The Use of Electrophysiological GIN Test in Diagnosing Temporal Resolution Problems in Patients with Sensorineural Hearing Loss

Abstract

Purpose

To evaluate the clinical validity of Acoustic Change Complex (ACC) measurements for diagnosing temporal resolution deficits in individuals with sensorineural hearing loss (SNHL).

Methods

This study included 100 participants, divided into two groups: 50 adults with normal peripheral hearing (NH), who served as the control group, and 50 adults with bilateral mild-to-moderate sensorineural hearing loss (SNHL). All participants completed both behavioral and electrophysiological Gap-in-Noise (GIN) assessments.

Results

Both behavioral and electrophysiological Gap-in-Noise (GIN) assessments demonstrated significantly elevated thresholds in the SNHL group compared with the NH group. Furthermore, electrophysiological GIN thresholds were consistently lower (better) than behavioral GIN thresholds in both groups. A strong positive correlation was observed between behavioral and electrophysiological GIN thresholds. The electrophysiological GIN threshold demonstrated a sensitivity of 84% and a specificity of 90% for identifying temporal resolution deficits.

Conclusion

These findings indicate that the electrophysiological Gap-in-Noise (GIN) test is a reliable objective measure for evaluating temporal resolution. Moreover, it may be particularly useful in clinical situations where behavioral assessment is not feasible.

1. Introduction

Much of the information conveyed by sound is encoded through temporal variations (Moore, 2003). Auditory temporal processing depends on both peripheral and central mechanisms involved in the perception of time-varying acoustic signals. Temporal resolution is defined as the minimum time required to segregate or resolve successive acoustic events (Shinn, 2003). Gap detection—the ability to perceive the shortest silent interval between sounds—is commonly used to assess the integrity of the central auditory nervous system (CANS) and its temporal processing capabilities (Phillips, 1999). Effective speech detection and comprehension are highly dependent on adequate temporal resolution. Individuals with sensorineural hearing loss (SNHL) often experience difficulty detecting brief pauses within continuous noise (Tyler et al., 1982). Because hearing-impaired listeners frequently report challenges in understanding speech, particularly in noisy environments, deficits in temporal resolution may contribute to auditory dysfunction impaired in auditory deficiency (de Oliveira Matos and Frota, 2013). However, temporal resolution abilities in individuals with hearing loss are variable, with some listeners exhibiting reduced performance and others demonstrating near-normal temporal processing (Fitzgibbons and Wightman, 1982).

The behavioral Gap-in-Noise (GIN) test evaluates a listener’s ability to identify the shortest silent interval between sounds. However, several factors including memory, cognition, motivation, hearing loss, and attention can influence behavioral performance, making it difficult to isolate the underlying auditory physiology using behavioral measures alone (Lister et al., 2007). Moreover, the behavioral GIN test cannot reliably differentiate perceptual deficits from cognitive limitations. Consequently, there is a clear need for objective methods to assess auditory temporal resolution that do not depend solely on behavioral responses (Angel, 2016).

The acoustic change complex (ACC) is a cortical auditory evoked potential elicited by changes within an ongoing acoustic stimulus. This response comprises an onset response, corresponding to the transition from silence to sound, and an ACC response, which is evoked by a subsequent change within the continuous stimulus. The onset response reflects sound detection without discrimination, whereas the ACC response reflects the neural discrimination of changes in the ongoing stimulus (Kim, 2015). These cortical potentials may be particularly valuable for estimating gap-detection thresholds in patients who have difficulty providing reliable behavioral responses.

Studies investigating cortical auditory evoked potentials (CAEPs) in gap-detection tasks are limited by small sample sizes and considerable variability in stimulus parameters and experimental protocols, which restricts the generalizability of their findings to the normal adult population. Although the P1–N1–P2 complex elicited by changes in frequency, intensity, and duration has been shown to correspond well with psychoacoustic and behavioral measures (Michalewski et al., 2005; Pratt et al., 2005; Atcherson et al., 2009; Palmer and Musiek, 2013, 2014), the studies demonstrating these relationships have generally involved limited sample sizes. Moreover, normative data for electrophysiological responses to gap-in-noise stimuli in adults are currently lacking.

This study aimed to establish normative data for an electrophysiological Gap-in-Noise (GIN) test and to evaluate whether cortical auditory evoked potentials (CAEPs) can reliably represent—and potentially substitute for—behavioral GIN detection when behavioral assessment is not feasible. Additionally, the study sought to investigate the clinical validity of using CAEPs to diagnose temporal processing deficits in individuals with sensorineural hearing loss (SNHL) by determining cutoff values that differentiate normal from abnormal temporal resolution thresholds.

