. 2026 Jul 2;18:54. doi: 10.1186/s11689-026-09718-9

Exploring the relationship between auditory processing characteristics and cognition in preschool children with developmental language disorder based on fNIRS

Wenmin Wang ^1,^2,^#, Panting Liu ^1,^4,^#, Bingzi Hao ^1,², Jia Zhou ^1,⁴, Bilan Chen ³, Yuying Sun ², Xinyue Yu ³, Tingyu Li ³, Yaqin Xu ¹, Liping Meng ¹, Lei Zhang ¹, Jun Qian ¹, Xia Chi ^1,^4,^✉, Qin Hong ^1,^4,^✉

PMCID: PMC13599336 PMID: 42387374

Abstract

Background

Auditory processing (AP) is a fundamental function in speech signal processing. The ability to process speech signals in noisy environments is commonly used to assess AP capabilities. Research on the AP characteristics in Chinese children remains scarce, and the relationship between AP and language/cognitive development has not been explored. This study investigated the AP performance of children with developmental language disorder (DLD) in daily life and under the standardized speech-in-noise (SIN) comprehension paradigm. Using functional Near-Infrared Spectroscopy (fNIRS) to monitor brain activation in real-time during noisy environments, the study analyzed the functional modes of targeted brain regions in Chinese DLD children under noisy environments, while also exploring the relationships between AP behaviors, brain activation patterns, and cognition.

Methods

This cross-sectional study enrolled 34 Chinese preschool children with DLD and 43 age- and gender-matched typically developing (TD) children. Parents completed basic demographic information, the Preschool Auditory Processing Assessment Scale (PAPAS), and the Behavior Rating Scale of Executive Function-Preschool Version (BRIEF-P). A self-developed SIN comprehension paradigm adapted for Chinese preschool children with DLD was utilized, and fNIRS was employed to record brain activation patterns while the children performed the SIN comprehension task.

Results

1) AP Behavior: The DLD group scored higher than TD group on all dimensions of the PAPAS (P < 0.05); main effects of the group (F = 54.303, P < 0.001) and listening condition (F = 11.83, P = 0.001) were significant for the accuracy (ACC). 2) Brain activation characteristics during the SIN comprehension task: The DLD group showed significantly lower activation than the TD group in the left dorsolateral prefrontal cortex (DLPFC), left Broca’s area, bilateral superior temporal gyrus and middle temporal gyrus (STG and MTG), and right Wernicke’s area (P-corrected < 0.05). 3) The AP total scores, auditory decoding scores, and hyperactivity impulse dimension scores of both groups of children were negatively correlated with the scores on each dimension of the DREAM-C and the WPPSI-IV/WISC-IV (P-corrected < 0.05); the scores on each PAPAS dimension were positively correlated with the scores on each dimension of BRIEF-P (P-corrected < 0.05). 4) The activation levels in the left DLPFC, left Broca’s area, right Wernicke’s area, and bilateral STG and MTG in both groups of children were positively correlated with all dimensions of the DREAM-C (P-corrected < 0.05). The activation levels in the left DLPFC, left Broca’s area, right Wernicke’s area, and right STG and MTG were positively correlated with the full-scale intelligence quotient and verbal comprehension index scores (P-corrected < 0.05).

Conclusions

Children with DLD may exhibit AP abnormalities as early as the preschool years, as reflected by poorer performance on AP tasks and atypical cortical activation patterns.

Keywords: Preschool children, Developmental language disorder, Auditory processing, Functional near-infrared spectroscopy, Cognitive-behavior

Introduction

Developmental Language Disorder (DLD) [1] refers to a condition occurring during human ontogeny, where an individual exhibits deficits in language acquisition, comprehension, production, or use (spoken or sign language) after excluding mental disorders or organic hearing impairments. In clinical terms, it often manifests as a child’s language abilities lagging significantly behind those of typically developing (TD) peers, resulting in markedly restricted communication skills. Data from the International Classification of Diseases, 11th Revision (ICD-11) [2], indicate a prevalence rate of DLD ranging from approximately 6% to 15%. In China, the prevalence among children is 8.5% [3]. Due to its high incidence and profound impact [4–6], academic interest in DLD continues to deepen.

Auditory processing (AP), also known as central auditory processing (CAP), refers to neurobiological activity within the central auditory nervous system (CANS) that involves the perception and processing of auditory information, ultimately generating electrophysiological auditory potentials [7]. AP plays a crucial role in the perception, comprehension and generation of auditory information. Models of Auditory-Cognitive Processing [8] set forth that when language signals are input, individuals first undergo AP—including feature extraction, selective attention, and object formation of auditory information—then match the extracted information with linguistic representations stored in memory, ultimately achieving language comprehension. Thus, well-developed AP skills form the foundation for language comprehension.

Previous studies have reported that children with DLD exhibit AP abnormalities, validated through behavioral [9, 10] and electrophysiological [11, 12] assessments. Disruptions in AP can lead to deficits in speech perception, sound localization, auditory discrimination, and the recognition of temporal features of sounds [13]. In real-life settings, children frequently need to be able to recognize speech in complex auditory environments with background noise (e.g., classrooms, playgrounds) or multi-speaker interference [14]. This is particularly crucial for language comprehension [15], classroom participation, and social interaction. Research indicates that children with DLD exhibit poorer speech perception abilities in noisy conditions [16–20], and this deficit is associated with impairments in AP [15]. Ellis-Weismer et al. [21] suggested that impaired speech perception may be one of the primary deficits in children with DLD, potentially explaining conditions such as poor phonological development, impaired vocabulary production, and deficient word comprehension. Therefore, investigating speech perception in noisy environments among children with DLD holds significant clinical importance.

As a typical tonal language, Chinese uses not only segmental information but also variations in fundamental frequency (F0) height and dynamic contours to distinguish lexical meaning. Previous studies have shown that children with DLD experience difficulties in perceiving tones [22], and the processing of Chinese tones requires individuals to finely perceive and neurologically process fluctuations in F0 [23, 24]. Therefore, for native Chinese-speaking children with DLD, the tonal characteristics of the language may further increase their auditory-linguistic processing load, presenting them with even greater challenges in complex auditory environments. However, current evidence regarding AP abilities in children with DLD primarily comes from Indo-European language contexts [25]. Given the significant differences between tonal and non-tonal languages in terms of acoustic characteristics, speech processing methods, and neural mechanisms, the generalizability of these findings to the Chinese language context is limited. Currently, there is a relative lack of research on AP in preschool-aged Chinese children with DLD in complex environments. In China, research on auditory function in DLD has been limited to peripheral hearing assessments [26], with no reports on CAP abilities. Therefore, evaluating and intervening in the auditory function of Chinese children with DLD within complex auditory environments is of particular importance.

Both the American Academy of Audiology (AAA) [27] and American Speech-Language-Hearing Association (ASHA) recommend using a comprehensive testing method to evaluate AP abilities. For AP assessments of preschool children under 7 years of age, they recommend using screening methods to identify children at high risk for AP disorders. Therefore, this study will assess the AP abilities of preschool-aged Chinese children with DLD based on the following two behavioral indicators.

The first one involves parents’ daily behavioral observations. Our research team previously developed the first Preschool Auditory Processing Assessment Scale (PAPAS) suitable for Chinese-speaking preschool children. This scale demonstrates excellent reliability and validity [28] and has established regional norms for preschoolers in Jiangsu Province, China. The second scenario is the standardized understanding of Speech-in-Noise (SIN). SIN recognition, via reducing the external redundancy of auditory stimuli in noisy environments, increases the difficulty of processing input information. It is considered a special case of auditory scene analysis—the ability to parse complex soundscapes into coherent objects or sources, involving the separation of target speech from irrelevant sounds [29]. Therefore, this study developed a SIN comprehension paradigm suitable for Chinese preschool-aged children. This paradigm employs a block design comprising 4 SIN perception blocks at 0dB signal-to-noise ratio (SNR) and 5 speech understanding blocks in quiet environments. The stimuli consist of disyllabic words, with four-person noise used as informational masking. Upon hearing a disyllabic word, children must respond by pressing a button to indicate whether the word matches the picture displayed on the screen. The primary behavioral outcome is the measure of button-press accuracy (ACC) across different listening conditions. The secondary outcome measures include reaction time (RT) and discriminability index (d-prime, d’) for correct trials under varying listening conditions.

Furthermore, we aimed to analyze the underlying brain functional characteristics involved in AP behaviors among children with DLD to understand the brain activation patterns observed during the SIN comprehension paradigm.

Functional Near-Infrared Spectroscopy (fNIRS) is an emerging brain non-invasive imaging technique that measures and analyzes optical signals transmitted between the skull and the cerebral cortex, enabling the detection of concentration changes in oxygenated hemoglobin (HbO, O₂Hb) and deoxy-hemoglobin (HbR) within the cerebral cortex at different time points, thereby reflecting the activity state of corresponding cortical regions [30]. Signals obtained through fNIRS exhibit minimal sensitivity to head movement, enabling data acquisition during motion and in open spaces [31]. Compared to fMRI, fNIRS offers moderate temporal and spatial resolution, relatively low cost, radiation-free operation, and reduced motion sensitivity [31], demonstrating significant advantages in studying the biological mechanisms underlying children’s cognitive development and neurodevelopmental disorders [32]. Therefore, this study applied fNIRS technology to monitor real-time brain activity patterns in Chinese preschool children with DLD during a hearing-processing-specific paradigm—the SIN comprehension task.

Research on speech processing emphasizes the close connection between auditory perceptual input and motor output [33, 34]. Early speech processing models proposed that perceptual processing occurs in the superior temporal gyrus (STG) (Wernicke’s area), while motor generation takes place in the inferior frontal gyrus (Broca’s area) [35, 36]. With recent advances in neuroimaging and neurophysiological research, traditional speech processing models no longer adequately explain the multidimensional complexity of language comprehension and generation. Contemporary theories generally consider that language processing relies on extensive networks distributed across the temporal lobe, frontal lobe and other brain regions [37–41], achieving tight structural and functional connectivity through connections such as the dorsal and ventral pathways [42].

Speech, as an auditory signal, is first transmitted along the ascending auditory pathway to the primary auditory cortex and the surrounding auditory association cortex, and then projects laterally to the STG [43]. The STG, comprising the non-primary auditory cortex, is a critical region for speech processing [44] that is essential for extracting meaningful linguistic features from speech input [45]. Meanwhile, the adjacent middle temporal gyrus (MTG) plays a vital role in semantic extraction and integration [46]. Furthermore, the role of the frontal lobe in contemporary language network models has become increasingly prominent. The dorsolateral prefrontal cortex (DLPFC), located within the middle frontal gyrus [47], was not traditionally considered a core language area in conventional language network models. However, increasing evidence underscores that the DLPFC plays a role in higher-level control, monitoring, and integration during language processing, serving as a key region for establishing functional connectivity between the language network and other functional networks, including inter- and subcortical circuits [48]. Based on the above theories, this study selected the bilateral Broca’s area, bilateral Wernicke’s area, bilateral STG and MTG, and bilateral DLPFC as the regions of interest.

Additionally, we aimed to investigate whether the AP abilities of Chinese preschool-aged children with DLD in noisy environments are influenced by cognitive factors.

Children with DLD exhibit significant deficits in core language skills, receptive language, and expressive language, while language comprehension and expression require deep cognitive engagement [49, 50]. According to the Ease of Language Understanding (ELU) model [51], when auditory input degrades under conditions like noise, listeners must mobilize additional cognitive and linguistic knowledge [52] to match and correct incomplete or ambiguous auditory information. This compensatory process relies heavily on the executive function (EF) system, particularly involving abilities such as working memory [51]. Children with DLD not only experience language difficulties but also often exhibit deficits in EF and cognitive abilities [53, 54]. This may impair their capacity to effectively mobilize cognitive resources to compensate for degraded auditory signals in noisy environments, thereby affecting their speech comprehension performance in noisy environments. Therefore, this study will employ a combined subjective and objective approach to investigate the relationship between AP and cognition in Chinese preschool-aged children with DLD.

We anticipated the following: 1) Children with DLD would score higher on the PAPAS scale and exhibit a greater risk of AP abnormalities. 2) They would demonstrate poorer behavioral performance on the SIN comprehension task, with worse ACC, RT and d' scores compared to TD children. 3) The DLD group would perform worse under noisy conditions than in quiet settings, and their performance in noisy environments would be worse than that of TD children. 4) Children with DLD would exhibit reduced activation in the frontal-temporal brain regions during the SIN comprehension task. 5) Children with DLD would demonstrate poorer performance in language abilities, intellectual levels, and EF compared to TD children. As deficits in cognition and EF among children with DLD correlate with PAPAS scores and the degree of regions of interest activation, this may further exacerbate their difficulties in speech recognition within noisy environments.

