiple regression analysis determined the degree to which the selected clinical correlates of walking performance explained variation in 6MWT distance. The P value was set at less than .05.

The final analysis was performed using results of the last trial of each measure that was examined more than once—that is, the third trial of the 6MWT (including 6MWT distance, RPP, PCI, pretest and posttest RPE, pain, fatigue), the second trial of the balance measure, and the second trial of the balance confidence questionnaire. Measure of corrected leg-length discrepancy was performed on 1 trial and used for final analysis. Measure of pulmonary function was performed 3 times, and the best score was used for the final analysis.

Participants 

Twenty-one subjects were recruited. Two subjects were excluded because of unstable cardiopulmonary conditions.

A summary of the 19 subjects (12 women, 7 men) who completed the study appears in table 1. The mean age and standard error of the mean (SEM) were 62.2±1.9 years. Subjects’ average height was 1.67±0.02m and their average weight was 77.2±2.7kg. Body mass index (BMI) averaged 27.63±0.78kg/m2. Seventeen subjects had confirmed diagnoses of PPS and 2 were indeterminate. Six subjects needed walking devices (ie, canes or crutches) during the 6MWT, and 14 used brace(s) or insole(s).

Table 1. Subject Characteristics

	Subject Code	Sex	Age (y)	Height (m)	Weight (kg)	BMI (kg/m2)	Known PPS	Walking Device	Brace/Insole
	1	F	56	1.59	77.2	30.54	Yes	No	Yes
	2	M	63	1.73	79.4	25.79	Yes	No	No
	3	F	71	1.61	76.6	29.55	Yes	No	Yes
	4	F	58	1.54	61.4	25.89	Yes	No	Yes
	5	M	72	1.70	81.2	28.10	Yes	No	No
	6	M	61	1.71	99.3	33.96	Yes	Yes	Yes
	7	M	79	1.78	79.9	25.22	No	Yes	No
	9	F	63	1.62	81.5	31.10	Yes	Yes	Yes
	10	M	59	1.77	90.9	29.04	Yes	Yes	Yes
	11	F	53	1.62	62.8	23.93	Yes	No	Yes
	12	F	62	1.63	78.3	29.47	Yes	Yes	Yes
	13	M	51	1.85	83.6	24.43	Yes	Yes	Yes
	14	F	69	1.62	63.5	24.20	Yes	No	Yes
	15	F	60	1.65	60.4	22.18	Yes	No	No
	16	F	68	1.69	77.8	27.24	Yes	No	Yes
	18	F	48	1.61	60.5	23.34	Yes	No	No
	19	F	54	1.64	83.4	31.01	Yes	No	Yes
	20	M	60	1.72	98.5	33.30	Yes	Yes	Yes
	21	F	74	1.62	69.8	26.60	No	Yes	Yes
	Mean ± SEM		62.2±1.9	1.67±0.02	77.2±2.7	27.63±0.78

Abbreviations: BMI, body mass index; F, female; M, male; SEM, standard error of the mean.

Fifteen subjects were taking prescribed medication regularly (average, 2 medications/subject). No change in medication occurred over the course of the study.

Descriptive Statistics 

Descriptive statistics on 6MWT distance and the dependent variables (PCI, RPP, RPE, pain pretest, pain posttest, fatigue pretest, fatigue posttest, muscle strength, standing balance, balance confidence, corrected leg-length discrepancy, and FEV1/FVC) appear in table 2.

Table 2. Descriptive Statistics of the Measures

	Measures	N	Min	Max	Mean ± SEM
	6MWT distance (m)	19	128.4	591.4	394.2±27.7
	PCI (beats/m)	19	0.3	1.47	0.6±0.0
	RPP (mmHg by beats/min)	19	83	237	153±9.4
	RPE posttest (range, 0–10)	19	1	7	4±0.3
	Pain pretest (range, 0–10)	19	0	3	1±0.3
	Pain posttest (range, 0–10)	19	0	7	3±0.5
	Fatigue pretest (range, 0–10)	19	0	3	1±0.2
	Fatigue posttest (range, 0–10)	19	1	7	4±0.4
	Muscle strength (kg)	19	51.9	292.6	178.5±16.5
	Standing balance (cm)	19	6	38	27±2.3
	Balance confidence (%)	19	33.8	94.3	65.9±4.4
	Corrected leg-length discrepancy (cm)	19	0	3	1±0.2
	FEV1/FVC	19	0.59	0.96	0.80±0.79
	Valid N⁎	19

Abbreviations: max, highest measured value; min, lowest measured value.