2. Materials and method

This study was conducted at the Audio-Vestibular Unit, Department of Otorhinolaryngology, Tanta University, following approval from the institutional ethics committee (Approval No. 33055/04/19).

Participants: A total of 100 participants aged 18–50 years were enrolled and divided into two groups. The control group (CG) comprised 50 individuals with hearing sensitivity within normal limits and no complaints regarding speech perception in noise. Their age ranged from 19-38 years with the mean of 26.73±3.41 years. The study group (SG) included 50 individuals diagnosed with bilateral mild-to-moderate heredofamilial sensorineural hearing loss (SNHL). They had a positive family history of hearing loss in the absence of identifiable risk factors. None of the participants used hearing aids. Their age ranged from 22–48 years, with a mean age of 32.63 ± 7.93 years.

Equipment: Pure-tone hearing thresholds and speech audiometry were estimated with a two channel Grasson-Stadler clinical audiometer (GSI, version 61) in a sound treated room (Tracoustic Model No RE. 241). Middle-ear function was evaluated with an Interacoustics AT235H immittancemeter. Standardized compact disc (CD) material for the behavioral Gap-in-Noise (GIN) test was used. A CD player (Sony CD Radio Cassette-Corder CFD-S05) was connected to the clinical audiometer to operate GIN CD material. Electrophysiological Gap-in-Noise (GIN) test (Acoustic change complex, ACC) was recorded using the Smart EPs system (Intelligent Hearing Systems, IHS).

Procedure: All participants underwent a comprehensive audiological evaluation, including a detailed audiological history, otological examination, and a complete battery of audiological assessments, which comprised the following:

i. Pure tone audiometry

ii. Speech audiometry

iii. Immittance testing

iv. The behavioral Gap-in-Noise (GIN) test was developed by Musiek et al. (2005). It consists of 6-s segments of broadband noise containing zero to three silent gaps per segment. Gap durations include 2, 3, 4, 5, 6, 8, 10, 12, 15, and 20 ms, with six repetitions of each gap duration, resulting in a total of 60 gaps per test. The interstimulus interval between consecutive 6-s noise segments is 5 s. A practice section is presented at the beginning of the recording. The test is administered monaurally via headphones at 50 dB sensation level (SL) relative to the speech reception threshold (SRT). A bone oscillator presented at very low intensity (−10 dB HL) is used to facilitate the examiner’s identification of gap occurrences. Participants are instructed to count the number of gaps detected in each noise segment and report the perceived number. The GIN threshold is defined as the shortest gap duration for which the participant correctly identifies at least four out of six gap presentations. The administration time for the GIN test is approximately 10 min per ear.

v. Electrophysiological Gap-in-Noise (GIN) Test – Stimulus Parameters: the stimuli were generated using Audacity software (version 2.3.1) and consisted of broadband noise containing silent gaps of 2, 4, 8, 12, and 20 ms. These five gap durations were selected as representative of the behavioral GIN gaps to reduce overall testing time. Each stimulus had a total duration of 500 ms, with the silent gap positioned 250 ms after stimulus onset. The rise–fall time was set to 1 ms around the gap and 10 ms at stimulus onset and offset (Figure. 1). Stimulus intensity was fixed at 70 dB nHL for the control group (normal-hearing participants) and ranged from 80 to 90 dB nHL for the study group (participants with SNHL), depending on the degree of hearing loss and the individual’s most comfortable listening level. Stimuli were delivered monaurally using insert earphones.

Figure 1.

Broadband noise stimulus (500 msec duration) with 20 msec gap introduced at 250 msec. The rise-fall time is 1 msec around the gap and 10 msec at the onset of the first marker and the offset of the second marker.

Figure 2.

(a & b) Traces for onset & ACC responses in cases of the CG at 70 dBnHL with gap detection threshold at 2 ms (ACC response could be recorded down to 2 ms).

Figure 3.

Traces for onset & ACC responses in cases of the SG at 90 dBnHL which was the upper comfortable level (UCL) (a) with gap detection threshold at 12 ms (the shortest gap duration that elicited ACC response), (b) with gap detection threshold at 8 ms.

Recording Procedure: For electrophysiological recording, three disposable surface electrodes were placed according to the international 10–20 system: a low frontal electrode at Fpz served as the ground, a high frontal electrode at Fz served as the active (positive) electrode, and a reference (negative) electrode was positioned on the mastoid ipsilateral to the stimulated ear. All electrodes were connected to the preamplifier of the Smart EP system. Recordings were obtained using 512 artifact-free sweeps within a time window of −50 to 700 ms and were band-pass filtered between 1 and 30 Hz.

