Int. Rehabil. Med., 7, 93-98 0 Eular Publishers
Exercise testing of leg amputees and the result of prosthetic training
J. A. VAN ALSTB, H. E. P. CRUTS~, K. H U I S M A N ~ and J. DE V R I E S ~
Twente University of Technology; a t Roessingh Rehabilitation Centre; bde Stadsmaten Hospital: Enschedc: The Netherlands
Revised piqier uccvpted.for publication: May 1985. Correspondence 10: Dr. J. 4. van AlstC, Twente University of Technology, Dept. of Electrical Engineering.
Biomedical Engineering Division. PO Box 2 17, 7500 AE Enschede, The Netherlands.
Key words disease.
Arm exercise - Cardiac condition - Exercise testing - Exercise electrocardiography - Leg amputees - Peripheral vascular
Summary Thirty-nine patients undergoing rehabilitation following leg amputation were examined to determine cardiac status, which included clinical examination and a graded exercise ECG test, using an arm ergometer. Results were compared to final walking ability. I t was found that the cardiac status of these patients was generally poor and that the exercise ECG results did co-relate to walking ability.
Resume Trente-neuf patients en reeducation fonctionnelle apres avoir subi Iamputation dune jambe ont ete evalues sur le plan cardiaque. Ce bilan comportait un examen clinique et un ECG a leffort a Iaide dun ergometre adapte au membre superieur. Les resultats enregistrb ont ete confrontes a la capacite de marche de chaque sujet en fin de la periode dentrahement. On a constate que la plupart de ces patients presentaient des fonctions cardiaques alterkes et que IECG a Ieffort etait un bon reflet de la capacite residuelle de marche.
Zusammenfassung 39 Patienten wurden wahrend ihrer Rehabilitation nach Beinamputation auf ihren Herzbefund untersucht. Diese Untersuchung bestand in einer klinischen Befundaufnahme und in einer Ergometrie, wobei ein Arm-Ergometer benutzt wurde. Die Ergebnissc wurden verglichen mit der Gehfahigkeit beim Austritt. Es wurde festgestellt, dass der kardiale Status dieser Patienten im allgemeincn zu wiinschen iibrig l i e s und dass die Resultate der Ergometrie mit der Gehfahigkeit korrelierten.
Resumen Se exaniinaron treinta y nueve pacientes sometidos a rehabilitacidn tras una amputacion de pierna, con vistas a determinar su estado cardiaco. Ello incluyo un reconocimiento clinico y la prueba del ejercicio graduado con ECG, empleando un ergdmetro de brazo. Los resultados se compararon con la capacidad final para caminar. Se encontro que el estado cardiac0 de estos pacientes era generalmente malo, y 10s resultados del ejercicio en el ECG estuvieron correlacionados con la capacidad para caminar.
Eighty to ninety per cent of the some 2000 leg amputations performed annually in the Netherlands are because of cir- culatory problems caused by peripheral atherosclerosis I. Combined with the relatively high age of the patients with a vascular indicated leg amputation (50-90 years) this gives rise to the expectation that the cardiac condition is dimin- ished in a significant number of them.
Despile this expectation quantitative data on the cardiac condition of this population are scarce. To our knowledge only Kavanagh? reported on the cardiac condition of a group of elderly Canadian amputees which he investigated by exercise testing. He reported a history of heart trouble in almost half of the patients and found a substantial depres- sion of the ST-T segment of the electrocardiogram (ECG) during exercise testing in two-thirds of the group. He stated that more attention should be paid to the cardiac problems of the elderly amputee.
It is important to detect subjects with a poor cardiac condition before starting a prosthetic rehabilitation pro- gramme aimed at the restoration of the walking ability. A poor cardiac condition may both (a) influence the progress
Acknowledgment This work has been supported by a grant from the Nederlandse Hartstichtung (The Dutch Heart Foundation).
of the rehabilitation and (b) increase the risk of coronary problems during the rehabilitation process 3 .4 . These as- pects are in addition interrelated; a high coronary risk imposes, or possibly should impose, certain restrictions in the rehabilitation programme in order to remain within safety limits. Patients with a poor cardiac condition are possibly less inclined to exert themselves to the same degree as non-cardiac subjects.
