Activity levels during inpatient stroke rehabilitation
Fig. 2D). No statistically significant differences could
be observed in healthy controls between weekdays and
weekends (Table III). There was no interaction effect bet-
ween leg activity and independence in walking (p > 0.55).
Activity levels in stroke compared with healthy
controls
Participants with stroke showed lower activity (z > 3.28,
p < 0.05, r > 0.61) in both limbs compared with controls,
except for the less-affected arm at weekdays (z = 1.61,
p = 0.11). The median value of the arm activity in healthy
controls was twice as large and leg activity 3 times as
large compared with the affected limbs in stroke. Also
the asymmetry between limbs (SMA ratio) in stroke was
larger between arms (z > 2.22, p < 0.05, r>0.40) and legs
(z > 2.73, p < 0.01, r > 0.51) than in controls.
Feasibility evaluation
Wearing sensors for 2 sessions of 48 h including the
night was perceived as acceptable by most of the parti-
cipants. Approximately half of the participants described
that they forgot that they were wearing the sensors at
some point. The trunk sensor was, however, perceived
uncomfortable or even disturbing by one-third and 8
reported problems with Velcro straps (too loose or tight,
skin irritation) and/or that the sensors were in the way
while getting dressed. Five expressed that it was reli-
eving to take the sensors off after a measurement period.
Two worried whether the sensors were always correctly
placed, e.g. when donned after showering. A few found
the measurement stimulating and were interested to
get feedback of the results. One participant described
a feeling of being watched over, and one preferred to
wear the sensors under the clothes, not visible to others.
The sensors were well tolerated overnight, but 5 indi-
viduals found the blinking from light-emitting diode
(LED) lights on the sensors, especially in the darkness,
disturbing. Two control participants working at hospital
needed to remove sensors on the arms occasionally due
to the hygiene rules. In rating their perception whether
the sensors were comfortable to wear, one agreed stron-
gly (3%), 11 agreed (30%), 19 agreed partly (53%), and
5 did not agree (14%). No adverse effects were repor-
ted. The management of multiple sensors and the need
for customized application for increased convenience
was perceived as a barrier for clinical feasibility by the
researcher collecting the data.
DISCUSSION
After stroke, people with hemiparesis in the inpatient
rehabilitation setting use not only their more-affec-
ted, but also their less-affected, arm and leg less at
431
weekends than on weekdays. Asymmetry in arm use,
indicated by a ratio of paretic to non-paretic arm, was
increased at weekends in people with stroke. Healthy
controls did not show differences between weekday
and weekend measurements in any of the activity
metrics. The findings of the current study, showing
differences in arm and leg activity levels between
weekdays and weekends, are novel. The results also
imply that the acceleration metrics used in this study
were sensitive enough to capture differences in activity
levels dependent on whether the measurement was
performed on a weekday or weekend.
There may be many factors that influence the activity
levels among people with hemiparesis after stroke in
the inpatient settings. The impairment level can vary
from very severe to mild, meaning that the need for
assistance in different activities varies. The hospital
environment itself, the number of staff available, and
decisions on rehabilitation routines might be a con-
tributing factor influencing activity levels (26–28).
Differences in activity at weekdays and weekends
can also be influenced by cultural, religious or social
norms, e.g. perceiving Sundays as a day of rest (29).
The current study did not reveal any differences
in trunk and limb activity between weekdays among
healthy controls. Similar results have been found for
working age and older community-dwelling adult
populations, revealing no difference in time spent
in sitting/standing/walking or number of steps taken
between weekdays and weekends (30, 31). Others
have shown that older adults took fewer steps and
spent less time in moderate PA on Sundays compared
with weekdays (29). In chronic stroke, the findings
are divergent regarding differences in PA between
weekdays. A study, using activity sensors on the non-
paretic leg, waist and upper arm, found differences for
stepping time and step counts, but not for time spent
in sitting, standing, or in light/moderate intensity PA
(17). Another study using an accelerometer worn on
the hip showed no difference in daily activity counts
between weekdays and weekends (32). In an inpatient
rehabilitation unit the time spent in task practice and
exercise was observed to be 14% and 2% on weekdays,
and only 7% and 1% at weekends, respectively (27).
In the current study, both the more-affected and less-
affected limbs were used less at weekends compared
with weekdays. This difference in activity cannot be
accounted for by changes in arm function, since the
weekday and weekend measurement sessions were
conducted within the same week, the weekday session
prior to the weekend session. The asymmetry between
arms was also greater at weekends, which may have
been caused by a combination of a generally lower
activity at weekends and that the less-affected arm
J Rehabil Med 51, 2019