C o p y r i a h t 0 Munksxaard 1998

Scand J Med Sci Sports 1998: 8: 222-228

Printed in Denmark All rights reserved

Scandinavian Joztrnai of
M E D I C I N E & SCIENCE
I N SPORTS
ISSN 0905-7188

The energy cost of level walking before and

after hydro-kinesi therapy in patients with
spastic paresis
Zamparo P, Pagliaro P. The energy cost of level walking before and after
hydro-kinesi therapy in patients with spastic paresis.
Scand J Med Sci Sports 1998: 8: 222-228. 0 Munksgaard, 1998
In this study the energy cost of level walking was measured in 23 patients
with stationary spastic paresis before and after a two-week treatment (45
min daily) of hydro-kinesi therapy, the latter consisting of passive and
active movements in warm (32C) sea water, free swimming and water immersion walking. Among the subjects (80.22 13.2 kg body mass;
56.02 14.6 years of age; 10.726.6 years of duration of spasticity), 12
were affected by hemiparesis, 4 by multiple sclerosis and 7 by spinal cord
injury. The energy cost of level walking (C,) was measured before and after
therapy from the ratio of the overall steady-state oxygen consumption to
the effective speed of progression. The differences in C , due to the treatment, at matched speeds, were found to be negligible at speeds higher
than 0.75 m . sP1 (less than 5%) but to increase, with decreasing speed, up
to about 17% at 0.1 m . s-'. The treatment was therefore effective in improving the gait characteristics of the subjects, through a decrease of their
C,, mainly at low speeds of progression.

Complex motor behaviours, such as locomotion, require the coordinated action of many muscles. The
neural circuits in the spinal cord coordinate the contractions of the active muscles, thus allowing the
higher centers to control complex movements with
relatively simple descending signals (1). The lesions
of the central nervous system (CNS) that interrupt
the descending motor pathways lead to abnormalities
of locomotion, since they affect the functional integrity of the pheripheral effectors (muscles and joints)
and the neural mechanisms of perception and control
of movement.
Whereas it is possible, by clinical examination, to
evaluate separately the spasticity and rigidity associated with the lesions of the CNS, more global information can be derived from the measure of the energy
cost of locomotion (2). This parameter corresponds
to the amount of energy needed by the subject to
cover one unit of distance and represents the energy
equivalent of all the single physiological mechanisms
whose final result is walking.
In healthy subjects, the energy cost of level walking
(C,) is a function of the speed of progression, the

222

P. Zamparo', P. Pagliaro*
'Dipattimento di Scienze e Tecnologie
Biomediche, Sezione di Fisiologia, Udine, Italy,
21stituto Elioterapico "G. Barellai", Grado (GO),
Italy

Key words: energy cost of walking; spastic

paresis; physical therapy; water immersion

Paola Zamparo, Dipartimento di Scienze e

Tecnologie Biomediche, Sezione di Fisiologia,
via Gervasutta 48, 33100 Udine, Italy
Accepted for publication 15 May 1998

relationship between C, and speed being curvilinear

with a minimum at about 1.1 m . s-l; at higher or
lower speeds C, is greater ( 3 ) .
In healthy subjects the self-selected speed is close
to the optimal one thus minimizing C, (3,4), whereas
patients with impaired walking function are reported
to choose lower speeds of progression, probably because of fear of falling, difficulties of coordination or
balance problems (2, 5).
It has also been reported that, following physical
(5) and surgical treatment (6-8) or physical exercise
(9), the improvement in walking ability in disabled
persons is reflected by an increase of the self-selected
speed and/or by a decrease of the energy cost of
walking.
Among the physical treatments adopted to reduce
hypertonus, a n aquatic exercise program was reported to induced positive changes in muscular
strength, fatigue, work and power in patients affected
by multiple sclerosis (10). These positive therapeutical
responses were attributed: i) to the use of dynamic
exercise (free-style swimming) as compared to traditional techniques of passive and active range of mo-

Hydro-therapy in spastic paresis

tion; and ii) to the buoyant nature of water which,
supporting the body weight, unloads the joints and
facilitates mobility.
The aim of this study was to evaluate the energy
cost of level walking in patients with stationary hemiparesis, paraparesis and multiple sclerosis before and
after a two-week treatment of passive and active
movements in water, free swimming and water immersion walking.

