Journal of Rehabilitation R&DVol . 20, No . 1 .1983 (B pn10a8)Pages 21–30

JOHN A . GRUNER, Ph . D ."ROGER M. GLASER, Ph. D .'STEVEN D. FEINBERG, M .D.STEVEN R. COLLINS, M .S.NOEL S. NUSSBAUM, Ph. D.

Department of PhysiologyWright State University School

of MedicineDayton, Ohio 45435

Rehabilitation Institute of OhioMiami Valley HospitalDayton, Ohio 45409

Physiology Research LaboratoryVeterans Administration Medical CenterDayton, Ohio 45428

a This work was supported in part by theRehabilitative Engineering Research andDevelopment Service of the Veterans Ad-ministration, and by the MontgomeryCounty Easter Seal Society.

Present affiliation : Department of Neuro-surgery, New York University Medical Cen-ter, 550 First Ave ., New York, N .Y . 10016.

Please address all correspondence to : Dr.Roger M . Glaser, Department of Physiol-»uv Wright State University School ofMedicine, Dayton, Ohio 45435 (Phone : 513-873-2742) dddd.

A System for Eva uation andExercise-Conditioning ~~ofPara Uyzed Leg Musc 0 esa

ABSTRACT

The purpose of this project was to develop instrumentation andprotocols in which electrical stimulation is used to induce exercise inparalyzed quadriceps muscles strength and endurance evaluationand conditioning . A computer-controlled m!eo1rioel stimulation oyo-tenn ' using surface electrodes, automatically regulates the bouts ofleg extension exercise . Load weights attached just above the anklescan be progressively increased over a number of training sessions insuch a manner that a measure of the fitness of the legs can beobtained. With three exercise sessions per week for 9 weeks, thestrength and endurance of the quadriceps muscles of two paraplegicand four quadriplegic subjects were gradually and safely increased.During exercise ate X load weight of 5 .4 kg, X heart rate did not riseabove rest, whereas systolic blood pressure increased about 20 mmHg, and skin temperature above the active muscles increased about1 .75°C. Such exercise conditioning appears to be safe and mayprovide important health benefits, including improved fitness of themuscles and bones, better circulation in the paralyzed limbs, andenhanced self-image . Conditioned electrically stimulated paralyzedleg muscles may be used for locomotion in conjunction with specialvehicles. ddSDLSDKJF:LKDJFDK:SJFDS:KLFJsldkjfdddd

INTRODUCTION

To improve fitness and locomotive capability of individuals withparalyzed legs, a wheelchair-like vehicle which can be propelled byelectrically stimulated knee-extensor muscle contractions has beenconstructed and successfully tested (1, 2) . Safe and effective use ofsuch a vehicle will, however, require certain minimal performancecapabilities of the paralyzed legs . Therefore, the strength and endur-ance of the muscles to be stimulated must be evaluated and, ifnecessary, conditioned by exercise prior to prescription of a legpropelled vehicle (LPV).

After a prolonged period (months to years) of disuse, a paralyzedendurance-type muscle tends to atrophy and undergo histochemicalchanges, a process in which twitch duration shortens and fatigabilityincreases when the muscle is electrically stimulated (3-5) . Previousstudies have shown that strength and/or endurance can be increasedin paralyzed muscles by electrical stimulation exercise programs (6-10) . This suggests that exercise protocols which are used to improvestrength and endurance via voluntary exercise may be adapted forconditioning paralyzed muscles via e!enirioal stimulation . Unfortu-nately, there are few studies in the literature describing instrumenta-tion or procedures for conditioning paralyzed muscles in humans viae!eotrioal stimulation .

DOiDOI;alskdjlakdjfg;lsadkj

22

GRUNER et al . : LEG MUSCLE CONDITIONING

Recent studies in our laboratories have shown thatsufficient leg strength can be obtained by electricalstimulation of paralyzed muscles to operate an LPV(11,12) . It has also been observed that strength andendurance of electrically stimulated paralyzed mus-cles can be increased in paraplegic and quadriplegicsubjects at a rather rapid rate when using progressiveloading exercise protocols. However, since the bonesof paralyzed limbs are usually osteoporotic (especiallywhere spasticity is low) and may not be restored at thesame rate as muscle, care must be taken during elec-trical stimulation to avoid excessive stresses whichmight lead to bone or joint injury.

