Cochlear Synaptopathy in Full-Term Neonates: A Case-Control Study of NICU Hospitalization Impact

Abstract

Aims

Cochlear synaptopathy, or hidden hearing loss, involves synaptic loss between inner hair cells and auditory nerve fibers. While linked to aging and excessive noise exposure in adults, its impact on neonates remains unclear. This study compared cochlear synaptopathy in full-term neonates hospitalized in the neonatal intensive care unit (NICU) versus healthy controls.

Methods

Sixty full-term neonates—30 hospitalized in the neonatal intensive care unit (NICU) and 30 healthy controls—were enrolled. Inclusion criteria comprised gestational age ≥ 37 weeks, chronological age < 1-month, normal auditory brainstem response (ABR) thresholds (≤ 25 dB nHL), transient evoked otoacoustic emissions, and tympanometry. ABR wave I and V amplitudes, wave I/V ratio, growth functions, acoustic reflex thresholds (ART), and adaptation tests (rapid-rate ABR) were analyzed.

Results

No significant differences emerged between NICU-hospitalized and control neonates in ABR wave I/V amplitudes (wave I: 0.05–0.15 µV vs. 0.06–0.11 µV; wave V: 0.12–0.21 µV vs. 0.14–0.20 µV), I/V ratio (0.57–0.75 vs. 0.54–0.67), ART (80.35–73.14 dB HL vs. 78.44–71.59 dB HL), or adaptation latencies (7.60 ms vs. 7.56 ms).

Conclusion

The results suggest that NICU hospitalization does not significantly increase the risk of cochlear synaptopathy in full-term neonates, even when evaluating synaptic integrity using ABR wave I/V ratios, a biomarker for hidden hearing loss, despite normal hearing thresholds.

Introduction

Hearing loss, a prevalent sensory impairment, arises from diverse disruptions in the auditory pathway, ranging from mechanical to neural dysfunctions. It affects a significant portion of the population [1]. Traditionally, sensorineural hearing loss has been attributed mainly to that the damage to inner hair cells (IHCs) and outer hair cells (OHCs) [2]. These hair cells play a vital role in that transmitting neural signals via ribbon synapses to afferent auditory nerve fibers (ANFs). Damage to the these structures typically results in elevated hearing thresholds [3]. However, recent animal studies have challenged this conventional view, showing that noise exposure can lead to acute degeneration of synapses between IHCs and ANFs- a condition referred to as cochlear synaptopathy or hidden hearing loss. This type of damage can occur alongside transient threshold shift (TTS), even when auditory thresholds remain within normal ranges [4]. However, such synaptic damage often is not detectable by standard threshold tests.

Approximately 95% of ANFs, particularly type-I fibers, synapse exclusively with IHCs _[4]. The functionality of these ANFs is largely dependent on IHC_S [5]. Type-I ANFs can be divided into three subgroups based on their spontaneous firing rates (SR): low-SR, medium-SR, and high-SR fibers. Low-SR ANFs are crucial for effective hearing in noisy environments and are integral to acoustic reflexes, as well as processing temporal and intensity cues. Consequently, impairments in low-SR ANFs can significantly affect auditory perception, even when hearing thresholds appear normal [4]. These fibers are particularly susceptible to damage from both noise exposure and aging [6]. Cochlear synaptopathy has been implicated as a contributing factor to auditory processing disorders and is recognized as a major cause of suprathreshold difficulties in humans [2].

Human studies have provided evidence supporting the existence of cochlear synaptopathy due to noise exposure. For example, Grose et al. reported cochlear synaptopathy revealed by decreased ABR wave I amplitude and wave I/wave V amplitude ratio in young adults frequently exposed to loud music compared to the control group [7]. Similarly, Ahmadpour et al. found evidence of cochlear synaptopathy in patients with tinnitus who had normal audiograms, as evidenced by decreased wave I amplitudes in auditory brainstem response (ABR) tests [5]. However, variability in ABR wave I amplitudes among subjects can arise from factors such as tissue conductivity and individual anatomical differences [6]. A more sensitive assessment of cochlear synaptopathy may involve evaluating wave I amplitude growth, which shows significant decrements at higher stimulus levels in individuals with cochlear synaptopathy [4, 8, 9].