Material and methods

Subjects
Twenty-three patients affected by stationary spastic
paresis able to walk for 5 to 6 min without assistance,
although with assistive devices, were the subjects of
the study. All patients underwent medical examination and gave their informed consent to the study.
Twelve were affected by hemiparesis (HP), 4 by
multiple sclerosis (MS) and 7 by spinal cord injury
(SCI). The study was approved by the hospitals ethics review board.
The patients spasticity was graded using the fivepoint ordinal scale described by Ashworth (1 1). In
order to quantify disability, the FIM scale (Functional Independence Measure) was also utilized (12).

This scale assesses independence in daily living with

scores from 1 (total assist) to 7 (complete independence) assigned to 13 motorial and 5 cognitive items.
The scores reported in this study refer to the 13 motorial items only (the highest score is therefore 91).
For SCI patients, the ASIA (American Spinal Injury
Association) motor score is also reported (13). This
score is based on the functional analysis of 10 muscular groups of the upper and lower limbs (right and
left side) with scores from 0 (total paralysis) to 5 (active movement against full resistance): the highest
cumulative score is therefore 100. The individual Ashworth, FIM and ASIA rates are reported in Table 1
along with the type and duration of injury, the subjects age, sex and weight.
Physical treatment

All patients underwent hydro-kinesi therapy for two

weeks (45 min daily). The treatment was carried out
in an indoor pool containing warm sea water (32C)
and consisted of: i) active and passive range of motion exercises in water (15 min each), ii) coordination
exercises and water immersion walking (15 min), and
iii) free-style swimming (whenever possible) at the
end of the session.

Table 1. Description of patients

~

Diagnosis
and site of lesion

Group A
MS
MS
HP (left hemisph.)
HP (left hemisph.)
HP (right hemisph.)
HP (right hemisph.)
HP (left hemisph.)
HP (right hemisph.)
HP (left hemisph.)
Group B
SCI (C5)
SCI
SCI
SCI (C5-C8)
SCI (T3-T4)
SCI (T10)
SCI
MS
MS
HP (left hemisph.)
HP (left hemisph.)
HP (left hemisph.)
HP (right hemisph.)
HP (left hemisph.)

Cause of illness

brain haemorrhage
brain haemorrhage
brain haemorrhage
ischaemic stroke
ischaemic stroke
ischaemic stroke
ischaemic stroke
spinal trauma
spinal cord embolia
spinal trauma
spinal trauma
spinal trauma
spinal trauma
spinal trauma
ischaemic stroke
brain haemorrhage
ischaemic stroke
ischaemic stroke
ischaemic stroke

~~

Duration of
spasticity
(years)

Age
(years)

Body mass
(kg)

10
10
13
18
9
3
5
30
5

56
33
62
68
44
49
80
65
69

78
59
65
78
106
77
85
90
102

11
15
10
12
4
4
10
20
20
3
13
10
8
4

31
72
55
59
19
54
60
49
55
65
71
71
46
55

80
95
75
86
62
77
66
55
78
90
76
96
82
87

Sex

Ashworth
rates
(0-4)

FIM scores ASIA scores

(13-91)
(0-100)

82
80
70
78
71
85
79
79
84

M
M
M
M

2
2
3
3

F
F
F
M
M
M
M
M
M

2
3
3
2
3
3
2
2
3

91
91
91
57
81
76
a4
73
82
74
56
86
50
83

Walking
aids

1 c
1 c

1c
2 FC
TR
1 c
1 c
75
78
75
56
50

50
68

2c
2 FC
2 FC
RGO
RGO
1 c
2 FC
1 FC
1c
1c
1c
1 c
1 c

Abbreviations: M, male; F, female; MS, multiple sclerosis; HP, hemiparesis; SCI, spinal cord injury; C, cane; FC, forearm crutches; TR, tripod; RGO,
Reciprocal gait orthosis; -: no walking aids. See text for details.