To facilitate evaluation and exercise conditioning ofparalyzed leg muscles in a clinical laboratory setting, astationary LPV simulator would be desirable . Further-more, safe and effective exercise protocols need to beestablished to minimize the risk of injuries to the para-lyzed Therefore, the purpose of this project wasto develop and evaluate an ergometer-type device andprotocols which could be used in conjunction withfunctional electrical stimulation (FES) for paralyzedmuscle evaluation and conditioning.

Polygraph

EKG

Force

Left Leg Position

Right Leg Position —or

METHODS

Subjects

Paraplegic (N = 2) and quadriplegic (N=4) subjectswith essentially no voluntary movement of their lowerlimbs and an absence of lower motor neuron involve-ment were chosen for this study . They were screenedby a physician for cardiopulmonary and musculoskel-etal conditions which would contraindicate participa-tion. All subjects were informed orally and in writingas to the nature and extent of testing, as well as to thepossible risks and the precautions taken . Understand-ing was expressed by the signing of consent forms.Screening and testing procedures were approved bythe human subjects research committees of WrightState University and Miami Valley Hospital.

Leg Conditioning System—LPV Simulator

Figure 1 illustrates a paraplegic test subject duringleg exercise using the leg conditioning system . Thisdevice is composed of three basic sections : (i) theframework, consisting of the ramp and backboardagainst which the wheelchair is positioned ; (ii) the

Left Leg Stimulation —or

Right Leg Stimulation

Force Transducer

or Weights ^

eart RateSkin Tempera ure

R

L

StimulusIsolationUnit

R

A/D

D/A

Computer

~at

23

"drive" system, which in this case consists of weightsattached above the ankles via a cable and puUeyaym-tono (Fig . 2) to simulate the type of repetitive limbmovement required for operating the LPV ; and (iii) theelectronic instrumentation consisting of the computer,the electrical stimulator, and feedback sensors . Duringexercise sessions the electrocardiogram, leg-position,and stimulus voltages are recorded on a Grass 7Ppolygraph, whereas the stimulus parameters, leg-ex-tension values, heart rate (as monitored by a GEDCOCT–2 cardiotachometer) and skin-surface tempera-ture (as monitored by a YS1 Telethermometer) over theactive muscles are printed out on an Epson MX–100printer with each extension-relaxation cycle . After ev-ery four contraction cycles, arterial blood pressure isdetermined by auscultation using a sphygmomanom-eter and stethoscope . A stimulation shut-off switch isheld in the subject's hand for safety.

During operation of this system, a computer pro-

FIGURE 1Leg conditioning system : LPV simulator inoperation with a paraplegic subject.

gram monitors leg position and controls stimulationto the muscles . Leg position during dynamic exerciseiadetermined bypotentiometers driven bytheweight-lifting cable attached to the ankle . Signals from thesefeedback sensors, after being converted to digital val-ues by analog-to-digital converters, are monitored bythe computer to form a closed-loop control system (cf13–16) . Errors in desired limb position are used tocalculate subsequent stimulus amplitudes to achievetracking . A schematic diagram of the computer systemis given in Figure 2.

With leg extension exercise, the computer graduallyincreases the stimulus amplitude until the leg extendsfrom its resting position to the maximum desired posi-tion (PMAX), and then gradually decreases the stimu-lation until the leg returns to resting position . How-ever, a trial run is first performed in which the stimulusthreshold for contraction is determined for each leg.Then, during continuous (cyclic) operation, the rate ofincrease of stimulus amplitude above threshold ioad-justed oneaohropatitionoothatPMAXvvil!be reachedin 7 sec. The stimulus amplitude will then graduallydecrease, allowing the leg to return to resting position.Although supra-threshold stimulation and musclecontractions occur for 14-sec periods, leg movementactually occurs when the contractile tension exceeds

FIGURE 2Schematic diagram of leg conditioning system /LPV simulator . Sub-ject is seated in wheelchair with legs attached just above the anklesto cable-and-weight system . Computer-controlled stimulus isola-tion unit provides stimulus to the appropriate muscle group throughsurface electrodes . Heart rate, skin temperature, and leg position aremonitored by the computer. The polygraph records EKG, leg posi-tion, and stimulus amplitude . A force transducer is placed in serieswith a cable attached to the leg to measure isometric leg strength orendurance periodically .

