deafness & Cognition

One astonishing consequence of cochlear implantation in congenitally deaf children is the outcome variability: even in absence of additional disabilities 30% of the implanted children, despite benefiting from the implant, do not meet the expectations of the clinicians. The reasons for this remain unknown.

Alexander Lurija observed that hearing loss has consequences outside of the auditory system. Due to the extensive  connections between all subsystems in the brain, loss of hearing must result in widespread adaptations in the brain. We suggested to view congenital deafness as a “connectome disease” (Kral et al., 2016) where the effective connectivity between the auditory system and other subsystems of the brain is modified in order to provide optimal adaptation to deafness. These modifications may be highly individual, depending on the strategy used in coping with sensory loss.

Connectome adaptations go beyond the well-known cross-modal reorganization (a take-over of the auditory system by the remaining sensory systems) and concern also the fine motor coordination, attentional processes, working memory, and executive functions (reviews in Kral et al., 2016; Kral & O‘Donoghue, 2010).  As an example, the sustained attention is reduced during first years of life in deaf children to the expense of attention to the periphery of the visual field. This may reduce the joint attentional timespan, i.e. the time spent focused on the same subject as the caretaker (mother), an important factor for learning. Furthermore, Prof. W. Kronenberger (Indiana University) and colleagues provided evidence that learning sequencing of events in the visual system is affected by absence of hearing (Conway et al., 2009; Kronenberger et al., 2014, reviews in Kral et al., 2016; Kral & O‘Donoghue, 2010). Our data from the congenitally deaf cat support a reorganization in the connectome, less at the anatomical level (Barone et al., 2013), more at the functional level, in oscillations  that are fingerprints of functional connectivity (Kral et al., 2017; Yusuf et al., 2017), as well as with functional and effective connectivity measures between primary and secondary auditory areas (Yusuf et al., 2021).

Each sensory system is specialized for a certain domain of the sensory stimuli: e.g. the auditory system outperforms the visual system by nearly a factor of 100 in temporal processing, whereas the visual system outperforms the auditory system in spatial processing by a similar factor. Hearing is specialized for detecting change, visual system complements this by reporting static objects and providing continuity in the perception - but consequently the temporal acuity of vision is poor. The sensory systems are complementary, not redundant. The auditory system represents an optimal reference structure for organizing events in time, whereas the visual system for organizing events in spatial relations. Loss of one sensory system cannot be fully compensated by the other systems. Indeed, in case of intersensory conflict the brain listens to hearing in temporal tasks and to vision in spatial tasks (“auditory capture” and “ventriloquist effect”). This explains why loss of hearing cannot be compensated by vision, since, among others, vision lacks the acuity of representing temporal sequences.

Congenital deafness therefore leads to both compensatory supranormal performance (due to cross-modal reorganization) and subnormal performance (due to loss of one sensory system) in behavior. These processes may have individual (subject-specific) components and significantly contribute to the large outcome variability following cochlear implantation. A personalized approach to each deaf child, analyzing these factors individually, may explain why some children do not optimally profit from cochlear implants.

These publications are the result of a collaborative effort together with Profs. W.G.Kronenberger and D.B.Pisoni (both Indiana University, USA) and G.M. O‘Donoghue (University of Nottingham, UK).

Hearing loss is associated with aging and cognitive decline, too, as the work of Dr. Frank Lin, Johns Hopkins Univ., USA, convincingly documents. How much this association is related to communication opportunities, to language, to hearing and to vascular issues that may affect both hearing and the brain remains to be investigated in the future.

The connectome model suggests that absence of hearing from birth changes the brain also outside of the auditory system, with adaptations in many non-auditory functions. These may be different in different individuals, leading to a high outcome variability after cochlear implantation. Figure modified from Kral et al., 2016, Lancet Neurol.
Using electroencephalographic recordings, early sensory processing and higher-order processing can be differentiated in the evoked response. For example, using cochlear-implanted single-sided deaf subjects we could show that both early sensory processing as well as subsequent higher-order cognitive processing differs in the same subject when the hearing and the deaf ear is stimulated. This demonstrates using objective measures that cochlear implants require more cognitive resources than normal hearing. The plot shows the sensory-dominated N1 component and the cognitive N2 component recorded using EEG in the same subjects (stimulation on hearing ear and deaf, implanted ear). From Finke et al. (2016).Deaf_Connectome_files/Finke_SSD_2016.pdfshapeimage_5_link_0
Parts of the connectome following congenital deafness in the cat model, shown using retrograde tracer analysis for two fields: A1 (primary auditory cortex) and DZ (secondary auditory cortex). Thickness of connecting lines corresponds to the number of projections. Orange lines denote ectopic projections (that have no correlate in hearing controls).

The results demonstrate generally preserved connection patterns within the auditory cortex, but also shifts in connection strength as well as new connections in deafness. Effective connectivity shows a more extensive effect of congenital deafness.

Figure from Kral et al. (2017); data from Barone et al. (2013).