223

Zamparo & Pagliaro

Experimental protocol

Statistical analysis

The subjects were asked to walk, for about 5-6 min,

along an indoor track 50 m long traced in an hospital
corridor. The steady-state oxygen uptake (V02,
1 min-') was measured by means of a telemetric,
portable system (Cosmed K4, I) that made it possible
to measure V02 from pulmonary ventilation and
from the O2 and C 0 2 fractions in mixed expired air.
During the walking tests the subjects wore a facial
mask equipped with the impeller expiratory flow
meter; they carried also the O2 and C 0 2 sensors, the
battery and the transmitting unit of the K4 system;
the overall carried mass did not exceed 0.9 kg. The
validity and the reliability of the measurements obtained with this method are reported in several papers
to which the reader is referred for further details (e.g.
14, 15). Oxygen uptake, heart rate, expired ventilation and gas exchange ratio were measured on a
time basis of 15 s and the values recorded during the
last minute of each test were averaged to yield the
steady-state values.
During level walking, at low speed of progression
and at steady state, the energy required is entirely delivered by aerobic sources (2, 16, 17). The energy cost
of walking (C,) was therefore determined from the
ratio of the overall (rest included) steady-state V02
(m102 . min-' . kg-') to the effective speed of progression (v, m * min-'):

The values are presented as means? 1 SD. The differences before and after therapy were investigated by
means of a Wilcoxon matched-pairs signed-ranks test
(SPSS for Windows). The statistical significance level
was set at P<0.05.

c, =VOz/v

Results
The overall energy cost of level walking is shown in
Fig. 1 as a function of the self-selected speed of progression in patients before the treatment. In the same
figure the values of C, as measured in control,
healthy, subjects at low speeds of progression, are also
reported. Data from control subjects were collected
in a previous study (2) on a group of 9 males and 8
females (54k 11 years of age, 78+ 14 kg body mass)
who were asked to walk at speeds ranging from 0.1
to 1.2 m s-'.
In the reported range of speeds, for both groups
(patients and controls) the relationship between the
energy cost of walking and the speed can be described
by a power function of the form: Cw=a . vPb (2). Indeed, the relationship between the natural logarithms
of C, and v was found to be linear, as described by:
In C,=1.557-0.816. In v (r2=0.906, n=23, in pa-

1)

and converted in joules per meter per kilogram body

mass on the assumption that 1 ml O2 consumed in
the human body yields 20.9 J (which is strictly true
only if the respiratory quotient equals 0.96). The average speed of progression (v), was measured by
means of a stopwatch from the time to cover 5 m
laps, marked along the walking track.

A
251

'

15-

Experimental procedure
The pre-therapy test was conducted during the week
preceding the start of the treatment. On that occasion
the energy cost of walking was determined at the selfselected speed in all subjects.
The post-therapy test was conducted during the
week following the completion of the physical treatment. On this second occasion, 14 patients were
asked to walk again at the self-selected speed (group
B), whereas 9 patients were asked to walk at the same
speed chosen in the first test (group A). This was set
by acoustic signals (Balise temporelle, Baumann
CEM, Switzerland) whose frequency was such that,
when walking at the appropriate speed, at each signal
the subject was passing equally spaced (1 m) marks
on the track.

224

ct:

lo5-

01
0

0.25

0.5

0.75

1.25

- -

v ( m s 1)
Fig. 1. The overall energy cost of walking (J . m-l . kg-') as a
function of the self-selected speed (m . s - l ) in patients affected
by spastic paresis before therapy (hemiparesis: unfilled squares;
multiple sclerosis: flnfilled circles; spinal cord injury: unfilled
triangles). The relationship between C, and speed as obtained
in a previous study (2) on healthy subjects is represented by the
r=0.93, n=56, dots). The
continuous line (C,=3.32. v
self-selected speed in control subjects (1.18 m . spl) is indicated
by the arrow pointing donwards (from Himann et al., 4).