24

snumEnet al . : LEG MUSCLE CONDITIONING

the load, typically during the middle 7 sec of theseperiods. (Figure 3 illustrates the relationship betweene!antrioel stimulation amplitude and leg position dur-ing cyclic contractions of the quadriceps muscles .) Atall times, the computer checks to determine whetherspecial conditions (including operator or subject com-mands) have occurred indicating that stimulationshould be discontinued. For example, if the stimuluscurrent is at maximum and the leg fails to extend morethan 20 deg from rest, the muscles are considered tobe fatigued and stimulation of that leg is terminated.

Electrical stimulation system—An Apple 11 Pluscomputer interfaced with AID and DIA converters(Analog Devices /4DCO808 andADC7524)with 8-bitresolution provides voltage-control signals to an elec-trically isolated, battery powered stimulator . The sti-mulator delivers constant-current pulses of 200 micro-seconds duration to two cathodal surface electrodes(3M Company) each located over a selected motorpoint of the quadriceps muscle group . Stimulation ofthe motor points is sequential (17) (180 deg out ofphase) at a frequency of 30 Hz . An anodal (common)electrode is placed just proximal to the patella of eachleg.

Figuna4 provides a functional block diagram of theelectrical stimulator . The circuitry employs CMOSdigital logic and single supply operational amplifierswhich operate from a small 9-volt battery . An externaldin,to'dic . converter powered by 6-volt Ael cells pro-vides the high voltage (+200 volts) output . The stimu-

AGURE 4Functional block diagram of

the computer-controlledelectrical stimulator for ex-

ercise eva!uationan g oonui-uoninnofvaralyzeomus-

o!eu . Only the right legchannel is illustrated .

Left legposition

Left legstimulation

level

Right legposition

Right legstimulation

levelFIGURE 3Recording of two extension-relaxation cycles for the right and leftlegs during exercise with a 2-kg load weight.

lator consists of three basic sections : pulse genera-tion, control, and high-voltage output . For pulsegeneration, a clock whose output is set at 120 Hz isused. This clock feeds into a divide-by-four counterwhich provides sequential 30-Hz pulses . Outputs 0 and2 are used for the right leg stimulator channel, where-as outputs 1 and 3ere used for the left leg stimulator

25

channel . (Because of the similarity in circuitry tor thesechannels, only the right leg channel is illustrated .)

To generate pulses of the desired width, the outputof the clock is also fed into a monostable multivibrator.By ANDing the multivibrator output with the counteroutput pairs, two 180 deg out-of-phase 30-Hz pulses of200 microsecond duration are generated . The ampli-tude of each stimulator channel output is controlled bythe computer via 0–5 volt DA signals with 256-unitresolution . These control signals are fed to opto-isola-tors which electrically isolate the stimulator from thecomputer and all other electrical equipment . They arethen fed by buffer amplifiers through analog switcheswhose switching rates are controlled by the outputs ofthe AND gates . Outputs from the analog switches passthrough buffer amplifiers, high voltage protection di-odes and current-limiting resistors (150 milliamperemax) to drive the high voltage switching transistors.The collector of each output transistor is connected toan active cathodel electrode which is placed over aselected motor point of the quadriceps, whereas thecommon anodal electrode, which is placed just proxi-mal tothopate!!m ' iaoonneoiedtotheposidveternnina!of the high voltage supply . Thus, computer controlledactivation of the analog switches causes the high volt-age switching transistors to conduct, which completesthe circuit between the high voltage supply and m!mo-trodao, resulting in curent flow through the skin andundodyin8dsouea . Notetha1theadnnu!utor^gruund"must be kept isolated from the subject and from allother e!ectrival circuitry to prevent potentially hazard-

IsolatedStimulator =Ground

Joumal of Rehabilitation R&D Vol . 20 No . l 1983 (e pn 10-38)

ous electrical shocks . At the established pulse width,using 30 cm 2 electrodes, the maximal transcutaneouscurrent density is 5 milliampere/cm 2 and average cur-rent is only 30 microamperelcm 2. No electrodeburns have been encountered with this system.

