Abstract
Keywords
Introduction
Patients with chronic obstructive pulmonary disease (COPD) have airflow limitation and chronic respiratory symptoms. The effects of COPD, however, are not limited exclusively to respiratory function, and these patients have frequent comorbidity, including cardiovascular disease, depression, and muscle wasting. Respiratory limitation and neuromuscular deficits reduce both function and mobility.1,2 Marquis et al 3 demonstrated that routine daily activities, such as walking, create a high physiological demand (93% of peak oxygen consumption) in patients with COPD and put these patients at risk for muscle fatigue in lower extremities. Consequently, patients with COPD have slower gait speeds and have subtle alterations in their walking pattern during 6-minute walk tests when compared with healthy subjects.1,2,4 They also have impaired balance and slower reaction times in response to perturbations, and these deficits increase the risk for falls and fractures.5-7 We analyzed the gait parameters in COPD patients using video recordings to determine if there were any characteristic changes that might contribute to a slow gait in these patients and to determine if they have alterations in gait that are not readily explained by chronic lung disease. These results were compared to normative spatiotemporal gait parameters in healthy older subjects. 8
Methods
This is a pilot study of 14 patients with the diagnosis of COPD based on Global Initiative for Chronic Obstructive Lung Disease (GOLD) criteria. The primary goal was to make quantitative measurements of gait characteristics using video recordings. The secondary goals were to evaluate balance confidence using a survey instrument and to make qualitative assessments of gait and balance performance using the Tinetti Get Up and Go test. Inclusion criteria included any patient with COPD attending pulmonary rehabilitation at University Medical Center in Lubbock, Texas, or attending the pulmonary clinics at Texas Tech University Health Sciences Center in Lubbock. Patients with a history of cerebral vascular disease affecting gait, with neuromuscular disorders causing lower extremity weakness, with significant joint disease other than medically treated degenerative joint disease, and with a body mass index (BMI) greater than 35 kg/m2 were excluded. We also excluded patients using walkers or canes but not using portable oxygen. Clinical and demographic data (age, sex, BMI), other respiratory diagnosis, comorbidities, history of coronary artery disease, congestive heart failure, chronic use of oral corticosteroids, history of prior rehabilitation, pulmonary function tests (absolute and percent predicted of forced vital capacity [FVC] and forced expiratory volume in 1 second [FEV1]), and 6-minute walk tests were reviewed and recorded. Patients were classified by the GOLD criteria into the following stages: stage 1—the FEV1 is ≥80% of predicted; stage 2—the FEV1 is 50% to 79% of predicted; stage 3—the FEV1 is 30% to 49% of predicted; and stage 4—the FEV1 is <30% of predicted, and the FEV1/FVC ratio is <0.70 in all stages. This project was approved by the institutional review board at Texas Tech University Health Sciences Center in Lubbock. All subjects gave written and informed consent.
We assessed gait characteristics using visual recordings. Two walking trials over a distance of six to ten meters at a self-selected comfortable pace were video recorded at 50 Hz in the sagittal plane. During the first trial, the right side faced the camera. During the second trial, the left side faced the camera. Average spatiotemporal gait parameters, including stride length, cadence, walking velocity, and double support time, were calculated using Dartfish video analysis software (Dartfish, Alpharetta, GA, USA).These results were compared to normal values determined in the Mayo Clinic Study of Aging. 8
Clinical assessment of gait was done using the Get Up and Go test, the Tinetti balance tool, and the Tinetti gait assessment tool.9,10 The balance and gait assessment were scored by 2 raters and used to calculate a final average score. The patients practiced walking, the Get Up and Go test, and the Tinetti balance and gait test at least once before formal testing to acquaint them with the testing process. The patients completed the Activities-Specific Balance Confidence Scale questionnaires regarding history of falls and a fear of falls. 11
We used descriptive statistics to summarize baseline characteristics; the software program was SPSS, version 16.0.