Hydro-therapy in spastic paresis

Table 2. Average values ( 2 1 SD) of the data collected before and after therapy

Group A

Group B

0.98
(0.22)

0.877
(0.112)

108.9
(15.1)

7.58
(2.03)

23.94
(6.71)

0.89
(0.21)

0.866
(0.081)

107.1
(17.9)

7.11
(2.45)

25.94
(5.68)

0.93
(0.27)

0.851
(0.088)

110.4
(22.8)

13.42
(7.73)

25.12
(4.82)

0.90
(0.28)

0.870
(0.067)

100.6
(11.1)

10.65*
(5.31)

26.04
(7.25)

0.576
(0.161)

0.419

pre therapy

(n=14)

(J . m-' . kg-')

(0.168)

0.578

post therapy

HR
(bpm)

VO,
( 1 . min-')

pre therapy

(n=9)

VE
(I . min-')

(m . s-l)

(0.250)

0.479'

post therapy

(0.278)

cw

~ _ _ _ _ _

Abbreviations v, speed of progression, VE, expired ventilation (BTPS), V02, oxygen consumption (STPD), R, gas exchange ratio, HR, heart rate, Cw,
energy cost of walking When underlined the speed values refer to the self-selected speed
* Statistical differences (Wilcoxon signed-rank test, P<O 05) with pretest values

Table 3. The energy cost of level walking at matched speed of progression

c,c*

(m . s - I )

0.1
0.2
0.3
0.4
0.5
0.6
0.7
0.8
0.9
1.o
1.1
1.2

(J

. m-1

. kg-')

14.39
9.26
7.15
5.95
5.16
4.60
4.1 7
3.83
3.55
3.32
3.1 2
2.96

CWP
pre therapy
(J . m-' . kg-')

CWP
post therapy
(J . m-' . kg-')

CWP
pre therapy vs. C
(%)**

CP, pre vs.

post therapy

31.05
17.64
12.67
10.02
8.35
7.20
6.35
5.69
5.17
4.74
4.39
4.09

26.60
15.65
11.48
9.21
7.76
6.75
6.00
5.42
4.95
4.57
4.25
3.97

115.73
90.56
77.22
68.32
61.73
56.54
52.28
48.68
45.58
42.86
40.44
38.27

16.74
12.69
10.38
8.77
7.54
6.55
5.71
5.00
4.37
3.81
3.30
2.85

(Yo) * * *

Abbreviations: ,C, energy cost of walking; v, speed of progression; P, patients affected by spastic paresis; C, control subjects.
Data from literature (2). * * Percentage difference between C, in patients (before physical theraphy) and controls; * * * Percentage difference between
C, in patients pre and post therapy.

tients); In C,=1.200-0.637 . In v (r2=0.858, n=56,

in controls). Therefore for C, in joules per meter per
kilogram body mass and v in meters per second: a =
4.74 and b=-0.816 in patients; a=3.32, b=-0.637
in controls. The power function referring to control
subjects is represented in Fig. 1 by the continuous
line.
As shown by Fig. 1, the self-selected speed (sss) in
patients (before physical therapy) is lower than the
sss in control subjects (indicated by the arrow pointing downwards), the average values for the different pathologies being: HP: 0.55920.186 m s-';
MS: 0.42520.207 m * s-'; SCI: 0.380-+0.290 m s-'
and the corresponding average values of C, being:
HP: 8.6453.33 J m-' kg-'; MS: 11.5928.59
J . m-' . kg-'; SCI: 15.15+-8.81 J . m-I . kg-'. The
value of speed (1.18 m . s-I) indicated by the arrow
in Fig. 1 corresponds to the sss of healthy men and

women over 40 years of age according to Himann et

al. (4), the corresponding C, amounting to about 3
J . rn-' kg-I.
The average values of speed of progression (v), expired ventilation (VE, BTPS), oxygen consumption
(VO,, STPD), gas exchange ratio (R), heart rate
(HR) and energy cost of walking (C,) are reported
in Table 2 for the patients who walked at the same
speed pre and post therapy (group A) and for the
ones who were asked to walk at the sss in both cases
(group B). Significant differences (P<0.005) in the
speed of progression and in the energy cost of walking were found in group B after the treatment. C,
was found to decrease after the treatment albeit not
significantly (P=0.066) also in group A, even if the
speed of walking was the same (P>0.5) in the two
experimental sessions.
As found for the pre-treatment data, the values col-

