Research paperAuditory filter width affects response magnitude but not frequency specificity in auditory cortex
Introduction
Spectral processing of sounds occurs at every level of the ascending auditory pathway. Processes at the auditory periphery have been linked to frequency selectivity at the level of the cochlea in the inner ear (Moore, 1986, Moore, 2003). Frequency selectivity here refers to frequency resolution, that is, to the ability to resolve sinusoidal components of complex sounds, which supports, for example, speech perception in noise (Moore, 1986, Moore, 2003). Frequency selectivity at the auditory periphery is often operationalized in terms of the width, or pass-band, of a filter-shaped function, referred to as auditory filter (Patterson et al., 1982, Moore, 1986, Moore, 2005, Glasberg and Moore, 1990). The pass-band of auditory filters has been linked to the performance on a number of psychophysical tasks requiring spectral stimulus processing, including frequency discrimination (Moore and Peters, 1992), tone detection (Schlauch and Hafter, 1991; but see Moore et al., 1996) and perception of speech in noise (Moore, 1986).
At the cortical level, processing the spectral properties of sounds has been investigated using electroencephalography (EEG) and event-related potentials (ERPs). In particular, the N1 component of the ERP, a negative deflection that peaks at around 100 ms after stimulus onset and is generated in auditory cortex (e.g., Hari et al., 1982, Näätänen and Picton, 1987, Pantev et al., 1988, Maess et al., 2007), has been shown to be modulated as a function of sound frequency (Butler, 1968, Picton et al., 1978, Näätänen et al., 1988, May et al., 1999, Herrmann et al., 2013). That is, the presentation of a tone reduces the responsiveness of neural populations in tonotopically organized auditory cortex responding to a succeeding tone, and this decline is strongest for small frequency separations between the two tones. This phenomenon has been labeled stimulus- or frequency-specific adaptation (Jääskeläinen et al., 2007, Jääskeläinen et al., 2011, Herrmann et al., 2013).
In addition, frequency-specific adaptation of the underlying neural population is not fixed, but has recently been shown to depend on the spectral variance in the acoustic stimulation. That is, spread of neural adaptation across tonotopically-organized regions of auditory cortex broadens for acoustic sequences with large spectral variance (Herrmann et al., 2013), and this finding resembles observations from single-neuron recordings in auditory cortex (Bitterman et al., 2008). It has furthermore been proposed that tuning of neural populations to sound frequency varies with stimulation intensity (Phillips et al., 1994), and intensity might thus be an additional factor affecting the spread of frequency-specific adaptation observed for N1 responses.
In an attempt to relate frequency selectivity at the auditory periphery and frequency specificity at the level of auditory cortex, Sams and Salmelin (1994) modeled frequency-specific N1m amplitudes (magnetic N1, i.e., measured using magnetoencephalography; MEG) with an auditory filter function, and observed filter shapes resembling those from previous psychophysical experiments (e.g., Patterson, 1976). Furthermore, in a series of studies, Soeta and colleagues observed increasing N1m amplitudes with increasing frequency separation between two simultaneously presented tones, but only for frequency separations greater than 100 Hz (e.g., Soeta and Nakagawa, 2007, Soeta et al., 2008). Based on this, they concluded that frequency specificity of the N1m amplitude resembles properties reported for auditory filter bandwidth (e.g., Fletcher, 1940). Thus, there are indications for a link between peripheral and cortical indices of frequency resolution.
Importantly, most of these previous studies were conducted in younger, normal-hearing participants. However, aging and hearing loss are associated with changes in the signatures of spectral processing at the peripheral level. In patients with cochlear hearing loss, frequency selectivity is impaired as reflected in wider pass-bands of the auditory filters, and thus in reduced frequency resolution at the auditory periphery (Glasberg and Moore, 1986, Peters and Moore, 1992). Furthermore, widening of the auditory filter pass-band has been ascribed to aging (Patterson et al., 1982), although it seems hearing loss has a much stronger impact on frequency selectivity than aging per se (Sommers and Humes, 1993, Sommers and Gehr, 1998).
Frequency specificity of neural responses in older participants has not been thoroughly investigated. This is unfortunate, given the indications for a likely link between peripheral and cortical spectral processing, and the cited deterioration of peripheral spectral processing with age. Further complicating the picture, previous research has produced diverging findings regarding the development of overall N1 amplitude with age. Some studies report larger amplitudes in older participants (Anderer et al., 1996, Amenedo and Díaz, 1999), others observe smaller N1 amplitudes (Harris et al., 2008, Schiff et al., 2008), and still others found no difference between younger and older participants (Ford et al., 1979, Czigler et al., 1992, McArthur and Bishop, 2002, Bennett et al., 2004).