As preliminary results of a study5 we observed relatively high cardiac loads during walking exercises by vascular amputees. This was not unexpected in view of the follow- ing.
(a) Comparison of trained amputees with leg prostheses with non-amputees walking at the same speed showed a higher energy consumption ranging from 10% in unilat- eral below-knee amputees to 40- 100/o in unilateral above-knee amputees6. 7 .
(b) An even higher energy consumption should be expected when untrained amputees are learning to walk again, while the uncoordinated gait pattern accomplished by fewer muscles requires additional effort.
(c) In the period before and after leg amputation the sub- jects have led a sedentary life which has diminished their physical condition *.
Since information concerning the vascular leg amputees is scanty, it was decided to evaluate the cardiac condition of
Int. Rehabil. Med., 1985. vol. 7 . no. 3
this group and to relate it to the degree of prosthetic walking attained at the end of the rehabilitation process, expressed in a three-level scale.
Graded exercise testing has been used with increasing frequency in the diagnosis of coronary artery disease. Exer- cise tests performed according to a suitable protocol provide valuable diagnostic information about the functional capac- ity of the cardiovascular system and form a rational basis for recommendations on physical activities in these pa- tients, l o . In this study we subjected all vascular leg ampu- tees admitted to our rehabilitation centre during a certain period to graded exercise testing in order to obtain an impression of their cardiac condition and physical work capacity.
Accordingly we defined cardiac condition as the maxi- mum heart rate obtained during graded exercise testing wilhout objective (as observed on the ECG) or subjective (as stated by the patient) cardiac problems. The heart rate is considered to be a good index of the myocardial oxygen consumption, and can therefore be used as an index for the cardiac energetic load. As a corollary the maximum heart rate is an index for the maximum cardiac load. By adopting the heart rate as reflecting cardiac energetic load we are able to make an explicit distinction between the cardiac and the total energetic load of the body. The latter can be calculated from the oxygen consumption.
The measure used for maximum physical work capacity was the maximum work load obtained during the test. There is no evidence of a linear relation between work load and heart rate in elderly leg amputees performing rowing ergo- metry (see Methods), as there is in bicycle ergometry in healthy subjects. Therefore we used both measures for eval- uation.
However, we were primarily interested in the cardiac load rather than in the patients physical work capacity.
The group of vascular leg amputees was examined in order to obtain an impression of
(a) the cardiac condition of amputees novitiating in a pros-
(b) the exercise-induced cardiac problems as visualized on
(c) the maximum heart rate provoked by exercise testing as
(d) the applicability of graded exercise testing in these
thetic rehabilitation programme;
a measure of the cardiac condition;
The level of prosthetic walking attained at the conclusion of the rehabilitation programme was established. The reha- bilitation result was compared with the findings of the exercise test performed at the initiation of the rehabilitation programme. This comparison was an attempt to correlate a patients exercise test performance at the beginning to the prosthetic walking result at the end of the rehabilitation programme.
Knowledge of a patients cardiac condition, and of the cardiac loads as induced by specific exercises, can be useful for the physiotherapist to keep the patient within safe levels of activity, while optimizing therapeutical intensity.
Patients The Roessingh Rehabilitation Centre in Enschede, The Netherlands, admits leg amputees for prosthetic rehabilita- tion programmes after their discharge from the hospital where they underwent amputational surgery. All patients with a leg amputation ensuing from peripheral vascular
disease who started a prosthetic rehabilitation programme at the Centre in the period January 1982-May 1983 were included in this study. The population consisted of 28 men and 11 women with a mean age of 67.4 years (range 43 to 92).
The level of the unilateral leg amputations in men (wom- en) was in seven (six) cases above-knee amputations, in nine (two) cases through-knee, and in eight (two) cases below- knee amputations. Of four male bilateral leg amputees two patients had both through-knee amputations, one patient had two below-knee and one patient had one above-knee and one below-knee amputation. One female patient had two below-knee amputations.
Additional diagnosis included 18 cases of diabetes melli- tus. Other non-cardiac factors present that could adversely affect the prosthetic rehabilitation were: cerebral vascular accident (three patients), blindness (two), deafness (one), arthritis (one), lung embolism (one), and Parkinsons dis- ease (one). A standard 12-lead ECG was obtained imme- diately before the exercise test.