225

Zamparo & Pagliaro

lected post therapy could be interpolated by a power
function of the form: Cw=a v -b where: a=4.57, b =
-0.765 (r2=0.793, n=23). On the basis of the values
of a and b obtained for patients (pre and post treatment) and controls, it was possible to calculate the
corresponding C, values at matched speed of progression. As shown in Table 3: i) the percentage difference between C, in patients (pre-treatment values)
and control subjects (C,P-C,C/C,C)
was found to
be a function of the speed of walking and to increase
from a minimum of 38.3% at 1.2 m . s- to a maximum of 115.7% at 0.1 m * s-. In addition, ii) the
percentage difference in the energy cost of walking
between pre- and post-hydro-kinesi therapy in patients was found to be negligible at speeds higher
(less than 5%) but to increase up to
than 0.75 m . SKI
about 17% at 0.1 m s- (see Table 3 and Fig. 2).
Finally, it was also observed that the differences
induced by the treatment were pathology dependent. Indeed, whereas the pre- and post-therapy differences of speed and energy cost of walking were
negligible in HP subjects (Av= -0.01 150.029 m . s-l
and AC,=0.4650.79 J * mK kg-), the decrease of
C, (and the increase of v) were appreciable in the
other two groups of patients: ACW=2.79L4.15
J * m- . kg- (MS) and 3.76k3.79 J * m- kg-
(SCI); Av=-0.062?0.099
m . s- (MS) and
-0.062?0.034 m . s- (SCI). However, only in SCI
patients were v and C, significantly different
(P<0.03) after the treatment.

Discussion
Subjects affected by pathologies of neurological origin reduce their physical activity not only for causes
which depend on the pathology itself but also for sociological and motivational reasons. This leads to a
decrease of their muscular and cardiovascular performances and to a poor physical fitness, a condition
partially or totally reversible by adequate physical
training.
Long-term muscular stretch of the hypertonic
muscles has been shown to reduce both spasticity and
co-contraction and to increase the range of motion in
the hip joint; as a consequence, the energy cost of
locomotion was found to decrease in patients with
stationary paraparesis (5). Treadmill training with
partial body weight support was found to be effective
with regard to restoration of gait ability and improvement of walking speed in patients with chronic hemiparesis and spastic paresis (18, 19).
However, the simplest way to unload the body is
water immersion: exercise in water has long been used
in patients with musculoskeletal disorders as it provides a favourable medium to exercise and relax their
muscles (20).
Therefore, dynamic exercise in an aquatic environment could be expected to improve the gait characteristics in patients with musculoskeletal disorders.
In the following paragraphs the data collected in
this study before therapy will be firstly compared
with those reported in the literature, and the effect of
the hydro-kinesi therapy will be subsequently discussed.

20

?n
!

Pre-treatment data
15

0
c
u)

10

B
0

?!
n

b
0

0
0

0.5

v (m

. s-1)

Fig. 2. The percentage differences between the values of C, pre

and post therapy (C, pre-C, post/C, pre, %) at matched
speeds are represented as a function of the speed itself (v,
m . s-). The data points were calculated by pooling all the
subjects (both groups) together (see text for details).