⁎ Valid number of data.

Reliability of the Measures 

6MWT distance 

Table 3 includes individual data on 3 trials of walking distance. The table does not include reliability results. A high test-retest reliability was found between the 6MWT distance on trials 1 and 2 (ICC=.93), between trials 2 and 3 (ICC=.98), and between trials 1 and 3 (ICC=.90) (P<.05). There was a tendency for distance walked to increase from visit 1 through visit 3, however, this did not reach significance (F=.20, P=.82).

Table 3. Individual Data of 6MWT Distance on 3 Trials

	Code No.	No. of Subjects	Trial 1 (m)⁎	Trial 2 (m)⁎	Trial 3 (m)†
	1	1	419.4	466.8	461.0
	2	1	469.8	518.9	589.6
	3	1	453.6	481.5	447.4
	4	1	244.2	220.9	229.4
	5	1	377.4	354.7	395.5
	6	1	352.0	396.1	389.9
	7	1	359.7	370.9	377.5
	9	1	346.7	356.6	389.2
	10	1	302.2	314.8	303.5
	11	1	492.7	442.7	448.0
	12	1	149.1	132.0	128.4
	13	1	499.2	556.0	591.4
	14	1	301.8	335.7	336.4
	15	1	384.0	387.7	379.0
	16	1	416.0	440.9	416.8
	18	1	360.4	457.3	478.2
	19	1	514.8	512.0	523.1
	20	1	380.0	398.7	413.0
	21	1	244.9	192.0	192.0
	Total	19
	Mean	19	372.0	386.1	394.2
	SEM	19	21.9	25.9	27.7

⁎ Practice trial. † Test trial.

Validity of the 6MWT 

Correlations between 6MWT distance and clinical correlates 

Correlations between 6MWT distance and the clinical correlates of walking capacity appear in table 4. The Pearson product-moment correlation between 6MWT distance and RPP was not significant (r=.16, P=.52), but significance was observed between 6MWT distance and PCI (r=−.48, P<.05). The pretest measures of pain correlated with 6MWT distance (r=−.50, P<.05), but pretest fatigue did not correlate (r=−.35, P=.15). Furthermore, the correlations between 6MWT distance and subjective posttest measures (RPE, pain, fatigue) were not significant. Relations were nonsignificant between 6MWT distance and RPE at the end of the walk (r=−.18, P=.47), between 6MWT distance and pain at the end of the walk (r=−.29, P=.23), and between 6MWT distance and fatigue at the end of the walk (r=−.35, P=.15). The relations were significant between 6MWT distance and lower-extremity muscle strength (r=.62, P<.01), standing balance (r=.64, P<.01), balance confidence (r=.61, P<.01), pulmonary function (FEV1/FVC) (r=.61, P<.05), and corrected limb-length discrepancy (r=−.56, P<.05).

Table 4. Correlations Between the 6MWT Distance and the Clinical Correlates of Walking Capacity

	Clinical Correlates	Correlations With 6MWT Distance (Pearson r)
	PCI (heart rate change/walking speed)	−.48⁎
	RPP (mmHg by beats/min)	.16
	RPE posttest (range, 0–10)	−.18
	Pain pretest (range, 0–10)	−.50⁎
	Pain posttest (range, 0–10)	−.29
	Fatigue pretest (range, 0–10)	.15
	Fatigue posttest (range, 0–10)	−.35
	Muscle strength (kg)	.62†
	Standing balance (cm)	.64†
	Balance confidence (%)	.61†
	Corrected limb-length discrepancy (cm)	−.56⁎
	Pulmonary function FEV1/FVC (%)	.61⁎

⁎ P<.05. † P<.01.