Leg Conditioning Program protocol

In designing a conditioning program for the legs inpreparation for using the LPV, both strength unden-durance development were taken into consideration.Thus, an individually paced progressive-intensity con-ditioning program using the leg conditioning systemwas developed . This program provides trainingthrough exercise at a high number of weight liftingrepetitions plus gradual incremental increases in load.The program consists of sessions approximately 1 .5hours in duration three times per week.

Figure 5 illustrates the load weight progression pro-tocol used for both evaluation and conditioning ofelectrically stimulated paralyzed quadriceps muscles.Euoha*oaionoonoiotouftvvo^atron0Uhuondidoning"trials during which the subjects attempt to lift a loadweight 20 times with each leg through a preset rangeof motion (typically 70deg) . The two trials are separat-ed by a 10'rninute rest period . Successful completionof both trials on two consecutive sessions at a giventest weight qualifies a subject to advance to the next-higher test weight (0 .5 kg greater) on the next session.Following the second trial of each session, an "aerobicconditioning" exercise routine is undertaken in whichthe load weight is decreased to no greater than 50percent of that in the previous exercise . Aerobic condi-tioning consists of more-rapid (1 per 3 sec) alternatingcontractions of both legs within the prescribed rangeof motion for 10 minutes or until the legs fatigue.

The maximal load weight which we anticipate usingwill be 18 kg . This value was selected on the basis ofboth practical and safety considerations . Since theLPV would require about 1 .5–2.0 kg of force to opepate, the 18-kg conditioning limit will most likely pro-vide adequate reserve for operating the LPV undervarious locomotive conditions . Furthermore, exceed-ing this limit increases unnecessarily the risk of injuryto subjects . Using the described protocol, approxi-mately 24 weeks of exercise is required to achieve themaximal load weight if all sessions are successfullycompleted. When the 18-kg limit is achieved by sub-jects, the conditioning program protocol will be modi-fied by keeping this load weight constant and increas-ing the number of lift repetitions . That shouldgradually increase endurance capabilities of thesemuscles.

CathodalElectrodes OverMotor Points

*vmwnrmno

Right LegProtection

TransistorsDiodes

26

GRUNER et al . : LEG MUSCLE CONDITIONING

PRE-EXERCISE ASSESSMENT

STRENGTH CONDITIONING

Trial 1 : Lift W 20 times through 70°

range of leg extension (alternating legs,

30 sec per cycle)

ENDURANCE CONDITIONING

Rapid contractions of legs (8 sec per

cycle) at 1/, W load for 10 min

End of session

FIGURE 5Diagram of protocol which was designed to increase the strengthand endurance of paralyzed quadriceps muscle in a nafo, p,0 0 res-sivomanner.

RESULTS

During the first few electrical stimulation exercisesessions, 4 of the 6 subjects (Fig . 6, subjects b, c, a, andf.) demonstrated spastic, jerky contractions, and/orflexor-withdrawal and crossed extensor reflex activity.With subsequent conditioning sessions, however,the contractions became smoother and morecoordinated.

Figure 6 provides individual subject information andtheir right leg load weight progression data for asmany as 30 exercise sessions. Left leg load weightprogression data are similar, although not identical.The exact starting load weight was determined duringpre-testing examination sessions . Individuals withgreater initial strength and endurance capabilities ofthe stimulated muscles (a,b,e,f) progressed regularlywith only an occasional extra session at a given loadweight. In contrast, subjects c and d, with their muchlower strength and endurance capabilities, remainedat the starting load weight for 10 and 14 exercise

sessions, respectively. By the time of this report, sub-ject c increased 6 increments in load weight, whereas dincreased 2 increments . Subject c was the only partici-pant to have the load weight reduced (sessions 16 and17) because of an inability to complete an exercise trialat a previously accomplished load weight . This subjectthen went on to successfully complete sessions at thenext 5 load weight increments.

