The Authors Respond [to: Dennison, Hirsch and Hammond]

We have read with interest the comments by Hirsch et al on our paper in their Letter to the Editor. The authors raise a number of interesting points that we would like to comment on. A first comment of Hirsch concerns our exclusion of patients with neglect. We decided to exclude these patients because clinical experience would suggest that patients with a left-sided neglect would not be able to easily practice while looking to their left in a mirror placed in the mid-sagital plane. While we agree with Hirsch that the study by Ramachandran et al on “mirror agnosia” is interesting, it is important to notice that in this study the mirror is used in a crucially different way. While the mirror in our study as other mirror therapy studies is placed mid-sagital in front of the patients, in the 1999 study by Ramachandran, the mirror is “on the patient’s right side in the parasagittal plane, so that when the patient rotates his head rightward and looks into the mirror, he sees the neglected side of the world reflected in the mirror.” While this concept is very interesting and warrants further study, it can not easily be combined with the mirror training in our study. However, it is interesting that in a recent randomized controlled trial on mirror therapy on upper-extremity motor recovery, Dohle et al found positive effects not only motor recovery, but also on the level of sensory and attentional deficits in 36 patients with severe hemiparesis after stroke.
As mentioned by Hirsch, in our study, we excluded patients with ideomotor apraxia, defined as a deficit in the temporal and spatial sequencing of an action, to obtain a homogeneous group. However, we agree that studying the effect of mirror training in people with apraxia may be worth further study. It is well known that action-execution and action-observation are encoded by the same brain regions and mirrors may be effective in people with apraxia for recruitment of areas involved in motor planning mechanisms. In a recent functional magnetic resonance imaging study, we reported that mirror training in healthy subjects activates areas in the brain related to the mirror neuron system. On the other hand, in a recent study, Pazzaglia et al demonstrated that patients with limb apraxia, a specific deficit in executing skilled limb movements or gestures, also display a deficit in recognizing observed gestures.
Another comment of Hirsch et al deals with a possible difference in the perception of credibility of the intervention in both groups. We agree that this is a valid point since practicing without a mirror will have been less interesting and will have raised less positive expectations than with a mirror. Unfortunately, this aspect is difficult to control in many therapy intervention investigation. Without further study, it is difficult to estimate size of this effect.
As a final comment, the authors comment on the use of parametric tests. While we agree that a nonparametric test may have been more suited, we did check that the choice of test did not have a major impact on the statistical significance of our findings.
We thank Hirsch for their compliments and interesting comments on our manuscript.