Saturday, December 16, 2017

DAF, stuttering, and central auditory processing

The study I want to discuss in this post is again about the effect of delayed auditory feedback (DAF) on stuttering. Luana Picoloto and her colleagues investigated the “Effect of delayed auditory feedback on stuttering with and without central auditory processing disorders.” The study was recently published in the Journal CoDAS – see here (free full text).

In the study, the fluency-enhancing effect of DAF (100ms delay) in two groups of individuals who stutter was compared: (i) a grroup without central auditory processing disorder (APD) and (ii) a group with APD. It came out that the DAF caused a statistically significant reduction of stuttering frequency in the group without APD, but not in the group with APD.

The reduction of the number of blocks and of repetitions of monosyllabic words was statistically significant in the group without APD. By contrast, in the group with APD, the DAF did not cause statistically significant effects, but there was a tendency towards heightened disfluency, particularly regarding part-word repetitions and prolongations (see Figures 1 and 2 in the paper).

There were also similarities between the groups: The number of blocks was reduced by DAF also in the group with APD, but the reduction did not reach statistical significance. Further, the number of intrusions was tendentially greater with DAF than with natural auditory feedback in both groups.

In their paper, the authors do not speculate about the way in which DAF reduces stuttering. But that’s an important question; I think, to understand the mechanism in which altered auditory feedback reduces stuttering is to understand the pathomechanism of the disorder.

My own hypothesis is: Altered auditory feedback (as long as it is unaccustomed) draws the speaker’s attention to the auditory channel, which improves the processing of auditory feedback and its involvement in speech control (see Section 3.1 on my theory website).

Basis of this hypothesis is the assumption that subtle deficits in central auditory processing cause individuals who stutter to turn away their attention from the auditory channel in order to prevent acoustic overstimulation. The core of the auditory processing deficit seems to be a less effective auditory gating, that is, the processing of redundant acoustic input is insufficiently suppressed (see Kikuchiet al., 2011).

Children with such an auditory processing deficit may early develop a compensatory habit of attentional misallocation, namely to always turn attention away from the auditory channel, except moments of active listening. Altered auditory feedback overcomes this habit because (and as long as) it sounds unfamiliar and odd. This hypothesis is supported by the fact that devices like SpeechEasy reduce stuttering even when delay and frequency shift are disabled (Foundas et al., 2013; Unger, Gl├╝ck, and Cholewa, 2012). Thus it seems not to be the delay or the frequency shift as such which acts, but the simple fact that it sounds anyway odd.

Is my theory consistent with the new findings obtained by Picoloto and colleagues? The first point is: Can my theory be true given that many stutterers have no (diagnosed) APD? There are two possible answers: (i) Deficits in central auditory processing may be very subtle in many a stutterer such that the scores in standardized tests are within normal range. (ii) All the tests applied for diagnosing APD (see Procedures) aim to detect something presented and not heard, but the problem of some stutterers may be that they hear too much too intensively because of a deficient auditory gating (see above).

An alternative possibility is: There are two groups in persistent stuttering, both with a deficit in attention control, but with versus without APD. I assume this because I don’t believe that APD immediately causes stuttering. If that was the case, then the disorder could not be as influenceable by emotions, situations, anticipations as it is. I think attention (i.e., the allocation of perceptual and processing capacity) is the interface between the mechanism of stuttering (which immediately causes the symptoms) and factors that negatively impact attention control, among them APD, but also fear or adverse communication situations.

A second point is: When I (i) assume that APD causes a misallocation of attention during speech, and this, in turn, causes stuttering (see Chapter 2 of my website),and when I (ii) assume that DAF overcomes the misallocation of attention, then one may expect a stronger (or at least a similar) effect of DAF in the group with APD, but the contrary was found. A possible explanation may be that the delay of 100ms was too much for the group with APD. In earlier studies. good effects were achieved with delays of 50ms and 75ms, and, as mentioned above, even with an active device in the ear, but without feedback alteration. On the other hand, a too long delay evokes disfluencies not only in nonstutterers, but also in stutterers. The increased number of intrusions perhaps indicates that the DAF was irritating for at least some participants in both groups.

If so, then we can assume that some participants attempted to ignore the DAF, that is, they all the more turned their attention away from the auditory channel, which, after my theory, results in more stuttering. And it would not be surprising if those with APD were more apt to behave in this way. They may be acoustically more sensitive and may more likely experience a 100ms delay as annoying.

If my assumption is true that the DAF effect on stuttering is an effect on attention, then there are always two possibilities: (i) the appropriate delay draws attention to the auditory channel and reduces stuttering, and (ii) an inappropriate, i.e., too long delay increases the person’s disposition to ignore the auditory channel, which can increase stuttering. The crucial thing is: Delayed auditory feedback must still be useful for the (automatic and unconscious) self monitoring of speech and for the mechanism of ‘audiophonatorycoupling’, which is not the case when the delay is too long. Therefore, it may be interesting to test the effect of various delays in a further study, or the effect of delays individually adapted.