Moreover, since ABR wave I is primarily influenced by onset responses- and low-SR ANFs exhibit slightly diminshed onset responses- it may not serve as the most sensitive indicator for detecting low-SR ANFs loss [10]. Previous research suggests that the acoustic reflex threshold (ART) could provide a more sensitive measurefor cochlear synaptopathy [11]. Low-SR fibers are critical in the afferent pathway of the ART, making it particularly responsive to losses in low-SR fibers associated with cochlear synaptopathy [11]. Adaptation tests, such as rapid-rate ABR, are sensitive to synaptic deficits caused by cochlear synaptopathy. Wynne et al. (2013) demonstrated that individuals with ribbon synapse disorders exhibit abnormal wave V latency shifts during adaptation tests, validating this method as a biomarker for synaptic integrity [12].

Cochlear synaptopathy raises significant public health concerns, especially as many individuals are at risk due to common non-occupational noise exposure [13]. Growing concern about environmental noise and its detrimental effects on individuals of all ages has emerged. Studies indicate that exposure to noise levels as low as 65 dBA for 4 h or 76 dBA for 8 h can induce mild transient threshold shifts (TTS) in humans [14]. According to Kujawa and Liberman (2009) study, cochlear synaptpathy often occurs alongside TTS [4]. Animal studies have further demonstrated that early noise exposure can lead to hair cell damage and delayed development of the primary auditory cortex due to such exposure [15], with ribbon synapses being particularly vulnerable to noise-induced damage [16].

Despite these insights, most human studies on cochlear synaptopathy have focused primarily on adults, leaving a notable gap regarding its implications for term neonates. Importantly, recent findings suggest that noise levels in neonatal intensive care units (NICU) often exceed internationally recommended limits [15, 17–19]. This raises the hypothesis that high noise levels during NICU hospitalization may contribute to cochlear synaptopathy in neonates- a phenomenon that has yet to be thoroughly investigated. Therefore, this study aimed to investigate cochlear synaptopathy in term neonates with and without NICU hospitalization using both ABR and ART methodologies. Investigating this hidden damage is essential for early diagnosis and timely interventions that could significantly enhance quality of life.

Methods

Participants

The study included a total of 60 neonates, divided into two groups: 30 term neonates who had hospitalized in the NICU- (15 boys and 15 girls) and 30 healthy control neonates who had not been hospitalized (15 boys and 15 girls). The inclusion criteria for both groups were normal results on ABR, transient evoked otoacoustic emission (TEOAE) and high-frequency tympanometry. All participants had a gestational age of at least 37 weeks and a chronological age of less than one month. For those in the NICU-hospitalized group, we added an additional requirement of a minimum hospitalization duration of 4 days. This group primarily consisted of neonates admitted for Transient tachypnoea of the newborn (TTN).

We took care to exclude any neonates with conditions associated with hearing loss such as severe hyperbilirubinemia, sepsis, or hematologic disorders [20]. This was crucial for isolating the effects of NICU hospitalization on cochlear synaptopathy. Furthermore, our exclusion criteria encompassed low birth weight, any history of otologic or neurological disorders, noise exposure—whether from acute or prolonged exposure to harmful noise levels outside the typical NICU environment (like during medical procedures such as MRI scans or loud alarms) or documented incidents (such as accidental exposure to loud sounds during transport). We also excluded cases of head trauma—whether from falling from falls during delivery or accidents at home—and any syndromes that could confound our results.

The study protocol received ethical approval from the university’s Ethics Committee and written informed consent was obtained from the parents of all participants.