Recording Conditions and Response Analysis: Participants were seated comfortably in a relaxed position to minimize restlessness, with adequate head support to reduce eye and head movements. They were instructed to relax their facial muscles and avoid excessive blinking during stimulus presentation. To maintain alertness while minimizing movement-related artifacts, participants viewed silent offline videos throughout the recording session. Blinking was permitted between stimulus runs but not during stimulus presentation. Continuous electroencephalographic (EEG) monitoring was performed throughout the recording.

For response analysis, recording began with the 20-ms gap stimulus, followed sequentially by shorter gap durations (12, 8, 4, and 2 ms) until the electrophysiological threshold was reached, at which point recording was terminated. Responses were identified by the presence of an onset response (P1, N1, and P2) elicited by the onset of the leading marker, and an acoustic change response (2P1, 2N1, and 2P2) elicited by the onset of the gap. The electrophysiological GIN threshold was defined as the shortest gap duration that elicited a measurable ACC response. Peak-to-baseline amplitudes and absolute latencies were measured for each waveform. ACC component latencies were calculated relative to the onset of the acoustic change. For example, ACC P1 (2P1) latency was calculated as follows: ACC P1 latency = absolute ACC P1 (2P1) latency − stimulus change onset (250 ms) (He et al., 2012).

2.1. Statistical analysis

Statistical analyses were performed using SPSS software (version 26; IBM Corp., Chicago, IL, USA). Data normality was assessed using the Shapiro–Wilk test and visual inspection of histograms. Quantitative parametric variables were expressed as mean ± standard deviation (SD) and compared between groups using the unpaired Student’s t-test. Quantitative nonparametric variables were presented as median and interquartile range (IQR) and analyzed using the Mann–Whitney U test and Friedman test, as appropriate. Qualitative variables were expressed as frequencies and percentages (%) and compared using the chi-square test or Fisher’s exact test, when indicated. Correlations between continuous variables were assessed using the spearman correlation coefficient. Receiver operating characteristic (ROC) curve analysis was performed to evaluate diagnostic performance, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV). A two-tailed p value < 0.05 was considered statistically significant.

3. Results

This study was conducted between December 2021 and April 2023 and included 100 participants categorized into two groups. The control group (CG) consisted of 50 individuals (21 males and 29 females) aged 19–38 years, with a mean age of 26.73 ± 3.41 years. All control participants exhibited normal peripheral hearing across the frequency range of 250–8000 Hz, with a mean pure-tone average (PTA) of 11 ± 2.8 dB HL in the right ear and 11 ± 2.6 dB HL in the left ear. The study group (SG) comprised 50 participants (27 males and 23 females) aged 22–48 years, with a mean age of 32.63 ± 7.93 years. These participants were diagnosed with bilateral heredofamilial sensorineural hearing loss (SNHL) of mild-to-moderate degree, with mean PTA thresholds of 40.21 ± 4.6 dB HL in the right ear and 41.32 ± 3.8 dB HL in the left ear. None of the participants in either group used hearing aids. Within-group comparisons between right and left ears for ACC waves latencies and amplitudes were performed using the Wilcoxon signed-rank test. These comparisons were performed for both CG & SG. They revealed no statistically significant difference between right and left ears within the same group as regard ACC waves latencies and amplitudes at all gaps durations (Tables 1–4). Accordingly, data from both ears were pooled for subsequent analyses. Comparison of behavioral Gap-in-Noise (GIN) thresholds between groups using the Mann–Whitney U test demonstrated a statistically significantly higher threshold in the SG compared with the CG (Table 5). For the electrophysiological GIN assessment, gap durations that elicited acoustic change complex (ACC) responses in 100% of participants were 20, 12, 8, and 4 ms in the CG, and 20 and 12 ms in the SG. An ACC response could not be recorded at a 2-ms gap duration in the SG, whereas it was present in 90% of the CG (Table 6). Furthermore, comparison of electrophysiological GIN thresholds revealed a significantly higher median threshold in the SG (8 ms) compared with the CG (2 ms) (Table 5).

Table 1. Comparison of ACC waves latencies between right and left ears of control group.

