Keywords
electrophysiology - auditory processing disorder - auditory evoked potentials - adults
Introduction
The interpretation of acoustic information is performed by the Central Auditory Nervous
System (CANS), through the occurrence of a cascade of mechanisms. For sound information
to be detected and interpreted properly, the anatomical and functional integrity of
the peripheral and central auditory pathways is necessary, so that the processing
takes place effectively.[1]
[2]
Auditory processing (AP) refers to the efficiency and effectiveness with which the
CANS uses verbal and nonverbal auditory information.[3] It is widely studied, mainly with the aim of identifying and clarifying the hearing
difficulties of children and adults in relation to sound perception, even while having
thresholds within the normal range.[4] The AP includes mechanisms underlying the abilities of sound localization and lateralization,
auditory discrimination and recognition, temporal aspects of hearing, such as temporal
integration and discrimination, temporal ordering and masking, auditory performance
in dichotic listening, and performance in degraded acoustic speech signals.[3]
[5]
The assessment of central auditory processing (CAP) consists of checking for one or
more altered auditory skills. It consists of behavioral tests capable of identifying
Central Auditory Processing Disorder (CAPD).[1]
[6] The APD refers to a deficit in the neural processing of acoustic stimuli, through
preserved cognitive and language skills. However, this disorder can be the cause or
coexist with specific alterations in language and learning, among other neurological
alterations.[3]
The American Speech-Language-Hearing Association (ASHA)[3] recommends that AP assessment be complemented by the electrophysiological assessment,
through the Auditory Evoked Potentials (AEPs). The use of middle- and, mainly, long-latency
auditory evoked potentials in AP alterations has been studied in recent years. Therefore,
it reinforces the need for further studies to establish the clinical utility of AEPs
in APD cases.
The AEPs assess the neuroelectric activity of the central auditory pathway, starting
in the auditory nerve up to the auditory cortex.[7]
The Brainstem Auditory Evoked Potential (ABR) evaluates the electrical activity of
the first neurons of the auditory system up to the brainstem. It is the most used
AEP.[7]
[8] The Middle Latency Auditory Evoked Potential (MLAEP) reflects cortical activity
related to the primary auditory skills of recognition, discrimination, and figure-ground
and non-primary skills, such as selective attention, auditory sequence, and auditory/visual
integration.[9]
[10] Long-Latency Auditory Evoked Potential (LLAEP) is composed of sequential waves P1,
N1, P2 and N2. The P1-N1-P2 complex evidences the arrival of the sound stimulus to
the auditory cortex and the beginning of cortical processing, being very important
to verify if the acoustic signal was received properly. The N2 wave is considered
a mixed component related to sound stimulus discrimination. Furthermore, the P300
cognitive component is between 300 and 500 ms post-stimulation. It reflects the activity
of cortical auditory areas related to discrimination, integration, and auditory memory
skills.[11]
[12]
[13]
In all age groups, the performance of the AP assessment is consolidated, as well as
the use of electrophysiological tests is highly recommended to complement the diagnosis.
There are several previous studies in the literature involving the AP and the AEPs,
especially regarding the P300 cognitive component.[11]
[14]
[15]
[16] Additionally, there are studies that sought to investigate the electrophysiological
activity of the central auditory pathway in cases of APD, correlating the objective
findings with the behavioral ones.[17]
[18] However, there are no in-depth studies on the accuracy of AEPs in AP alterations
in adults, without other associated pathologies.
Accuracy is considered in epidemiology a measure of high validity, being widely applied
in studies on the evaluation of diagnostic or screening tests. Its investigation makes
it possible to verify the degree to which the data measure what they should measure
or how much the results of an assessment correspond to the true state of the phenomenon
being measured.[19]
The aim of this study was to measure the accuracy of the middle- and long-latency
auditory evoked potential in adults with central auditory processing disorders.
Methods
Participants
This is a case-control study. The sample was recruited through an invitation directed
by e-mail to the academic community of State University of Londrina (UEL).