Mean heart rate and artoriol blood pressure re-sponses elicited during maximal strength-condition-ing exercise of the last exercise session completed (20contractions of each leg in 10 min at a X ~ SD loadweight of5 .4± 2.9 kg) are illustrated in figures 7 and 8,respectively. Since the results for the two 10-mm exer-cise trials were essentially the same, the mean valuesfor these trials are presented . With the onset of exer-cise, there was a drop in heart rate of 10 BPM, whereassystolic and diastolic blood pressure rose approxi-mately 20 mm Hg and 5 mm Hg respectively. Through-out the remainder of the exercise, heart rate and arteri-al blood pressure remained relatively constant.

Mean skin temperatures over the right quadricepsmuscle group during the tvi,io successive 10-minstrength conditioning exercise trials is illustrated inFigure 9 . Skin temperature increased approximately1 .75 deg C over this 30-mm period . Skin temperatureincreased throughout the first 10-min exercise trial,and continued to increase during the 10-mm rest peri-od. During the second 10-min exercise trial, however,skin temperature remained essentially constant.

DISCUSSION

The particular individually-paced exercise protocolused in this study permits both evaluation andcondi-tioning ofa!aotrioa!!yatinnu!a1od paralyzed quadricepsmuscles . Because of its progressive nature over nu-merous exercise sessions, the risks of injury to theparalyzed limbs appears to be greatly diminished . Thismay not be true with exercise protocols which incor-porate maximal contractions during each of the exer-cise sessions . Thus, with the present protocol, infor-mation concerning maximal strength and endurancecan only be obtained during exercise sessions thatcannot be successfully completed—which may takeseveral weeks to ascertain . Although we currentlyhave no objective data on bone density, we feel thatgradual progressions in load weights will permit moreeven gains in muscle and bone strength . We thereforesuggest caution to anyone considering the use ofelectrical stimulation exercise protocols aimed at rap-id development of muscle strength.

Another potential benefit of using this progressiveprotocol is that it permits a period of orientation atlight loads which seems essential for smooth, coordi-nated muscle contractions. During the first severalexercise sessions, a decline in spastic, reUexiveaotiv-

ere all lifts successful?

W-v5kg for

fatigued leg

0 min rest period

STRENGTH CONDITIONING

Trial 2 : Same as trial 1

La_Se

eight (W o be lifted by each leg If trials 1 and 2 of last

two sessions were

successful, then:

w ^ omm

27

Journal of Rehabilitation R&D Vol . 20 No . 1 1983 (BPR 10-38)

Individual Subject - Load Weight Progression (Right leg)

Subject Information (N=6)

Subject a(•) b(A) c(0) d(X) e(0) f(~Ir)

Sex M M F F M M

Age 25 29 28 38 22 22 .Date of injury 7/80 1/81 8/80 7/70 8/78 12/81

Class Para . Para . Quad . Quad . Quad . Quad.

Level of injury T-5-6-8 T-11 C-7 C-5 C-7 C-6

A

A A • •

A • •

A A • •

10 .0

9 .0

8 .0

7.0

6 .0

5 .0

4 .0

3.0

2.0q

1 .0

q q

A

q q

A Q , •

q q

A 8

•q

AA*

*

-A A *

0 0 0

Q

Q Q Q Q Q Q Q Q x x x I I 1 1 1 1 1 1 1 1 1 1

2

4

6

8

10 12

0 0

0 0

0 0

0 0

o

x x x x x x x

x Q Si x

x 111111111111111E11'4

16

18

20 22 24 26 28 30

•

q q

q q

A

q q

A nq q

A n

•q q

A

•

0 0

Exercise Session

FIGURE 6Individual subject load weight progression data for the right leg . The conditioning protocol in Figure 5 was followed.

ity during electrical stimulation was observed in mostof the subjects tested . This gives the appearance thatsome form of "learning" is taking place—possibly atthe spinal-cord level . Passive stretching exercises pri-or to active exercise sessions may also contribute toimproved performance.