Procedure

Following the collection of demographic data, TEOAE and ABR screening were conducted in a quiet room outside the NICU for the NICU-hospitalized group and in the maternity ward for the control group. We also performed clinical TEOAE in the audiometric room to confirm cochlear health and ensure normal outer hair cell function. Before testing, we verified proper probe fit using the device’s probe-fit function and ensured a tight acoustic seal. The results were considered normal if the signal-to-noise ratio (SNR) was ≥ 6 dB and reproducibility was ≥ 80%, confirming the results with two repeatable recordings.Tympanometry, ART testing (using a 1000 Hz probe tone) and ABR were also performed in the audiometric room. Normal middle ear function was indicated by a single-peaked admittance tympanogram in infants (21). For ABR testing, electrodes were placed at frontal (F_z) and temporal (T₁ and T₂₎ regions using a two-channel ICS EP200 system (Denmark). The ABR protocol involved a 100–3000 Hz band-pass filter for click stimuli, with rarefaction polarity, a presentation rate of 21.1 Hz, 2000 sweeps, and a time window of 15 ms. Stimuli were presented at three intensity levels: 60, 70, and 80 dB nHL. The 21.1 Hz rate aligns with neonatal ABR protocols to ensure reliable waveform resolution, as recommended previously [21]. An adaptation test was also conducted using 4000 sweeps at an intensity level of 80 dB nHL to determine wave V latency of the ABR. The adaptation test utilized a 74.5 Hz stimulation rate, selected to optimize synaptic deficit detection while avoiding 50 Hz electrical noise interference.

ART and tympanometry measurements utilized the Madsen Zodiac system (Denmark). The ipsilateral ART was assessed using a 1000 Hz probe tone at frequencies of 0.5, 2, and 4 kHz with intensity levels ranging from 60 to 90 dBHL in 1 dB increments [22]. The threshold was determined by gradually increasing the intensity in 5 dB steps from 60 to 90 dB HL. Testing was stopped at a high level (i.e., > 90 dB HL) because the infant may be disturbed, resulting in behavioral responses such as startle and blink eyes, and may produce a sudden change in admittance in synchrony [23]. When responses were noted, intensity levels were decreased by 2 dB steps; if no response occurred, intensity was increased by 1 dB steps until responses were recorded twice. The threshold in the ART was considered as the point where the response was repeated twice. An active 1 kHz stimulus was not used because of the possible interaction with the probe tone 1000 Hz [23].

Additionally, ambient noise levels in the NICU were monitored using the “Sound Meter” mobile application, a validated tool for environmental noise measurement in clinical research [24]. Noise measurements were performed at a distance of 10 cm from the neonate’s head [18]. Measurements were taken over two days at five intervals daily for 15 min each time while avoiding any interference from crying or sneezing [18]. The recorded noise was analyzed using the Audacity 2.3.3 software. The results showed that noise frequencies ranging from 100 to 3000 Hz, with levels between 52.3 and 67.8 dBA and an average level of 60.4 dBA— exceeding recommended limits.

Data Analysis

The Shapiro–Wilk test confirmed that all variables exhibited normal distribution. To evaluate the effects of group and stimulus intensity level on various ABR measures—including wave I amplitude, wave V amplitude, the ratio of wave I/wave V amplitude and their growth—a two-way analysis of variance (ANOVA) was conducted with group (hospitalized vs. control) as the between-subject factor and stimulus intensity level (60 vs. 70 vs. 80 dB HL) as the within-subject factor. An independent t-test was done to compare ART results and wave V absolute latency with rapid stimulus rate between the two groups. The significance level was set at 0.05 for all statistical analyses.

Results

Thirty neonates hospitalized in the NICU (15 boys, and 15 girls; mean ± SD gestational age of 38.05 ± 0.85 weeks; mean ± SD chronological age of 14.87 ± 5.94 days; and mean ± SD weight of 2.99 ± 0.49 kg) and 30 control neonates (15 boys, and 15 girls; mean ± SD gestational age of 38.73 ± 0.88 weeks; mean ± SD chronological age of 11.15 ± 3.55 Days; and mean ± SD weight of 3.19 ± 0.49 kg) were recruited. The NICU-hospitalized group had an average hospitalization duration of 9.11 ± 4.97 days.

The descriptive statistics for ABR measures including wave I amplitude, wave V amplitude, and wave I/wave V amplitude ratio are presented in Table 1. Analysis showed no significant main effect of group or interaction effect of the group by stimulus intensity level regarding growth of wave I and wave V amplitude or wave I/wave V amplitude ratio across all presentation levels, revealing that there was no significant difference between the NICU-hospitalized group and the control group in the wave I amplitude, wave V amplitude or wave I/wave V amplitude ratio at any presentation level (Table 2; Fig. 1). However, significant main effect of stimulus intensity level was found for both wave I amplitude (F = 20.51, P = 0.000) and wave V amplitude (F = 11.32, P = 0.001), indicating that by increasing the intensity of stimulus, the amplitudes of both wave I and wave V similarly increased in both the hospitalized group and the control group. To account for individual differences, the ratio of the amplitude of wave I to wave V was determined for each ABR. Neither the main effect of the group level nor the interaction effect of the group by presentation level was significant for the wave I/wave V ratio. However, the main effect of stimulus intensity level was significant for this measure (F = 3.14, P = 0.047) (Table 2).