The community was informed about the objective and justification of the study, the
inclusion and exclusion criteria, the place where the exams were performed, as well
as the researchers' telephone number and email address, in case they were interested
in participating. Only individuals without otoscopic alterations, with normal hearing
thresholds, according to the criteria of the World Health Organization (WHO)[20] for the adult population, tympanometry with peak of maximum compliance around atmospheric
pressure of 0 daPa and equivalent volume of 0.3 and 1.3 ml for both groups, ipsilateral
and contralateral stapedial acoustic reflexes present for the control group[21] integrity of the auditory pathway of the brainstem verified by the ABR, with or
without complaints of difficulty in understanding speech in silence and in noise,
difficulty in auditory memory, and complaint of inattention. Individuals with an otologic
history of alteration or pathology in the middle ear, previous diagnosis of type I
or II diabetes, neurological or neurodegenerative diseases, previous auditory training
for APD intervention, and drug users were excluded.
The present study was approved by the Human Research Ethics Committee, CAAEE: 95467918.2.0000.5231.
Data were collected at an audiological clinic specializing in hearing and balance,
in the city of Londrina, Paraná, Brazil, between August 2018 and August 2019. All
participants were instructed and signed the informed consent (IC) form.
Study Design
In the first stage, the volunteers underwent a basic audiological assessment to define
audibility thresholds and conditions of the middle ear, and a complete CAP exam to
identify individuals with altered AP. In the second stage, the electrophysiological
assessment was performed, consisting of ABR, MLAEP and LLAEP. The ABR was performed
before the other potentials, to verify the integrity of the brainstem auditory pathway.
After the two steps described, the exams were evaluated by an examiner experienced
in audiology and the volunteers were divided into two groups. One group consisting
of controls (n = 30) with normal hearing thresholds and no changes in the CAP exam and a group of
cases (n = 43), composed of individuals with normal auditory thresholds and with alterations
in the AP exam.
Procedures
Immitanciometry, Audiometry and Logoaudiometry
Tympanometry was performed using the Otometrics OTOFLEX 100 (Natus Medical Inc., Middleton,
WI, USA) equipment and a probe with a 226Hz tone. The ipsilateral and contralateral
acoustic reflexes were investigated in both ears at sound frequencies of 500, 1,000,
2,000 and 4,000Hz.
In the pure tone audiometry, a two-channel MADSEN ITERA II (Natus Medical Inc., Middleton,
WI, USA) audiometer calibrated to the ANSI-69 standard and TDH39 supra-aural headphones,
was used as a stimulus transducer. Hearing thresholds were surveyed via air at frequencies
of 250, 500, 1,000, 2,000, 3,000, 4,000, 6,000, and 8,000Hz. The speech audiometry
was composed by the speech recognition threshold (SRT), which was performed live through
a list of trisyllables and the intensity in which the participant hit 50% of the presented
words was adopted as a result. To perform the percentage index of speech recognition
(PISR), 30dB were added above the tonal threshold of the average of 500, 1,000, and
2,000Hz. A list of phonetically balanced monosyllabic words was used, which were presented
to the individual by means of recording.[22] A percentage of correct answers between 88 and 100% was considered normal.
Assessment of Central Auditory Processing
The battery of tests for the CAP assessment consisted of non-verbal stimuli, except
for the dichotic digit test, presented through CDs, according to the literature.[23]
[24]
[25] The test selection procedures followed the standards suggested by the Clinical Guide.[26] The assessment consisted of the following tests: speech-in-noise (SIN) test, binaural
interaction and separation, frequency pattern test (PPS), Random Gap Detection Test
(RGDT), and Masking Level Difference (MLD). The normality standard considered for
each test was the one proposed in the literature.[23]
[24]
[25]
[27]
Electrophysiological Assessment
The electrophysiological assessment was performed with the SMART – EP (Intelligent
Hearing Systems, Miami, FL, USA) equipment and the Insert ER – 3A transducers (Natus
Medical Inc., Middleton, WI, USA), in an acoustically and electrically prepared room.