Methods

Participants

This study adheres to the Declaration of Helsinki and was approved by the Medical Ethics Committee of Nanjing Maternity and Child Health Hospital (Nanjing, China; Ethics Approval No.2022KY-052). Written informed consent was obtained from the parents or legal guardians of all participants.

The case group included a total of 53 children (14 girls) diagnosed with DLD at the Language Clinic or Psychological Behavior Clinic of the Department of Child Healthcare at Nanjing Women and Children’s Healthcare Hospital. The inclusion criteria were: children who met the diagnostic criteria for DLD in the ICD-11; age: 5 to 6 years and 11 months; the Wechsler Preschool and Primary Scale of Intelligence-Fourth Edition (WPPSI-IV) / The Wechsler Intelligence Scale for Children-The Fourth Vision (WISC-Ⅳ) full-scale intelligence quotient (FSIQ) score ≥ 70; right-handed; those who could understand the task. The exclusion criteria were: as diagnosed by a clinician, children suffering from other neurodevelopmental or psychiatric disorders (e.g., Attention-Deficit/Hyperactivity Disorder, autism spectrum disorders, mental retardation, etc.), organic neurological disorders (epilepsy, sequelae of craniocerebral trauma, etc.), hearing impairments (binaural average pure tone air-conducted hearing thresholds of 0.5–4.0 kHz > 20 dBHL), and visual impairments.

Inclusion criteria for the TD group: fifty-nine (17 girls) TD children matched to the DLD group for gender and age were recruited from kindergarten as a control group, all of whom were right-handed, with WPPSI-IV/WISC-IV FSIQ scores of ≥ 70, able to comprehend the task. The exclusion criteria are the same as detailed above.

During the experiment, the quality of the collected signals was poor due to the low cooperation of some children with excessive head movement, etc. Finally, 34 cases (11 girls) were included in the DLD group, with an average age of 71.0 (65.5, 75.3) months, and 43 cases (14 girls) in the TD group, with an average age of 69.0 (66.0, 73.0) months. All of the included children were able to understand the task and cooperate.

Materials and procedure

Child participants in both groups completed pure-tone audiometry, the Diagnostic Receptive and Expressive Assessment of Mandarin-Comprehensive (DREAM-C) [55], fNIRS testing, and WPPSI-IV/WISC-IV assessments. During testing, parents were guided to complete demographic questionnaires while waiting for their child’s initial assessment and filled out paper-based PAPAS and Behavior Rating Scale of Executive Function-Preschool Version (BRIEF-P) forms based on their child’s behavioral performance. The entire assessment process took 3 h, with most participants (n = 68) completing it during a single weekend visit. For some less cooperative children (7 DLD children and 2 TD children), any remaining assessments were scheduled for a second weekend visit. Children received approximately 10 min of rest time after completing each assessment task.

Basic demographic information

The basic demographic information was collected via parent-completed questionnaires and included child participants’ gender, age, birth weight, gestational age, parents’ age at child’s birth, neonatal hyperbilirubinemia, parental education, household socioeconomic status (SES), caregiver status, history of otitis media, and co-occurring neurodevelopmental disorders. Daily screen time, daily parent-child interactive game time, and daily parent-child shared reading time were also recorded.

Pure Tone Audiometry (PTA)

The PTA assessment was conducted in a soundproof room that met the GB/T16403-1996 standards, with a background noise level below 30 dBA. A Conera diagnostic audiometer (Danish Erthomi) and plug-in earphones were used to test hearing thresholds at 500, 1000, 2000, and 4000 Hz. Children with ≥ 20 dB HL thresholds were excluded.

DREAM-C

The DREAM-C (Mandarin version), as a diagnostic language development assessment tool for children aged 2 years and 6 months to 7 years and 11 months, is able to test children’s language in a comprehensive manner and includes five parts: total language, receptive language, expressive language, semantics, and syntax development. Scoring was conducted using standardized scores based on the mainland Chinese norm [55]. All tests were administered by graduate students from our department who had received specialized training from speech therapists and passed the assessment. In Nanjing, China, Mandarin Chinese is widely spoken. Likewise, for all child participants in this study, Mandarin was the primary language used by their families when interacting with them, serving as the main language for their daily communication.

Subjective AP assessment: PAPAS

The PAPAS scale assesses subjective AP abilities and includes the AP total score and five subscales: auditory decoding, auditory attention, communication, hyperactivity impulse, and visual attention. It comprises 30 items, each scored on a 5-point Likert scale: “Never” (1), “Rarely” (2), “Sometimes” (3), “Often” (4), and “Always” (5). Higher scores indicate a greater risk of AP abnormalities. This scale was independently developed by our research team in an earlier phase and has demonstrated good reliability and validity [28]. The relevant norms have been established for the Jiangsu Province, China.

Objective AP assessment: SIN comprehension

Objective AP assessment was performed using fNIRS to record brain activation under the SIN comprehension paradigm. The task was programmed via E-prime 3.0 software, and the acoustic stimuli were delivered binaurally via RadioEar IP30 plug-in headphones [56], at a loudness of 60 dB HL.

Two-syllable words from a graded word list for Mandarin preschoolers that meet the cognitive level of preschoolers were selected as stimulus materials. The test audio was recorded by a speech rehabilitator with standard Mandarin in a soundproof shielded room, and the recorded waveforms were edited into a separate file, using GoldWave audio processing software to add 500 ms of silence before and after each two-syllable word. After calculating the root mean square of the audio and equalizing it, four-person speech noise was added as information masking, and the SNR was set to 0 dB, a level equivalent to the noise level in a typical kindergarten environment [57]. Visual images appropriate for children’s cognitive levels were selected, and 20 preschool children were recruited separately (9 children with DLD, 11 TD children) to perform picture recognition tasks, ensuring that the children could identify the images proficiently before testing.

The SIN comprehension task was designed in blocks and was repeated 4 times. The total procedure consisted of 4 SIN comprehension chunks as a task and 5 under-quiet speech comprehension chunks as a reference, with a total duration of 3 min 52 s. Each SIN block consisted of 7 bisyllabic speech stimuli under noise, and each quiet under speech comprehension block consisted of 6 bisyllabic stimuli without background noise (e.g., Fig. 1).

Fig. 1 — Flowchart of the SIN comprehension task

Acquisition of NIRS data

This study employed a functional near-infrared brain imaging system (ETG-4100, Hitachi Medical Co., Tokyo, Japan) developed and manufactured by Hitachi, Japan, utilizing a self-designed SIN comprehension paradigm as the task. The device utilizes near-infrared wavelengths of 695 nm and 830 nm, with a detection depth of 20–30 mm beneath the scalp. We arranged 44 channels (16 near-infrared light-emitting probes and 14 near-infrared light-receiving probes, with adjacent emitting and receiving probes forming one channel spaced 3 cm apart) in a 2 × 3 × 5 data acquisition pattern, placing the optodes on the children’s frontal and temporal lobes (e.g., Fig. 2). All channels were sampled synchronously at 10 Hz. All signal sensors were fixed to a soft, elastic fabric sleeve and secured to the children’s heads with elastic straps. The probe placement followed the international 10–20 system. The MNI coordinates of the corresponding brain area of each measurement channel were spatially registered with the aid of a 3D localizer (FASTRAK, Polhemus, Colchester, VT, USA), and the NIRS_SPM platform was used to output the MNI coordinates of the localized brain areas and the corresponding Brodmann area of each channel. During the test, the near-infrared light penetrates and scatters through the cerebral cortex tissues and is received by the light source detector. It indirectly yields the hemoglobin concentration in the cerebral cortex based on the Modified Beer-Lambert Law calculation [58], which is further used to obtain the activation of the cerebral cortex.

Fig. 2 — A 2D schematic of the source emitter, source-detector distribution and corresponding channels. B 3D schematic of the 44 channel locations. ROI_A corresponds to the left DLPFC, ROI_B to the left Broca’s area, ROI_C to the left Wernicke’s area, ROI_D to the left STG and MTG, ROI_E to the right DLPFC, ROI_F to the right Broca’s area, ROI_G to the right Wernicke’s area, and ROI_H to the right STG and MTG

In practice, due to variations in subject head shape, probe cap placement, spacing between light-emitting and optode pairs, and positioning errors, the experimenters inevitably introduce deviations when placing fNIRS probes on subjects’ heads. This leads to differences in the cortical locations corresponding to the same “channel” across individuals [59, 60]. The greater the number of channels, the more challenging it becomes to maintain consistent channel positioning across subjects, thereby reducing channel-level reproducibility. To mitigate this issue, region of interest (ROI) analysis can be employed [61]. This approach involves selecting broad brain regions based on anatomical or functional hypotheses and then merging/weighting multiple channels to this region. This enhances consistency in group-level analysis and reduces the interference of head shape and positioning errors on the results.

Behavioral data

The fNIRS task required all child participants to sit in front of a computer monitor at a distance of 50 cm without external distractions, wear a optode cap, to minimize movement or changes in head posture and to keep their eyes on the monitor in front of them. Before the start of the task, each child was informed of the rules of the game in detail and practiced verbally to ensure that the child knew the rules of the task. Following the practice session, a formal assessment was conducted, during which the children cooperated to complete the SIN comprehension task. The children heard one word at a time and had to decide whether the word they heard and the picture they saw were the same; in the latter case, they had to press the blue key, otherwise they had to press the red key. During the testing phase, we utilized the E-Prime 3.0 program to present experimental stimuli while simultaneously recording children’s behavioral response data. The primary metrics collected were task ACC and RT.

The primary outcome measure for the SIN task behavior was the key press ACC under quiet and noisy listening conditions. The secondary outcome measures included RT for correct trials and signal detection metrics under both quiet and noisy listening conditions. Signal sensitivity under noisy and quiet conditions was calculated based on the signal detection theory. The ability to correctly identify target stimuli was measured using the d’, calculated as (zFA - zHR).

Cognitive behavioral assessment

WPPSI-IV/WISC-IV

This study included children aged 5 to 6 years and 11 months. Preschool children completed the WPPSI-IV, while elementary school children completed the WISC-IV. Both scales are grounded in the same theoretical framework of intelligence and align closely with the contemporary Cattell–Horn–Carroll model [62]. Their core indices include the Verbal Comprehension Index (VCI), Working Memory Index (WMI), Processing Speed Index (PSI), and FSIQ, all measured on the same standard score scale (M = 100, SD = 15). The FSIQ and primary indices are composite measures, and those of both scales demonstrate good internal consistency and test-retest reliability. The WPPSI-IV FSIQ is composed of the VCI, visual-spatial index, fluid reasoning index, WMI, and PSI, while the WISC-IV FSIQ is composed of the VCI, perceptual reasoning index, WMI, and PSI.

Taking the aforementioned theoretical and measurement foundations as bases, this study selected only the VCI, WMI, PSI, and composite FSIQ scores—which exhibit complete conceptual equivalence across both scales—for subsequent analysis. This approach enhances the statistical comparability of intelligence indicators across different age groups. All assessments were administered and scored one-on-one by professionals certified in WPPSI-IV/WISC-IV, strictly adhering to the test manuals.

BRIEF-P

The BRIEF-P scale evaluates children’s executive functioning in daily life and is assessed by parents. Measures such as inhibit, shift, emotional control, working memory, plan/organize, inhibitory self-control index (ISCI), flexibility index (FI), and Emergent Metacognition Index (EMI) were analyzed. The higher the score is, the more severe the impairment in executive function.

Preprocessing of fNIRS data

The data were preprocessed and analyzed on MATLAB 2013a platform using the NIRS_SPM (Near-Infrared Spectroscopy-Statistical Parametric Mapping) program package [63]. The data were filtered and baseline-corrected; the high-pass cutoff frequency was set to 0.01 Hz, and the low-pass cutoff frequency to 0.1 Hz. Signal drift was removed using the Wavelet minimum description length (Wavelet-MDL) method [64], and low-pass filtering was performed by haemodynamic response functions (HRF) to eliminate signal distortion in the fNIRS data caused by high-frequency noise and artifacts, etc. Given that changes in HbO exhibit higher sensitivity to alterations in cerebral blood flow [65], HbO concentration data were selected for subsequent analysis.