To evaluate risk encountered by subjects duringelectrical stimulation exercise sessions, heart rate,blood pressure, and the skin temperature above theactive muscles were monitored . With electrical stimu-lation induced exercise in paraplegic and quadriplegicsubjects, heart rate does not appear to provide anindex of exercise stress as commonly used for able-bodied subjects during voluntary exercise. The lack ofincrease in heart rate when going from rest to exercisemay be due to loss of afferent pathways from recep-tors in the periphery, and lack of vagal inhibition andsympathetic stimulation of the heart . In contrast, arte-rial blood pressure in spinal-cord-injured subjectsdoes respond to the exercise stress by increases inboth systolic and diastolic values. This response maybe due to intact spinal reflexes which result in splanch-

nic vasoconstrictions, and increases in cardiac outputvia increases in stroke volume. Similar patterns ofheart rate and blood pressure responses have beenreported for bicycle ergometry and isometric leg exer-cises using electrical stimulation of paralyzed muscles(11,12) . The level of injury may, however, influence themagnitude of these cardiovascular responses.

We consider it essential to evaluate blood pressureresponses to this type of exercise to minimize risk tothe subjects . The importance of this became clearwhen one of our prospective subjects (C-8 quadriple-gic) experienced greatly elevated arterial blood pres-sure (in excess of 175 mm Hg systolic) on two separateattempts to exercise via electrical stimulation . Thiswas considered hazardous, and the individual wasexcluded from participation in this study.

Because of the loss of sympathetic pathways to thelower extremities in many spinal-cord-injured pa-tients, which could diminish central control of arter-ioles and sweat glands in the affected limbs, thermo-regulation during exercise could present a potentialproblem. Our measurements of a 1 .75 deg C increase

28

GRUNER et al . : LEG MUSCLE CONDITIONING

FIGURE 7Mean heart-rate response during the two strength-conditioning trials of thelast exercise session completed.

FIGURE 8Mean arterial blood pressure responses during the two strength-condition-ing trials of the last exercise session completed .

12

16

20

Contractions (each leg)

60

40 _N=6

Exercise Time =1 0 minX— Load Weight= 5 .4 ± 2 .9 kg_20

140 Systolic

120CO+1

0) 100][

EE 80 _,.-.-

60 Diastolic

40 N=6

20 _

Exercise time

10 min

— Load Weight =5 .4±2 .A kg

0

Rest

4

8

12

16

20

Contractions (each leg)

29

Journal of Rehabilitation R&D Vol . 20 No . 1 1983(BPR10-38

FIGURE 9Mean skin temperature over the right quadriceps muscle group during twosuccessive 10-min strength-conditioning trials of the last exercise sessioncompleted .

N=6I )-Z~ Load VVeighd= 5 .4 ±2 .A kg

cc 25X.

0 ORest

8

12

16

20 1

1 4

8

12

16

20 1

Trial I

Trial II10 min

10 min

Contractions

O °U9*0 28 _

w

0~~ 28 _

al .2 27 _a) -0

Ea) 0 26 —

10 min rest

in skin temperature above the active muscles indicatesthat these muscles are not being over-heated at theparticular exercise intensities used . The leveling off ofthese temperatures during trial II suggests that steady-state thermoregulation was achieved . It is apparent,however, that muscle temperature is influenced by theexercise intensity and duration, the clothing worn, andthe environmental conditions . Therefore, muscle tem-perature monitoring is advisable to avoid hazardousthermal stresses.

Unexpected effects of the electrical stimulationconditioning programs were observed in two sub-jects.Subje/t ois an incomplete C-7 quadraplegic whowas capable of weak voluntary contractions of thequadriceps when beginning the program . After sever-al weeks of exercise, however, he reported that thesevoluntary movements were getting stronger, and thathe had better leg functions . Thus, electrically inducedconditioning programs, which permit strong contrac-tions, may promote training effects in muscles beyondthose achieved with low levels of voluntary exercise . Inanother unexpected situation, subject f who had been

diagnosed as a C-6 quadraplegic with complete motorparalysis (absence of E .M.G . activity when attemptingto move voluntarily and some sensory sparing) dem-onstrated the ability to assist leg extensions voluntar-ily, and to control the rate of leg lowering during per-iods of weak electrically stimulated contractions.Thus, this subject may actually be incomplete, andmotorneurons to the quadriceps may be facilitated byaupraopinul pathways during electrical stimulationwhich permits this voluntary aspect of movement.(These effects were not observed in the other foursubjects .)