The subjects were accommodated in a reclining chair in a comfortable position. Before
starting the collection, the skin of each subject was cleaned using a Nuprep abrasive
paste (Weaver and Company, Aurora, CO, USA) in the places where the Solidor disposable
electrodes (São Paulo, SP, Brazil) were fixed. Then, they were fixed using the Tem
20 electrolytic paste (Weaver and Company, Aurora, CO, USA) to improve the electrical
conductivity.
Subjects were instructed to keep their eyes closed during the assessment to avoid
artifacts, while awake. All assessments were performed monaurally under two conditions:
assessment of the right ear and assessment of the left ear.
The assembly of the electrodes followed the standards established by the International
Electrode System (IES) 10 to 20 for its correct use. The electrode impedance remained
below 3 KΩ and the difference between the electrodes was below 2 KΩ for all exams.
MLAEP
The electrodes were arranged as follows: ground electrode on the forehead (A); the
active (positive) electrodes in the right and left coronal region (C4 and C3); the
reference electrodes (negative) on the right and left ear lobes (A2 and A1), using
the two channels of the equipment. A jumper was used to connect the inputs of the
reference electrodes of channel A and B.
In the acquisition of the MLAEP, two collections were performed containing 1,000 intermediated
stimuli and free of artifacts, and the responses were recorded twice in each condition
(C3A1, C4A1, C3A2, C4A2) to increase reliability. The components were identified and
marked by the researcher, following the baseline. The Na component was the first negative
peak identified between 16 and 30ms; Pa was the next highest positive peak observed
between 30 and 45ms; Nb was the second negative peak located between 46 and 56ms;
and Pb was the second negative peak identified between 55 and 65ms.[28]
The functional analysis of the CANS was performed by comparing the interamplitude
of Na and Pa between the ears and between the cerebral hemispheres. Each response
on one side and the other should not be less than 50% in the same individual. The
presence of electrode effect and ear effect configured a functional abnormality of
the CANS.[29]
LLAEP
The active electrodes were positioned at the vertex (Channel A - Cz) and (Channel
B - Fpz), the reference electrode at the right (A2) and left lobes (A1) and the ground
electrode at Fpz. A jumper was used to connect the inputs of the reference electrodes
of channel A and B.
The subjects were instructed to count aloud the number of rare stimuli so that the
assessment could be performed correctly. Only the tracing of the rare stimulus captured
in Cz in both ears was considered for the analysis and for presenting better morphology
in relation to Fz. The collections considered were those with artifact values lower
than 10%. The following components were identified and manually marked by the researcher:
P1, N1, P2, N2, and P300. The P1 component was identified between 54 and 73ms; N1
was the first negative peak found between 83 and 135ms; P2 was the second positive
peak located between 137 and 194ms; and N2 was the second negative peak observed between
188 and 231ms. The P300 cognitive component was the third positive peak identified
between 225 and 365ms for individuals between 17 and 30 years, and between 290 and
380ms for individuals between 30 and 50 years.[30] However, the presence of positive double deflection in P300 was verified, to correctly
identify the presence of the P3a and P3b component. According to the literature, P3a
occurs around 280ms and P3b has latency equal to or above 300ms. Thus, we consider
the third positive peak with latency equal to or greater than 300ms as cognitive P300.[30]
The parameters for acquiring the MLAEP and LLAEP are described in [Table 1].