Gaps	CG Latency	Rt Median (IQR)	Lt Median (IQR)	P Value
Wilcoxon signed ranks test, Significant P value ≤ 0.05.
20 msec	2P1	74.00 (24)	74.00 (24)	0.925
	2N1	110.00 (13)	110.00 (13)	0.994
	2P2	159.00 (13)	159 (13)	0.596
12 msec	2P1	70.00 (16)	70.00 (16)	1.000
	2N1	107.00 (14)	107.00 (14)	1.000
	2P2	158.00 (13)	158.00 (13)	1.000
8 msec	2P1	61.0000 (17)	64.0000 (17)	0.975
	2N1	98.00 (14)	98.00 (14)	0.933
	2P2	147.00 (30)	147.00 (30)	0.773
4 msec	2P1	56.00 (22)	50.00 (22)	0.884
	2N1	89.00 (16)	88.00 (16)	0.869
	2P2	132.00 (30)	132.00 (30)	0.956
2 msec	2P1	45.00 (30)	42.00 (30)	0.829
	2N1	80.00 (14)	80.00 (14)	0.879
	2P2	132.00 (52)	115.00 (53)	0.990

Table 4. Comparison of ACC waves amplitudes between right and left ears of study group.

Gaps	Amplitude	Rt SG Median (IQR)	Lt SG Median (IQR)	P Value
Wilcoxon signed ranks test, Significant P value ≤ 0.05.
20 msec	2P1	0.8000 (0.60)	0.8000 (0.27)	0.618
	2N1	2.2000 (1.20)	1.5000 (1.25)	0.520
	2P2	0.9000 (1.10)	0.7000 (0.95)	0.531
	2N1P2	3.6000 (2.30)	2.0000 (1.78)	0.504
12 msec	2P1	0.7000 (0.45)	0.6000 (0.45)	0.903
	2N1	1.4000 (1.40)	1.4000 (1.40)	0.843
	2P2	0.8500 (0.80)	0.5000 (0.80)	0.743
	2N1P2	2.5000 (2.35)	1.8000 (2.10)	0.843
8 msec	2P1	0.4000 (0.40)	0.4000 (0.40)	0.739
	2N1	1.3000 (1.75)	1.3000 (1.80)	0.612
	2P2	0.5500 (1.15)	0.5000 (1.20)	0.517
	2N1P2	1.8000 (2.88)	1.8000 (3.00)	0.612
4 msec	2P1	0.4000 (0.05)	0.4000 (0.05)	1.000
	2N1	0.800 (0.6)	0.800 (0.6)	1.000
	2P2	0.5000 (0.20)	0.5000 (0.20)	1.000
	2N1P2	1.2000 (0.75)	1.2000 (0.75)	1.000

Table 5. Comparison of behavioral and electrophysiological GIN thresholds between control group and study group.

	CG	SG	U value	P value
	Median (IQR)	Median (IQR)
Data is presented as median & IQR, CG: Control Group, SG: Study group, GIN: Gap in Noise, Mann-Whitney U test, IQR: interquartile range, *: Significant P value ≤ 0.05.
behavioral GIN threshold	4.00 (1)	10.00 (5)	0	< 0.001*
electrophysiological GIN threshold	2.00 (0)	8.00 (4)	50	< 0.001*

Table 6. Detectability of ACC response to gap in noise in control and study groups.

Groups	20 msec	12 msec	8 msec	4 msec	2 msec
Data is presented as percentage, CG: Control Group, SG: Study group.
CG	100%	100%	100%	100%	90%
SG	100%	100%	66.67%	16.67%	Zero%

Table 2. Comparison of ACC waves amplitudes between right and left ears of control group.

Gap	CG ACC Wave (amplitude)	Rt	Lt	P value
		Median (IQR)	Median (IQR)
Wilcoxon signed ranks test, Significant P value ≤ 0.05.
20 msec	2P1	0.9000 (0.54)	0.8000 (0.56)	0.459
	2N1	2.1800 (1.07)	2.1800 (0.90)	0.958
	2P2	1.4500 (0.45)	1.4000 (0.32)	0.263
	2N1P2	3.5700 (0.96)	3.5700 (0.96)	0.524
12 msec	2P1	0.6000 (0.40)	0.5400 (0.30)	0.423
	2N1	1.8000 (0.43)	1.8450 (0.60)	0.947
	2P2	0.9900 (0.55)	0.9900 (0.55)	0.446
	2N1P2	2.8900 (1.05)	2.8450 (1.03)	0.427
8 msec	2P1	0.5400 (0.30)	0.5400 (0.30)	0.994
	2N1	1.4000 (0.50)	1.5000 (0.50)	0.701
	2P2	0.9000 (0.50)	0.9000 (0.50)	0.992
	2N1P2	2.3450 (0.81)	2.5450 (0.86)	0.855
4 msec	2P1	0.5000 (0.30)	0.5000 (0.21)	0.611
	2N1	1.150 (0.4)	1.200 (0.4)	0.593
	2P2	0.8000 (0.33)	0.8000 (0.30)	0.691
	2N1P2	2.3500 (0.78)	2.5000 (0.69)	0.747
2 msec	2P1	0.5000 (0.24)	0.5000 (0.22)	0.769
	2N1	0.900 (0.40)	0.900 (0.50)	0.510
	2P2	0.700 (0.30)	0.700 (0.40)	0.602
	2N1P2	1.600 (0.50)	1.600 (0.70)	0.526

Table 3. Comparison of ACC waves latencies between right and left ears of study Group.