Table 1.

Descriptive data (mean ± SD) of different measures of auditory brainstem response (ABR) test in the control and hospitalized groups

Variable	Stimulus intensity level (dBnHL)	Control group (n = 30)	NICU-hospitalized group (n = 30)
Wave I amplitude (µV)	60	0.06 ± 0.04	0.05 ± 0.04
	70	0.11 ± 0.11	0.11 ± 0.08
	80	0.11 ± 0.08	0.15 ± 0.09
Wave V amplitude (µV)	60	0.14 ± 0.08	0.12 ± 0.07
	70	0.18 ± 0.12	0.16 ± 0.09
	80	0.20 ± 0.13	0.21 ± 0.13
Wave I/Wave V amplitude ratio	60	0.54 ± 0.49	0.57 ± 0.50
	70	0.73 ± 0.52	0.82 ± 0.64
	80	0.67 ± 0.66	0.75 ± 0.43

Table 2.

Summary of analysis of variance (ANOVA) for different measures of auditory brainstem response (ABR) test in the control and NICU-hospitalized groups: F ratios, P values and effect sizes by variable

Variable	Main effect						Interaction effect
	Group			Stimulus intensity level			Group × Stimulus intensity level
	F	P		F	P		F	P
Wave I amplitude (µV)	0.27	0.61	0.005	19.88	0.000	0.27	1.42	0.25	0.02
Wave V amplitude (µV)	0.13	0.72	0.002	21.97	0.000	0.28	1.43	0.24	0.02
Wave I/Wave V amplitude ratio	0.47	0.50	0.008	3.14	0.047	0.05	0.06	0.94	0.001

Fig. 1.

ABR wave I/ V amplitude ratio as the function of stimulus intensity level (60, 70, and 80 dBnHl) in NICU-hospitalized group and control group

Table 3 shows the descriptive data of the adaptation test (i.e. ABR wave V absolute latency with rapid rate (74.5 Hz) stimulus) and ART (at 0.5, 2, and 4 kHz, as well as averaged ART (i.e. mean of ART at 2 and 4 kHz)). The results showed that there was no significant difference between the NICU-hospitalized group and the control group in the adaptation test (P = 0.74). In addition, no significant difference was found between the two groups regarding the ART at 0.5 kHz (P = 0.18), 2 kHz (P = 0.09), 4 kHz (P = 0.46), and averaged ART (P = 0.30).

Table 3.

Comparison of the results of adaptation test (absolute wave V latency) acoustic reflex threshold (ART) at 0.5, 2 and 4 khz and averaged 2 and 4 khz in the control and hospitalized groups

Variable		Control group (n = 30)	NICU-hospitalized group (n = 30)	t	P
Adaptation test (ms)	Absolute wave V latency	7.56 ± 0.34	7.60 ± 0.55	0.33	0.74
ART (dB HL)	0.5 kHz	78.44 ± 7.38	80.35 ± 6.15	1.34	0.18
	2 kHz	71.23 ± 5.82	73.66 ± 5.93	1.72	0.09
	4 kHz	71.59 ± 6.31	73.14 ± 6.20	0.74	0.46
	Averaged ATR at 2 and 4 kHz	71.45 ± 5.44	72.65 ± 4.60	1.03	0.30

Discussion

Our study provides insights into cochlear synaptopathy in term neonates in the NICU. We didn’t find any significant differences in ABR wave amplitudes, wave I/V ratios, or acoustic reflex thresholds between NICU-hospitalized neonates and healthy controls. This suggests that NICU hospitalization, under the conditions we studied, doesn’t seem to increase the risk of cochlear synaptopathy in full-term neonates.