Table 1
Parameters used to acquire the MLAEP and LLAEP[28]
[31]
Parameters
|
MLAEP
|
LLAEP
|
Stimulated ear
|
OD / OE
|
OD / OE
|
Stimulus type
|
Click
|
Nonverbal/tone burst
|
Presentation Rate
|
9.8 / sec
|
1.1 / sec
|
Number of scans
|
1,000
|
300
|
Polarity
|
Rarefied
|
Alternate
|
Intensity
|
70dB
|
75dB
|
Frequency of frequent stimulus
|
−
|
1,000Hz
|
Percentage of frequent presented stimulus
|
−
|
80%
|
Frequency of rare stimulus
|
−
|
2,000Hz
|
Percentage of rare presented stimulus
|
−
|
20%
|
High pass acquisition filter
|
20Hz
|
10Hz
|
Low pass acquisition filter
|
1,500Hz
|
300Hz
|
High pass analysis filter
|
10Hz
|
−
|
Low pass analysis filter
|
100Hz
|
−
|
Analysis time
|
70 ms
|
533 ms
|
Abbreviations: dBNA - decibels; ms - milliseconds; HZ - Hertz; LLAEP- Long latency auditory evoked
potentials; MLAEP- Middle Latency Auditory Evoked Potential; OD- Right ear; OE- Left
ear.
Statistical Analysis
The sample was calculated considering a difference in the percentage of presence of
alteration in MLAEP and LLAEP of 40% between the group with normal and altered auditory
processing. With a significance level of 5% and a power of 80%, the need for 23 individuals
per group was determined. An addition of 7 subjects per group was made to increase
the accuracy of secondary analyses.
The accuracy of the tests was verified through diagnostic tests of sensitivity, specificity,
positive predictive value, and negative predictive value. The chance of change in
CAP due to changes in electrophysiological tests was calculated by logistic regression.
Categorical variables were analyzed using the Fisher exact test. Furthermore, p-values < 0.05 were considered significant. Data were analyzed using the Statistical
Package for the Social Sciences (SPSS, IBM Corp. Armonk, NY, USA), version 20.0.
Results
Among the 147 individuals who agreed to participate, 73 could be included in the study.
The control group was composed of 63% of female subjects and the study group of 65%
of female subjects. Most participants had completed or ongoing university education,
aged 18 to 55 years, of both sexes, with normal hearing thresholds, and the demographic
characteristics between volunteers with CAP alteration and controls were matched ([Table 2]).
Table 2
Demographic characteristics of participants by study group
Variables
|
Categories
|
Controls (n = 30)
|
APD (n = 43)
|
Age (mean, SD)
|
≥ 18 years ≤ 55 years
|
29.4 (7.9)
|
29.3 (6.9)
|
Sex (%)
|
Men
|
36.7
|
34.8
|
Women
|
63.3
|
65.2
|
Race (%)
|
White
|
100
|
100
|
Non-White
|
0
|
0
|
Education (%)
|
< Highschool
|
0
|
0
|
Highschool
|
50
|
27.9
|
Undergraduate degree or Higher
|
50
|
72.1
|
Abbreviations: APD- auditory processing disorder; SD- standard deviation.
The MLAEP showed low sensitivity and high specificity to detect individuals with AP
alterations. It also presented an accuracy of 51.4% for APD cases. Individuals with
altered MLAEP were 1.78 times more likely to have APD (odds ratio, OR: 1.78, 95% confidence
interval, CI: 0.6–5.4, p > 0.42), that is, it is not a good test to aid in the diagnosis of APD ([Table 3]).