Gaps	Latency	Rt SG	Lt SG	P Value
		Median (IQR)	Median (IQR)
Wilcoxon signed ranks test, Significant P value ≤ 0.05.
20 msec	2P1	63.00 (20)	63.00 (20)	0.658
	2N1	107.50 (8)	107.50 (8)	0.811
	2P2	150.00 (23)	140.00 (23)	0.544
12 msec	2P1	59.00 (32)	59.00 (23)	0.994
	2N1	100.00 (14)	100.00 (8)	0.976
	2P2	141.00 (16)	138.00 (16)	0.976
8 msec	2P1	67.5000 (33.75)	69.0000 (30.00)	0.432
	2N1	91.00 (19)	98.00 (19)	0.663
	2P2	129.00 (22)	128.00 (19)	0.613
4 msec	2P1	71.00 (1)	71.00 (1)	1.000
	2N1	104.00 (3)	104.00 (3)	1.000
	2P2	122.00 (3)	122.00 (3)	1.000

The Friedman test was employed to examine variations in the latencies and amplitudes of acoustic change complex (ACC) waves across different gap durations. Regarding latency, a significant shortening of all ACC component latencies was observed with decreasing gap duration in both groups (Tables 7 and 8). In terms of amplitude, a significant reduction in ACC wave amplitudes was also noted as gap duration decreased in both groups, with more pronounced changes observed at larger gap step sizes (Tables 9 and 10). Between-group comparisons across gap durations were performed using the Mann–Whitney U test. Comparison at the 2-ms gap duration was not feasible due to the absence of ACC responses in the study group. With respect to latency, no statistically significant differences were observed between the control and study groups at any gap duration (Table 11). In contrast, amplitude analysis revealed significantly smaller amplitudes of the 2P2 and 2N1–P2 components in the study group compared with the control group across all gap durations. No statistically significant differences were found for the remaining components (2P1, 2N1, and 2P1–N1) (Table 11).

Table 7. Comparison of ACC waves latencies among all gap durations in the control group.

Wave latency	Gaps	Median (IQR)	P value of Friedman’s Test	Pairwise comparison
				Gaps	P value
Friedman’s test, *statistically significant p value ≤ 0.05.
2P1	20 msec	74.00 (53-77)	0.000*	20-12	0.058
				20-8	0.000*
	12 msec	70.00 (58-74)		20-4	0.000*
				20-2	0.000*
	8 msec	64.00 (50-67)		12-8	0.000*
				12-4	0.000*
	4 msec	50.00 (45-67)		12-2	0.000*
				8-4	0.001*
	2 msec	43.50 (32.25-59)		8-2	0.000*
				4-2	0.000*
2N1	20 msec	110.00 (98-111)	0.000*	20-12	0.732
				20-8	0.000*
	12 msec	107.00 (96-110)		20-4	0.000*
				20-2	0.000*
	8 msec	98.00 (86-100)		12-8	0.000*
				12-4	0.000*
	4 msec	88.00 (84-100)		12-2	0.000*
				8-4	0.002*
	2 msec	80.00 (74-88)		8-2	0.000*
				4-2	0.000*
2P2	20 msec	159.00 (150-163)	0.000*	20-12	0.011*
				20-8	0.000*
	12 msec	158.00 (149-161)		20-4	0.000*
				20-2	0.000*
	8 msec	147.00 (128-158)		12-8	0.008*
				12-4	0.000*
	4 msec	132.00 (125-148)		12-2	0.000*
				8-4	0.008*
	2 msec	123.50 (100-132)		8-2	0.000*
				4-2	0.001*

Table 8. Comparison of ACC waves latencies among all gap durations in the study group.

Wave latency	Gaps Median (IQR)				P value of Friedman’s Test	Pairwise comparison
						Gaps	P value
Friedman’s test, *statistically significant p value ≤ 0.05.
2P1	63.00 (60-80)	59.00 (50-80)	55.00 (50-78)	50.00 (42-71)	0.000*	20-12	0.500
						20-8	0.003*
						20-4	0.000*
						12-8	0.500
						12-4	0.003*
						8-4	0.500
2N1	106.00 (102-110)	104.00 (95-108)	100.00 (85-105)	91.00 (83-104.5)	0.000*	20-12	0.500
						20-8	0.001*
						20-4	0.000*
						12-8	0.226
						12-4	0.011*
						8-4	1.000
2P2	140.00 (139-162)	138.00 (134-150)	135.00 (125-150)	130.00 (121.5 -145)	0.000*	20-12	1.000
						20-8	0.011*
						20-4	0.000*
						12-8	0.226
						12-4	0.001*
						8-4	0.500

Table 9. Comparison of ACC waves amplitudes among all gaps durations in the control group.

