Mirror therapy: (how) does it work?

Some 50-70% of stroke patients suffer from a paretic arm. Optimizing upper
extremity functioning is therefore a central issue in post-stroke
rehabilitation. A recently developed intervention is mirror therapy, in which a
patient exercises with the affected limb behind a mirror while the non-affected
limb is mirrored. Using the mirror reflection, visual feedback is provided as
if the paretic hand is functioning normally. Several studies have shown the
effectiveness of mirror therapy, for acute, sub-acute and chronic stroke
patients. However, there are a number of ways in which mirror therapy can be
administered. The manner in which one chooses to administer mirror therapy is
partly dependent on ideas on the underlying working mechanism. It is
conceivable that the effectiveness of mirror therapy will be largely dependent
on the manner in which it is administered. However, research that has been
performed so far does not allow for inferences on the most effective
administration of mirror therapy. As a result, some patients receive mirror
therapy in a suboptimal format.

The aim of the present study is to increase the understanding of the optimal
administration of mirror therapy and thereby improve the quality and
effectiveness on this therapy. In order to realise this several different
applications of mirror therapy will be compared. In order to realise this, we
will use a movement task which is known to have short-term learning effects in
stroke patients. By having patients practice the movement task under different
conditions, similar to the different ways in which mirror therapy can be
performed, we can learn the most effective administration of mirror therapy.

Ninety stroke patients will be assigned randomly to six training groups. Every
group will train the movement task in a different way, similar to the ways in
which mirror therapy can be administered.

The intervention consists of performing a motor task 70 times. Dependent on the
group to which participants will be allocated they perform the task with the
affected arm, with the unaffected arm or with both arms together. Two groups
will just watch a video from someone else performing the motor task. Before and
after the intervention all participants will perform the motor task 5 times
with the affected arm. These measurements will serve as baseline and effect
measurement respectively.

The burden for the participants consists of a single visit to the Erasmus MC.
During this visit, patients will undergo three intake measurements, and will
have to perform a reaching task 80 times with either their affected or
unaffected arm. Intake measurements will take around half an hour, whereas
performing the motor task will take around one and a half hour. As far as we
know no risks are associated with either the intake tests or with performing
the motor task.