ORIGINAl ARTIClE
Optimisation of isometric
Wendy IUmIJw
Rob Herbert strength training intensity
The aim of this study was to determine the he provision of effective strength increases in strength than subjects who
optimal intensity for isometric strength training. training programs involves trained at target intensities of either 25
Isometric elbow flexor strength of 51 subjects prescription of exercise per cent or 100 per cent of MVC,
was measured before and after a six-week parameters such as the intensity, indicating that the optimal training
training period. Subjects were randomly duration and frequency of muscle intensity is submaximaL In the present
assigned to target training intensities between contractions to be employed in study we attempted to extend these
o and 100 per cent of a maximal voluntary training. A large body of research has findings by providing a more precise
contraction, and undertook supervised isometric investigated the optimal exercise estimate of the optimal isometric
elbow flexor training at the allocated target parameters for increasing strength training intensity.
intensity three times each week for six weeks. (reviewed in Atha 1981 and Clarke The design of the present study also
Strength increases at the trained angle of the 1973) and provides some clear made it possible to further investigate
trained limbwere linearly or near-linearly related guidelines, at least for dynamic exercise two related phenomena. Numerous
to training intensity, with the greatest increases involving the repeated lifting of studies have shown that isometric
in strength occurring at maximal ornear-maximal weights. For example, it has been training at one joint angle produces
intensities. These findings suggest that the shown that the optimal training load is strength increases that are larger at
optimal intensity for isometric strength training about six repetition maximum (RM) trained than untrained angles (Kitai
is maximal or near maximal. (the largest load the subject can lift six and Sale 1988, Lindh 1979, Thepaut-
[Khouw Wand Herbert R: Optimisation of times; Berger 1962). Fatiguing Mathieu et a11988, Weir et aI1994). A
isometric strength training intensity. Australian contractions performed at this secondary aim of the present study was
Journal of Physiotherapy 44: 43-46] intensity produce greater increases in to determine whether isometric
strength than non-fatiguing training intensity influences the
contractions performed at the same magnitude of strength increases at
intensity, so it is probably preferable, untrained joint angles. It has also
where other considerations permit, to commonly been reported that
continue lifting the 6 RM load to unilateral training produces increases
exhaustion (Rooney et aI1994). in strength of the contralateral limb
While the literature on dynamic (Kannus et aI1992; see Enoka 1988
strength training provides useful and Smith 1970 for references). The
guidelines for exercise prescription, present study investigated whether the
Key words: Exercise; Isometric there has been relatively little carry-over of training effects to the
Contraction; Muscles investigation of the optimal parameters contralateral limb is influenced by
for isometric strength training. To our training intensity.
knowledge, only one study has used a
true experimental design to investigate Methods
the optimal training parameters for
isometric strength training. Szeto and Fifty-one subjects, 18 men and 33
colleagues compared the effects of women, participated in the study. All
W Khouw BAppSc(phty), GradDipBiomedE and three weeks of isometric training at were university students between 18
and RHerbert BAppSc(Phty), MAppSc(ExSpSc) target intensities of 25 per cent, 50 per and 35 years of age who were not
are lecturers in the School of Physiotherapy at cent or 100 per cent of a maximal currently undertaking, or who had not
The University of Sydney. voluntary contraction (MVC; Szeto et recently undertaken, strength training
al 1989). It was found that subjects exercise for their upper limbs. The
Correspondence: RD Herbert, School of
who trained at target intensities of 50 procedures used in the study were
Physiotherapy, The University of Sydney, PO
Box 170, Lidcombe, New South Wales 2141. per cent MVC obtained greater
ORIGINAl ARTICLE
Page 43
A B
approved by the Ethics Review
Committee of The University of Trained Angle (140°) Untrained Angle (gOo)
Sydney.
60 60
Isometric elbow flexor strength of all ~
~
subjects was measured in the following CD e.....
way. Subjects were seated with the 'tJ
u;
<!) 40
° ° 40
.
~
(f)
t1l
upper arm supported on an inclined 'tJ e:0 °
~
20 20 eo' • " •
~
bench and with the forearm free. They CD .5
c:
grasped a handle connected by a chain
to the floor. The angle the chain made
'e
I-
tc:
<!)
• • ".0fIt'_
0 .. .... ...............
0
~
~.o
......."....
... ... .........
~'
~
-.- ~~
with the forearm was adjusted to 90 ~
degrees by varying the length of the -20 -20
chain. The handle incorporated a load 0 25 50 75 100 0 25 50 75 100
cell which measured the longitudinal
C D
force applied to the chain. The signal
60 60
from the load cell was amplified and
displayed on a chart recorder or CD ~
"0 e.....
sampled at 20 Hz and stored on a
computer.