Wave amplitude	Gaps	Median (IQR)	P value of Friedman’sTest	Pairwise comparison
				Gaps	P value
Friedman’s test, *statistically significant p value ≤ 0.05.
2P1	20 msec	0.80 (0.49-1.03)	0.000*	20-12	0.000*
				20-8	0.000*
	12 msec	0.60 (0.30-0.68)		20-4	0.000*
				20-2	0.000*
	8 msec	0.54 (0.30-0.60)		12-8	1.000
				12-4	0.381
	4 msec	0.50 (0.30-0.59)		12-2	0.047*
				8-4	1.000
	2 msec	0.50 (0.30-0.54)		8-2	0.990
				4-2	1.000
2N1	20 msec	2.18 (1.80-2.87)	0.000*	20-12	0.000*
				20-8	0.000*
	12 msec	1.80 (1.47-1.90)		20-4	0.000*
				20-2	0.000*
	8 msec	1.50 (1.20-1.70)		12-8	0.000*
				12-4	0.000*
	4 msec	1.20 (0.90-1.30)		12-2	0.000*
				8-4	0.008*
	2 msec	0.90 (0.70-1.20)		8-2	0.000*
				4-2	0.067
2P2	20 msec	1.40 (1.30-1.70)	0.000*	20-12	0.000*
				20-8	0.000*
	12 msec	0.99 (0.85-1.40)		20-4	0.000*
				20-2	0.000*
	8 msec	0.90 (0.70-1.20)		12-8	0.043*
				12-4	0.000*
	4 msec	0.80 (0.60-0.90)		12-2	0.000*
				8-4	0.451
	2 msec	0.70 (0.50-0.80)		8-2	0.001*
				4-2	0.504
2N1P2	20 msec	3.57 (3.31-4.27)	0.000*	20-12	0.000*
				20-8	0.000*
	12 msec	2.89 (2.30-3.29)		20-4	0.000*
				20-2	0.000*
	8 msec	2.45 (1.96-2.80)		12-8	0.000*
				12-4	0.017*
	4 msec	2.50 (2.00-2.70)		12-2	0.000*
				8-4	0.381
	2 msec	1.60 (1.28-1.90)		8-2	0.000*
				4-2	0.325

Table 10. Comparison of ACC waves amplitudes among all gap durations in the study group.

Wave latency	Gaps Median (IQR)				P value of Friedman’s Test	Pairwise comparison
	20 msec	12 msec	8 msec	4 msec		Gaps	P value
Friedman’s test, *: statistically significant p value ≤ 0.05, ACC: Acoustic Change Complex, IQR: interquartile range.
2P1	0.80(0.80-1.40)	0.70(0.50-0.95)	0.40(0.40-0.80)	0.40(0.40-0.43)	0.000*	20-12	0.500
						20-8	0.000*
						20-4	0.000*
						12-8	0.034*
						12-4	0.092
						8-4	1.000
2N1	2.20(1.50-2.70)	1.40(1.10-2.50)	1.30(0.70-2.50)	0.80(0.65-1.03)	0.000*	20-12	0.146
						20-8	0.002*
						20-4	0.000*
						12-8	0.995
						12-4	0.146
						8-4	1.000
2P2	0.90(0.50-1.60)	0.85(0.50-1.30)	0.50(0.30-1.50)	0.50(0.38-0.53)	0.000*	20-12	0.226
						20-8	1.000
						20-4	0.003*
						12-8	0.092
						12-4	0.995
						8-4	1.000
2N1P2	3.60(2.00-4.30)	2.50(1.60-3.70)	1.80(1.00-4.00)	1.20(1.10-1.58)	0.000*	20-12	0.015*
						20-8	0.026
						20-4	0.000*
						12-8	1.000
						12-4	0.846
						8-4	0.599

Table 11. Comparison of ACC waves’ latencies and amplitudes between control group and study group.