226

Data reported in Fig. 1 show that the sss in patients is

substantially lower than in healthy subjects. Subjects
affected by spastic paresis are therefore at a disadvantage because they chose speeds which are, per se,
characterized by a higher C, (i.e. they chose to
walk in a portion of the C, versus speed relationship that is not the optimal one).
Moreover, i) they have a higher C , in comparison
with control subjects also at paired speed, and ii) this
difference increases the lower the speed of walking
(see Fig. 2 and Table 3).
The values of C, in subjects affected by hemiparesis were found to be higher (2), and in subjects
affected by multiple sclerosis lower (21,22) than those
reported in the literature: this can be attributed to a
different health state of the subjects, to the different
use of walking aids and to the experimental protocol
adopted.
Indeed, the use of forearm crutches was recently
reported to increase C, due to an increase of the
mechanical work and a reduced efficiency of positive

Hydro-therapy in spastic paresis

work production (17). Whereas in this study all the
subjects but one affected by hemiparesis used walking
aids, in the previous one (2) the majority of the subjects walked freely.
Olgiati et al. (21) reported that the energy cost of
treadmill walking, in subjects affected by multiple
sclerosis, is significantly larger (up to 2-3 times) than
in controls at all speeds. Piperno et al. (22) reported
lower differences of C, (up to 1.5 times) in mildly
disabled MS patients (walking on a treadmill without
assistive devices) and suggested that even lower C,
values are to be expected in these patients if the test is
carried out on the ground rather than on a treadmill.
Indeed, in this study, the energy cost of walking on
the ground in MS subjects was found to follow the
same C, vs. speed relationship of H P and SCI subjects, probably due to a better exercise tolerance of
MS subjects to walking on the ground and to a lower
level of disability in comparison with subjects of previous studies.
Finally, Mattsson et al. ( 5 ) reported similar values
of C, in subjects with spastic paraparesis: from
4.8 J - m - . kg- (at 0.83 m - s - ) to 12.12
J . m-I . kg- at 0.52 m . s-.
Post-treatment data
Fig. 1 shows that, before the treatment, the relationship between the energy cost of walking and the speed
of progression, in patients, could be described by a
continuous function regardless of the neurological
pathology. Indeed: i) the coefficient of correlation of
the power function with which the data were interpolated is quite good (r=0.952); and ii) the data referring to patients affected by different lesions are
equally scattered and do not show a specific illness
dependent trend.
Also the data collected after the treatment are well
interpolated by a continuous function (r=0.890) and
do not show any specific illness and/or protocol
dependent trend since, in all patients, the therapy
had the effect of decreasing the energy cost of level
walking at matched speed of progression (Table 3).
The ratio of the energy cost of walking before, to
that after, the treatment is reported in Fig. 2 as a
function of the speed. This figure shows that the effect of the therapy is speed dependent: it is greater
the slower the speed. So, the decrease of C, after
treatment was particularly evident for the patients of
group B who, before the treatment, were characterized by a rather slow self-selected speed. As a consequence, since after the treatment their sss increased,
their corresponding C , decreased substantially and
significantly (Table 2). Indeed they moved to the
right along the steeper portion of the curvilinear C,
versus speed relationship. Therefore, in addition to
the decrease in the energy cost of walking at matched

speed of progression, a further decrease of C, was

brought about in this group by the increased sss.
In contrast, the patients of group A were characterized by a large sss before the treatment. As a consequence, the decrease of C, brought about by the
treatment was lower (see Fig. 2) and remained below
the level of significance; this was so also because they
were asked to walk at the same absolute speed (see
Table 2). Indeed, even if the subjects of group A
would have been allowed to move to the right along
the curvilinear C, versus speed relationship, rather
low differences in C, would have been detected since,
before the treatment, these subjects were already
walking in a relatively flat portion of this relationship.
Therefore, the effect of the treatment was greater
in subjects of group B and negligible in subjects of
group A essentially because they walked in a different portion of the C, versus v relationship.
In conclusion, the subjects who benefit more from
this treatment are the ones who are characterized by
slow self-selected speeds of progression (in our case
the SCI patients). This finding is consistent with the
observation that the effect of physical training is
likely to be less effective in subjects who, like subjects
affected by hemiparesis, exercised daily in the form
of level walking.
The outcome of this study is therefore that the
more debilitating the illness (the lower the sss) the
greater the benefits of the hydro-kinesi therapy.

Conclusions
A treatment of active and passive mobilization in
warm water, free swimming and immersed walking
is suggested in patients affected by stationary spastic
paresis since it improves their gait characteristics by
decreasing the energy cost of walking (mainly at low
speeds of progression).