U;
"0
<!)
(f)
t1l
e:
40
°
o • o. ,
40
.. .,. 0
. . ..
.
After a few submaximal warm-up
CD
c:
'ec:
.5
0 20
0
:::" ' .' ••~ ... 0
20
~
• 0 0
- 0.
... . ........
eo • •
contractions, subjects performed three tc: ~
•_ _ _ _- _ 0
•
~. . . . . 0
MVCs. They were loudly exhorted to
perform truly maximal efforts. The
::l
e:
(j)
0 ..... ~ .. ~ . ..... .......
'
0 . ....................... .
.•
-20 -20
largest force produced in the three
0 25 50 75 100 0 25 50 75 100
efforts was taken as the subject's
isometric strength. Measurements of Training Intensity (%) Training Intensity (%)
isometric strength were made at 90
degrees and 140 degrees (ie 40 degrees
from the extended position) on both figure 1: Effect of tl·ail1ing intensity on strength increases of (A). the trained <Ingle of the
the right and left arms. The order of trained side, (B), the untrained angie of the trained side, (C), tlte trained angle of the
testing sides and angles was untrained side, and (I)), the untrained allgh~ 011 the untrained side. lines are linear
randomised. regressions am! 95 per cent confidence intervals. Training intensity significantly affected
After testing, each subject was strength oilly at the trained angle 011 tile trained side (slope = 0.19, [1= 0.0(6). The '1-
randomly assigned an arm to be intercept (strength increase associated with zero intellsity trainillg) was significant at
trained (left or right) and a target tl'le untrained 811gle on the trained side and at both angies on tile tmtrained side
training intensity between 0 and 100 (intercepts of 10.6, 9:1 ami 9.2 per cent, and p = 0.11111, 0.02 and 0.02 respectively).
per cent MVC. As there were 51
subjects, the range between 0 and 100
per cent was divided into fifty-one
MVC, as well as one subject who absolute training intensity was
2 per cent increments. One subject was trained at 78 per cent MVC, were progressed in proportion with the
randomly allocated to each 2 per cent
unable to sustain the target intensity subject's increase in strength. Subjects
increment of intensity, and each for the full 10 seconds of all six were asked to refrain from other forms
subject trained at his or her allocated contractions because of fatigue. Thus of upper limb strength training for the
intensity for the full duration of the
these nine subjects actually trained at duration of the study. At the end of the
training period. All training
intensities that were, at least for part of training period, all four measures of
contractions were performed at an
each training session, less than the isometric strength (ie those made at 90
elbow angle of 140 degrees. Training target intensity (although these sub- degrees and 140 degrees on the trained
consisted of six lOs contractions at the
target intensities involved maximal and untrained limbs) were repeated
target intensity, with a one-minute rest
effort). All training was carried out using identical procedures.
between contractions. Feedback of
under supervision in the laboratory
actual force production and the target Data analysis
three times each week for six weeks.
intensity was provided on a chart
To accommodate increases in strength, To normalise for inter-subject
recorder, oscilloscope or computer
the isometric strength of the trained differences in initial strength, increases
screen. All subjects who trained at or
limb was re-measured at the end of the in strength were expressed as a
above a target intensity of 86 per cent
first, third and fifth weeks, and the percentage of pre-training values. The
ORIGINAl ARTIClE
relationship between training intensity Strength was also increased at the training effects could have been
and increases in each of the four untrained angle of the trained limb produced by intensities of less than 100
measures of isometric elbow flexor (y-intercept 11 per cent; p = 0.001 ) per cent. When second and third-order
strength were assessed with and at the trained and untrained angles polynomials are fitted to the data, they
incremental polynomial regression. on the untrained limb (y-intercept 9 show maxima at intensities of 87 per
This involved fitting high order per cent and p = 0.02 at both angles; cent and 80 per cent respectively.