Results
This is a pilot study with 14 patients; 13 patients attended pulmonary rehabilitation at University Medical Center, and 1 patient attended pulmonary clinics at Texas Tech University Health Sciences Center (Table 1). These COPD patients were elderly with mean age of 69.7 years and a mean FEV1 of 34.1% of predicted. Six patients were classified as GOLD stage 3 COPD, and 6 were classified as GOLD stage 4 COPD. The mean 6-minute walk distance was 320 m with a calculated gait speed of 0.9 m/s based on pulmonary rehabilitation center records. Fifty percent had coronary artery disease as a comorbidity.
Clinical and Demographic Characteristics of Study Patients.
FEV1, forced expiratory volume in 1 second; FVC, forced vital capacity; COPS, chronic obstructive pulmonary disease; BMI, body mass index; GOLD, Global Initiative for Chronic Obstructive Lung Disease.
GOLD stages are defined in the Methods section.
No patients routinely required the use of a cane or walker during daily activities. Three participants (21%) had falls within the past 6 months with minor injuries which did not require an emergency room visit, surgery, or hospitalization. The lowest percent of the Activities-Specific Balance Confidence Scale questions was 70.7% (on the question “How confident are you that you will not lose your balance or become unsteady when you walk outside on an icy sidewalks”) (Table 2). The mean total score was 91.8%. Overall, most patients rated their confidence in the range of 80% to 95%. Only 1 subject had an average confidence rating lower than 50%.
Balance Confidence.
The mean Get Up and Go time was 11.9 ± 2.0 seconds. The mean Tinetti balance assessment was 14.6 ± 1.1 (maximum score 16), and the mean of Tinetti gait assessment was 11.9 ± 0.4 (maximum score 14). No patients were at high risk for falls using a Tinetti balance and gait score total of less than 19 (maximum possible score 30). Gait characteristics based on visual recordings were step length 0.65 m, double support time ratio 30.4%, average free walking speed 65.3 m/min, and cadence 100 steps per minute (Table 3). These results were compared with normative gait parameters reported by the Mayo Clinic Study on Aging. These comparisons suggest that gait speed is slower and that the double support time and the double support time ratio are longer in patients with COPD.
Gait Analysis.
Healthy adult age 70-74 years; 27 men; 33 women, reference 12.
Speed, double support time, and double support time ratio are outside the 95% confidence interval calculated for 70- to 74-year-old men.
Discussion
These patients were elderly and had severe COPD. More than 50% had cardiovascular comorbidity. They had completed pulmonary rehabilitation at the time of this testing. Their timed Get Up and Go test results and the Tinetti assessments were essentially normal for this age group. They generally had relatively high scores on the Activities-Specific Balance Confidence Scale, but there was significant variability. One participant consistently had a low confidence rating. Gait analysis using visual recording indicated that they had a reduced step length and a reduced gait speed. They also had increased time in double support. This gait pattern might increase stability and reduce the risk of falls.
We used video recordings for gait analysis in this project. This approach can provide information about gait recorded in the sagittal plane and in the frontal plane. It allows relatively precise calculations of step length, cadence, and gait speed. It also provides information on the percent time spent in various gait phases and calculation of joint movement and angles. Stolze and coworkers 12 have demonstrated that test-retest variability in gait parameters in adults is low. In addition, the variability within a test is low. Video analysis has been used to make more precise measurements in individuals with very abnormal gaits, such with cerebral palsy, and in patients with neurologic disorders who have no obvious clinical disability.13-15 We speculated that video analysis of the gait in patients with COPD might identify abnormal gait parameters that were not apparent to simple inspection of the gait or identified by measurement of gait speed, and, in fact, these patients did have an increase in the time spent in double support.
