3.3.3.a Introduction

Tinnitus consists in a phantom auditory perception in the absence of any corresponding external acoustic stimulus. It is estimated that nearly 10 % of the population [Coles, 1996] experience tinnitus. Disabling chronic tinnitus raises the question why abnormal and non-relevant neuronal signals associated with tinnitus are discriminated from background noise without habituation. Habituation is defined as an adaptation process of the Central Nervous System (CNS), namely a decrease or a complete disappearance of the perception of a stimulation that is either continuous in the environment and/or non-positively/negatively reinforced. Actually, most tinnitus patients (75 %) make a successful adaptation to the presence of these phantom sounds, but when it fails, tinnitus may become a source of significant disability, sometimes leading to a depressive state or even to suicide [Lewis et al., 1994]. In this context, it has been suggested [Hallam et al., 1984; Jastresboff, 1990; Andersson, 2002] that habituation to tinnitus is prevented by the patient’s emotional response to the perceived threat represented by tinnitus. In other words, in the case of chronic disabling tinnitus, one may hypothesize a negative reinforcement due to the frequent negative associations of tinnitus sufferers [Jastreboff and Jastreboff, 2000]. Consequently, the highly emotional connotation of disabling tinnitus could lead to a higher level of selective attention directed toward the tinnitus signal, that could either potentially increase the distress induced by tinnitus, or prevent habituation from happening, or both. Indeed Jacobson et al. [1995] found, for the first time, convincing evidence of an abnormal attention participation in tinnitus sufferers. In their study, they analyzed auditory selective attention indexed from the measurement of event related potential called the negative difference wave. The results showed that this index of early selective auditory attention was of significant greater magnitude in tinnitus patients, between 110 and 145 ms. The authors suggest that a greater attention processing of sensory information is involved in tinnitus patients compared to non-tinnitus subjects.

However, if selective attention was to be modified in tinnitus patients, we could also expect another cognitive system, at a lower level – involuntary attention, to participate in detecting the neuronal signals related to tinnitus. Indeed, Goodwin and Johnson [1980], by comparing reaction times to auditory stimuli in hearing-impaired tinnitus and normal-hearing non-tinnitus participants, found that the tinnitus group exhibited shorter reaction times (RTs) to auditory stimulation than did the non-tinnitus group. The results indicated that RTs of tinnitus and non-tinnitus participants differed only at tinnitus frequency but not at other frequencies. The authors suggested a particular cognitive process at the tinnitus frequency, related to the likely generation process, namely the edge effect [Hazell and Jastreboff, 1990], in tinnitus patients affected by presbycusis. A further study of RTs in tinnitus and non-tinnitus participants confirmed the tendency of tinnitus patients to have a reduction of RTs at sensation levels near the threshold [Nieschalk et al., 1998]. The results of these studies are in favor of an implication of low-level mechanisms in the processing of tinnitus signal.

In the present study, we hypothesized that the automatic direction of attention toward the tinnitus ear could be one of the factors improving the detection of tinnitus. Thus, automatic attention was studied in tinnitus and non-tinnitus participants by measuring the effect of involuntary switching to task-irrelevant sound change on performance in a dichotic listening paradigm (oddball paradigm). Previous studies have demonstrated an impaired processing of targets following task-irrelevant changes [Schröger, 1996]. This effect of impairment will be called “attention capture”.

Three experiments were carried out. Experiment 1 was aimed at testing the oddball protocol to ensure we could observe an attention capture effect. Control participants should present an attention capture effect as revealed by less accuracy and less rapidity to targets following task-irrelevant changes.

Experiment 2 was conducted to investigate the influence of chronic tinnitus on attention processes. Indeed, we aimed at determining whether the presence of tinnitus could influence the attention capture. We reasoned that if tinnitus could automatically attract and direct attention on its signal, tinnitus patients would have more difficulties to ignore the ear affected by tinnitus. Thus, performing a categorization task in the tinnitus ear while ignoring the non-tinnitus ear might be easier than performing that task in the opposite situation. In particular, we hypothesized that less attention capture effect may be shown in the ear where tinnitus was perceived. However, it was critical to control for differences of performance due to the possible masking effects of tinnitus at the perceptual level, i.e., a possible reduction of hearing sensation caused by the presence of tinnitus. Consequently, in order to control for such differences, when necessary, we had to compensate the perceptual difference between left- and right-ear stimuli (see “Method” section).

Finally, in Experiment 3, the same paradigm as in Experiment 1 and 2 was conducted on participants on whom a tinnitus was simulated by presenting a continuous narrow-band noise that had the same acoustic characteristics as a common tinnitus (“tinnitus-simulated” group). Our objective was to address whether the effects of tinnitus on attention processes might be equivalent to those of the presence of a continuous stimulus in one ear. We reasoned that performance observed in tinnitus-simulated participants should only reflect perceptual differences due to the presence of an interfering stimulus and consequently should be different from those observed in tinnitus participants. Thus, tinnitus-simulated participants should present attention capture, but also some differences according to the side of the artificial-like tinnitus signal, since perceptual interference might occur in the ear where the tinnitus-simulation is delivered.