polynomials to the data, assessing the Figure 1b-d). These strength increases Interestingly, these maxima correspond
goodness of fit as measured by the did not appear to be influenced by quite closely to the highest target
adjusted coefficient of determination, training intensity because the slope of intensity which subjects could sustain
and then decreasing the order of the the regression relating training for a whole training session. This is
polynomial by one. The process was intensity to increases in strength at the consistent with the hypothesis of Szeto
repeated until there was a significant untrained angle of the trained limb or et al (1989) that training at a target
decline in the goodness of fit. It was at the trained or untrained angles on intensity of 100 per cent MVC is sub-
found that there was no significant the untrained limb did not differ optimal because fatigue limits the
explanatory power in the second or significantly from zero (slopes of 0.07, actual training intensity (even though
higher-order terms, so ultimately, p = 0.20; 0.04,p = 0.59 and 0.03, it has recently been shown that fatigue
simple linear regression was used in all p = 0.6 respectively). is a necessary component of an optimal
analyses. Analysis of the residuals of dynamic strength training program;
the linear regression suggested that the Discussion Rooney et a11994). However, higher-
assumptions of normally-distributed order polynomials do not explain any
error and homoscedasticity were met. In the present study, a regression more of the total variance than linear
The use of robust curve-fitting design was used to investigate the regression, so the optima at
techniques was considered, but analysis optimal intensity of isometric strength submaximal intensities may be a result
of the residual distributions suggested training. As subjects were randomised of "over-fitting" the data. A
that these more complex models to different intensities, this design conservative interpretation of these
provided no better fit to the data, so provides the same degree of data is that the optimal training
they were not used. experimental control as if subjects had intensity is maximal or near-maximal.
been randomised to groups that Future studies with larger sample sizes
Results trained at different intensities (as in the
earlier study by Szeto et al 1989).
might be able to provide a more
precise estimate of the optimal training
With the exception of one subject who Moreover, as some subjects trained at intensity, although the nature of
withdrew from the study early in the low intensities which would generally sampling error is such that a much
training program, all subjects attended be considered to have no training larger sample size would provide only a
all training sessions within the six-week effect, these subjects effectively modest increase in precision.
training period. provided the same degree of control as
This conclusion appears to differ
Figure 1 shows the effect of training a sham control group would provide in
from that of Szeto et al (1989), as those
intensity on strength at the trained a randomised groups design. The
authors concluded that training at
angle of the trained side. The slope of regression approach used here has the
intensities of 50 per cent MVC
the regression was 0.19 (p < 0.006), additional benefit of potentially
produced substantially greater
indicating that subjects who trained at providing information about the shape
increases in strength than training with
intensities near maximum typically of the relationship between training
maximal contractions. The Szeto study
attained strength increases about 19 intensity and increase in strength, and
differed from the present study in a
per cent greater than subjects who thus it potentially provides a more
number of respects including muscles
trained with near zero intensities. The precise estimate of the optimal training
groups trained (knee extensors versus
y-intercept of the regression line, intensity than a group design using
elbow flexors), number and duration of
which provides the best estimate of the comparable numbers of subjects.
contractions (30 5s contractions five
training response of subjects who As the relationship between training times weekly versus six 10s
trained near 0 per cent MVC, was 5.3 intensity and increase in strength was contractions thrice weekly), duration of
per cent (p = 0.19). As the relationship apparently linear, the data suggest that training (3 weeks versus 6 weeks) and
between training intensity and the the optimal training intensity for total strength increases (46 per cent
increase in strength at the trained isometric strength training is maximal. versus 15 per cent when training at 50
angle on the trained limb is apparently However, given the high degree of per cent MVC), but it is not clear how
linear, and as the slope of the between-subject variability, it may not these differences might influence the
regression is positive, the data suggest have been possible to detect small non- optimal training intensity.
that training with maximal linearities in the relationship between
As in many previous studies, we
contractions produces the greatest training intensity and strength, so it is found that when subjects trained with
increase in isometric strength. possible that the greatest strength
~
--
ORIGINAl ARTIClE
from Page 45 are an experimental artefact, rather for Field Settings. Boston: Houghton Mifflin.
one limb at one joint angle they than a training effect produced by Enoka R (1988): Muscle strength and its
forceful muscle contractions. development: new perspectives. Sports
experienced angle-specific increases in Medicine 6: 146-168.
strength; increases in strength at the While contralateral training effects Garfinkel S and Cafarelli E (1992): Relative changes
untrained angle were smaller than at have been widely reported (Enoka in maximal force, EMG, and muscle cross-
the trained angle, even though the 1988, Smith 1970) and frequently cited sectional area afrer isometric training. Medicine
increase in strength was expressed as a as evidence of neural training andScienceinSportsandExercise24: 1220-1227.
