Journal of Rehabilitation Medicine 51-4inkOmslag | Page 16

250
R. Stock et al.
electronic dynamometer is valid (intra-class correlation (ICC)
0.98–0.99) compared with the Jamar hydraulic dynamometer
and has excellent test-retest reliability (ICC 0.98–0.99). The
advantages of electronic dynamometers are the sensitivity to
record low grip force and the possibility to assess force-time
characteristics. Pinchmeters can be used to obtain a reliable
assessment of pinch strength in stroke patients (17, 18).
All grip strength measurements were performed according to
the recommendations of the American Society of Hand Thera-
pists (ASHT) (19). The patients were seated with their shoulder
in a neutral position, the elbow flexed to 90°, the wrist in a
neutral position; the same chair was used for all measurements.
The examiner explained and demonstrated the testing procedure.
First, 2 trials with submaximal isometric contractions were
performed to familiarize the participant with the equipment.
Each MVC was performed 3 times in the 5 handle positions.
The hands were tested alternately, with 30 s rest between the
trials, i.e. 60 s rest before the same hand was tested again, as
recommended by Watanabe et al. (20).
Pinchmeter recordings were performed in the same manner
in 3 different grip positions (see Fig. 1, F–H): key grip (holding
the pinchmeter between the lateral side of the 2nd phalanx of the
index finger and the tip of the thumb), 3-finger grip (holding the
pinchmeter between the fingertips of the index finger, middle
finger and thumb) and 2-finger grip (between the fingertips of
the index finger and thumb). If the patient was not able to hold
the instrument in a stable position, it was placed and gently held
by the examiner in the correct position. During actual testing, the
participants were instructed to grip as hard as possible, and were
encouraged verbally, as follows: “Harder...Harder...Relax” (21).
Rate of force development and sustainability of grip force
A 0–0.5-s time interval was chosen to evaluate rate of force
development (22). Sustainability of grip force can be measured
in absolute values (23) or as the percentage or ratio of the mo-
mentary force value relative to maximum force (24, 25), which
makes it possible to express how much the individual force
curves drop during a given period. Both absolute and relative
values are reported in the current study. The measurements of
rate of force development and sustainability of grip force were
performed once in both the affected and the non-affected hand
with a hand dynamometer (position 2) and pinchmeter (key
grip). The participants were instructed to increase grip force as
fast as possible, followed by the instruction “hold as hard as
you can” for 15 s.
was assessed by visual inspection of quantile-quantile (Q-Q)
plots. Due to multiple comparisons between the 2 hands during 5
time-points and 5 grip positions, the possibility for Type II error
was high. p-values < 0.05 were therefore regarded as indicating
a possible difference between the hands.
RESULTS
Of the 47 patients included in the NORCIMT study, 14
were recruited from Trondheim University Hospital. Of
these, 11 participants had available data on maximum
grip force, rate of force development and sustainability
of grip force of the affected and non-affected hands.
Table I presents the baseline characteristics of the 11
patients included in the study. The participants were
middle-aged to elderly and mostly men. The Fugl-
Meyer score for the upper extremity indicates that
the patients had mild to moderate reduction in motor
function. Disability (Modified Rankin Scale) ranged
from slight disability to moderate disability.
One patient missed the follow-up assessments
at W28, W30, and W54 for the non-affected side
because of pain due to overload of the non-affected
hand during walking with walking aids. More than 3
s was needed to reach maximum force during some
recordings. As a result, 7% (15/214) of the sustained
curve recordings were shorter than 12 s (mean 9.3 s
(SD 2.1)). One patient had an additional minor stroke
after 28 weeks, which did not result in a pronounced
difference in grip strength parameters, except that the
force curve dropped markedly faster during sustained
grip on the non-affected side at W30, but not at W54.
The assessment of the onset of the force-time curves
by 2 independent raters showed excellent agreement:
ICC (3,1) = 0.98.
Power grip strength
Patients reached the highest maximum force values
in hand grip position 2 for both the affected and
Statistical analysis
Stata (StataCorp. 2017. Stata Statistical Software: Release 15.
College Station, TX: StataCorp LLC) was used for the statistical
analyses. Background variables were reported as mean (standard
deviation (SD)) or median (range) when non-normally distri-
buted. Differences between the affected and non-affected hand
were analysed by independent t-test, or by the Mann˗Whitney U
test when the data were non-normally distributed. The onset of
the force-time curve was visually determined by 2 independent
raters as the point where the curve starts to rise after stable base-
line measurements. In case of disagreement on the onset point,
the raters reached consensus through discussion. ICC (3,1) was
used to determine the degree of agreement between the 2 raters.
Last observation carried forward was applied where observa-
tions were missing in the non-affected hand. Normal distribution
www.medicaljournals.se/jrm
Table I. Baseline characteristics of participants ( n  = 11)
Characteristics Values
Age, years, mean (SD) [range]
Females, n (%)
Days post-stroke, mean (SD) [range]
National Institutes of Health Stroke Scale (0–42), mean
(SD) [range]
Fugl-Meyer Assessment of the upper extremity (0–66),
mean (SD) [range]
Modified Rankin Scale (0–6), mean (SD) [range]
Affected side, right, n (%)
Dominant side affected, n (%)
Ischaemic stroke, n (%)
New stroke after inclusion, n (%) 59.1 (11) [44–78]
3 (27)
16.4 (7) [7–29]
SD: standard deviation.
3 (2) [0–6]
48.7 (7) [32–61]
2.6 (1) [2–4]
6 (55)
7 (64)
11 (100)
1 (9)