Acknowledgement
We are grateful to Prof. I? E. di Prampero for his useful comments.

References
Gordon J. Spinal mechanisms of motor coordination. In:
Principles of Neural Sciences. New York: Elsevier, 1990:
56680.
Zamparo P, Francescato MP, De Luca G, Lovati L, di
Prampero PE. The energy cost of level walking in patients
with hemiplegia. Scand J Med Sci Sports 1995: 5: 348-52.
Margaria R. Sulla fisiologia e specialmente sul consumo
energetic0 della marcia e della corsa a varia velocita ed
inclinazione del terreno. Atti Acc Naz Lincei 1938: 7: 299368.
Himann JE, Cunningam DA, Rechnitzer PA, Paterson
DH. Age related changes in speed of walking. Med Sci
Sports Exerc 1988: 290: 161-6.

227

Zamparo & Pagliaro

5. Mattsson E, Brostrom LA, Borg J, Karlsson J. Walking
efficiency before and after long term muscle stretch in patients with spastic paraparesis. Scand J Rehab Med 1990:
22: 55-9.
6. Mattsson E, Brostrom LA, Linnarsson D. Walking efficiency after cemented and non cemented total hip arthroplasty. Clin Orthop 1990: 254: 170-9.
7. Brown M, Hislop H, Waters R, Pore11 D. Walking efficiency before and after total hip replacement. Phys Ther
1980: 60: 1259-63.
8. McBeath AA, Bahrke MS, Balke B. Walking efficiency before and after total hip replacement as determined by oxygen consumption. J Bone Joint Surg 1980: 62: 807-10.
9. Olgiati R, di Prampero PE. Effect de lexercice physique
sur ladaptation a leffort dans la sclerose en plaques.
Schweiz Med Wschr 1986: 116: 374-7.
10. Gehlsen GM, Grigsby SA, Winant DM. Effects of an
aquatic fitness program on the muscular strength and endurance of patients with multiple sclerosis. Phys Ther
1984: 64: 653-7.
11. Ashworth B. Preliminary trial of carsoprodal in multiple
sclerosis. Practitioner 1964: 192: 540-2.
12. FIM: Functional Independence Measure. Strumento di
misura della disabilita. In: Ric Riabil 2 (2). Milano: S 0 . GE.COM., 1992: 1 4 4 .
13. American Spinal Cord Injury Association. Standards for
neurological classification of spinal injuried patients. Chicago: ASIA, 1982.

228

14. Kawakami Y, Nozaki D, Matsuo A, Fukunaga T. Reliability of measurement of oxygen uptake by a portable
telemetric system. Eur J Appl Physiol 1992: 65: 409-14.
15. Lucia A, Fleck SJ, Gotshall RW, Kearney JT. Validity and
reliability of the Cosmed K2 instrument. Int J Sports Med
1993: 14: 380-6.
16. di Prampero PE. Energetics of muscular exercise. Rev Physiol Biochem Pharmacol 1981: 89: 143-222.
17. Thys H, Willems PA, Saels F? Energy cost, mechanical
work and muscular efficiency in swing-through gait with
elbow crutches. J Biomech 1996: 11: 1473-82.
18. Hesse S, Bertlet C, Jhanke MT, et al. Treadmill training
with partial body weight support compared with physiotherapy in non ambulatory hemiparetic patients. Stroke
1995: 26: 97G81.
19. Visintin M, Barbeau H. The effects of body weight support
on the locomotory pattern of spastic paretic patients. Can
J Neurol Sci 1989: 16: 315-25.
20. Frangolais DD, Rhodes EC. Metabolic responses and
mechanisms during water immersion running and exercise.
Sports Med 1996: 22: 38-53.
21. Olgiati R, Jaquet J, di Prampero PE. Energy cost of walking and exertional dispnea in multiple sclerosis. Am Rev
Respir Dis 1986: 134: 1005-10.
22. Piperno R, Amadori L, Tonini A, Betti L. Costo energetico del cammino in pazienti affetti da sclerosi multipla.
Eur Med Phys 1995: 31: 183-91.