Table 3
Comparison between electrophysiological tests in the diagnosis of central auditory
processing disorders
Electrophysiological tests
|
N
|
Sensitivity
|
Specificity
|
PPV
|
NPV
|
Accuracy
|
OR (95% CI)
|
p-value
|
MLAEP
|
70
|
32%
|
80%
|
68%
|
45%
|
51%
|
1.78 (0.65.4)
|
0.42
|
LLAEP
|
73
|
56%
|
84%
|
83%
|
57%
|
67%
|
6.3 (2–19.6)
|
0.01*
|
P1-N1-P2 Complex
|
73
|
32%
|
93%
|
87%
|
49%
|
57%
|
6.76 (1.4–32.5)
|
0.01*
|
N2-P300
|
73
|
42%
|
83%
|
78%
|
50%
|
58%
|
3.6 (1.16–11.2)
|
0.039*
|
P300
|
73
|
35%
|
83%
|
75%
|
49%
|
55%
|
2.63 (0.85–8.43)
|
0.112
|
Abbreviations: APD- Auditory Processing Disorder; CI- Confidence interval; LLAEP- Long latency auditory
evoked potentials; MLAEP- Middle Latency Auditory Evoked Potential; N- number of individuals;
NPV- Negative Predictive Value; OR- Odds Ratio; P300 event related potential; PPV-
Positive Predictive Value. Notes: The LLAEP, as well as the analysis of its complexes, demonstrated high accuracy in
the detection of neurophysiological alterations, at the level of the primary auditory
cortex, in adult individuals with APD. p-value <0.05; Fisher exact test; * Significant.
The LLAEP, as well as the subcomponents P1-N1-P2 and N2-P300, demonstrated showed
low sensitivity and high specificity. The accuracy of the LLAEP encompassing all components
was 67.1%, a 15.7% higher rate than the MLAEP. Individuals with altered LLAEP were
six times more likely to have APD, which confirms that it is a good exam to complement
the diagnosis (OR = 6.3, 95% CI: 2–19.6, p < 0.01) ([Table 3]).
The P1-N1-P2 complex accuracy was 57.5%. Individuals with alterations in this complex
were six times more likely to have APD (OR = 6.76, 95% CI: 1.4–32.5, p < 0.010). The N2-P300 complex obtained an accuracy of 58.9%. Individuals with altered
N2-P300 were three times more likely to have the AP test altered (OR = 3.60; 95% CI:
1.16–11.20, p < 0.039). Finally, the cognitive component P300 did not obtain significant results,
presenting an accuracy of 55% (OR = 2.63; 95% CI: 0.85–8.43, p > 0.11) ([Table 3]).
Secondary results (supplementary material).
Discussion
The present study demonstrates that individuals with altered LLAEP are six times more
likely to have APD. Thus, the LLAEP is an efficient electrophysiological method to
be associated and used to confirm the diagnosis of APD ([Table 3]). However, due to their low sensitivity, we emphasize that electrophysiological
tests should not be used alone, as a screening or diagnostic method, in adult individuals
with APD complaints.
In our study, we observed that among all the analyses, the MLAEP was the one with
the lowest accuracy ([Table 3]), indicating that it is not a good evaluation method to aid in the diagnosis of
APD (51%, OR = 1.78). The low accuracy found in our study may justify a previous study
that did not observe a correlation between the AEPs and temporal pattern tests[32] and others that identified a weak and moderate correlation between the results of
the MLAEP and the behavioral tests of the AP, respectively.[17]
[18]
Regarding the analysis of the latency of the Pa component and the interamplitude of
Na-Pa, we did not observe any difference between the group with and without APD, considering
the leads C3A1/C3A2/C4A2/C4A1 ([Supplementary Tables S1] and [S2], supplementary material). However, we numerically observed a decrease in Na-Pa interamplitude in individuals
with APD in our data, in line with a study[33] that observed lower latencies of the Na and Pa components, as well as the Na-Pa
interamplitude for individuals with APD compared with controls. Perhaps, a larger
sample could have statistically demonstrated this difference.
The analysis of the Pa wave amplitude demonstrates the presence of the electrode effect
and/or ear effect. Presence indicates alteration and is one of the main ways of evaluating
the results of the MLAEP.[32] In our results, there was the presence of the ear effect ([Supplementary Figure S1], supplementary material), consistent with a study that concluded that the presence of the ear effect is more
compatible with cases of APD, to the detriment of the presence of the electrode effect,
which is more evident in cases of neurological injuries.[34] When evaluating three different cut-off points (50, 40, and 30%), we found that
the lower the cut-off point, the greater the percentage of altered MLAEP ([Supplementary Figure S2], supplementary material). This finding corroborates a study that demonstrated that the 30% cutoff point is
more reliable to identify neurological lesions and APD.[34]
The LLAEP, as well as the P1-N1-P2 and N2-P300 complexes analyzed separately, showed
good accuracy, in relation to MLAEP ([Table 3]). This finding justifies the use of LLAEP in studies with different populations.