From the data obtained thus far, we conclude thatthe instrumentation described and the fitness evalua-tion/exercise conditioning protocol used are effectiveand safe . No adverse effects of the e!ooirioal stimula-tion exercise to the limbs have been observed and theintermittant progressive intensity exercise permits ad-aptation to the stress and continuous monitoring ofphyoio!ogioul responses . The heart rate, blood prmo'oure ' and skin temperature data presented suggestthat physiological responses to this type of exerciseare not excessive for most individuals, under the envi-

30

GRUNER et al . : LEG MUSCLE CONDITIONING

ronmental conditions and exercise intensities em-ployed . Potential benefits of this form of exercise in-clude improvement in : muscle strength andendurance ; bone strength ; circulation ; cardiopulmon-ary fitness ; and locomotive capability using a leg-propelled vehicle . More research, however, is neces-sary to substantiate and quantify benefits

1. Garcia EE, Glaser RM, FichtenbaumBM, Ruchman TJ, and Petrofsky JS:Stress analysis for operating a multi-propulsion mode wheelchair. Physiolo-gist 25 :201, 1982.

2. Glaser RM, Gruner JA, Feinberg SD,and Collins SR : Locomotion via para-lyzed leg muscles : feasibility study fora leg-propelled vehicle . Bull Prosth Resunder new name, Journal of Rehabilita-tion R&D. BPR 10-38 Vol . 20 No . 1, 1983.

3. Peckham PH, Mortimer JT, and Marso-lais EB : Alteration in force and fatigabil-ity of skeletal muscle in quadriplegichumans following exercise induced bychronic electrical stimulation . ClinOrthop Re! Res 114 :326-334, 1976.

4. Riley DA and Allin EF : The effects ofinactivity, programmed stimulation,and denervation on the histochemistryof skeletal muscle fiber types. ExpNeurol 40 :391–413, 1973.

5. Salmons S, and Vrbova G : The influ-ence of activity on some contractilecharacteristics of mammalian fast andslow muscles . J . Physiol . 201 :531–539,1969.

6. Cooper EB, Bunch WH, and Campa JF:Effects of chronic human neuromuscu-lar stimulation. Surg Forum 24 :477–479, 1973.

7. Kralj A, Bajd T, and Turk R : Electricalstimulation providing functional use ofparaplegic patient muscles . Med ProgrTechnol 7 :3–9, 1980.

8. Kralj A and Grobelnik S : Functionalelectrical stimulation : A new hope forparaplegic patients . Bull Prosth Res,BPR 10-20, Fall 1973, pp . 75–102 .

9. Mortimer JT : Motor Prostheses . In:Handbook of Physiology—The NervousSystem II (Brookhart JM, MountcastleVB, Brooks VB, and Geiger SR, eds).Bethesda, Maryland, America! Physio-logical Society, 1981, pp . 155-187.

10. Munsat TL, McNeal D, and Waters R:Effects of nerve stimulation on humanmuscle . Arch Neurol 33 :608-617, 1976.

11. Glaser RM, Petrofsky JS, Gruner JA,and Green BA : Isometric strength andendurance of electrically stimulated legmuscles of quadriplegics. Physiologist25 :253, 1982.

12. Petrofsky JS, Glaser RM, Phillips CA,and Gruner JA : The effects of electrical-ly induced bicycle ergometer exerciseon blood pressure and heart rate.Physiologist 25 :253, 1982.

13. Vodovnik L, Crochetiere WJ, and Res-wick JB : Control of a skeletal joint byelectrical stimulation of antagonists.Med Biol Eng 5 :97–109, 1967.

14. Trnkoczy A : Variability of electricallyevoked muscle contractions with spe-cial regard to closed-loop controlledorthosis . Ann Biomed Eng 2 :226–238,1974.

15. Petrofsky JS, and Phillips CA : Con-stant-velocity contractions in skeletalmuscle by sequential stimulation ofmuscle efferents . Med and Biol Eng andComput 17 :583–592, 1979.

16. Crago PE, Mortimer JT, and PeckhamPH : Closed-loop control of force duringelectrical stimulation of muscle . IEEETrans Biomed Eng BME-27 :306–312,1980.

17. Peckman PH, Van Der Meulen JP, andReswick JB : Electrical activation ofskeletal muscle by sequential stimula-tion . In : The Nervous System and Elec-trical Currents . (Wulfson N and SancesA, Jr., eds) New York : Plenum, 1970, pp.45– 50 .