Prevalence of mismatch negativity with tonal stimuli in normal-hearing individuals

Electrophysiological measures are one of the objective modes of assessment to check the integrity of the auditory function. The endogenous potentials like mismatch negativity (MMN) are responses which are due to internal events such as cognition or perception. In clinical practice and researches it was observed that MMN not present in all normal hearing individual. So, there is a need to study prevalence of MMN in normal population, which helps the researchers and clinicians in interpreting clinical findings. 50 participants recruited in the age range of 18-25 years. MMN was recorded with pair of stimuli. The pair was having /1000Hz/ and /1100Hz/ with /1000Hz/ as frequent stimulus and /1100Hz/ as the infrequent stimulus. Out of 50 normal hearing subjects, MMN was present only in 33 normal hearing subjects (66%). So clinician should be cautious during interpretation of clinical findings using MMN in abnormal population.


Introduction
Electrophysiological measures are one of the objective modes of assessment to check the integrity of the auditory function. Th ese measures complement the information provided by the behavioral measures. An auditory evoked potential is one of the electrophysiological measures that describe a series of electrical changes occurring in the peripheral and central nervous system, usually related to the sensory pathways. Th e auditory evoked potential is further classifi ed as endogenous and exogenous potentials. Th e exogenous potentials are primarily evoked by some external event-related dimension of the stimulus. Th e endogenous potentials such as mismatch negativity (MMN) are responses that are due to internal events such as cognition or perception. MMN was fi rst described by Näätänen et al. [1] in 1978. Th e brain is able to perceive even a minute change in acoustic environment. MMN has been gaining impetus as a measure to assess discrimination. Näätänen and Escera [2] defi ned MMN as 'an electric brain response, a negative component of the event-related potential, elicited by any discriminable change (deviant) in some repetitive aspect of auditory stimulation (standard), usually peaking around at 100-200 ms from onset'. MMN is elicited when a discriminable sound changes in intensity, duration, frequency, or phase of tone burst stimuli. It is also observed for complex change in phonemes [3]. MMN is the only objective measure of central auditory processing that may accurately correlate with behavioral perceptual measures [4]. It is an objective measure of the duration of echoic memory [5]. MMN is also an objective index of general brain degeneration, and the gross functional state of the brain can be obtained using MMN [6]. MMN can be evoked even in the absence of attention and is easy to administer [7]. Th e MMN refl ects the central code of stimulus change. Its amplitude and latency are related to the degree to which the deviant stimuli diff er from the standard stimuli, not the absolute levels of the deviant or standard stimuli. Generally, the larger the acoustic diff erence, the earlier and larger is the MMN, although there may be a ceiling eff ect in amplitude with larger diff erence [8]. Kasai et al. [9] recorded MMN for tonal and phonetic stimuli, which showed no eff ect in amplitude, latency, or laterality in relation to sex for either tonal or phonetic MMN.
In clinical practice and studies, it was observed that MMN is not present in all normal-hearing individuals. Th us, there is a need to study the prevalence of MMN in the normal population, which would help the researchers and clinicians in interpreting clinical fi ndings.

Participants
A total of 50 participants in the age range of 18-25 years were recruited. Participants were selected

Prevalence of mismatch negativity with tonal stimuli in normal-hearing individuals
Himanshu Kumar Sanju, Prawin Kumar Electrophysiological measures are one of the objective modes of assessment to check the integrity of the auditory function. The endogenous potentials like mismatch negativity (MMN) are responses which are due to internal events such as cognition or perception. In clinical practice and researches it was observed that MMN not present in all normal hearing individual. So, there is a need to study prevalence of MMN in normal population, which helps the researchers and clinicians in interpreting clinical ndings. 50 participants recruited in the age range of 18-25 years. MMN was recorded with pair of stimuli. The pair was having /1000Hz/ and /1100Hz/ with /1000Hz/ as frequent stimulus and /1100Hz/ as the infrequent stimulus. Out of 50 normal hearing subjects, MMN was present only in 33 normal hearing subjects (66%). So clinician should be cautious during interpretation of clinical ndings using MMN in abnormal population. from a private science college. Informed written consent was obtained from all participants selected for the study.

Participant selection criteria
All participants had normal hearing thresholds as defi ned by pure-tone thresholds of less than 15 dBHL at 250-8000 Hz. Furthermore, they had normal middle ear function as revealed with the middle ear analyzer. Participants with any other otological, neuromuscular, and neurological problems were excluded from the study.

Testing environment
Electrophysiological tests were carried out in a soundtreated room where the noise level was as per the guidelines in ANSI S3. 1 (1999). Th e testing rooms were well illuminated and air conditioned for the comfort of the experimenter, as well as the participant.

Instrumentation
A calibrated double-channel clinical audiometer (Orbitor-922, GN Otometrics, North America) was used for pure-tone audiometry. A calibrated GSI-Tympstar (Grason-Stadler, 7625 Golden Triangle Drive, Suite F, Eden Prairie MN 55344) Immittance meter was used for tympanometry and refl exometry. Th e Intelligent Hearing System with smart electrophysiology (EP) was used to record MMN.