Correlations between physiologic measures during 6MWT and clinical correlates 

The relations of RPP, PCI, RPE, and the following dependent variables were examined: pain, fatigue, lower-extremity muscle strength, standing balance, balance confidence, corrected limb-length discrepancy, and pulmonary function. RPP did not correlate with any of these variables. Walk intensity (PCI) correlated with balance confidence (r=−.50, P<.05) and pulmonary function (r=−.72, P<.01). The PCI did not correlate with RPP, RPE, pretest and posttest level of pain and fatigue, lower-extremity muscle strength, balance, and corrected limb-length discrepancy (table 5). RPE correlated with posttest fatigue (r=.67, P<.01) and balance confidence (r=−.54, P<.05) but not with RPP, PCI, pain and fatigue pretest, pain posttest, lower-extremity muscle strength, standing balance, corrected limb-length discrepancy, or pulmonary function.

Table 5. Correlations Between the PCI and Clinical Correlates of Walking Capacity

	Clinical Correlates	Correlations With PCI (Pearson r)
	6MWT distance (m)	−.48⁎
	RPP (mmHg by beats/min)	.38
	RPE posttest (range, 0–10)	.31
	Pain pretest (range, 0–10)	.08
	Pain posttest (range, 0–10)	−.17
	Fatigue pretest (range, 0–10)	−.04
	Fatigue posttest (range, 0–10)	.05
	Muscle strength (kg)	−.39
	Standing balance (cm)	−.32
	Balance confidence (%)	−.50⁎
	Corrected limb-length discrepancy (cm)	.28
	Pulmonary function (FEV1/FVC) (%)	−.72†

⁎ P<.05. † P<.01.

Participants 

Subjects were men and women who had a confirmed history of chronic spinal poliomyelitis. Our convenience sample was considered representative of people with chronic spinal poliomyelitis seen in a postpolio clinic for whom a walking assessment would be indicated.34, 39, 40 All except 2 had confirmed PPS7, 41 with varying degrees of neuromuscular and musculoskeletal dysfunction and ranging degrees of fatigue, weakness, and pain. The remaining 2 subjects had an indeterminate diagnosis of PPS.

The mean 6MWT distance in our sample was less than that reported by Noonan et al6 in people with PPS, but the criteria for inclusion in their study were an ability to walk on a treadmill with minimal support at a speed of at least 1.6km (1mph) and an ability to achieve a steady-state heart rate between 60% and 80% of the age-predicted maximum. In the present study, almost half of the subjects needed a cane or walker for walking indoors and had more severe walking disability. Muscle strength was impaired, and the pattern of weakness was asymmetric, which is consistent with the distribution pattern of the effect of chronic spinal poliomyelitis.11, 42 One side of the body was not considered necessarily more involved than the other. The variable muscle weakness and its asymmetric pattern are consistent with the literature on muscle status of survivors of poliomyelitis.5, 11, 31

Balance was impaired in our cohort. Although no study of polio survivors has previously assessed balance using functional reach, our results are comparable with those reported for people (men and women) of similar age with Parkinson’s disease43 and lower than those reported for healthy older adults.43 Similarly, balance confidence is reduced in people with chronic spinal poliomyelitis. In a study by Silver and Aiello44 on fear of falling, falls, and subsequent injuries in polio survivors, 179 of 233 reported fear of falling and 145 of those had made lifestyle changes. Fallers were more fearful of falling than nonfallers.44 Although falls in people with chronic spinal poliomyelitis are directly related to muscle weakness (ankle dorsiflexion, knee extension, knee flexion),3 anecdotally we observed no relation between lower-extremity muscle strength and number of falls in the last 6 months. Number of falls was based on subjects’ recall in our study; thus, further study is warranted. Balance confidence, however, was associated with both lower-extremity muscle strength and balance. These results lend support for a treatment focus on balance as well as strength.