Cochlear synaptopathy, characterized by the loss of ribbon synapses connecting IHCs and ANFs, is often associated with excessive noise exposure. Previous studies have shown that reductions in ABR wave I amplitude can serve as a reliable proxy for cochlear synaptopathy, particularly in older populations. For example, Song et al. (2021) reported decreased ABR wave I amplitudes alongside reduced ribbon synapses in guinea pigs exposed to moderate-to-high intensity noise [25]. Similarly, Grose et al. (2017) highlighted discrepancies between ABR measures and actual auditory perception among young adults frequently exposed to loud music, emphasizing that synaptic integrity is crucial for effective auditory processing even when hearing thresholds appear normal [7].

A significant aspect of our research is its focus on term neonates, an area that has received insufficient attention in studies related to cochlear synaptopathy. Much of the existing literature has concentrated on adults, leaving a significant gap in understanding how early noise exposure in environments such as NICU might affect auditory development. For example, Kujawa and Liberman (2009) demonstrated that cochlear synaptopathy can arise from noise exposure without accompanying changes in audiometric thresholds [3]. Our study extends this understanding to neonates, suggesting they too may be at risk for similar hidden auditory deficits.

Interestingly, our findings suggest that NICU hospitalization may not significantly contribute to cochlear synaptopathy, at least as measured by ABR and ART. Specifically, we found no significant differences in wave I and wave V amplitudes or the wave I/V ratio across various stimulus intensity levels. This absence of measurable change could imply that short-term exposure to NICU environments does not lead to detectable cochlear damage.

One possible explanation for our findings is that the mean duration of hospitalization (9.11 days) may not have been long enough to induce measurable cochlear synaptopathy detectable through ABR and ART assessments. This aligns with previous findings indicating that prolonged exposure to elevated noise levels can lead to gradual neural damage [3]. Additionally, While the average noise levels in the NICU (60.4 dBA) were higher than recommended, they’re still much lower than the noise levels known to cause cochlear synaptopathy in experimental models [4]. This likely explains why we didn’t see significant synaptic deficits in our group– the NICU noise intensity might not be high enough to cause that kind of damage. The lack of significant differences between the NICU-hospitalized group and control group in adaptation test results further supports our findings. If cochlear synaptopathy were present, we would expect to see delayed wave V latencies during rapid stimulus presentations [12]; however, our results showed no such delays between groups. This suggests that both NICU-hospitalized and control neonates exhibited similar neural adaptation responses.

Despite these findings, it is essential to recognize the potential for hidden auditory deficits even when traditional measures do not indicate damage. Future research should prioritize longitudinal evaluations of auditory function following NICU hospitalization to better understand how early noise exposure impacts auditory development over time. Additionally, exploring strategies to minimize noise exposure in the NICU could greatly benefit infants hospitalized in such environments.

We recommend that future studies incorporate 500 Hz tone burst stimuli and other parameters of ABR and ART, alongside tests like Envelope Following Response and Electro Cochleo Graphy for a comprehensive evaluation of cochlear synaptopathy.

Moreover, while our current study focused solely on term neonates, preterm neonates- who exhibit impaired autonomic- and heightened susceptibility to environmental stressors [15, 26]- represent a critical population for further investigation. Future studies should compare cochlear synaptopathy in preterm neonates hospitalized in the NICU compared with term controls to elucidate the compounded effects of prematurity and noise exposure.

Conclusion

In summary, while our study did not find significant differences in conventional measures of auditory function between NICU-hospitalized neonates and control neonates, it highlights an important area for ongoing research into the effects of cochlear synaptopathy on this vulnerable population. Understanding these issues is crucial for developing effective screening and intervention strategies that could enhance auditory health and overall quality of life for affected infants.

Authors Contribution

Zeinab Hasanpour: Conceptualization, investigation, writing original draft, writing-review and editing; Malihah Mazaheryazdi: Conceptualization, investigation, methodology, writing original draft, writing-review and editing; Hassan Haddadzade Niri: Methodology, validation, writing-review and editing; Mohamad Maarefvand: Conceptualization, formal analysis; Hajar Mehdizadeh: Formal analysis, writing original draft, writing-review and editing.

Funding

This study was supported by the School of Rehabilitation Sciences, Iran University of Medical Sciences.

Data availability

Not applicable.

Declarations

Competing Interests

The authors have no relevant financial or non-financial interests to disclose.

Ethics Approval

The study protocol received ethical approval from the Ethics Committee of the Iran University of Medical Sciences [IR.IUMS.REC.1401.471].

Consent

Written informed consent was obtained from the parents of all participants.