percentage of pre-training strength at mechanisms (eg Sale 1988), existing Kannus P, Alosa D, Cook L,J ohnson RJ, Renstrom
that angle. The effects of training evidence of contralateral effects of P, Pope M, Beynnon B, Yasuda K, Nichols C
and Kaplan M (1992): Effect of one-legged
intensity on strength at untrained training (in contrast with an effect of exercise on the strength, power and endurance
angles was small and appeared to be testing) is not particularly strong. of the contralateral leg. A randomized,
independent of training intensity, Production of definitive evidence of controlled study using isometric and
because the slope of the relationship contralateral training effects would concentric isokinetic training. European
necessitate demonstrating that the Journal ofApplied Physiology 64: 117-126.
between training intensity and increase
in strength did not differ significantly increases in strength on the Kitai TA and Sale DG (1989): Specificity of joint
angle in isometric training. European Journal
from zero. Likewise, the apparent contralateral limb are greater than the ofApplied Physiology 58: 744-748.
contralateral effects of training were increases in strength experienced by a
Lindh M (1979): Increase of muscle strength from
independent of training intensity - control group that does not train. isometric quadriceps exercises at different
training at near-zero intensities However, to our knowledge only six knee angles. Scandinavian Journal of
appeared to have the same effect on studies of isometric training have used Rehabilitation Medicine 11: 33-36.
the strength of the contralateral limb no-training controls to investigate McDonagh MJN and Davies CTM (1984):
as training at high intensities. contralateral training effects (Cannon Adaptive response of mammalian skeletal
and Cafarelli 1987, Garfinkel and muscle to exercise with high loads. European
There are two possible Journal ofApplied Physiology 52: 139-155.
interpretations of these findings. First, Cafarelli 1992, Kannus et aI1992,
Rooney KJ, Herbert RD and Balnave RJ (1994):
it is possible that the effects on Komi et al1978, Laughman et a11983,
Fatigue contributes to the strength training
strength at untrained angles and on the Sale et al 1992, Smith 1970) and none stimulus. Medicine and Science in Sports and
untrained limb were too small to be have demonstrated significandy greater Exercise 26: 1160-1164.
detected, given the variable response to increases in strength in untrained limbs Sale DG (1988): Neural adaptation to resistance
training. We consider this to be a of trained subjects than in untrained training. Medicine and Science in Sports and
control subjects. Exercise 20: S135-145.
likely explanation of the non-
Smith LE (1970): Facilitatory effects of myotatic
significant effect of training intensity at In conclusion, short-duration
stretch training upon leg strength and
the untrained angle of the trained limb, isometric strength training is most contralateral transfer. American Journal of
as it is hard to conceive of mechanisms effective when it is performed at Physical Medicine 49: 132-140.
whereby training would increase maximal or near-maximal intensities. Szeto G, Strauss GR, de Domenico G and Lai HS
strength at the trained angle but have Angle-specific and contralateral effects (1989): The effect of training intensity on
no effect on strength at untrained of isometric training are at least pardy voluntary isometric strength improvement.
_independent of training intensity, and AustralianJournalofPhysiotherapy 35: 210-217.
angles. However, an alternative
may have been exaggerated by Thepaut-Mathieu C, Van HoeckeJ and Maton B
explanation may also be correct: the (1988): Myoelectric and mechanical changes
y-intercept of the regressions on experimental artefact. linked to length specificity durng isometric
strength increases at the untrained training. Journal of Applied Physiology
References 64: 1500-1505.
angle of the trained limb and both
angles on the untrained limb differed Atha J (1981): Strengthening muscle. Exercise and Weir JP, Housh TJ and Weir LL (1994):
Sports Science Reviews 9: 1-74. Electromyographic evaluation of joint angle
significantly from zero, indicating that
Berger RA (1962): Optimum repetitions for the specificity and cross-training after isometric
while training intensity had no training. Journal of Applied Physiology
development of strength. Research Quarterly
detectable effect on strength, strength in Exercise and Sports Science 33: 334-338. 77: 197-201.
increased nonetheless. If the response Cannon RJ and Cafarelli E(1981): Neuromuscular
was intensity-independent it could not adaptations to training. Journal of Applied
be due to training, at least in so far as Physiology 63: 2396-2401.
training implies forceful muscle Clarke DH (1973): Adaptations in strength and
contractions. The most likely source of muscular endurance resulting from exercise.
the increase in strength is a placebo or Exercise and Sports Science Reviews 1: 73 -1 0 1.
Hawthorne effect, or familiarity with Cook TD and Campbell DT (1979): Quasi-
the test protocol (Cook and Campbell experimentation. Design and Analysis Issues
1979). If this interpretation is correct,
the contralateral training effects
commonly reported in training studies