Patients with COPD have reduced gait speeds when measured over both short distances and during 6-minute walks. 16 These reductions likely reflect respiratory limitation associated with muscle wasting, reduced muscle strength, and reduced physical activity resulting in the loss of conditioning.1,2,17 Patients with COPD have reduced muscle strength secondary to the systemic effects of chronic inflammation, reduced physical activity, weight loss, and possibly chronic corticosteroid use. Yentes and coworkers 18 reported that 10% of COPD patients evaluated in the National Health and Nutrition Examination Survey III survey had abnormal gaits consisting of a shuffle or limp or another walking abnormality identified by a physician evaluator. These investigators suggested that this may be explained by decreased physical activity in COPD patients. Pulmonary rehabilitation usually increases both 6-minute walk distances and gait speeds and can increase lower body strength with specific attention to muscle strengthening. Most of our subjects had completed rehabilitation; this program possibly changed (normalized) their gait patterns, but we do not have prerehabilitation information.
Patients with COPD also have impaired balance.5,17,19 Beauchamp et al 5 demonstrated that these patients have abnormalities in all subsystems related to balance and that these abnormalities were most noticeable in biomechanical constraints, anticipatory adjustments and transitions, and stability in gait. Identifying these balance impairments requires relatively sophisticated testing. The explanation(s) for these deficits in balance is uncertain and may reflect alterations in truncal mechanics associated with changes in the thorax, peripheral muscle weakness, and somatosensory deficits. 20 Balance is often worse after exercise, and consequently information obtained from a rested patient may not reflect the patient’s impairments after routine activities, such as using the upper arms. Annegarn and coworkers 4 used an accelerometer in patients with COPD doing a 6-minute walk test and found that the patients had a lower walking intensity, a lower cadence, and higher variability in mediolateral direction. The latter result was attributed to lateral foot placement to compensate for balance disturbances during forward propulsion. Balance can be improved with balance training during pulmonary rehabilitation.21,22 Our study demonstrates that COPD patients have a slow gait speed and spend more time in a double support stance. These changes in pattern may reflect subtle compensation for reduced balance.
Patients with COPD had more frequent falls and are at risk for important complications, such as fractures. 7 These falls are likely explained by reduced lower body strength and impaired balance. 6 Falls create a vicious cycle effect which results in reduced physical activity due to fear of falling and consequently more muscle deconditioning and atrophy. A history of falls and concerns about falling can be obtained easily from patients with short questionnaires. Patients with a history of falls or a fear of falling likely need more assessment. Twenty percent of our patients did have a history of minor falls, but this group did not differ from the other patients in their gait characteristics.
This study has several limitations. This was a pilot project that involved a relatively small number of patients who had completed pulmonary rehabilitation. Most of the patients had severe to very severe COPD, and these results may not be applicable to COPD patients in earlier GOLD stages. There was no age-matched control group, but results were compared with normal values of healthy older adults from the literature. Patients walked at a comfortable self-selected pace. It is possible that more abnormalities in gait would become apparent at a faster pace or following a period of exercise or if patients with a wider range of pulmonary function impairment were studied. Also it would be useful to know if participation in pulmonary rehabilitation changes gait parameters.
Conclusion
The spatiotemporal relationships or characteristics of the gait in these patients included slow gait speeds as expected and increased time spent in double support. This would contribute to the reduced speed and may compensate for subtle changes in balance. Slow gait speeds likely reflect complex interactions among aging, respiratory limitation, cardiovascular comorbidity with possibly decreased cardiac outputs and decreased perfusion to large muscle groups, muscle wasting, and subtle impairments in balance. The average patient with COPD probably does not need sophisticated assessment of balance or gait characteristics. These patients should complete short questionnaires about falls and concerns about specific activities which might result in falls and a timed Get Up and Go test, which measures lower body strength and gait speed. Patients with abnormal results on these preliminary assessments would need more evaluation by a physical therapist in an effort to modify their pulmonary rehabilitation programs to improve gait stability and reduce falls. Video recordings provide a method for analyzing gait and storing records for comparisons over time. In addition, these methods are applicable to many medical disorders, such as patients with diabetic neuropathy, and clinical problems, such as frequent falls, monitoring responses to treatment, and the natural progression of diseases affecting gait. Consequently, the methods and underlying constructs are relevant to primary care physicians during patient evaluation and care.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.