Influence of auditory rehabilitation on reaction time.

Although no specifically bDLF-related change in RT was observed, a general alteration was found at 3 months’ auditory rehabilitation. This effect was subject-dependent: in two subjects (subjects D.P. and A.D.), RT was significantly longer after HA fitting, and in three subjects (subjects F.P, Y.M. and C.V.) it was significantly shorter. Furthermore, even response times at the reference frequency fref, which was located one octave before the beginning of the hearing loss and might consequently be expected not to be affected by HA fitting, were altered by the rehabilitation.

How can this change in mean RT between the two experimental sessions be explained? Repeated practice of an RT task affects the resultant mean RT, generally reducing it (e.g., Welford, 1980). One of the present authors has shown that change in mean RT from session to session does not involve the sensitivity factor of RT, but merely results from a change in the response criterion used by the subject, whether RT decreases or increases (Bonnet & Dresp, 2001). Another factor that might, more hypothetically, be involved in the variations in reaction time is the type of HA fitting. The two subjects in whom RT was significantly longer after rehabilitation had had monaural HA fitting. The type of HA is now known to be extremely important in the rehabilitation of hearing-impaired listeners, and monaural rehabilitation is known to have a negative impact on the unaided ear (Gelfand et al., 1987; Hurley, 1999). Consequently, this negative effect of monaural rehabilitation may be reflected in an overall increase in RT.

An intriguing result concerns the linear relationship between reaction time and frequency discrimination limens at three months post-rehabilitation. Except in one subject in whom the linear adjustment model remained stable and non-significant, this relationship tended to increase, was statistically significant in two subjects and marginally significant in one. RTs were faster when DLF performance was better. How could auditory rehabilitation have a positive influence on this linear relationship? One possible explanation could be linked to the improved discrimination performance in our subjects. This local improvement did not alter the reaction time, but may have had a negative impact on the linear adjustment as a whole. At 3 months, with the supposed disappearance of the local bDLF improvement (Gabriel et al., submitted), the relationship between RT and DLF may thus have strengthened. However, it is hard to imagine how data for one frequency out of seven could be sufficient to elicit such an increase in the linear adjustment model. Another factor linked to procedural learning may be involved. At three months post-rehabilitation, subjects knew the tasks, both RT and DLF, they had to perform, which may have affected their degree of attention, for example, so modifying the results of at least one of the tasks.

The present study emphasised on two points. Firstly, RT is certainly more dependent on peripheral than on central parameters. Furthermore, a remarkable inter-individual variability occurs in hearing-impaired subjects. In our subjects with high-frequency steeply sloping hearing loss, RT were increased or decreased at 3 months post rehabilitation, but certainly reflects more a variation in the response criterion than an effect of HA fitting.

Secondly, if the DLF improvement corresponds to a cortical reorganisation of the representation of lesion-edge frequencies, then the behavioural consequences of this plasticity seem to be difficult to observe and reaction time tasks may be inadequate for this purpose. Nevertheless, HA fitting does enhance the correlation between DLF and RT, so that a straightforward relationship may be present between these two factors. Any such relationship, however, remains to be determined.