		CG	SG	P
Mann-Whitney U test, Data are presented as median (IQR). * Significant p value <0.05, CG: control group, SG: Study group, *: Significant P value ≤ 0.05.
Waves latency
20 msec	2P1	74.0(53-77)	63.0(60-80)	1.000
	2N1	110.0 (98-111)	106.0(102-110)	1.000
	2P2	159.0(150-163)	140.0(139-162)	0.137
12 msec	2P1	70.0(58-74)	59.0(50-80)	0.537
	2N1	107.00(96-110)	104.0 (95-108)	0.160
	2P2	158.0(149-161)	138.0 (134-160)	0.524
8 msec	2P1	64.0(50-67)	55.0(50-78)	0.062
	2N1	98.0(86-100)	100.0 (85-105)	1.000
	2P2	147.0(128-158)	135.0(125-155)	0.536
4 msec	2P1	50.0(45-67)	50.0(42-71)	1.000
	2N1	88.0(84-100)	91.0(83-104.5)	1.000
	2P2	132.0(125-148)	130.0(121.50-145)	0.511
Waves’ amplitudes
20 msec	2P1	0.80(0.49-1.03)	0.80(0.80-1.40)	0.538
	2N1	2.18(1.80-2.87)	2.20(1.50-2.70)	0.184
	2P2	1.40(1.30-1.70)	0.90(0.50-1.60)	0.000*
	2N1P2	3.57(3.31-4.27)	3.60(2.00-4.30)	0.019*
12 msec	2P1	0.60(0.30-0.68)	0.70(0.50-0.95)	0.521
	2N1	1.80(1.47-1.90)	1.40(1.10-2.50)	0.404
	2P2	0.99(0.85-1.40)	0.85(0.50-1.30)	0.000*
	2N1P2	2.89(2.30-3.29)	2.50(1.60-3.70)	0.028*
8 msec	2P1	0.54(0.30-0.60)	0.40(0.40-0.80)	1.000
	2N1	1.50(1.20-1.70)	1.30(0.70-2.50)	0.123
	2P2	0.90(0.70-1.20)	0.50(0.30-1.50)	0.000*
	2N1P2	2.45(1.96-2.80)	1.80(1.00-4.00)	0.002*
4msec	2P1	0.50(0.30-0.59)	0.40(0.40-0.43)	0.894
	2N1	1.20(0.90-1.30)	0.80(0.65-1.03)	0.064
	2P2	0.80(0.60-0.90)	0.50(0.38-0.53)	0.000*
	2N1P2	2.50(2.00-2.70)	1.20(1.10-1.58)	0.000*

Correlation analysis demonstrated a strong positive association between behavioral and electrophysiological GIN thresholds in both groups (Spearman’s r_s = 0.826 in the CG and r_s = 0.959 in the SG, the confidence interval for r_s is 95%). Furthermore, receiver operating characteristic (ROC) curve analysis was used to evaluate the diagnostic performance of the electrophysiological GIN threshold for assessing temporal resolution. The analysis revealed high sensitivity (84%) and specificity (90%) for detecting temporal resolution deficits. An optimal cutoff value of 4 ms was identified, with thresholds exceeding this value considered indicative of impaired temporal resolution (Table 12).

Table 12. Sensitivity and specificity of electrophysiological GIN threshold in evaluation of temporal resolution.

AUC	P	Cut off point	Sensitivity	Specificity	95%	CI
Friedman’s test, *statistically significant p value ≤ 0.05.
0.903	<0.000*	4 msec	84%	90%	0.857	0.949

4. Discussion

This study examined the limitations of relying solely on behavioral Gap-in-Noise (GIN) testing to assess auditory temporal resolution and explored the potential of electrophysiological measures as a complementary and objective alternative. Specifically, the study aimed to evaluate the clinical feasibility of using cortical auditory evoked potentials (CAEPs) to diagnose temporal processing deficits in individuals with sensorineural hearing loss (SNHL). In addition, the study investigated the anticipated ear-related performance differences based on the presumed left-hemisphere dominance for temporal auditory processing (Sulakhe et al., 2003). Contrary to this expectation, no significant interaural differences were observed, a finding consistent with previous reports in the literature (Efron et al., 1985; Baker et al., 2000).

Regarding behavioral GIN performance, the gap-detection threshold for individuals with normal hearing was 4 ms, which is in agreement with some previous studies (He et al., 1999; Samelli and Schochat, 2008) but differs from others (Phillips et al., 1997; Lister et al., 2007). These discrepancies may be attributed to variations in stimulus characteristics, particularly the type of noise employed. Broadband noise, as used in the present study, provides greater high-frequency information and reduced temporal fluctuation, which may account for the lower gap-detection thresholds compared with studies utilizing narrowband noise (Phillips et al., 1997; Lister et al., 2007). Participants with SNHL demonstrated significantly elevated gap-detection thresholds compared with normal-hearing participants, despite testing at equivalent loudness levels. The effect of hearing loss on duration discrimination remains controversial, with some studies reporting no significant impact (Fitzgibbons and Gordon-Salant, 1994; Grose et al., 2004), while others have documented degraded temporal discrimination abilities (Irwin and Purdy, 1982). Individuals with hearing loss are known to exhibit elevated speech reception thresholds in noisy environments (Bronkhorst and Plomp, 1990) and reduced ability to benefit from temporal dips in masking noise, partly due to reduced audibility of speech during silent intervals of the masker (Rawool, 2006).