Statistical analysis

Statistical analysis was performed using SPSS 26.0 software. The normality of continuous data was assessed using a combination of the Shapiro-Wilk test (appropriate for the sample size in this study) and the quantile-quantile plot (Q-Q plot). If the Shapiro-Wilk test result was P > 0.05 and the data points on the Q-Q plot were generally close to the reference line, the data were considered to follow a normal distribution and were described as mean ± standard deviation (x ± s), with between-group comparisons performed using the two-sample t-test; otherwise, the data were considered non-normally distributed and described as median (interquartile range) [M (P25, P75)], and the Mann-Whitney U test was used for between-group comparisons. Categorical data are presented as frequencies (n) and percentages (%), and the chi-square test is used for between-group comparisons. For non-normally distributed continuous variables, Spearman’s rank correlation analysis was used. To account for multiple comparisons, the resulting P-values were adjusted using the Benjamini-Hochberg procedure, controlling the false discovery rate (FDR). All tests were two-sided; P < 0.05 was considered statistically significant.

The fNIRS data were analyzed by the NIRS_SPM software package based on a general linear model (GLM). The β-value of the magnitude of the response of the task in each channel relative to the baseline was derived based on the modeling of oxyhemoglobin in each channel. The obtained channel results were corrected for multiple comparisons by the FDR. The average β-value of each ROI was calculated and the activation level of each ROI was analyzed by one-sample t-test (test value of 0). Finally, the β-values of the TD and DLD groups were compared using the independent samples t-test. P < 0.05 was considered a statistically significant difference.

Visualization

Visualization was performed using the BrainNet package in MATLAB 2013a.

Results

Comparison of basic demographic information

No statistically significant differences were found between the two groups of children in terms of gender, age, gestational age, age of parents at birth, and whether or not they had neonatal hyperbilirubinemia (P > 0.05). Furthermore, there were statistically significant differences between the groups in terms of birth weight, daily screen time, parent-child playtime, parent-child shared reading, parental education, and average monthly household income (P < 0.05), as shown in Table 1.

Table 1.

Comparison of basic demographic information

	DLD (n = 34)	TD (n = 43)	χ²/t/Z	P
Sex, n (%)
male	23(67.6)	29(67.4)	< 0.001^a	0.985
female	11(32.4)	14(30.2)
Age (months)	71.0(65.5, 75.3)	69.0(66.0, 73.0)	-1.120^c	0.263
Gestational age, n (%)
≥ 37 weeks	29(85.3)	41(95.3)	1.265^a	0.261
< 37 weeks	5(14.7)	2(4.7)	1.265^a	0.261
Birth weight (g)	3019.1 ± 641.7	3334.9 ± 533.9	-2.357^b	0.021*
Age of the mother at the time of the child’s birth	29(26, 32.5)	30(28, 34)	-1.913^c	0.056
Age of the father at the time of the child’s birth	29.5(27, 35)	31(29, 34)	-1.434^c	0.151
Neonatal hyperbilirubinemia
no	30(88.2)	40(93.0)	0.107^a	0.744
yes	4(11.8)	3(7.0)	0.107^a	0.744
Daily screen time, n (%)
< 0.5 h	8(23.5)	23(53.5)	7,195^a	0.027*
0.5–1 h	17(50.0)	14(32.6)
> 1 h	9(26.5)	6(14.0)
Daily parent-child playtime, n (%)
< 0.5 h	9(26.5)	2(4.7)	7.772^a	0.021*
0.5–1 h	12(35.3)	23(53.5)
> 1 h	13(38.2)	18(41.9)
Daily parent-child shared reading time, n (%)
< 0.5 h	22(64.7)	11(25.6)	11.985^a	0.002**
0.5–1 h	10(29.4)	25(58.1)
> 1 h	2(5.9)	7(16.3)
Mother’s educational attainment, n (%)
high school and below	10(29.4)	1(2.3)	23.518^a	< 0.001***
post-secondary	10(29.4)	3(7.0)
undergraduate	10(29.4)	20(46.5)
postgraduate student	4(11.8)	19(44.2)
Father’s educational attainment, n (%)
high school and below	9(26.5)	1(2.3)	19.755^a	< 0.001***
post-secondary	10(29.4)	6(14.0)
undergraduate	12(35.3)	16(37.2)
postgraduate student	3(8.8)	20(46.5)
Average monthly household income, n (%)
< 10,000 yuan	12(35.3)	1(2.3)	18.906^a	< 0.001***
10,000–30,000 yuan	18(52.9)	23(53.5)
> 30,000 yuan	4(11.8)	19(44.2)

Open in a new tab

DLD developmental language disorder, TD typical development

*P < 0.05; **P < 0.01; ***P < 0.001

^a: x² test, ^b: t test, ^c: Mann-Whitney U test

Language and related cognitive performance in children with DLD

Since the data did not follow a normal distribution, the Mann-Whitney U test was used to compare the results of the DREAM-C, WPPSI-IV/WISC-IV, and BRIEF-P between groups. The results showed that children with DLD scored lower than TD children on all dimensions of the DREAM-C, WPPSI-IV/WISC-IV (P < 0.001), while children with DLD scored higher than TD children on all dimensions of the BRIEF-P (P < 0.05). (See Figs. 3B, C, and D.)

Fig. 3 — A Comparison of PAPAS results between the DLD and TD groups, where a1 is hyperactivity impulse, a2 is communication, a3 is visual attention, a4 is auditory decoding, a5 is auditory attention, and a6 is the AP total scale. B Comparison of DREAM-C results between children in the DLD and TD groups, where b1 is syntax development, b2 is receptive language, b3 is semantics, b4 is expressive language, and b5 is total language. C Comparison of WPPSI-IV/WISC-IV results between the DLD and TD groups, where c1 is FSIQ, c2 is PSI, c3 is VCI, and c4 is WMI. D Comparison of the results of the BRIEF-P for children in the DLD and TD groups, where d1 is working memory, d2 is plan/organize, d3 is FI, d4 is emotional control, d5 is inhibit, d6 is ISCI, d7 is EMI, and d8 is shift. DLD, developmental language disorder, TD, typical development; Wilcoxon rank-sum test; *P<0.05; **P<0.01; ***PP<0.001

AP characteristics in children with DLD

Behavioral performance of AP in children with DLD

Results of the PAPAS scale

Since the data across all dimensions of the PAPAS scale did not follow a normal distribution, Mann-Whitney U test was used for intergroup comparisons of PAPAS. The results showed that the DLD group had higher scores on the total AP score (z=-5.265, P < 0.001), auditory decoding (z=-5.665, P < 0.001), auditory attention (z=-2.646, P = 0.008), communication (z=-3.961, P < 0.001), hyperactivity and impulse (z=-4.458, P < 0.001), and visual attention (z=-2.556, P = 0.011) than TD children, and the difference was statistically significant, as shown in Fig. 3A.

Behavioral performance on the SIN comprehension task

The behavioral results of children in the DLD and TD groups on the SIN task are shown in Tables 2, 3, 4, 5, 6 and 7. A two-way ANOVA was conducted to examine the effects of the two independent variables—group (DLD vs. TD) and listening condition (quiet vs. noisy)—on the dependent variables ACC, RT, and d'. To assess the effect size, we reported effect measures (e.g., η²) quantifying the influence of group and listening condition on dependent variables. When the main effects were significant, we conducted Bonferroni-corrected pairwise comparisons and reported 95% confidence intervals (CI) for differences. When interactions were significant, simple effects analyses were performed to further explore their specific contributions.

Table 2.

Two-way ANOVA for ACC

Source	Sum of squares	df	Mean square	F	P	η²
Corrected model	2.369	3	0.790	22.676	< 0.001***	0.312
Groups	1.891	1	1.891	54.303	< 0.001***	0.266
Listening condition	0.412	1	0.412	11.830	0.001**	0.073
Group × listening condition	0.032	1	0.032	0.931	0.336	0.006
Error	5.223	150	0.035
Corrected total	83.028	154

Open in a new tab

Groups: DLD and TD; Listening conditions: Quiet and Noisy; F: Two-way ANOVA

**P < 0.01; ***P < 0.001

Table 3.

ACC main effect pairwise comparisons

(I) Groups	(J) Groups	Mean difference (I-J)	SE	P-value^a	95% CI^a
DLD	TD	-0.223	0.03	< 0.001***	(-0.283, -0.163)
TD	DLD	0.223	0.03	< 0.001***	(0.163, 0.283)
quiet	noisy	0.104	0.03	0.001**	(0.044, 0.164)
noisy	quiet	-0.104	0.03	0.001**	(-0.164, -0.044)

Open in a new tab

^a: Multiple comparison adjustment: Bonferroni method

**P < 0.01; ***P < 0.001

Table 4.

Two-way ANOVA for RT

Source	Sum of squares	df	Mean square	F	P-value	η²
Corrected model	1124159.986	3	374719.995	2.642	0.052	0.053
Groups	1085342.360	1	1085342.360	7.654	0.006**	0.051
Listening condition	8439.297	1	8439.297	0.060	0.808	0.000
Group × listening condition	34650.502	1	34650.502	0.244	0.622	0.002
Error	20136633.960	142	141807.281
Corrected total	761,988,662	146

Open in a new tab

Groups: DLD and TD; Listening conditions: Quiet and Noisy; F: Two-way ANOVA

**P < 0.01

Table 5.

RT main effect pairwise comparisons

(I) Groups	(J) Groups	Mean difference (I-J)	SE	P-value^a	95% CI^a
DLD	TD	-174.431	63.051	0.006**	(-299.071, -49.792)
TD	DLD	174.431	63.051	0.006**	(49.792, 299.071)
quiet	noisy	15.381	63.051	0.808	(-109.258, 140.021)
noisy	quiet	-15.381	63.051	0.808	(-140.021, 109.258)

Open in a new tab

^a: Multiple comparison adjustment: Bonferroni method

**P < 0.01

Table 6.

Two-way ANOVA for d'

Source	Sum of squares	df	Mean square	F	P-value	η²
Corrected model	85.521	3	28.507	22.308	< 0.001***	0.322
Groups	68.388	1	68.388	53.517	< 0.001***	0.275
Listening condition	11.442	1	11.442	8.954	0.003**	0.060
Group × listening condition	3.454	1	3.454	2.703	0.102	0.019
Error	180.180	141	1.278
Corrected total	486.798	145

Open in a new tab

Groups: DLD and TD; Listening conditions: Quiet and Noisy; F: Two-way ANOVA

**P < 0.01, ***P < 0.001

Table 7.

d' Main effect pairwise comparisons

(I) Groups	(J) Groups	Mean difference (I-J)	SE	P-value^a	95% CI^a
DLD	TD	-1.391	0.19	< 0.001***	(-1.767, -1.015)
TD	DLD	1.391	0.19	< 0.001***	(1.015, 1.767)
quiet	noisy	0.569	0.19	0.003**	(0.193, 0.945)
noisy	quiet	-0.569	0.19	0.003**	(-0.945, -0.193)

Open in a new tab

^a: Multiple comparison adjustment: Bonferroni method

**P < 0.01, ***P < 0.001

A two-way ANOVA (group × listening condition) was conducted on accuracy rates. The results indicated no significant interaction effect (F = 0.931, P = 0.336). A significant main effect of group was observed (F = 54.303, P < 0.001). Post-hoc pairwise comparisons revealed statistically significant differences between the DLD and TD groups in ACC (P < 0.001). A significant main effect of listening condition was observed (F = 11.83, P = 0.001). Post-hoc pairwise comparisons revealed a statistically significant difference in ACC between quiet and noisy conditions (P = 0.001). See Tables 2 and 3.

A two-way ANOVA (group × listening condition) was conducted on RT, revealing no significant interaction effect (F = 0.244, P = 0.622). The main effect of group was significant (F = 7.654, P = 0.006). Paired comparisons revealed a statistically significant difference in RT between the DLD and TD groups (P = 0.006). The main effect of the listening condition was not significant (F = 0.06, P = 0.808). See Tables 4 and 5.

A two-way ANOVA was conducted on d' (group × listening condition). The results indicated no significant interaction effect (F = 2.703, P = 0.102). A significant main effect of group was observed (F = 53.517, P < 0.001). Post-hoc pairwise comparisons revealed a statistically significant difference in d' between the DLD and TD groups (P < 0.001). The main effect of listening condition was significant (F = 8.954, P = 0.003). Post-hoc pairwise comparisons revealed a statistically significant difference in d' between quiet and noisy conditions (P = 0.003). See Tables 6 and 7.

Brain activation characteristics during the SIN comprehension task in children with DLD

The brain activation levels of each ROI under the SIN comprehension task vs. the quiet speech comprehension task are shown in Table 8. Compared with the quiet condition, DLD children showed decreased activation in the left DLPFC, left Broca’s area, and right Wernicke’s area under the noisy condition, with a statistically significant difference (P < 0.05); TD children showed increased activation in the left STG and MTG, the right DLPFC, and the right STG and MTG under the noisy condition, compared with the quiet condition, with a statistically significant difference (P < 0.05). To control for the multiple comparison problem, FDR correction was applied. After correction, compared with the quiet condition, DLD children still exhibited reduced activation in the left DLPFC (P-corrected = 0.009), left Broca’s area (P-corrected = 0.007), and right Wernicke’s area (P-corrected = 0.016) under the noisy condition, as shown in Fig. 4A. Compared with the quiet condition, children in the TD group showed significantly increased activation in the left STG and MTG (P-corrected < 0.001) and right STG and MTG (P-corrected < 0.001) under the noisy condition, as shown in Fig. 4B; activation in the right DLPFC was no longer significant (P-corrected = 0.087).

