音声ブラウザご使用の方向け: SKIP NAVI GOTO NAVI

COMPARISON OF OXYGEN UPTAKE AND PHYSIOLOGIC COST INDEX (PCI) AT SUBMAXIMAL WALKING VELOCITIES

James J. Carollo, Jon Williamson, and Patricia Winchester Mobility Research and Assessment Laboratory / Department of Physical Therapy University of Texas Southwestern Medical Center at Dallas Dallas, Texas 75220 USA

ABSTRACT

The purpose of this project was to compare the inter-day reliability and effect of varying walking velocity on two measures of overall gait performance; oxygen uptake rate (VO2) and the physiologic cost index (PCI). Thirteen healthy subjects (6 male, 7 female) walked on a treadmill at four randomly ordered, submaximal walking velocities (3.0, 3.7, 4.4, and 5.1 mph) for approximately five minutes at each speed. At steady state, VO2, heart rate (HR), and PCI were recorded from each subject at all four speeds. The identical experiment was repeated approximately one week later, to determine inter-day reliability. Intra-class correlation coefficients (ICC1,1) were calculated for inter-day observations of VO2, HR, and PCI, grouping all speeds and all subjects (n = 52). Linear regression was used to determine the relationship between the three dependent variables (VO2, HR, and PCI) and walking velocity. The results indicated that VO2 and HR are more reliable than PCI for observations taken several days apart, primarily because of the need to obtain accurate resting heart rates when calculating PCI. However, PCI does demonstrate the same trend as VO2 and HR; all three increase with increasing walking velocity.

BACKGROUND

When attempting to characterize an individual's ability to walk, it is helpful to measure some quantity that can be used as an index of overall gait performance. Having such an index can be useful for determining the merits of a therapy or conditioning program targeted to improve a subject's gait. One good candidate for this index is to record oxygen uptake (VO2) during walking. Conventional methods for obtaining VO2 recordings on a breath-by-breath basis during gait usually require treadmill ambulation, and a face mask or mouth piece connected to a stationary instrument to capture and analyze the expired gases. Such methods are unreasonable or unavailable in many clinical environments, especially if the goal is to obtain an easily acquired index of overall gait performance rather than to quantify the cardiovascular system. Furthermore, the physical restrictions imposed by the measurement equipment may interfere with the subject's normal walking pattern, especially for handicapped individuals who often rely more heavily on visual input to substitute for reduced proprioception and kinesthesia. A more practical index would show a similar response to changes in workload as oxygen uptake, but be obtainable without the cumbersome and potentially disruptive data collection methods common to VO2.

While ambulatory recordings of average heart rate (HR) are easy to obtain and correlate well with VO2 at submaximal work intensities (1), use of HR as a gait performance index may not be ideal. This is because HR by itself does not include any functional measure describing the ability of the subject to accomplish the primary goal of walking; to efficiently move the body forward. The Physiologic Cost Index (PCI), first introduced by MacGregor (2), is the ratio of heart rate increase above resting to walking velocity. Algebraically,

PCI = HRwalking - HRresting/ average walking velocity

in units of beats/meter. PCI may be useful as a measure of overall gait performance, since it combines a physiologic measurement (HR), with a functional measure of gait (average walking velocity) in a single, easily obtainable index. It has been used clinically to evaluate the gait performance of children with cerebral palsy (3), elderly subjects enrolled in a walking program (4), and paraplegics using the ORLAU ParaWalker (5) and functional electrical stimulation (6). However, no study has compared PCI to VO2 prospectively in the same subjects.

RESEARCH QUESTION

This project was designed to substantiate the use of PCI as an index of overall gait performance, by comparing PCI to VO2 in terms of inter-day reliability and each measureÕs relationship to changes in walking velocity. If PCI was found to be equally reliable and displayed a similar relationship as VO2, this would support the continued use of the index in clinical environments where the use of traditional gas collection apparatus was impractical.

METHODS

Subjects: Thirteen healthy adult subjects (6 male, mean age 27.3 years, 7 female, mean age 30.0 years) without history of previous gait pathology volunteered for the study. Any individuals with significant health problems or a history of orthopedic, cardiovascular, or pulmonary dysfunction were excluded from the study. All subjects read and signed an informed consent form approved by the Institutional Review Board of the University of Texas Southwestern Medical Center prior to beginning the study.

Experimental Procedure: Each subject reported to the laboratory wearing comfortable workout clothing and running shoes. Disposable ECG surface electrodes were applied in a lead II configuration over the mid clavicular region bilaterally, and close to the left 7th intercostal space. These were connected to a Hewlett Packard model 78101 ECG telemetry system, whose receiver was connected via a signal conditioner to one of two counters on a National Instruments data acquisition board in a 286 class PC. After affixing a mouthpiece and nose clip to the subject for expired air collection, resting HR was recorded while the subject was quietly standing on the Pacer motorized treadmill used for the experiment. After allowing a brief warmup period, the treadmill was set to the first of four speeds (3.0, 3.7, 4.4, and 5.1 mph) selected in random order. Each subject walked or ran at the selected speed for a period of approximately five minutes, while continuously recording HR and expired gases breath-by-breath. Gases were collected using a Perkin Elmer model MGA-1100 mass spectrometer and ventilation rate was recorded using a Sensormedics VMM. Both devices were synchronized and recorded in real time using a custom hardware/software interface running on the same PC used for HR recording. Thirty second averages of HR, VO2, and PCI were obtained at two and five minutes into the selected speed. If the five minute sample was within ±5% of the two minute sample, it was assumed that steady state had been achieved, and the five minute recording was used for analysis. If not, additional 30 second averages were obtained until this criteria was met, at which point the treadmill speed was changed immediately to the new speed. This was repeated until all four speeds were tested, and the entire procedure was repeated approximately one week later.