Electrophysiological GIN testing using ACC responses has been previously investigated in normal-hearing adults (Michalewski et al., 2005; Pratt et al., 2005; Lister et al., 2007; He et al., 2012; Palmer and Musiek, 2013, 2014; Angel, 2016; Morse and Vander Werff, 2019). To the best of our knowledge, this study is the first to examine ACC-based GIN thresholds in individuals with SNHL. Although limited studies have evaluated ACC thresholds in SNHL using alternative stimulus paradigms (Kang et al., 2018; Vonck et al., 2021, 2022), the morphology of the P1–N1–P2 complex and the detectability of gap durations observed in the present study were generally consistent with earlier findings. Notably, the ACC threshold obtained for the normal-hearing control group was lower than that reported in some previous studies (Lister et al., 2007; Atcherson et al., 2009; He et al., 2012). These differences may be related to methodological variations, such as the minimum gap duration employed by (He et al. 2012) (5 ms) and the relatively small sample size used by (Lister et al. 2007). Furthermore, the observed increase in ACC thresholds in individuals with hearing loss is consistent with findings reported by (Vonck et al. 2022), suggesting reduced sensitivity to brief acoustic changes, which may ultimately contribute to diminished speech intelligibility (Kılıç and Kara, 2023).

In the present study, decreasing gap durations were associated with systematic reductions in both ACC latency and amplitude. Findings related to latency have varied across studies. Our results are consistent with those reported by previous investigators (Aly, 2019; Shalaby et al., 2024), whereas (He et al. 2012) described a more complex pattern, with latency decreasing up to a 20-ms gap and increasing thereafter. Other studies have reported either increased latency with shorter gap durations (Pratt et al., 2005; Lister et al., 2007) or no significant latency effects (Michalewski et al., 2005; Palmer and Musiek, 2013). With respect to amplitude, our findings align with earlier studies demonstrating reduced ACC amplitudes in individuals with hearing loss across gap durations (He et al., 2012; Vonck et al., 2019). Although our latency findings differed from those of (Vonck et al. 2022), agreement was observed regarding amplitude reductions in the hearing-impaired group. It should be noted that (Vonck et al. 2022) examined ACC responses to frequency changes rather than gap-in-noise stimuli.

Several mechanisms may account for the observed electrophysiological patterns. Hearing loss may lead to prolonged neural latencies due to broadened frequency tuning, resulting in increased overlap among neuronal populations responding to target and reference stimuli (Seki and Eggermont, 2002). Conversely, reduced peripheral input to the central auditory nervous system (CANS) has been shown to induce hyperactivity within subcortical and cortical circuits, potentially leading to shortened response latencies (Knipper et al., 2022). These opposing effects may offset one another, yielding comparable latencies in individuals with SNHL and normal hearing, as observed in the present study. This interpretation is further supported by the absence of latency differences between groups when stimuli were presented at comparable loudness levels.

Finally, a strong positive correlation was observed between behavioral and electrophysiological GIN thresholds in both groups, consistent with previous findings by . (Palmer and Musiek 2014). However, our results differ from those reported by (He et al. 2012) with respect to the diagnostic sensitivity of electrophysiological gap-detection measures. Receiver operating characteristic (ROC) analysis demonstrated high sensitivity and specificity of the electrophysiological GIN threshold for identifying temporal resolution deficits, with an optimal cutoff value of 4 ms. These findings support the clinical utility of electrophysiological GIN testing as an objective tool for assessing auditory temporal resolution, particularly in individuals who are unable to provide reliable behavioral responses.

5. Conclusion

The acoustic change complex (ACC) response may serve as an objective electrophysiological substitute of behavioral gap-detection (temporal discrimination) thresholds, with threshold measures demonstrating greater clinical reliability than latency or amplitude parameters. Sensorineural hearing loss is associated with temporal resolution deficits that cannot be adequately addressed by sound amplification alone. In this context, the electrophysiological Gap-in-Noise (GIN) test may provide valuable clinical insights for hearing-aid selection and verification, evaluation of cochlear implant performance, and monitoring of auditory rehabilitation outcomes. Ultimately, incorporation of electrophysiological GIN assessment may contribute to improved communication abilities across diverse patient populations.

Funding Statement

Not applicable.