Limb-length discrepancy is a common sequela of chronic spinal poliomyelitis.45, 46 Muscle weakness and atrophy in childhood interfere with growth, resulting in deformity and limb-length discrepancy.47 In a sample of 144 patients with chronic poliomyelitis, Ahmadi et al45 reported an average limb-length discrepancy of 5.2cm. A limb-length discrepancy of 3cm has implications for oxygen cost and RPE in older adults.13 We examined the relation between corrected limb length and walking distance by having subjects wear shoes and orthotics. Although 14 subjects used braces or insoles, 9 had a persisting limb-length discrepancy—that is, corrected limb-length discrepancy. Typically, the footwear and/or orthosis had not been modified for many years.

The FEV1/FVC, an index of restrictive impairment of the lungs associated with neuromuscular deficits, in our subjects was similar to predicted values (mean, 102.2%; range, 75.4%–124.3%). The mean lung function was higher than that reported by Dean et al,4 but in their study, subjects included those with spinal and spinal-bulbar poliomyelitis, whereas we excluded those with bulbar involvement. Nonetheless, Dean4 reported that subjects with chronic spinal poliomyelitis in the absence of bulbar involvement could exhibit marked restrictive lung pathology.

Test-Retest Reliability of the Measures 

Measures of lower-extremity muscle strength, balance, and balance confidence had high test-retest reliability.48 The reliability of muscle strength measures of individual lower-extremity muscle groups using the HHD was somewhat higher than that reported by Klein et al49 for individual lower-extremity muscle groups in survivors of poliomyelitis (mean ICC, .85; range, .52−.99). This may be explained by our strictly standardized measurement protocols. The same tester performed all measures, whereas in the study by Klein,49 3 physical therapists performed muscle testing. Also, the use of a fixed calculated position of the transducer likely increased the reliability of our measures. Our results of manual muscle testing using an HHD were reliable and comparable with those of other studies.5, 25 Use of the HHD has potential in clinical assessment in people with prior poliomyelitis. Further research is needed to refine the limits of its clinical application.

The test-retest reliability of the FRT26 in our study was comparable with that reported for elderly people (ICC range, .72−.92).26, 50 The FRT discriminates multiple fallers (≥2 falls in the 6mo before the study), nonmultiple fallers (<2 falls in the 6mo before the study), and a control group (with no known conditions affecting their balance or mobility) (F2.78=10.46, P<.001). The cutoff point between the multiple fallers and nonmultiple fallers was 25cm. The number of falls reported by our subjects was less than 2 falls over 12 months before the study (mean, 1.68), and the forward reach distance of our subjects was greater than the cutoff point reported by Giorgetti et al.50 This may reflect adaptation by people with chronic spinal poliomyelitis. Despite impaired balance their functional ability, such as functional reach, may be superior to that expected of people with such disability. Studies that investigate the strategies that people use during the test and the fall histories and patterns of people with chronic poliomyelitis are needed. This information will further refine assessment, diagnosis, treatment prescription, and outcome evaluation.

The reliability of balance confidence scores was high and consistent with that reported for older people37 and for people with lower-limb amputation.51

The test-retest reliability of 6MWT distance was high among the 3 trials. When the testing protocol is tightly standardized (based on updated criteria combining the original testing standards and the current literature on exercise testing), a single test may be representative of walking capacity and hence suffice as a valid assessment or outcome measure. However, because of the challenge of controlling extraneous variables in patient populations, other studies—including those from our clinic52, 53, 54—have advocated performing at least 1 trial of the 6MWT as a precaution. Our data support this contention. Although one can argue that 1 test may be sufficient statistically, examination of the raw data (see table 3) shows marked differences in the distances walked. The question arises about what distance is clinically versus statistically significant. The clinical significance of the change between the three 6MWT trials was not determined in this study, but the distance walked in all 3 trials ranged from 128.4 to 591.4m. The most difference observed between trials (1 and 2, 2 and 3, or 1 and 3) for each subject ranged from 8.7 to 119.8m, with a mean change of 43.3m. The percentage change ranged from 2.1% to 24.6%, with a mean of 10.6%. Although the difference between trials was not significant, the clinical significance has to be considered.

