**5. Psychoacoustic / behavioural measurements**

The first behavioural audiogram of the gerbil was determined by Ryan (1976) using a shock avoidance procedure. Subsequently, Sinnott et al. (1997) developed a go-nogo procedure

The Mongolian Gerbil as a Model for

**5.2 Temporal integration** 

**5.3 Gap detection** 

and varied as a function of threshold elevation.

the Analysis of Peripheral and Central Age-Dependent Hearing Loss 83

There is not a simple, straight-forward explanation for the age–dependent, increasing discrepancy between behavioural and ABR thresholds (Hamann et al., 2002). One factor is a decreased synchronisation of the neural responses that will lead to reduced amplitudes of evoked potentials and reduced slopes of the CAP and ABR growth functions. In addition, a specific loss of auditory nerve fibres with low spontaneous activity, that typically have higher thresholds than those with high spontaneous activity could contribute to reduced amplitudes of CAP and ABR without associated elevation of threshold (Schmiedt et al., 1996; Lin et al., 2011). The advanced age-dependent cochlear pathologies discussed above

Thresholds for tones increased by more than 10 dB in both normal hearing, young gerbils and old gerbils, as signal duration was reduced from 300 to 10 ms (Gleich et al., 2007b). Like in humans and other species, temporal integration was reduced in gerbils with hearing loss

In the presence of fixed-level modulated and un-modulated speech like maskers, threshold shift due to the masker was inversely related to threshold in quiet: sensitive gerbils showed more masking compared to gerbils with slightly elevated thresholds. Consequently, the temporal integration functions (plots of the masked threshold as a function of duration) became very similar for all 13 gerbils with an age varying between 7 and 43 months and the functions were independent of peripheral hearing. Compared to the unmodulated masker, thresholds for short signals (10 and 30 ms) showed slightly more masking , while those for long signals (300 and 1000 ms) showed slightly less masking in the presence of the modulated masker, suggesting that long signals can be detected in the troughs while detection of short signals interferes with the peaks of the modulated masker. These data

(see heading 3) will eventually lead to behavioural manifestation of hearing loss.

suggest that temporal integration in normal hearing gerbils is not affected by age.

The gap detection paradigm determines the minimum duration for the detection of a short period of silence (gap) embedded in a broadband noise pulse. It has been widely used to characterise the temporal resolution of the auditory system. By selecting old human subjects with no or minimal peripheral hearing loss (determined by pure tone audiometry), Snell (1997) demonstrated that mean gap detection thresholds increased with age even in the absence of peripheral hearing loss: some old subjects showed impaired performance, while others retained good temporal resolution, resulting in an increased inter-individual variability of gap detection thresholds in old human subjects. Very similar results were obtained in gerbils (Hamann et al., 2004). When tested with a noise carrier presented 30 dB above the threshold for the carrier, the minimum audible gap in young gerbils was below 4 ms, while approximately 50% of the old gerbils had gap detection thresholds above 4 ms. The variation of the threshold for the noise carrier explained less than 20% of the variation of the gap detection threshold in gerbils. This suggests that peripheral hearing loss was not the dominant cause of impaired temporal resolution. These data point to central auditory processing deficits that result in increased gap detection thresholds in normal hearing old humans and gerbils and are consistent with results obtained by ABR in gerbils (Boettcher et

where gerbils initiate a test trial by jumping onto an observation platform and indicate the perception of the test stimulus by jumping off the platform. In this procedure, correct responses are rewarded by a small food pellet. By repeated presentation of a fixed set of test stimuli, where the parameter under investigation varies over a given range (e.g. sound pressure level), a psychometric function is constructed by plotting the correct response probability as a function of the stimulus parameter. The derivation of threshold and other parameters from psychometric functions of gerbils are described in detail in Gleich et al. (2006). In addition to measuring threshold for the detection of signals in quiet, more complex tasks require the detection of a stimulus that deviates form a constantly and repeatedly presented background stimulus. This approach has been used to characterise the ability to discriminate between synthetic speech-like stimuli (Sinnott & Mosqueda, 2003), determine the minimum audible gap duration in a broadband noise pulse (Hamann et al., 2004) and characterise forward masking (Gleich et al., 2007a) in gerbils.

#### **5.1 The audiogram and age-dependent threshold elevation**

Behaviourally–determined thresholds in up to 3–year-old gerbils using positive reinforcement (food reward; Hamann et al., 2002) resembled those previously reported by Ryan (1976) using shock avoidance. Behavioural thresholds of 30-36 month-old gerbils showed no significant elevation for broadband noise and 10 kHz and only a small degree of hearing loss at 2 kHz (mean 7.2 dB) compared to gerbils up to 1 year of age. Inter-animal variability of behavioural thresholds in gerbils older than 3 years increased and showed a higher mean loss. However, in clinical terms, the losses were only mild, typically less than 40 dB (Hamann et al., 2002). This pattern differed considerably from the description of hearing loss based on previous CAP and ABR measurements where the mean hearing loss was 10-20 dB at 2 years of age and increased to 25-30 dB for frequencies above 4 kHz in 3 year old gerbils (Hellstrom & Schmiedt, 1990; Mills et al., 1990). To address the question whether the different patterns of hearing loss observed by ABR and behavioural testing were due to methodology or the breeding line, Hamann et al. (2002) determined ABR thresholds in a group of 5 gerbils at the age of 28-29 months. On average, these gerbils showed a 14 dB hearing loss, whereas thresholds in 4 of these gerbils determined at 18 months of age were not elevated. Based on the ABR, these 5 gerbils developed more than 10 dB hearing loss between 18 and 28-29 months of age. Behavioural thresholds of these same gerbils were obtained a few months following ABR testing at an age of 31-33 months and were not found to be elevated compared to the behavioural thresholds of young gerbils. Thus, although ABR thresholds were clearly elevated in these gerbils at 28-29 months of age, the behavioural thresholds determined a few months later showed no hearing loss, the elevation of ABR thresholds in these old gerbils was not reflected in the behavioural thresholds. The difference between behavioural thresholds and thresholds determined by ABR in the frequency range of 1-8 kHz increased from 13-14 dB in young (< 12 months) to 25-30 dB in approximately 2 year old gerbils. A similar observation was described in Boettcher (2002a) for humans. The difference between ABR and behavioural thresholds was around 8 dB at 2 and 4 kHz in a group of young human subjects and increased to around 20 dB in a group of old human subjects. Thus, ABR based thresholds in old humans and gerbils may lead to an over-estimation of threshold loss compared to pure tone audiometry.

There is not a simple, straight-forward explanation for the age–dependent, increasing discrepancy between behavioural and ABR thresholds (Hamann et al., 2002). One factor is a decreased synchronisation of the neural responses that will lead to reduced amplitudes of evoked potentials and reduced slopes of the CAP and ABR growth functions. In addition, a specific loss of auditory nerve fibres with low spontaneous activity, that typically have higher thresholds than those with high spontaneous activity could contribute to reduced amplitudes of CAP and ABR without associated elevation of threshold (Schmiedt et al., 1996; Lin et al., 2011). The advanced age-dependent cochlear pathologies discussed above (see heading 3) will eventually lead to behavioural manifestation of hearing loss.