Procedure
Pure-tone thresholds were obtained using a modifi ed version of Hughson and Westlake procedure across octave frequencies from 250 to 8000 Hz for air conduction and frequencies from 500 through 1000, 2000, and 4000 Hz for bone conduction. Th e middle ear analyzer (GSI-Tympstar) was used to carry out tympanometry using a probe-tone frequency of 226 Hz and to obtain ipsilateral and contralateral acoustic refl exes thresholds at 500, 1000, 2000, and 4000 Hz. . Th e wave fi le was then converted to stimulus fi le for AEPs using the software 'Stimconv' provided by the Intelligent Hearing System (6860 SW 81st Street, Miami, FL 33143, USA). MMN was recorded in a vertical montage with 'Fz' as the positive electrodes referenced to the nape of the neck. Th e ground electrode was placed on the lower forehead. A second channel was used to record the eye blink response. Th e sweeps with large eye blink artifacts were eliminated from the averaging. Stimuli were presented in the oddball paradigm, with the probability of standard and deviant stimulus as 80 and 20% at 70 dBnHL, respectively. Th e stimuli were presented in the rarefaction polarity with a repetition rate of 1.1/s. Th e response was averaged for 150 sweeps (150 infrequent stimuli+the corresponding number of frequent stimuli) from -50 to 500 ms (with reference to stimulus onset). Th e band pass fi lter was set to a frequency between 0.1 and 30 Hz, and it was amplifi ed up to 50 000 times. Stimuli were presented binaurally. Th e participants were seated comfortably to avoid muscular artifacts and were made to watch a silent movie to promote passive listening. Th e participants were instructed not to pay attention to the auditory stimuli. Th e skin surface of the target electrode sites was cleaned, and disc electrodes were placed. Th e absolute impedance was less than 5 kΩ, and interelectrode impedance was less than 2 kΩ when recording MMN. Apart from recording MMN in the conventional paradigm for each stimulus pair, late latency responses (LLRs) were also recorded for the infrequent stimulus for 150 presentations maintaining the same recording parameters as it was for MMN.

Response analysis
Conventional MMN recording was obtained in the oddball paradigm, which consisted a waveform for the frequent and the infrequent stimulus. Th is was followed by a second recording, which was the conventional LLR for the infrequent stimulus at the rate of 1.1/s, averaged for 150 sweeps. Th e LLR obtained for the infrequent stimulus was later used to analyze MMN by comparing it with the infrequent waveform of the conventional oddball paradigm. Th is paradigm was adopted to rule out any chance of error in MMN parameters due to diff erence in the LLRs elicited by the two stimuli of the oddball paradigm and also to reduce the N1 eff ect [10]. MMN was located in the diff erence wave to obtain its onset latency, peak latency, off set latency, peak amplitude, and the area under curve.
For the identifi cation of the MMN true response through visual detection, MMN should be the fi rst negative trough in the latency range of N 1-P 2 or P 2-N 2 complex of LLR of amplitude more than -0.3 V and a positive peak should follow the negative peak. If an extra negativity occurred in the P 1 area, it was ignored.

Statistical analysis
Descriptive statistics was performed to determine the mean and SD for all parameters of MMN (onset latency, off set latency, peak latency, peak amplitude, and area under the curve) using SPSS, version 17.

Results
Out of 50 participants, MMN was present only in 33 participants (66%). Hence, further statistical analysis was carried out only for 33 participants. Th e diff erent measures of MMN -that is, onset latency, off set latency, peak latency, peak amplitude, and area under curve -were extracted from the MMN waveform through visual inspection for each participant (Fig. 1, Tables 1 and 2).

Discussion
The present study showed that MMN was present only in 66% of the participants. The absence of MMN in 34% of the population may be due to poor signal-to-noise ratio and attention to stimuli. MMN responses also depend on the cognition of the participant. Heterogeneity in the cognition of normal population can also be a reason for the absence of MMN in 34% of the population. Pulvermüller and Shtyrov [11] reported MMN as a tool for studying higher cognitive processes. The present study highlights that the finding of MMN in clinical population should be interpreted cautiously, as it was seen that there is a possibility of absence of MMN even in normal-hearing individuals. Similarly, MMN was studied by Koelsch et al. [12] on professional violinists and nonmusicians. The results showed that a distinct MMN was evoked in professional violinists but MMN was absent in nonmusicians. Previous studies have also reported MMN to be robust at the group level, but identification of MMN was difficult at the individual level [13][14][15][16]. Dalebout and Fox [16] also reported that MMN identification rate was too low (29%) to allow reliability to be evaluated. Lang et al. [13] reported that various physiological factors (attention, alertness, and topographic distribution) can also effect MMN in normal individuals, which made the interpretation of results difficult among audiologists in clinical population. In a similar study conducted by Kurtzberg et al. [17], they showed that unfavorable signal-to-noise ratio of individual MMN data limits its clinical applicability. The present study showed the need of an alternative technique for the identification of MMN.

Conclusion
Th e present study showed that the prevalence of MMN is 66% only with tonal stimuli even in normal-hearing individuals. Th erefore, the clinician should be cautious during interpretation of clinical fi ndings in abnormal population.