To sum up, while evidence from frequency selectivity in the auditory periphery shows clear results related to aging and/or hearing loss (i.e., widened pass-band of the auditory filter), evidence from N1 cortical responses is less clear on frequency-specific adaptation and the effects of age. Furthermore, evidence that frequency selectivity at the auditory periphery is related to the frequency specificity of auditory cortical responses only exists indirectly thus far. Hence, the current EEG study aimed to investigate (i) frequency selectivity at the auditory periphery in younger and older adults, (ii) frequency specificity of N1 responses and their dependence on spectral variance and overall sound level, (iii) whether aging affects frequency specificity in tonotopically organized regions of auditory cortex, and (iv) how frequency selectivity at the auditory periphery is linked to frequency specificity of auditory cortical responses.
Section snippets
Participants and experimental sessions
Fifteen younger adults aged 20–31 (median: 24 years; 7 female) and fourteen older adults aged 49–63 (median: 56.5 years; 7 female) took part in this study. All participants were right-handed (Oldfield, 1971), did not report any history of neurological diseases, gave written informed consent prior to the experiment, and were paid 7 €/hour for their participation. Audiometry was acquired for each participant, to ensure that all participants had normal hearing (i.e., <20 dB HL) up to 3 kHz (Jerger
Auditory filter shape and predicted thresholds
Fig. 1 depicts the observed and predicted tone thresholds as a function of normalized frequency from the psychophysical experiment. Predicted thresholds reflect the average across normalized frequency levels of the fitted value at each point estimated from the auditory filter (roex) function.
Younger participants had a significantly smaller predicted threshold than older participants (t27 = 2.06, P = 0.050), reflecting overall better detection sensitivity. Auditory filter pass-band as reflected
Discussion
The current study investigated auditory frequency selectivity at the auditory periphery and auditory frequency specificity at the level of the auditory cortex in younger and older normal-hearing adults. Frequency-specific neural responses were examined under different spectral variance and intensity conditions. The main findings of the current study are as follows: (i) Frequency selectivity as measured by auditory filter shape was comparable between age groups, (ii) N1 auditory cortex responses
Conclusions
The current study investigated frequency selectivity at the auditory periphery and frequency specificity in auditory cortex in younger and older participants. Critically, auditory filter pass-bands associated with frequency selectivity at the level of the cochlea were similar between age groups, as was frequency-specific adaptation of auditory cortex responses. Yet, overall neural responses were larger in older compared to younger adults. Importantly, auditory filter pass-band and spread of
Acknowledgments
The authors were supported by the Max Planck Society (Max Planck Research Group grant to J.O.). We thank Nadine Schlichting for her help with EEG data acquisition and analyses, Nancy Grochol for her help with setting up the psychophysical experiment, and two anonymous reviewers for their helpful comments.
References (70)
- et al.
Multichannel auditory event-related brain potentials: effects of normal aging on the scalp distribution of N1, P2, N2 and P300 latencies and amplitudes
Electroencephalography and Clinical Neurophysiology
(1996) - et al.
Age-related differences in auditory event-related potentials during a cued attention task
Clinical Neurophysiology
(2004) - et al.
Frontal auditory evoked potentials and augmenting-reducing
Electroencephalography and Clinical Neurophysiology
(1985) - et al.
Age and inter-stimulus interval effects on event-related potentials to frequent and infrequent auditory stimuli
Biological Psychology
(1992) - et al.
Adaptation to stimulus statistics in the perception and neural representation of auditory space
Neuron
(2010) - et al.
Derivation of auditory filter shapes from notched-noise data
Hearing Research
(1990) - et al.
Interstimulus interval dependence of the auditory vertex response and its magnetic counterpart: implications for their neural generation
Electroencephalography and Clinical Neurophysiology
(1982) - et al.
Electrophysiologic correlates of intensity discrimination in cortical evoked potentials of younger and older adults
Hearing Research
(2007) - et al.
Age-related differences in sensitivity to small changes in frequency assessed with cortical evoked potentials
Hearing Research
(2008) - et al.
Short-term plasticity as a neural mechanism supporting memory and attentional functions
Brain Research
(2011)