Data Analysis: PCI, VO2, and HR data from all subjects were compiled in an Excelª spreadsheet on a Macintosh computer. The absolute (non-normalized) values were exported to another program for calculation of intra-class correlation coefficients (ICC1,1 - n = 52) for determining inter-day reliability of the three dependent variables. Normalized data were calculated by dividing each subject's steady state values by that individual's maximum value obtained over the four speeds for that testing day. These data were then exported into IGORª analysis software where least square linear regressions were calculated using all subjects at all speeds on both days (n = 104). Normalizing data prior to regression permitted direct comparison of the resulting curves.

RESULTS

The results of the inter-day reliability analysis using ICC1,1 are shown in Table 1.

Table 1: Inter-day Reliability Results

VO2 HR PCI
n 50 52 52
ICC1,1 0.885 0.789 0.292

The results for the regression analysis of VO2, HR, and PCI are shown in Graphs 1, 2, and 3 respectively.

Graph 1: Normalized VO2 vs. Velocity Graph 2: Normalized HR vs. Velocity Graph 3: Normalized PCI vs. Velocity All ICC's were calculated using both observations at each speed for all subjects. VO2 data had an n = 50 because the results for the two intermediate speeds on the second trial for one subject were lost due to a mass spectrometer failure. Despite this, the inter-day reliability was the highest among the three measures.

The linear regression graphs include the individual data points that were used to calculate the regression lines shown. The Pearson's linear correlation coefficients (r) for each of the graphs are compiled in Table 2.

Table 2: Correlation Coefficients for Regressions

VO2 HR PCI
r 0.919 0.8 76 0.663

DISCUSSION

Increasing numbers of clinical investigators are turning to PCI as a simple outcome measure that can describe the gait performance of their patients when other methods, such as oxygen uptake, are deemed inappropriate. By combining a physiologic measure (HR) with a fundamental gait performance measure (average walking velocity), PCI has the potential to satisfy this need. However, any measure considered as a replacement for traditional VO2 recording must not only demonstrate a similar relationship, but must also possess similar reliability. In this experiment, PCI met only the first criteria.

Insight into why the inter-day reliability of PCI is considerably less than VO2 can be obtained by considering the other variables comprising the PCI equation. Since the reliability of HR alone approaches the reliability of VO2 (0.789 vs. 0.885), and the walking velocities were fixed, the resting HR recorded at the beginning of each day must have a major influence on the reliability of PCI. In this experiment, resting HR was obtained while the subject was standing on the treadmill, with all expired gas collection devices attached. While this is desirable from the standpoint that the resulting increase in HR seen during the experiment is completely due to the gait task performed, it was apparent that there was a wide variation in the subjectÕs ability to control their HR prior to testing. Some subjects were able to lower their HR substantially on command, while others seemed more anxious, resulting in a temporarily elevated HR. In addition, outside factors such as caffeine level and time of day were not strictly controlled, which may have lead to inaccurate recordings during the resting HR period. While it is clear the reduced inter-day reliability of PCI cannot be completely explained by variations in the resting HR, we feel that great care must be taken when recording this reference if PCI is to be used in the clinic.

Despite this lower than expected inter-day reliability, the regression analysis illustrates that the relationship between PCI and velocity mimics the VO2 and HR graphs; increased workload yields an increased PCI. While definitely not as good a fit as either VO2 or HR (r = 0.663 vs 0.919 and 0.876 respectively), it appears that with enough samples PCI may prove satisfactory in those situations where more robust procedures are unavailable.

ACKNOWLEDGMENTS

The authors wish to thank Jack Wrobbel for developing the data acquisition system used to record real time oxygen uptake from the mass spectrometer, and physical therapy students Debra Aaron, Diane Blumenau, Robert Prater, and Sherri Rankin for all their efforts during the course of this project.

REFERENCES

1. Rose, J., Gamble JG, Medeiros , JM , et al, Energy Cost of Walking in Normal Children and in Those With Cerebral Palsy; Comparison of Heart Rate and Oxygen Uptake, J. Pediatr Orthop 9::276-279, 1989.

2. MacGregor, J, The Objective Measurement of Physical Performance with Long-Term Ambulatory Physiological Surveillance Equipment (LAPSE). Proc 3rd Int Symp on Ambulatory Monitoring, London, 29-39, 1979.

3. Nene, AV, Simultaneous Multiple Operations for Spastic Diplegia. Outcome and Functional Assessment of Walking in 18 Patients, J Bone Joint Surg (British), 75: 488-94, 1993. 4. Humerickhouse, R, Cassety, TM, Davis, CM, A Comparison of Endurance As Meas

ured by the Physiological Cost Index Among Elderly People, Physical and Occupational Therapy in Geriatrics, 11: 35-46, 1993.

5. Nene, AV, Jennings, SJ, Physiological Cost Index of Paraplegic Locomotion Using the ORLAU ParaWalker, Paraplegia 30: 246-252, 1992.

6. Winchester, P, Carollo, JJ, Habasevich, R, Physiological Costs of Reciprocal Gait in FES Assisted Walking, Paraplegia 32: 680-686, 1994.

Please address correspondence to:

James J. Carollo, M.S., P.E. Mobility Research & Assessment Laboratory 9705 Harry Hines Blvd., Suite 105 Dallas, TX 75220 - 214-351-2041