Validity of the 6MWT and Clinical Implications 

Our results indicate that neuromuscular and musculoskeletal impairments in people with chronic spinal poliomyelitis is a primary limitation to the distance walked during the 6MWT, which is consistent with previous findings.6 As we expected, there were no relations between the physiologic (aerobic) measures and the 6MWT distance and the index of myocardial work was not related to distance walked, although the PCI was inversely related. The lack of a relation between the PCI and RPE and RPP, and the association of the PCI with balance confidence and FEV1/FVC, indicate that walking performance in people with chronic spinal poliomyelitis reflects biomechanic factors to a greater extent than cardiovascular determinants.

There were no relations between subjective measures and the 6MWT distance in that we observed no association between the distance walked and fatigue pretest and posttest, posttest pain, and RPE. Pain before the walk test, lower-extremity muscle strength, balance, balance confidence, corrected limb-length discrepancy, and FEV1/FVC all were independently associated with the distance walked in the 6MWT. Such structural and functional impairments are commonly associated with walking disability in neuromuscular and musculoskeletal conditions9, 15, 55 and have implications for the assessment and management of these people to optimize walking performance and social participation, which are priorities in contemporary rehabilitation. Specifically, these results support optimizing muscle strength through modified strength training or rest,34, 35, 55 balance training, and correction of limb-length discrepancy and optimizing lung function as justifiable treatment priorities that may enhance walking capacity and walking training itself. Further studies are needed, however, to establish the degree to which each of these foci of management contributes to enhanced walking capacity. Vice versa, the relative contribution of walking training to enhanced strength, balance, balance confidence, lung function, and walking capacity needs to be elucidated. Optimizing even minimal limb-length discrepancy, for example, may disproportionately enhance walking performance. Whether there is a critical threshold needs to be established.

Balance and pain before walking predicted 67.9% of the distance walked in the test. Balance and balance confidence were impaired in our subjects, and 8 of the 19 subjects used walking aids. Postural stability has not been examined to a great extent in people with chronic spinal poliomyelitis, although commonly these people use walking aids and devices for support, have structural and functional muscular asymmetry, and deformity. Lord et al3 examined postural sway as an index of balance in people with chronic spinal poliomyelitis and reported that impaired postural sway was associated with weakness. Balance in sitting and standing activities has been reported to be associated with distance walked in people with stroke.15, 56 In the present study, balance was the main predictor of the 6MWT distance, which supports a primary clinical focus on balance to maximize walking performance in patients with chronic spinal poliomyelitis.

Pain before walking was a secondary predictor of walking performance, which supports the marked disability associated with pain as a primary complaint and risk factor for reduced quality of life secondary to reduced ambulation in patients with PPS.14 Both joint and muscle pain are common in people with chronic spinal poliomyelitis14 and are reported more frequently by women than men as contributing to a reduced health-related quality of life. Vasiliadis et al14 examined joint and muscle pain combined and as separate entities (whether or not subjects were experiencing muscle or joint pain) in people with PPS. They observed that joint pain reported by 78% of subjects was associated with initial motor unit involvement whereas muscle pain was not associated with such involvement, although it was reported by 68% of subjects. Also, the results showed that muscle and joint pain was related to BMI. Being overweight may contribute to overload on muscles and joints and, in turn, to pain.14 Furthermore, Trojan et al57 reported that recent weight gain is associated with PPS and can be used to distinguish patients with new weakness and fatigue from control patients without such complaints. Subjects in our study were overweight (BMI, 27.6±0.8kg/m2), and most subjects had reduced muscle mass. The BMI, however, was not associated with pain before or after testing or 6MWT distance. Nonetheless, nutrition and weight control are justifiable foci of treatment for energetic and biomechanic reasons.9

Overall, our results indicated that there are relations between the objective clinical correlates of walking distance we selected based on the literature and 6MWT distance but not between subjective measures and walking distance. The lack of association between walking performance and subjective parameters—namely, RPE and posttest pain and fatigue—supports the capacity of people with PPS to minimize their subjective complaints, which has been proposed as an adaptive coping strategy used by survivors of poliomyelitis over the years.58, 59 Thus, self-reports need to be interpreted cautiously.