Table 8.

Results of brain activation scores in SIN comprehension task in the two groups

ROI	β Mean ± SD	t	P	P-corrected
DLD (n = 34)
ROI_A	-0.034 ± 0.061	-3.297	0.002**	0.009**
ROI_B	-0.042 ± 0.067	-3.654	< 0.001^***	0.007**
ROI_C	-0.025 ± 0.080	-1.801	0.081	0.146
ROI_D	0.015 ± 0.067	1.281	0.209	0.278
ROI_E	-0.008 ± 0.058	-0.838	0.408	0.408
ROI_F	-0.012 ± 0.058	-1.187	0.244	0.278
ROI_G	-0.056 ± 0.110	-2.942	0.006**	0.016*
ROI_H	0.021 ± 0.071	1.739	0.091	0.146
TD (n = 43)
ROI_A	0.013 ± 0.045	1.879	0.067	0.134
ROI_B	0.012 ± 0.053	1.515	0.137	0.218
ROI_C	-0.004 ± 0.080	-0.335	0.739	0.738
ROI_D	0.065 ± 0.070	6.126	< 0.001^***	< 0.001^***
ROI_E	0.019 ± 0.058	2.208	0.033*	0.087
ROI_F	0.015 ± 0.071	1.397	0.170	0.226
ROI_G	0.011 ± 0.072	1.044	0.302	0.346
ROI_H	0.079 ± 0.079	6.565	< 0.001^***	< 0.001^***

Open in a new tab

ROI_A, left DLPFC; ROI_B, left Broca’s area; ROI_C, left Wernicke’s area; ROI_D, left STG and MTG; ROI_E, right DLPFC; ROI_F, right Broca’s area; ROI_G, right Wernicke’s area; ROI_H, right STG and MTG; t, one-sample t-test with a test value of 0

*P < 0.05, **P < 0.01, ***P < 0.001

Fig. 4 — ROI activation maps for children in the DLD and TD groups. A Activation levels in the DLD group during the SIN comprehension task compared to the quiet speech comprehension task. Under the noisy condition, the DLD group showed reduced activation in the left DLPFC, left Broca’s area, and right Wernicke’s area. B Activation patterns in the TD group during the SIN comprehension task compared to the quiet verbal comprehension task. Under the noisy condition, the TD group showed increased activation in the left STG and MTG and the right STG and MTG. C During the SIN comprehension task, activation levels in the left DLPFC, left Broca’s area, left STG and MTG, right Wernicke’s area, and right STG and MTG were lower in the DLD group than those in the TD group

The comparison of the brain activation degree between the DLD group and the TD group in the SIN comprehension task is shown in Table 9. The activation degrees of left DLPFC, left Broca’s area, left STG and MTG, right DLPFC, right Wernicke’s area, and right STG and MTG in the DLD group were significantly lower than those of the TD group, with a statistical significance (P < 0.05). After FDR correction, the DLD group showed significant differences in the left DLPFC (P-corrected < 0.001), left Broca’s area (P-corrected < 0.001), left STG and MTG (P-corrected = 0.003), right Wernicke’s area (P-corrected = 0.003) and right STG and MTG (P-corrected = 0.003) were significantly lower than those in the TD group. Activation in the right DLPFC showed no statistically significant difference (P-corrected = 0.053), as shown in Fig. 4C, Fig. 5.

Table 9.

Comparison of brain activation in the SIN comprehension task between the two groups of children

	DLD	TD	t	P	P-corrected
	Median (Q1, Q3)	Median (Q1, Q3)	t	P	P-corrected
ROI_A	-0.034 ± 0.061	0.013 ± 0.045	-3.921	< 0.001***	0.001**
ROI_B	-0.042 ± 0.067	0.012 ± 0.053	-3.964	< 0.001***	0.001**
ROI_C	-0.025 ± 0.080	-0.004 ± 0.080	-1.124	0.265	0.265
ROI_D	-0.015 ± 0.067	0.065 ± 0.070	-3.194	0.002**	0.003**
ROI_E	-0.008 ± 0.058	0.019 ± 0.058	-2.092	0.040*	0.053
ROI_F	-0.012 ± 0.058	0.015 ± 0.071	-1.790	0.077	0.088
ROI_G	-0.056 ± 0.110	0.011 ± 0.072	-3.222	0.002**	0.003**
ROI_H	-0.21 ± 0.071	0.079 ± 0.079	-3.331	0.001**	0.003**

Open in a new tab

*P < 0.05; **P < 0.01; ***P < 0.001

Fig. 5 — Waveform of ROI O₂Hb concentration in the two groups of children. TD: Typically Developing; DLD: Developmental Language Disorder; SD: Standard Deviation; TOI: Time of Interest

Correlations between behavioral performance on the SIN comprehension task and ROI activation levels

In the SIN comprehension task, Spearman’s correlation was used to analyze the relationship between the activation levels of various brain ROIs and behavioral scores in the SIN task. The activation levels of ROI_A and ROI_B were positively correlated with the ACC (P-corrected < 0.001) and d' (P-corrected < 0.05) scores under the quiet listening condition, respectively; the activation levels of ROI_A, ROI_B, and ROI_E were positively correlated with the ACC and d' scores under the noisy listening condition, respectively (P-corrected < 0.05). The results are presented in Fig. 7(D).

Fig. 7 — Correlation Analysis Between Activation Levels in Brain ROI During Speech Recognition Under Noise and Cognitive Behavioral Assessment Results. A Correlation between the activation degree of each ROI in speech recognition tasks under noise and DREAM-C, where a1 is Total language, a2 is Receptive language, a3 is Expressive language, a4 is Semantics, and a5 is Syntax development. B Correlation analysis of brain ROI activation in the SIN comprehension with the WPPSI-IV/ WISC-IV. C Correlation analysis of the degree of activation of brain ROIs in the SIN comprehension with BRIEF-P scores, where b1 is Inhibit, b2 is Shift, b3 is Emotional control, b4 is Working memory, b5 is Plan / organize, b6 is FI, b7 is EMI, and b8 is ISCI. D Correlation analysis between activation in the brain ROI identified by SIN and behavioral outcomes, where c1 is quiet_d', c2 is quiet_ACC, c3 is quiet_RT, c4 is noisy_d', c5 is noisy_ACC, c6 is noisy_RT. Correlation analysis: Spearman’s correlation analysis. *P < 0.05; **P < 0.01; ***P < 0.001

Associations between AP characteristics and language and cognitive performance in the two groups

Correlations between PAPAS scores and language and cognitive behavioral performance

Spearman’s correlation analysis was used to examine the relationship between PAPAS scores and DREAM-C scores, as shown in Fig. 6(A). The results showed that the AP total scores, auditory decoding, communication, and hyperactivity impulse scores were significantly negatively correlated with the scores for total language, receptive language, expressive language, semantics, and syntax development (P-corrected < 0.05).

Using Spearman’s correlation, we analyzed the relationship between PAPAS scale scores and WPPSI-IV/WISC-IV scores, as shown in Fig. 6(B). The results showed that the AP total scores, auditory decoding, and hyperactivity impulse scores were significantly negatively correlated with FSIQ, VCI, WMI, and PSI (P-corrected < 0.05); auditory attention scores were significantly negatively correlated with WMI (P-corrected = 0.038); communication scores were significantly negatively correlated with FSIQ, VCI, and WMI (P-corrected < 0.05); and visual attention scores were significantly negatively correlated with VCI scores (P-corrected = 0.030).

Using Spearman’s correlation, we analyzed the relationship between PAPAS scale scores and BRIEF-P scores, as shown in Fig. 6(C). The results showed that the AP total scores, auditory decoding, auditory attention, communication, hyperactivity-impulse, and visual attention were found to be significantly positively correlated with scores on inhibit, shift, emotional control, working memory, plan/organize, FI, EMI, and ISCI (P-corrected < 0.05).

Correlations between ROI activation levels during the SIN comprehension task and language and cognitive behavioral performance

In the SIN comprehension task, Spearman’s correlation analysis was used to examine the relationship between activation levels in various ROI and DREAM-C scores. The activation levels of ROI_A, ROI_B, ROI_D, ROI_G, and ROI_H were significantly positively correlated with the scores on the total language, receptive language, expressive language, semantics, and syntax development dimensions (P-corrected < 0.05). The results are presented in Fig. 7(A).

In the SIN comprehension task, Spearman’s correlation analysis was employed to examine the relationship between activation levels in various brain ROI and WPPSI-IV/WISC-IV scores. The activation levels of ROI_A, ROI_B, ROI_G, and ROI_H were significantly positively correlated with FSIQ and VCI (P-corrected < 0.05); The activation levels of ROI_A, ROI_B, ROI_D, ROI_E, ROI_F, ROI_G, and ROI_H were significantly positively correlated with WMI scores (P-corrected < 0.05); the activation levels of ROI_B and ROI_D were significantly positively correlated with PSI scores (P-corrected < 0.05). The results are presented in Fig. 7(B).

In the SIN comprehension task, Spearman’s correlation analysis was used to examine the relationship between activation levels in various brain ROI and BRIEF-P scores. The activation levels of ROI_A, ROI_B, and ROI_D were significantly negatively correlated with scores on the inhibit, working memory, plan/organize, and EMI (P < 0.05); the activation level of ROI_D was significantly negatively correlated with the score on the ISCI (P = 0.019). After FDR correction, none of the P-values were significant (P-corrected > 0.05). The results are shown in Fig. 7(C).

Regression analysis of AP characteristics and language performance

Variables showing differences between groups in the basic demographic data, indicators with statistically significant differences in the PAPAS, and ACC under noisy listening conditions were included in the regression model. A stepwise analysis was used to identify factors influencing total language ability in the DREAM-C results. The results showed that ACC under noisy listening conditions in the SIN comprehension task, the level of activation in ROI_H, and the duration of daily parent-child shared reading time had a significant impact on total language ability (P < 0.05). See Table 10.

Table 10.

Factors influencing total language ability in the two groups of children

		Unstandardization coefficient		Standardization coefficient	t	p
		B	SE	Beta
Constant		90.826	10.714		8.477	＜0.001***
ROI_H		40.437	16.863	0.261	2.398	0.020*
Daily parent-child shared reading time	0.5-1h	7.176	2.953	0.287	2.430	0.019*
Daily parent-child shared reading time	＞1h	11.630	4.881	0.288	2.383	0.021*
Noisy_ACC		13.465	6.470	0.234	2.081	0.042*

Open in a new tab

Dependent variable: Total language; Independent variables: activation levels in ROIs A, B, D, G, and H during the SIN comprehension tasks; ACC under noise listening conditions; the AP total score; birth weight; father’s educational attainment; mother’s educational attainment; average monthly household income; daily screen time; daily parent-child playtime; daily parent-child shared reading time. *P < 0.05; ***P < 0.001

Building on the findings in Table 10, FSIQ was included in the regression model. The results indicate that FSIQ and the daily parent-child shared reading time at home have a significant impact on total language ability (P < 0.05). See Table 11.

Table 11.

Factors influencing total language ability in the two groups of children

		Unstandardized coefficient		Standardization coefficient	t	P
		B	SE	Beta
Constant		51.861	10.950		4.736	＜0.001***
FSIQ		0.407	0.069	0.556	5.917	＜0.001***
Daily parent-child shared reading time	0.5-1h	4.802	2.344	0.192	2.049	0.046*
Daily parent-child shared reading time	＞1h	9.166	3.841	0.228	2.386	0.021*

Open in a new tab

Dependent variable: Total language; Independent variables: activation levels in ROIs A, B, D, G, and H during the SIN comprehension tasks; ACC under noise listening conditions; the AP total score; FSIQ; birth weight; father’s educational attainment; mother’s educational attainment; average monthly household income; daily screen time; daily parent-child playtime; daily parent-child shared reading time. *P < 0.05; ***P < 0.001

Associations between AP characteristics and language and cognitive performance in children with DLD

The activation levels of ROI_A in the DLD group were positively correlated with ACC activity under the quiet listening condition, while the activation levels of ROI_E and ROI_G were positively correlated with d' under the noisy listening condition; however, none reached post-hoc significance (P-corrected > 0.05).

There was no statistically significant correlation between the scores on each dimension of the DLD group’s PAPAS scale and the scores on each dimension of the DREAM-C and WPPSI-IV/WISC-IV (P > 0.05). In the DLD group, scores for auditory decoding and auditory attention were positively correlated with scores for inhibit, working memory, plan/organize, the ISCI, and the EMI (P-corrected < 0.05). Communication scores were positively correlated with scores for inhibit, working memory, the FI, the ISCI, and the EMI (P-corrected < 0.05). Hyperactivity impulse scores were positively correlated with scores for inhibit and the ISCI (P-corrected < 0.05), as shown in Table 12.

Table 12.

Correlation analysis between the PAPAS scale and the BRIEF-P scale for children with DLD

	inhibit	shift	emotional control	working memory	plan/organize	ISCI	FI	EMI
AP total scores	0.658***	0.226	0.376*	0.613***	0.509**	0.639***	0.368*	0.649***
auditory decoding	0.559**	0.122	0.238	0.601***	0.454*	0.521**	0.234	0.604***
auditory attention	0.701***	0.262	0.372*	0.518**	0.466*	0.666***	0.374*	0.557**
communication	0.409*	0.345	0.365*	0.488*	0.272	0.426*	0.420*	0.474*
hyperactivity impulse	0.419*	0.132	0.222	0.198	0.188	0.394*	0.221	0.254
visual attention	0.226	-0.101	-0.063	0.296	0.324	0.152	-0.092	0.324

Open in a new tab

Spearman’s correlation analysis

*P < 0.05, **P < 0.01, ***P < 0.001

The activation level of ROI_F in the DLD group was positively correlated with total language scores, but did not reach statistical significance after correction (P-corrected > 0.05). The activation level of ROI_F in the DLD group was positively correlated with FSIQ and WMI scores. However, neither correlation reached the level of post-correction significance (P-corrected > 0.05). The correlations between the activation levels of each ROI in the DLD group and the scores on each dimension of the BRIEF-P were not statistically significant (P > 0.05).

Discussion

This cross-sectional investigation utilized fNIRS technology to analyze and compare the behavioral characteristics and brain activation patterns during AP tasks between children with DLD and TD children. It further assessed the cognitive behavioral differences between the two groups and preliminarily explored the correlation between fNIRS features and language/cognitive behavioral performance. This section focuses on discussing the following aspects: (1) Demographic factors and cognitive behavioral performance in children with DLD; (2) Behavioral performance of AP in children with DLD; (3) Brain activation patterns in children with DLD during the SIN comprehension task; (4) Correlations between AP behaviors, brain activation patterns in the SIN comprehension task, and language/cognitive behavioral performance in children with DLD.

Comparison of basic demographic information

This study included 77 children and found significant differences in birth weight between the DLD and TD groups, with low birth weight identified as a key risk factor for DLD [66]. As is known, parent-child interaction is crucial for language development. A lack of diverse parent-child interactive behavior is associated with poorer language abilities [3], and parent-child reading time is closely related to language development [67]. The American Academy of Pediatrics recommends that children aged 2–5 have no more than one hour of screen time daily [68]. Indeed, excessive screen exposure is negatively correlated with language development [69]. Furthermore, low parental education is associated with reduced interaction and social engagement, impacting language growth. These factors may contribute to differences in daily screen time, parent-child interaction, and education levels between the DLD and TD groups. Consistent with prior studies [70], there was also a significant difference in the average monthly household income between the two groups of children; children from families with low socioeconomic backgrounds exhibited slower vocabulary development and a higher likelihood of persistent language difficulties.

Language and related cognitive performance in children with DLD

Studies employing WPPSI-IV/WISC-IV to assess intelligence and cognitive abilities have shown that children with DLD score significantly lower than children with TD in FSIQ and cognitive areas (VCI, WMI, PSI). Children with DLD exhibit significant deficits in core language abilities, including receptive and expressive language [71]. The lower VCI scores in DLD children are partly attributed to poor syntactic awareness, which impedes syntactic-based predictive processing and spoken sentence comprehension [72]. Meanwhile, working memory deficits in DLD children have been linked to language performance [73, 74]. Impairments in central executive function and the phonological loop subsystem hinder the retention, retrieval, and processing of speech signals [75, 76]. Additionally, slower processing speeds in both verbal and non-verbal tasks correlate with lower language abilities [77, 78]. However, the causal relationship between cognitive deficits and language disorders remains unclear [79] and requires further investigation.

Executive function involves higher cognitive processes for goal setting, planning, and effective execution, essentially regulating other cognitive functions [80]. Compared to TD children, those with DLD exhibit significant impairments in all BRIEF-P scale factors, which has been verified in previous studies [81]. These difficulties persist even after controlling for language ability [82]. Some researchers have suggested that EF deficits limit the development of verbal skills in DLD children, though this relationship remains controversial [83, 84]. Nevertheless, both perspectives confirm that children with DLD experience EF deficits.

AP characteristics in children with DLD

Behavioral performance of AP in children with DLD

Results of the PAPAS scale

The PAPAS scale was used to screen for AP abnormalities in preschoolers with DLD, demonstrating good reliability and validity. The DLD group scored significantly higher than the TD group on both total and dimensional PAPAS scores, indicating a higher risk of AP abnormalities in DLD children, consistent with previous studies [85]. These results suggest that children with DLD exhibit AP deficits in behavioral performance. CAP abnormalities impair speech comprehension, particularly in challenging listening environments [86]. To further validate this notion, this study used a SIN comprehension behavioral paradigm and fNIRS to assess brain activation during AP tasks.

Behavioral performance on the SIN comprehension task

Speech comprehension under noisy conditions better reflects real-life communication ability, involving both bottom-up and top-down processing [87]. Bottom-up processing decodes speech content by matching it in working memory with long-term memory. Meanwhile, top-down processing uses existing language knowledge and contextual cues to understand speech in noise. SIN recognition involves mechanisms such as auditory perception, selective attention, speech decoding, and cognitive resource allocation. Since four-person noise more closely resembles everyday noise, this study used it as background noise. Individual cognitive differences significantly predict SIN accuracy [88]. To control for cognitive variability, this study implemented disyllabic words as test materials, ensuring that they align with preschool children’s cognitive levels. A pre-test assesses children’s understanding of the test requirements and their ability to identify relevant information.

At the group level, children with DLD demonstrated significantly poorer overall performance than TD children on the speech recognition task. This was reflected in lower ACC and d' for DLD children compared to TD children, consistent with our expectations. Notably, despite the lower ACC and d' scores, DLD children exhibited shorter RT than the TD group, which may stem from more impulsive or less deliberate responses during auditory tasks, potentially due to weakened attentional control. In continuous performance tasks, DLD children demonstrated greater impulsivity and poorer attentional focus than their peers [89], or a tendency to rush responses before fully processing auditory content.

In terms of the listening condition, both groups of children demonstrated lower accuracy rates and d’ values in noisy environments compared to quiet settings, reflecting the widespread challenge noise poses to children’s speech perception [90]. However, the difference between RT in quiet versus noisy conditions was not statistically significant.

Nevertheless, the lack of significant group × listening condition interactions found for ACC, RT and d’ in this study contradicted our expectations. This may stem from noise similarly impairing TD children, while DLD children exhibit fundamental difficulties in perceiving and processing speech signals—even under optimal auditory conditions [17]—thereby nullifying the interaction effects. Nevertheless, these findings underscore the importance of supporting children with DLD in noisy environments and further reveal the critical role of the relevant auditory perception mechanisms.

This study investigated a bimodal language comprehension task involving auditory and visual integration, where children were asked to recognize two-syllable words, comprehend meanings, and match them to visual stimuli. Children with DLD may face more difficulty in lexical decision-making. Therefore, future research could explore cross-modal tasks and speech types (words vs. sentences).

Brain activation characteristics during the SIN comprehension task in children with DLD

Language and speech processing involves three core processes: speech processing, semantic processing [91], and syntactic processing [92]. These functions engage brain regions in the left frontal, temporal and parietal cortex [93–96], which are connected by subcortical white matter fiber tracts [97]. The dual-pathway model of speech processing comprises the dorsal speech pathway and the ventral semantic pathway [98]. While both pathways are activated in concept-to-hearing and concept-to-articulation tasks, like speech production [99], the ventral pathway is activated in auditory-to-conceptual tasks, such as speech comprehension [100]. The dorsal pathway is mainly located in the left prefrontal cortex and STG [94], while the ventral pathway spans the bilateral temporal and parietal lobes [101]. Studies using fMRI have shown that higher background noise activates more frontal regions during SIN recognition [102]. Working memory capacity mediates speech processing in noise, with increased noise level requiring more reliance on working memory [103]. fNIRS technology assesses brain activation patterns in children during SIN auditory processing, engaging multiple brain regions, particularly those involved in AP, language, attention, and working memory [104–106], mainly in the frontal and temporal lobes.

In our study, compared to speech comprehension under quiet conditions, children with TD showed elevated activation levels in the bilateral STG and MTG during SIN comprehension. This finding is largely consistent with previous studies investigating activation patterns in healthy adults, which demonstrated significant activation in the right frontal lobe and bilateral temporal lobes during the comprehension of Mandarin speech sentences in noisy environments [107]. Children with DLD exhibited reduced activation in the left DLPFC, left Broca’s area, and right Wernicke’s area under noisy conditions, indicating that these children are unable to mobilize the normal network of prefrontal and temporal brain regions required to complete complex AP tasks.

Under the noisy conditions, compared to TD children, DLD children exhibited significantly lower activation levels in the left DLPFC, Broca’s area, STG and MTG, and right Wernicke’s area, STG and MTG. Badcock et al. [108] observed similar results, noting lower activation in the left subfrontal cortex and bilateral STG in DLD children, consistent with our findings. The DLPFC is critical for linguistic working memory as it filters and monitors auditory and visual information [109–111]. Broca’s area is crucial for language [112, 113] and the STG processes AP [114], which highlights a fundamental defect in the bottom-up AP of children with DLD. Insufficient activation of these areas in DLD children impairs speech recognition under noise. A plausible explanation is that listening in noisy environments places high demands on the language system. Due to the weaker or less stable representations of speech and vocabulary, DLD individuals cannot fully activate the neural circuits required to recognize words in impaired signals [15]. In contrast, TD individuals, possessing stronger language processing abilities, can mobilize additional neural resources—including the right auditory cortex and frontal control areas—to overcome these challenges.

Correlations between behavioral performance on the SIN comprehension task and ROI activation levels

Under quiet listening conditions, higher activation in the left DLPFC and the left Broca’s area was associated with ACC and higher d' scores. Under noisy listening conditions, this positive correlation extended to the right DLPFC, suggesting that noisy environments require additional frontal control resources and may lead to compensatory bilateralization. This is consistent with the findings of previous fNIRS studies, which showed that during speech perception in noise within naturalistic scenes, as the SNR decreased and task difficulty increased—and as participants reported greater difficulty understanding speech—activity levels in the right DLPFC significantly increased [115]. The findings of this study further support the notion that noise reduces the reliability of speech and increases category competition, causing the auditory speech system to rely more heavily on top-level attention and working memory to maintain a stable representation of the task goal. This mechanism contributes to increased decision sensitivity and reduced error rates, consistent with the Framework for Understanding Effortful Listening (FUEL), which posits that difficult listening requires the active allocation of mental resources [116].

Associations between AP characteristics and language and cognitive performance in the two groups

Correlations between PAPAS scores and language and cognitive behavioral performance

The PAPAS scores showed a significant negative correlation with the DREAM-C language assessment, as PAPAS is reverse-scored. This indicates that preschool children with DLD exhibit poorer language ability if their AP abilities are weaker. A longitudinal study further demonstrated that early deficits in rapid AP can predict subsequent language delay in children [117].

The PAPAS scores showed a negative correlation with WPPSI-IV/WISC-IV scores, indicating that children’s AP abilities positively correlate with intellectual levels, consistent with previous research findings [118, 119]. The encoding and decoding of speech signals within the central auditory nervous system are prerequisites for language comprehension.

The various dimensions of the PAPAS showed significant and moderate-to-high positive correlations with multiple dimensions of the BRIEF-P, suggesting that the more pronounced a child’s AP difficulties are, the more prominent their executive function issues become, which is consistent with previous research [120]. When the efficiency of AP declines, children require more cognitive resources to compensate for perceptual deficits, which in turn affects their self-regulation and metacognitive abilities.

Correlations between ROI activation levels during the SIN comprehension task and language and cognitive behavioral performance

In the SIN comprehension task, the activation of the left frontal lobe and bilateral temporal lobes was significantly positively correlated with DREAM-C dimensions. A significant positive correlation was also observed between right DLPFC activation and DREAM-C receptive language and semantics. These findings suggest that effective language ability may rely on bottom-up AP in the temporal lobe and top-down replenishment mechanisms in the frontal lobe [121].

The activation degree of both frontal and temporal lobes was positively correlated with WPPSI-IV/WISC-IV: the higher the cognitive level of children was, the higher the activation degree of frontal and temporal lobes. Differences in individual cognitive levels have a greater predictive effect on the accuracy of the overall SIN [88], which may be related to the cognitive demands of behavioral tasks.

In the SIN comprehension task, the activation of the left prefrontal lobe was positively correlated with the inhibition of BRIEF-P, working memory, plan/organize, and EMI. The prefrontal cortex plays a key role in executive processes. Frontal-parietal mediated inhibitory control is crucial for recognizing speech in challenging auditory conditions [83]. As background noise increases, the ability to focus attention on target speech and suppress distractions also engages this region [83]. Response inhibition and perceptual decision-making are guided by the frontotemporal network, where perceptual information from the middle temporal lobe is converted into control information in the right inferior frontal cortex via the superior temporal sulcus, enabling response inhibition [122]. Speech comprehension under noise requires suppressing external interference and focusing on sentence meaning, involving the frontotemporal network, which aligns with our findings. Although preliminary observations suggest a possible association between brain activation patterns in children with DLD and their BRIEF-P scores, this association was not statistically significant after p-value correction, indicating that the relationship may not be robust or may be due to chance.

Regression analysis of AP characteristics and language performance

When FSIQ was not included in the model, ACC under noisy listening conditions and activation in the right STG and MTG were significantly associated with total language ability. This finding suggests that speech recognition in complex auditory environments and auditory–linguistic processing in the temporal lobe may be closely related to children’s language performance. It also indicates that clinical assessments of speech perception under noisy conditions may, to some extent, reflect children’s underlying language abilities [123]. However, after FSIQ was incorporated into the model, ACC under noisy listening conditions and activation in the right STG and MTG were no longer independently significant, whereas FSIQ became a significant predictor of total language ability. This change may be attributable to the fact that FSIQ encompasses multiple components, including verbal comprehension, working memory, processing speed, and reasoning. Therefore, using FSIQ as a predictor of total language ability may introduce construct overlap or result in overadjustment. Accordingly, these findings do not imply that AP ability is unrelated to language ability; rather, they suggest that after controlling for general cognitive ability, ACC under noisy listening conditions and activation in the right STG and MTG did not explain additional independent variance in total language ability. Future studies may conduct sensitivity analyses using nonverbal IQ scores instead of FSIQ to reduce measurement overlap between the verbal components of intelligence tests and total language ability as measured by the DREAM-C.

Associations between AP characteristics and language and cognitive performance in children with DLD

The level of activation in the left DLPFC among children with DLD showed a positive correlation with ACC under quiet listening conditions, while the level of activation in the right DLPFC and the right Wernicke’s area showed a positive correlation with d' under noisy listening conditions. This suggests that children with DLD may need to rely more heavily on DLPFC-mediated attentional control, working memory, and task monitoring, as well as supplementary involvement of the right language areas, to maintain relatively good performance on the SIN task. Future studies need to increase the sample size and incorporate multiple measures—such as task difficulty, ACC, d', and RT—to further verify whether children with DLD indeed exhibit compensatory brain activation patterns related to task performance.

No significant correlations were observed between the PAPAS scale scores and the scores on the various dimensions of the DREAM-C in the DLD group. This may be attributed to the group’s overall impaired language abilities and sample size limitations; it may also suggest that language difficulties in children with DLD are not solely attributable to AP issues, but are instead influenced by a combination of factors, including auditory perception, phonological representation, attentional control, working memory, and language experience.

Auditory decoding and auditory attention in the DLD group were positively correlated with inhibit, working memory, plan/organize, the ISCI, and the EMI. Communication skills were positively correlated with inhibit, working memory, the FI, the ISCI, and the EMI. These findings suggest that AP difficulties in children with DLD may be closely related to language information retention, behavioral control, and metacognitive regulation abilities. This is also consistent with previous evidence that children with DLD often exhibit impaired executive function, particularly deficits in verbal working memory [124]. It should be noted that both the PAPAS and the BRIEF-P rely on caregiver reports; the significant correlation between the two may be partially influenced by common-rater bias or overlap in behavioral manifestations. Therefore, the findings of this study are best interpreted as indicating a consistent association between AP-related behavioral difficulties and executive function problems, rather than as direct evidence that AP abnormalities cause executive function impairment.

The level of activation in the right Broca’s area in the DLD group was positively correlated with the total language score, but this correlation did not reach statistical significance after correction for multiple comparisons. The right Broca’s area in children with DLD may play a role in compensatory or supplementary language processing; that is, children with DLD who have relatively better language abilities may rely more heavily on resources in the right inferior frontal gyrus during speech recognition in noise [125]. However, since this result has not been corrected for multiple comparisons, it should be regarded as an exploratory finding rather than a definitive conclusion.

None of the other correlations reached statistical significance. This result may be attributed to the reduced sample size following grouping, decreased statistical power, and limited score ranges within each group; it also suggests that the significant correlations observed in the combined sample may be partially influenced by group differences.

Children with DLD experience difficulties in language expression and understanding, and are often associated with AP and cognitive behavioral abnormalities in complex environments. This study focuses on preschool-aged children, a critical period for cognitive and language development, where the early identification of AP issues can significantly impact future language abilities and cognitive behavior. Additionally, fNIRS technology, being safe, convenient, and non-invasive, offers unique advantages in detecting AP abnormalities. These considerations point to potential directions for intervention measures, such as enhancing auditory discrimination abilities and cognitive strategy training, to help these children bridge the cognitive gap in adverse listening environments.

This study has certain limitations: First, constrained by sample size and research design, we were unable to systematically assess the moderating effects of other demographic variables (such as children’s exposure to bilingualism and dialect history) on the results. Consequently, our study may have failed to control for these potential confounding factors, limiting the generalizability of the findings. Future research should hence expand sample sizes and fully account for demographic variables to further explore AP characteristics in children with DLD. Second, we employed our self-developed SIN comprehension paradigm, where the SNR was fixed at 0 dB and a four-person conversation served as the informational masking measure. Subsequent research can systematically investigate how different noise types, broader SNR ranges, and speech-related tasks influence the SIN comprehension abilities of children with DLD. Third, this study employed fNIRS data to compare group and condition differences using multiple t-tests, but this analytical strategy has shortcomings. We extracted only one beta value per ROI across all conditions, followed by single-sample t-tests for each ROI and independent-samples t-tests for group comparisons, with multiple comparisons corrected using the FDR method. This repeated t-test approach fails to detect interaction effects between task conditions and group membership. Without a mixed-design statistical model (e.g., mixed ANOVA or linear mixed-effects model), we could not directly examine whether condition effects significantly varied across groups. Despite FDR correction, multiple comparisons may still increase the risk of Type I errors (false positives). Finally, while fNIRS was used to assess AP in the cortex, it is important to investigate how subcortical networks interact with cortical language pathways. Future research could implement EEG or fMRI for multimodal studies to further explore this aspect.

Conclusion

The findings of this study suggest that preschool children with DLD may exhibit AP abnormalities, characterized by poor performance in AP tasks and atypical cortical activation. Both bottom-up AP and top-down cognitive regulation may contribute to the speech recognition difficulties DLD children face in noisy environments. Therefore, early comprehensive assessments, including AP evaluations, and targeted interventions are recommended.

Acknowledgements

The authors gratefully acknowledge the children and parents who participated in this research for their commitment and cooperation.

Clinical trial number

Not applicable.

Abbreviations

DLD: Developmental Language Disorder
TD: Typically developing
AP: Auditory processing
AAA: American Academy of Audiology
PAPAS: Preschool Auditory Processing Assessment Scale
SIN: Speech-in-Noise
ACC: Accuracy
RT: Reaction time
fNIRS: Functional Near-Infrared Spectroscopy
HbO: Oxygenated hemoglobin
HbR: Deoxy-hemoglobin
STG: Superior temporal gyrus
MTG: Middle temporal gyrus
DLPFC: Dorsolateral prefrontal cortex
EF: Executive function
WPPSI-IV: Wechsler Preschool and Primary Scale of Intelligence-Fourth Edition
WISC-Ⅳ: Wechsler Intelligence Scale for Children-The Fourth Vision
CPT: Continuous Performance Test
DREAM-C: Diagnostic Receptive and Expressive Assessment of Mandarin-Comprehensive
BRIEF-P: Behavior Rating Scale of Executive Function-Preschool Version
SNR: Signal-to-noise ratio
ROI: Region of interest
VCI: Verbal Comprehension Index
WMI: Working Memory Index
PSI: Processing Speed Index
FSIQ: Full-Scale Intelligence Quotient
ISCI: Inhibitory self-control index
FI: Flexibility index
EMI: Emergent Metacognition Index
FDR: False discovery rate
CI: Confidence interval

Authors' contributions

Wenmin Wang: Conceptualization (Leading), Methodology (Leading), Data curation (Leading), Writing–original draft (Leading). Panting Liu: Investigation (Leading), Validation (Leading), Software (Leading), Formal analysis (Leading). Bingzi Hao: Investigation (Supporting), Validation (Supporting), Data curation (Supporting), Visualization (Equal). Jia Zhou: Investigation (Supporting), Validation (Supporting), Data curation (Supporting). Bilan Chen: Investigation (Supporting), Validation (Supporting), Data curation (Supporting), Visualization (Equal). Yuying Sun: Investigation (Supporting), Validation (Supporting), Data curation (Supporting), Visualization (Equal). Xinyue Yu: Investigation (Supporting), Validation (Supporting), Visualization (Equal). Tingyu Li: Investigation (Supporting), Validation (Supporting). Yaqin Xu: Writing–review & editing (Supporting), Supervision (Supporting). Liping Meng：Writing–review & editing (Supporting), Supervision (Supporting). Lei Zhang: Writing–review & editing (Supporting), Supervision (Supporting). Jun Qian: Writing–review & editing (Supporting), Supervision (Supporting). Xia Chi: Writing–review & editing (Supporting), Supervision (Supporting), Resources (Equal). Qin Hong: Writing–review & editing (Leading), Supervision (Leading), Project administration (Leading), Funding acquisition (Leading), Resources (Equal).

Funding

This research was supported by the Nanjing International/Hong Kong, Macao and Taiwan Science and Technology Cooperation Program Project [grant numbers 202401037]; the Nanjing Health Science and Technology Development Special Fund Project [grant numbers YKK24151, YKK23152]; the Jiangsu Province Science and Technology Program Special Fund (Key R&D Program for Social Development) [BE2023666]; the Maternal and Child Health Association of Jiangsu Province [grant numbers FYX202340]; and the General Project of Science and Technology Development Fund of Nanjing Medical University [grant numbers NMUB20220102].

Data availability

The data and materials supporting the findings of this study are available from the corresponding author upon reasonable request.

Declarations

Ethics approval and consent to participate

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.

Footnotes

Publisher’s note

Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.

Wenmin Wang and Panting Liu are co-first authors.

Xia Chi and Qin Hong are co-corresponding authors.

Contributor Information

Xia Chi, Email: chixia2001@njum.edu.cn.

Qin Hong, Email: rambler_hq@163.com.

References

1.Bishop DVM, Snowling MJ, Thompson PA, Greenhalgh T. Phase 2 of CATALISE: a multinational and multidisciplinary Delphi consensus study of problems with language development: Terminology. Journal child Psychol psychiatry allied disciplines. 2017;58(10):1068–80. 10.1111/jcpp.12721. [DOI] [PMC free article] [PubMed] [Google Scholar]
2.The L. ICD-11. Lancet (London, England). 2019;8(10188):2275. 10.1016/s0140-6736(19)31205-x. [DOI] [PubMed]
3.Wu S, Zhao J, de Villiers J, et al. Prevalence, co-occurring difficulties, and risk factors of developmental language disorder: first evidence for Mandarin-speaking children in a population-based study. The Lancet Reg health Western Pacific. 2023;34:100713. 10.1016/j.lanwpc.2023.100713. [DOI] [PMC free article] [PubMed] [Google Scholar]
4.Andreou G, Aslanoglou V. Written Language Production in Children With Developmental Language Disorder. Front Psychol. 2022;13:833429. 10.3389/fpsyg.2022.833429. [DOI] [PMC free article] [PubMed] [Google Scholar]
5.Graham S, Hebert M, Fishman E, Ray AB, Rouse AG. Do Children Classified With Specific Language Impairment Have a Learning Disability in Writing? A Meta-Analysis. Journal of learning disabilities. 2020;53(4):292–310. 10.1177/0022219420917338. [DOI] [PubMed]
6.Young AR, Beitchman JH, Johnson C et al. Jul. Young adult academic outcomes in a longitudinal sample of early identified language impaired and control children. Journal of child psychology and psychiatry, and allied disciplines. 2002;43(5):635 – 45. 10.1111/1469-7610.00052. [DOI] [PubMed]
7. ASHA. Central auditory processing disorder. Accessed 2025-08-18, https://www.asha.org/Practice-Portal/Clinical-Topics/Central-Auditory-Processing-Disorder/.
8.Protopapas A. From temporal processing to developmental language disorders: mind the gap. Philosophical Trans Royal Soc Lond Ser B Biol Sci. 2014;369(1634):20130090. 10.1098/rstb.2013.0090. [DOI] [PMC free article] [PubMed] [Google Scholar]
9. Goswami U, Cumming R, Chait M, et al. Perception of Filtered Speech by Children with Developmental Dyslexia and Children with Specific Language Impairments. Front Psychol. 2016;7:791. 10.3389/fpsyg.2016.00791. [DOI] [PMC free article] [PubMed] [Google Scholar]
10.Corriveau K, Pasquini E, Goswami U. Jun. Basic auditory processing skills and specific language impairment: a new look at an old hypothesis. Journal of speech, language, and hearing research: JSLHR. 2007;50(3):647 – 66. 10.1044/1092-4388(2007/046). [DOI] [PubMed]
11.Chinn LK, Zhukova MA, Kroeger RJ, Ledesma LM, Cavitt JE, Grigorenko EL. Auditory brainstem response deficits in learning disorders and developmental language disorder: a systematic review and meta-analysis. Scientific reports. 2022;22(1):20124. 10.1038/s41598-022-20438-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
12.Elmahallawi TH, Gabr TA, Darwish ME, Seleem FM. Children with developmental language disorder: a frequency following response in the noise study. Brazilian J otorhinolaryngology. 2022;88(6):954–61. 10.1016/j.bjorl.2021.01.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
13.Griffiths TD. Mar. Human complex sound analysis. Clinical science (London, England: 1979). 1999;96(3):231-4. [PubMed]
14.Shield BM, Dockrell JE. The effects of environmental and classroom noise on the academic attainments of primary school children. The J Acoust Soc America. 2008;123(1):133–44. 10.1121/1.2812596. [DOI] [PubMed] [Google Scholar]
15.Magimairaj BM, Nagaraj NK, Champlin CA, Thibodeau LK, Loeb DF, Gillam RB. Speech Perception in Noise Predicts Oral Narrative Comprehension in Children With Developmental Language Disorder. Front Psychol. 2021;12:735026. 10.3389/fpsyg.2021.735026. [DOI] [PMC free article] [PubMed] [Google Scholar]
16.Ziegler JC, Pech-Georgel C, George F, Alario FX, Lorenzi C. Deficits in speech perception predict language learning impairment. Proceedings Natl Acad Sci United States America. 2005;27(39):14110–5. 10.1073/pnas.0504446102. [DOI] [PMC free article] [PubMed] [Google Scholar]
17.Ziegler JC, Pech-Georgel C, George F, Lorenzi C. Noise on, voicing off: Speech perception deficits in children with specific language impairment. Journal experimental child psychology. 2011;110(3):362–72. 10.1016/j.jecp.2011.05.001. [DOI] [PubMed] [Google Scholar]
18.Stollman MH, van Velzen EC, Simkens HM, Snik AF, van den Broek P. Assessment of auditory processing in 6-year-old language-impaired children. International J audiology. 2003;42(6):303–11. 10.3109/14992020309101322. [DOI] [PubMed] [Google Scholar]
19.Vandewalle E, Boets B, Ghesquière P, Zink I. Auditory processing and speech perception in children with specific language impairment: relations with oral language and literacy skills. Research in developmental disabilities. 2012;33(2):635–44. 10.1016/j.ridd.2011.11.005. [DOI] [PubMed]
20.Gordon KR, Lewis D, Lowry S, Smith M, Stecker GC, McCreery RW. Remote Microphones Support Speech Recognition in Noise and Reverberation for Children With a Language Disorder. language, speech, and hearing services in schools. 2025;9(1):225–33. 10.1044/2024_lshss-24-00018. [DOI] [PMC free article] [PubMed]
21.Ellis Weismer S, Evans J, Hesketh LJ. Oct. An examination of verbal working memory capacity in children with specific language impairment. Journal of speech, language, and hearing research: JSLHR. 1999;42(5):1249-60. 10.1044/jslhr.4205.1249. [DOI] [PubMed]
22.Cheng YY, Wu HC, Shih HY, Yeh PW, Yen HL, Lee CY. Deficits in Processing of Lexical Tones in Mandarin-Speaking Children With Developmental Language Disorder: Electrophysiological Evidence. Journal of speech, language, and hearing research: JSLHR. 2021;64(4):1176–1188. 10.1044/2021_jslhr-19-00392. [DOI] [PubMed]
23.Xu Y, Sun X. Maximum speed of pitch change and how it may relate to speech. The J Acoust Soc America. 2002;111(3):1399–413. 10.1121/1.1445789. [DOI] [PubMed] [Google Scholar]
24.Krishnan A, Xu Y, Gandour J, Cariani P. Encoding of pitch in the human brainstem is sensitive to language experience. Brain Res Cogn brain research. 2005;25(1):161–8. 10.1016/j.cogbrainres.2005.05.004. [DOI] [PubMed] [Google Scholar]
25.Rocha-Muniz CN, Befi-Lopes DM, Schochat E. Sensitivity, specificity and efficiency of speech-evoked ABR. Hearing research. 2014;317:15–22. 10.1016/j.heares.2014.09.004. [DOI] [PubMed] [Google Scholar]
26.QIN WH. Clinical Experience in Auditory Brainstem Response Testing and Care for Children with Language Disorders. J Chengde Med Univ. 2012;29(3):283–283. 10.15921/j.cnki.cyxb.2012.03.036. [DOI] [Google Scholar]
27.Audiology AAo. Clinical Practice Guidelines: Diagnosis, Treatment, and Management of Children and Adults with Central Auditory Processing Disorder. Accessed 18 Nov 2025. https://www.audiology.org/practice-guideline/clinical-practice-guidelines-diagnosis-treatment-and-management-of-children-and-adults-with-central-auditory-processing-disorder/.
28.Liu P, Lin H, Xiao Z, et al. The development, validity, reliability, and norm of a preschool auditory processing assessment scale in China. Research Dev disabilities. 2022;128:104272. 10.1016/j.ridd.2022.104272. [DOI] [PubMed] [Google Scholar]
29.Sussman ES. Auditory Scene Analysis: An Attention Perspective. Journal of speech, language, and hearing research: JSLHR. 2017;17(10):2989–3000. 10.1044/2017_jslhr-h-17-0041. [DOI] [PMC free article] [PubMed]
30.Guo Y, Li Y, Liu F, et al. Association between neural prosody discrimination and language abilities in toddlers: a functional near-infrared spectroscopy study. BMC pediatrics. 2024;12(1):449. 10.1186/s12887-024-04889-7. [DOI] [PMC free article] [PubMed] [Google Scholar]
31.Rahman MA, Siddik AB, Ghosh TK, Khanam F, Ahmad M. A Narrative Review on Clinical Applications of fNIRS. Journal Digit imaging. 2020;33(5):1167–84. 10.1007/s10278-020-00387-1. [DOI] [PMC free article] [PubMed] [Google Scholar]
32.Pinti P, Tachtsidis I, Hamilton A, et al. The present and future use of functional near-infrared spectroscopy (fNIRS) for cognitive neuroscience. Annals New York Acad Sciences. 2020;1464(1):5–29. 10.1111/nyas.13948. [DOI] [PMC free article] [PubMed] [Google Scholar]
33.Liberman AM, Cooper FS, Shankweiler DP, Studdert-Kennedy M. Perception of the speech code. Psychological review. 1967;74(6):431–61. 10.1037/h0020279. [DOI] [PubMed] [Google Scholar]
34.Galantucci B, Fowler CA, Turvey MT. The motor theory of speech perception reviewed. Psychonomic Bull & review. 2006;13(3):361–77. 10.3758/bf03193857. [DOI] [PMC free article] [PubMed] [Google Scholar]
35.Cogan GB, Thesen T, Carlson C, Doyle W, Devinsky O, Pesaran B. Sensory-motor transformations for speech occur bilaterally. Nature. 2014;6(7490):94–8. 10.1038/nature12935. [DOI] [PMC free article] [PubMed] [Google Scholar]
36. Geschwind N. The organization of language and the brain. Science (New York, NY). 1970;170(3961):940-4. 10.1126/science.170.3961.940. [DOI] [PubMed]
37.Skeide MA, Friederici AD. The ontogeny of the cortical language network. Nature reviews Neuroscience. 2016;17(5):323–32. 10.1038/nrn.2016.23. [DOI] [PubMed] [Google Scholar]
38.Hertrich I, Dietrich S, Ackermann H. The role of the supplementary motor area for speech and language processing. Neuroscience biobehavioral reviews. 2016;68:602–10. 10.1016/j.neubiorev.2016.06.030. [DOI] [PubMed] [Google Scholar]
39.Bornkessel-Schlesewsky I, Schlesewsky M, Small SL, Rauschecker JP. Neurobiological roots of language in primate audition: common computational properties. Trends Cogn sciences. 2015;19(3):142–50. 10.1016/j.tics.2014.12.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
40.Poeppel D, Emmorey K, Hickok G, Pylkkänen L. Towards a new neurobiology of language. The J neuroscience: official J Soc Neuroscience. 2012;10(41):14125–31. 10.1523/jneurosci.3244-12.2012. [DOI] [PMC free article] [PubMed] [Google Scholar]
41.DeWitt I, Rauschecker JP. Phoneme and word recognition in the auditory ventral stream. Proceedings Natl Acad Sci United States America. 2012;21(8):E505–14. 10.1073/pnas.1113427109. [DOI] [PMC free article] [PubMed] [Google Scholar]
42.Dick AS, Bernal B, Tremblay P. The language connectome: new pathways, new concepts. The Neuroscientist: Rev J bringing Neurobiol Neurol psychiatry. 2014;20(5):453–67. 10.1177/1073858413513502. [DOI] [PubMed] [Google Scholar]
43.Scott SK. From speech and talkers to the social world: The neural processing of human spoken language. Science (New York, NY). 2019;366(6461):58–62. 10.1126/science.aax0288. [DOI] [PMC free article] [PubMed]
44.Bhaya-Grossman I, Chang EF. Speech Computations of the Human Superior Temporal Gyrus. Annual Rev psychology. 2022;4:73:79–102. 10.1146/annurev-psych-022321-035256. [DOI] [PMC free article] [PubMed] [Google Scholar]
45.Yi HG, Leonard MK, Chang EF. The Encoding of Speech Sounds in the Superior Temporal Gyrus. Neuron. 2019;19(6):1096–110. 10.1016/j.neuron.2019.04.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
46.Petrides M. On the evolution of polysensory superior temporal sulcus and middle temporal gyrus: A key component of the semantic system in the human brain. The J Comp neurology. 2023;531(18):1987–95. 10.1002/cne.25521. [DOI] [PubMed] [Google Scholar]
47.Mylius V, Ayache SS, Ahdab R, et al. Definition of DLPFC and M1 according to anatomical landmarks for navigated brain stimulation: inter-rater reliability, accuracy, and influence of gender and age. NeuroImage. 2013;78:224–32. 10.1016/j.neuroimage.2013.03.061. [DOI] [PubMed] [Google Scholar]
48.Hertrich I, Dietrich S, Blum C, Ackermann H. The Role of the Dorsolateral Prefrontal Cortex for Speech and Language Processing. Front Hum Neurosci. 2021;15:645209. 10.3389/fnhum.2021.645209. [DOI] [PMC free article] [PubMed] [Google Scholar]
49.Hahn M, Futrell R, Levy R, Gibson E. A resource-rational model of human processing of recursive linguistic structure. Proceedings Natl Acad Sci United States America. 2022;25(43):e2122602119. 10.1073/pnas.2122602119. [DOI] [PMC free article] [PubMed] [Google Scholar]
50.Fedorenko E, Shain C. Similarity of computations across domains does not imply shared implementation: The case of language comprehension. Current Dir Psychol science. 2021;30(6):526–34. 10.1177/09637214211046955. [DOI] [PMC free article] [PubMed] [Google Scholar]
51.Rönnberg J, Lunner T, Zekveld A, et al. The Ease of Language Understanding (ELU) model: theoretical, empirical, and clinical advances. Front Syst Neurosci. 2013;7:31. 10.3389/fnsys.2013.00031. [DOI] [PMC free article] [PubMed] [Google Scholar]
52.McCreery RW, Miller MK, Buss E, Leibold LJ. Cognitive and Linguistic Contributions to Masked Speech Recognition in Children. Journal of speech, language, and hearing research: JSLHR. 2020;63(10):3525–3538. 10.1044/2020_jslhr-20-00030. [DOI] [PMC free article] [PubMed]
53.Acosta RodríguezV, Ramírez Santana GM, Hernández Expósito S. Executive functions and language in children with different subtypes of specific language impairment. Neurologia (Barcelona, Spain). 2017;32(6):355–362. Funciones ejecutivas y lenguaje en subtipos de niños con trastorno específico del lenguaje. 10.1016/j.nrl.2015.12.018. [DOI] [PubMed]
54.Vermeij BAM, Wiefferink CH, Scholte RHJ, Knoors H. Language development and behaviour problems in toddlers indicated to have a developmental language disorder. International J Lang & communication disorders. 2021;56(6):1249–62. 10.1111/1460-6984.12665. [DOI] [PubMed] [Google Scholar]
55.Liu XL, de Villiers J, Ning C et al. Research to Establish the Validity, Reliability, and Clinical Utility of a Comprehensive Language Assessment of Mandarin. Journal of speech, language, and hearing research: JSLHR. 2017;60(3):592–606. 10.1044/2016_jslhr-l-15-0334. [DOI] [PubMed]
56.Fedtke T, Bug MU. Equivalent hearing threshold levels for the RadioEar IP30 insert earphone and short-term stimuli: comparison of peak-equivalent and RMS-based measures. International J audiology. 2020;59(5):398–402. 10.1080/14992027.2019.1682690. [DOI] [PubMed] [Google Scholar]
57.Avivi-Reich M, Roberts MY, Grieco-Calub TM. Quantifying the Effects of Background Speech Babble on Preschool Children’s Novel Word Learning in a Multisession Paradigm: A Preliminary Study. Journal of speech, language, and hearing research: JSLHR. 2020;63(1):345–356. 10.1044/2019_jslhr-h-19-0083. [DOI] [PubMed]
58.Tak S, Ye JC. Statistical analysis of fNIRS data: a comprehensive review. NeuroImage. 2014;85 Pt 1:72–91. 10.1016/j.neuroimage.2013.06.016. [DOI] [PubMed]
59.Sun PP, Tan FL, Zhang Z, Jiang YH, Zhao Y, Zhu CZ. Feasibility of Functional Near-Infrared Spectroscopy (fNIRS) to Investigate the Mirror Neuron System: An Experimental Study in a Real-Life Situation. Front Hum Neurosci. 2018;12:86. 10.3389/fnhum.2018.00086. [DOI] [PMC free article] [PubMed] [Google Scholar]
60.Tsuzuki D, Dan I. Spatial registration for functional near-infrared spectroscopy: from channel position on the scalp to cortical location in individual and group analyses. NeuroImage. 2014;85 Pt 1:92–103. 10.1016/j.neuroimage.2013.07.025. [DOI] [PubMed]
61.Yanagisawa H, Dan I, Tsuzuki D et al. Acute moderate exercise elicits increased dorsolateral prefrontal activation and improves cognitive performance with Stroop test. NeuroImage. 2010;50(4):1702-10. 10.1016/j.neuroimage.2009.12.023. [DOI] [PubMed]
62. Reverte IG, Favez P, Rossier N, Lecerf J. Structural validity of the Wechsler Intelligence Scale for Children (WISC-IV) in a French-speaking Swiss sample. Learn Individual Differences. 2014;29:114–9. 10.1016/j.lindif.2013.10.013. [DOI] [Google Scholar]
63.Ye JC, Tak S, Jang KE, Jung J, Jang J. NIRS-SPM: statistical parametric mapping for near-infrared spectroscopy. NeuroImage. 2009;15(2):428–47. 10.1016/j.neuroimage.2008.08.036. [DOI] [PubMed] [Google Scholar]
64.Jang KE, Tak S, Jung J, Jang J, Jeong Y, Ye JC. Wavelet minimum description length detrending for near-infrared spectroscopy. Journal biomedical optics. 2009;14(3):034004. 10.1117/1.3127204. [DOI] [PubMed] [Google Scholar]
65.Strangman G, Culver JP, Thompson JH, Boas DA. A quantitative comparison of simultaneous BOLD fMRI and NIRS recordings during functional brain activation. NeuroImage. 2002;17(2):719–31. [PubMed] [Google Scholar]
66.Whitehouse AJ, Shelton WM, Ing C, Newnham JP. Aug. Prenatal, perinatal, and neonatal risk factors for specific language impairment: a prospective pregnancy cohort study. Journal of speech, language, and hearing research: JSLHR. 2014;57(4):1418-27. 10.1044/2014_jslhr-l-13-0186. [DOI] [PubMed]
67.Collisson BA, Graham SA, Preston JL, Rose MS, McDonald S, Tough S. Risk and Protective Factors for Late Talking: An Epidemiologic Investigation. The J pediatrics. 2016;172:168–e1741. 10.1016/j.jpeds.2016.02.020. [DOI] [PubMed] [Google Scholar]
68. Media, Minds Y. Pediatrics. 2016;138(5). 10.1542/peds.2016-2591. [DOI] [PubMed]
69.Kim KM, Chung US. Associations among exposure to television or video, language development, and school achievement in childhood: a prospective birth cohort study. Social psychiatry psychiatric epidemiology. 2021;56(5):847–56. 10.1007/s00127-020-01967-w. [DOI] [PubMed] [Google Scholar]
70.Sultana N, Wong LLN, Purdy SC. Natural Language Input: Maternal Education, Socioeconomic Deprivation, and Language Outcomes in Typically Developing Children. Language, speech, and hearing services in schools. 2020;51(4):1049–1070. 10.1044/2020_lshss-19-00095. [DOI] [PubMed]
71.Parks KMA, Hannah KE, Moreau CN, Brainin L, Joanisse MF. Language abilities in children and adolescents with DLD and ADHD: A scoping review. Journal communication disorders. 2023;106:106381. 10.1016/j.jcomdis.2023.106381. [DOI] [PubMed] [Google Scholar]
72.Jones SD, Westermann G. Predictive Processing and Developmental Language Disorder. Journal of speech, language, and hearing research: JSLHR. 2021;14(1):181–5. 10.1044/2020_jslhr-20-00409. [DOI] [PubMed]
73.Acosta-Rodríguez VM, Ramírez-Santana GM, Hernández-Expósito S. Intervention for oral language comprehension skills in preschoolers with developmental language disorder. International J Lang & communication disorders. 2022;57(1):90–102. 10.1111/1460-6984.12676. [DOI] [PubMed] [Google Scholar]
74.Jackson E, Leitão S, Claessen M, Boyes M. Working, Declarative, and Procedural Memory in Children With Developmental Language Disorder. Journal of speech, language, and hearing research: JSLHR. 2020;14(12):4162–78. 10.1044/2020_jslhr-20-00135. [DOI] [PubMed]
75. Delage H, Frauenfelder UH. Relationship between working memory and complex syntax in children with Developmental Language Disorder. Journal child language. 2020;47(3):600–32. 10.1017/s0305000919000722. [DOI] [PubMed] [Google Scholar]
76.Zebib R, Tuller L, Hamann C, Ibrahim LA, Prévost P. Syntactic complexity and verbal working memory in bilingual children with and without Developmental Language Disorder. J First Lang. 2020;40(4):461–84. 10.1177/0142723719888372. [DOI] [Google Scholar]
77.Park JS, Miller CA, Sanjeevan T, van Hell JG, Weiss DJ, Mainela-Arnold E. Bilingualism and Processing Speed in Typically Developing Children and Children With Developmental Language Disorder. Journal of speech, language, and hearing research: JSLHR. 2020;63(5):1479–1493. 10.1044/2020_jslhr-19-00403. [DOI] [PubMed]
78.Ebert KD. Revisiting the Influences of Bilingualism and Developmental Language Disorder on Children’s Nonverbal Processing Speed. Journal of speech, language, and hearing research: JSLHR. 2021;14(9):3564–70. 10.1044/2021_jslhr-21-00156. [DOI] [PMC free article] [PubMed]
79.Kapa LL, Plante E. Executive Function in SLI: Recent Advances and Future Directions. Current Dev disorders reports. 2015;2(3):245–52. 10.1007/s40474-015-0050-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
80.Miyake A, Friedman NP, Emerson MJ, Witzki AH, Howerter A, Wager TD. The unity and diversity of executive functions and their contributions to complex Frontal Lobe tasks: a latent variable analysis. Cognitive psychology. 2000;41(1):49–100. 10.1006/cogp.1999.0734. [DOI] [PubMed] [Google Scholar]
81.Wittke K, Spaulding TJ, Schechtman CJ. Specific language impairment and executive functioning: parent and teacher ratings of behavior. American J speech-language pathology. 2013;22(2):161–72. 10.1044/1058-0360(2012/11-0052). [DOI] [PubMed] [Google Scholar]
82.Vissers C, Koolen S, Hermans D, Scheper A, Knoors H. Executive functioning in preschoolers with specific language impairment. Front Psychol. 2015;6:1574. 10.3389/fpsyg.2015.01574. [DOI] [PMC free article] [PubMed] [Google Scholar]
83.Katsos N, Roqueta CA, Estevan RA, Cummins C. Are children with Specific Language Impairment competent with the pragmatics and logic of quantification? Cognition. 2011;119(1):43–57. 10.1016/j.cognition.2010.12.004. [DOI] [PubMed] [Google Scholar]
84.Schuchardt K, Bockmann A-K, Bornemann G, Maehler C. Working memory functioning in children with learning disorders and specific language impairment. Top Lang Disorders. 2013;33(4):298–312. 10.1097/01.TLD.0000437943.41140.36. [DOI] [Google Scholar]
85.Kujala T, Leminen M. Low-level neural auditory discrimination dysfunctions in specific language impairment-A review on mismatch negativity findings. Developmental Cogn neuroscience. 2017;28:65–75. 10.1016/j.dcn.2017.10.005. [DOI] [PMC free article] [PubMed] [Google Scholar]
86.Crandell CC. Individual differences in speech recognition ability: implications for hearing aid selection. Ear hearing. 1991;12(6 Suppl):s100–8. 10.1097/00003446-199112001-00003. [DOI] [PubMed] [Google Scholar]
87.Skoe E, Karayanidi K. Bilingualism and Speech Understanding in Noise: Auditory and Linguistic Factors. Journal Am Acad Audiology. 2019;30(2):115–30. 10.3766/jaaa.17082. [DOI] [PubMed] [Google Scholar]
88.Blomberg R, Danielsson H, Rudner M, Söderlund GBW, Rönnberg J. Speech Processing Difficulties in Attention Deficit Hyperactivity Disorder. Front Psychol. 2019;10:1536. 10.3389/fpsyg.2019.01536. [DOI] [PMC free article] [PubMed] [Google Scholar]
89.Finneran DA, Francis AL, Leonard LB. Sustained attention in children with specific language impairment (SLI). Journal of speech, language, and hearing research: JSLHR. 2009;52(4):915 – 29. 10.1044/1092-4388(2009/07-0053 ). [DOI] [PMC free article] [PubMed]
90.ETHGTADMESF M. Specific language impairment: speech perception in relation to noise. Egypt J Otolaryngol. 2021. 10.1186/s43163-021-00115-8. [DOI] [Google Scholar]
91.Binder JR, Desai RH, Graves WW, Conant LL. Where is the semantic system? A critical review and meta-analysis of 120 functional neuroimaging studies. Cerebral cortex (New York, NY: 1991). 2009;19(12):2767-96. 10.1093/cercor/bhp055. [DOI] [PMC free article] [PubMed]
92. Matchin W, Hickok G. The Cortical Organization of Syntax. Cerebral cortex (New York, NY: 1991). 2020;30(3):1481–1498. 10.1093/cercor/bhz180. [DOI] [PMC free article] [PubMed]
93.Fedorenko E, Thompson-Schill SL. Reworking the language network. Trends Cogn sciences. 2014;18(3):120–6. 10.1016/j.tics.2013.12.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
94.Fridriksson J, Yourganov G, Bonilha L, Basilakos A, Den Ouden DB, Rorden C. Revealing the dual streams of speech processing. Proceedings Natl Acad Sci United States America. 2016;27(52):15108–13. 10.1073/pnas.1614038114. [DOI] [PMC free article] [PubMed] [Google Scholar]
95.Hesling I, Labache L, