The effects of internal and external attentional focus instructions on motor learning after stroke

Dual-task performance is often impaired after stroke. One way to target this
problem is to enhance patients' automaticity of movement, as more automatized
movement should require less attentional control, and hence allow for better
dual-task performance. Recent evidence suggests that enhancing automaticity of
movement can best be achieved by triggering an external focus of attention (on
movement effects) when relearning movements, rather than an internal one (on
movement execution itself). In healthy adults, it has consistently been found
that an external focus accelerates skill acquisition, results in enhanced
movement automaticity, and hence facilitates dual-task performance. However, it
remains to be seen whether these results also generalize toward the stroke
patient population, and whether an external focus can indeed enhance motor
recovery post-stroke.

The primary objective of this study is to assess whether motor learning using
external attentional focus instructions results in superior single-task motor
performance, movement automaticity, and dual-task performance in inpatient
stroke patients when compared with internal focus instructions.

Intervention study:
This study is a single-blind randomised trial concerning the effect of
different attentional focus instructions on motor learning after stroke.

Sixty patients will be included, and randomly assigned to an internal or
external focus group. Patient characteristics regarding demographics, stroke
characteristics, and motor/cognitive functioning will be recorded from the
patients medical file before start of the study.

Wearing a full-body safety harness, patients will practice to stabilize a
balance board (75 x 75 cm) that can only rotate along a horizontal axis in the
patients* frontal plane and has a maximal range of motion of 30 degrees.
Patients* goal is to stand as still as possible for 30 seconds, with their feet
placed at hip width. Task difficulty can be manipulated by adjustment of the
board's rotational stiffness.

Baseline measurements are spread out across two sessions. In the first session,
the Timed-up-and-Go, posturography, 10 meter walk test, MSRS, and USER will be
determined.
On the second baseline session, we will use a 2-down-1-up modified staircase
procedure to determine the threshold rotational stiffness at which patients are
just able to remain in balance (i.e., keep the deviations of the board below
2.5 degrees) for 70% of the trial. The procedure is as follows. Patients
perform 16 single-task trials (of 30 seconds each). Every patient performs the
first trial at a rotational stiffness of 150 Nm/rad. Patients* performance is
then evaluated using the criteria outlined below:

(1) SUCCES = Board deviations are within 2.5 degrees for > 70% of the trial &
patient does not grab the handrail
(2) REDO TRIAL = Board deviations are within 2.5 degrees for > 70% of the trial
& patient grabs the handrail 1 or 2 times
(3) FAILURE = Board deviations are within 2.5 degrees for > 70% of the trial &
patient grabs handrail > 2 times
(4) FAILURE = Board deviations are within 2.5 degrees for < 70% of the trial
(regardless of how many times the handrail was grabbed)
Further, the following rules will be applied:
#1: 2x success in a row = reduce rotational stiffness with -50 Nm/rad
#2: 1x failure = increase rotational stiffness with +40 Nm/rad
#3: Every reversal = halve step size (down to a minimum of -6.25 Nm/rad & +5
Nm/rad)
#4: 4x consecutive successful/failed trials = double step size (up to a maximum
of -50 Nm/rad & +40 Nm/rad)
#5: 2x consecutive ""REDO TRIAL"s = counts as failure (i.e. rotational
stiffness for next trial will be increased)

Afterwards, a regression line is fit through the obtained 16 data points, to
establish the RStiff2.5 (i.e., threshold rotational stiffness at which the
patient successfully manages to keep the board*s deviation below 2.5 degrees
for 70% of the trial). This RStiff2.5 value will serve as the benchmark value
with which the individual patient will start his/her first training session.
Finally, patients will perform 2 single- and two dual-task trials at this
individually determined RStiff2.5 value. No specific attentional focus
instructions will be given throughout this baseline assessment ("Try to stand
as still as possible"). For dual-task trials, patients will also be instructed
to prioritize balance performance over cognitive performance.

Patients will practice the balancing task in fifteen 30-second trials (within
30-minute sessions), 3 times a week, for a period of 3 weeks (135 practice
trials in total). Trials will be separated by 90-second rest intervals. Before
the start of each trial, patients will receive an internal or external focus
instruction. As stated earlier, patients will perform the first block of 5
trials in the first practice session with their personal RStiff2.5. Depending
on the patients' average performance in the preceding block, the rotational
stiffness for the next block of 5 trials can either be reduced with -20%
(performance > 70% & < 2x grabbing the handrail), unchanged (performance > 70%
& >2x grabbing handrail | performance 60-70% & <2x grabbing handrail), or
increased with 20% (performance<60% | performance 60-70% & > 2x grabbing
handrail). The stiffness of the balance board is adjusted in this manner from
each practice session to the next to ensure that the task remains sufficiently
challenging throughout practice.

After one week of practice, a post-test is performed at which the RStiff2.5 is
again determined (following the same procedure as at baseline). Also, after the
complete training has been completed a second post-test is performed, at which
the RStiff2.5 is determined. Also, at this point the TUG, posturographic
measurements and USER are administered again as well. Post-tests are done by an
experimenter who is blind to group allocation.

Assessment of reliability and validity of balance board assessment:
After inclusion for the RCT has been concluded, we will include an additional
22 patients to assess the balance board assessment's concurrent validity and
test-retest reliability. These 22 patients will only perform the baseline
balance board measurement twice, on two consecutive days. Also, they will
perform a subset of the clinical tests described above. Specifically, they will
complete the posturographic measurements, NSA, and MSRS, while the scores on
the BBS, FAC, neuropsychological tests, and USER will be obtained from their
medical files. The clinical tests will be conducted on the first measurement
day, prior to the first balance board assessment. Concurrent validity is then
assessed by correlation the RStiff 2.5 with the BBS score and posturographic
performance measures. Test-retest reliability is assessed by comparing the
balance board scores of both measurement occasions.

The experimental intervention is the manipulation of attentional focus when
learning a balancing task.
During practice, the internal focus group will receive instructions that direct
their attention toward movement execution ("Try to keep your feet as still and
horizontal as possible"). The external focus group will receive instructions
that direct their attention toward the outcome/effects of their movements ("Try
to keep the balance board as still and horizontal as possible").

The (extra) risks associated with participation are deemed negligible.
Participants will practice a balancing task that is often practiced within
conventional rehabilitation therapy. Patients will wear a safety harness, and
performance is continuously monitored by the investigator. Task difficulty will
be adjusted to the patients' abilities, trials are of short duration (30
seconds), and rest breaks in between trials are of sufficient duration.
Therefore, the risks and psychological/physical burden associated with
participation are considered minimal, and certainly will not exceed those
associated with regular conventional rehabilitation therapy,