The exercise test The graded exercise test was performed on a specially designed ergometerI*, as shown in Fig. 1. It consists of a load unit on which the seated patient exercises by turning two cranks against an adjustable torque. Both cranks point in the same direction so the patient pushed or pulled with both arms simultaneously during exercise. The resul ting rotation permitted coordinated trunk movement with the use of large body muscles. We called the apparatus provi- sionally a rowing ergometer.
Figure 1 The rowing ergometer in operation.
The patient was seated on the ergometer chair which was positioned at wheelchair height. The patient was secured with a hip seat belt and swivelled into the exercise position. The height of the chair was adjusted and fixed so that the fulcrum of the cranks was at shoulder height and the load- unit position was adjusted to the length of the patients arm. The patients were asked to maintain a speed of 50 revolu- tions per minute although the load controlled by the load unit (Lode instruments, type Angio) is in a wide range independent of the rotational speed. Speed was indicated by means of a green lamp (when within the accepted h i t s , 4 6 - 5 4 revolutions per minute) and two red lamps (when outside these limits). This speed indication was also easily interpreted by patients who were visually handicapped.
The test protocol was as follows. First the load was sct at
VAN ALSTE et a/.: Exercise testing of lee. amDutees
zero to allow the rotating parts of the ergometer to attain working speed, followed by an initial load value of 30 W. The load was increased every minute by 10 W until the patient was unable to continue the exercise or the physician stopped the test.
The electrocardiogram was recorded using the bipolar leads CC5 and CM5 1 3 . The baseline wander, as well as signal and noise components over 100 Hz, were removed by a hardware linear phase bandpass filter
As stated before we used the ECG to detect an upper limit in the attainable cardiac load. If a significant ST-segment depression or rhythm disturbance occurred, the heart rate involved was considered the maximum cardiac load. The detection of ST segment depression, abnormal beats and rhythms were required for this analysis of the ECG. If no ECG abnormalities were encountered, maximum cardiac load was identified as the heart rate obtained at maximum work load. The patient was then unable to continue the test, as a result of general or local fatigue, dyspnoea, etc.
A specially developed hardware device detected QRS complexes and measured interbeat intervals which were transmitted to an Apple-I1 computer. The computer also controlled the ergometer work load protocol and recorded and displayed the heart rate versus work load. The comput- er programme was written in Basic. The unfiltered and filtered ECGs were recorded on a Siemens Elema Oscillom- ink for 10 seconds every minute, also under computer control.
Arterial blood pressure was measured before and within 15 seconds after conclusion of the exercise. The tests were performed under supervision of a cardiologist, who also took the patients history, performed the physical examina- tion and judged the rest and exercise ECG.
Quantitative evaluation of the electrocardiogram Besides the simple on-line processing and monitoring as described previously, the ECG was further analysed from an analog recording. The processing was performed by a PDP 11/23 computer system extended with a special-purpose processor called VIPERiJ for some time-consuming mathe- matical operations. It consisted of digital filtering, for removal of base wander, line frequency disturbances and componcnts over 100 Hz. All ECG beats were detected and then classified according to their morphology. This was done by continously calculating the correlation coefficient R(t) between the incoming ECG and a number of templates representing the actual QRS morphologies. When R(t)>0.92, then the ECG beat was assigned to the appro- priate beatclass and the point of time t where R(r) was maximum, was used as fiducial point for averaging. As known from a pilot study, ectopic beats are often present in the E(Gs of elderly leg amputees. Averaged beats were obtained from each morphological class in order to remove noise duc to exercise, and updated to enable measurement of morphological parameters throughout the test. The heart rate was plotted versus exercise time, as was the percentage of ECG heats incorporated in a specific morphological beat class. Morphological and time parameters were plotted after computer measurement from the most prominent beat class, containing the normal beats (by definition). Fig. 2 shows a typical example of averaged beats and presents the morphological development of a beat class with the exercise time. The morphological parameters were measured on this type of averaged beat.
Prosthetic rehabilitation result The result of the completed rehabilitation programme was classified according to the level of wa...