Page 38 - Winter 2020

PSYCHOACOUSTICS OF TINNITUS
high doses has been shown to reliably induce tinnitus in humans but is also usually reversible and again limited to high doses; a baby aspirin is unlikely to cause tinnitus.
Given that tinnitus is a phantom auditory perception, how can it be measured in animals? The simple answer is that patients cannot perceive quiet while tinnitus is present, and neither can animals. Across studies, animals are trained to exhibit one set of behaviors (e.g., pressing levers, moving from one side of the chamber to another, climbing a pole) when there is no sound in the environ- ment and another set of behaviors when sound is on in order to obtain food or avoid punishment. Among the animal models (Brozoski and Bauer, 2016), the most common approach is to have animals (usually rodents) detect a gap in a continuous sound. When tinnitus is present, animals make more errors detecting gaps in continuous sound, especially if the frequency of the con- tinuous sound is similar in pitch to their tinnitus.
Several of these animal studies have shown that the pat- tern of results supports the presence of tinnitus after high doses of sodium salicylate, quinine (an antimalarial drug known to induce tinnitus in humans), and noise expo- sure. Importantly, the pitch of the tinnitus is consistent with the adjusted frequency range (relative to peripheral HL) reported in humans.
To effectively test animals for the presence of tinnitus, several fundamental features are necessary for rigorous investigation. These include the use of well-established behavioral response paradigms for determining the phan- tom sound of tinnitus, known and reliable inducers of tinnitus, and/or reliable physiological responses consistent with the presence of tinnitus. Psychophysical assessment of tinnitus is typically categorized either as an interrogative model, which evaluates changes in behavioral outcomes as a function of tinnitus, or as a reflexive model, which assess changes in automatic, lower-order processing responses consistent with the perception of a phantom sound.
Interrogative models require that the animal voluntarily respond to the acoustic environment indicating the pres- ence of silence or the presence of an auditory stimulus. Early preclinical behavioral measures of tinnitus used interrogative methods, operant conditioning, and response suppression to detect and characterize the presence of tin- nitus (Jastreboff et al., 1988). In the first animal model,
rats were conditioned to associate a mild but unavoidable foot shock that occurred after a continuous sound was turned off. This resulted in suppressed licking from a water spout in preparation of the imminent shock. Following conditioning, rats in the experimental group were given a high dose of sodium salicylate, whereas the control group received a placebo. During this phase, the foot shock was eliminated but the sound conditions remained. Rats in the control group continued to suppress licking when the sound was turned off because the lack of sound was associated with foot shock. In contrast, rats treated with sodium salicylate continued to lick even when the sound was turned off. Simply put, the animals could not tell that the sound was turned off (presumably due to presence of tinnitus) and continued to lick from the waterspout.
A number of subsequent animal models have shown results consistent with the presence of tinnitus and consis- tent with Jastreboff ’s lick suppression model (Eggermont and Roberts 2015). Other models have used either avoid- able shock or positive reinforcement with food whereby animals have to differentiate between trials with sound and trials with no sound. Although interrogative assess- ments in animal models are crucial for investigating perceptual correlates of tinnitus, it is important to note the considerable challenges in interrogative models because behavioral conditioning requires lengthy and consistent training schedules (Brozoski and Bauer, 2016), and even then, some animals may not respond as expected due to inability to do the task or lack of motivation.
Given the challenges associated with interrogative models, reflexive models for tinnitus assessment have been widely used for determining the presence of tin- nitus. The acoustic startle reflex (ASR) is a large-motor response akin to a jumping/jolt-like response that can be readily elicited in rodents using a loud startling acoustic stimulus. The ASR can be easily measured in rodents using pressure sensitive platforms to record the ampli- tude and duration of the reflex (Turner et al., 2006).
Interestingly, the ASR can be attenuated by presenting an acoustic cue before the startling acoustic stimulus. For example, a 50-ms tone before the loud startling stimulus will result in a reduction in the ASR. Because of the com- pressed time frame, the changes in the ASR are believed to involve rapid lower level auditory processing before the startle elicitor; in other words, the animal did not
38 Acoustics Today • Spring 2021