Kumar et al.[35] observed higher latencies and reduced amplitudes of P1, N1, and P2 components in
individuals with type II diabetes. Oliveira et al.[36] identified a relationship between LLAEP and cognitive performance in the elderly
population. Prestes et al.[37] identified that adults who stutter have worse performance in auditory temporal processing
skills and increased latencies of the N2 and P300 components. In addition to these,
another study identified alterations in LLAEP components in children with APD.[38] Our results also demonstrated that the potential has good specificity, both for
the joint analysis of all components and for the analysis of the complexes ([Table 3]). Furthermore, by evaluating the latency values of the LLAEP components ([Supplementary Table S3], supplementary material), we identified the increase for N1, P2, and N2, in subjects diagnosed with APD.
As for the amplitudes, we observed numerically smaller amplitudes for individuals
with APD ([Supplementary Table S4], supplementary material). The increase in latency and decrease in the amplitude of the components is expected
in cases of APD, as a neurobiological alteration is observed in the SNAC, which directly
affects the auditory abilities.[39]
Given the above, we can say that the LLAEP is the best electrophysiological method
to assess CANS at the cortical level, and thus complement the assessment of the AP.
Regarding the P300 cognitive potential, our results indicated that it does not show
good accuracy for APD cases ([Table 3]). The test is often performed in clinical routine for the evaluation of AP, especially
in school-age children and adolescents who may or may not have other pathologies or
associated complaints.[18]
[40]
[41]
[42] Considering P300 captures the potentials related to the executive functions of memory
and attention, in our study, we excluded complaints phonological, reading, writing
difficulties, among others. That could influence the AP assessment. Thus, even having
presented specificity above 80%, the chance of an individual with APD having altered
P300 was not significant.
It is important to emphasize that there is a lot of variation in the results in the
literature, usually due to the protocol used, small sample sizes, and the form of
analysis used, which are mostly correlations or just descriptions of results through
cases. Additionally, it was found that studies on auditory processing are mostly performed
with school-age children and adolescents[43]
[44] with reading and writing difficulties, learning[45] phonological alterations, and associated pathologies, such as attention deficit,
dyslexia, and autism.[46]
[47] Thus, we point out the difficulty of finding studies with samples composed only
of adults with characteristic complaints of APD to compare with our results. Another
point is the fact that the studies did not use behavioral tests with a low linguistic
load as a protocol for execution or did not use the complete minimum protocol recommended
by ASHA.[3]
The present study has some limitations. First, it was performed only with adult subjects.
There are few studies involving AP and electrophysiological assessment in adults with
normal hearing thresholds in the literature. One hypothesis would be the lack of knowledge
about the AP and its abnormalities, and consequently the nonidentification of changes,
in addition to the ability to create strategies to address the complaints. This hypothesis
is consistent since in our study most the adult individuals were university students
or had already graduated and not had complaints, but had alterations in the AP exam,
which made it difficult to find a healthy individual. Additionally, it was not easy
to obtain a homogeneous sample regarding gender, to perform an analysis separately,
since there was little male adherence to the research.
Another relevant point is the type of stimulus used. We used the click stimulus in
MLAEP and the oddball stimulus for LLAEP since we wanted to eliminate the interference
of speech processing in the results. Nevertheless, it is important to point out that
speech stimuli are already used in the assessment of LLAEP and P300 because they are
more complex to be processed through the auditory pathway.[48]
Conclusion
Individuals with altered LLAEP were more likely to have APD and, therefore, the test
can be used as a complementary tool to confirm the diagnosis. The MLAEP did not prove
to be a good test to aid in the diagnosis of APD in adults.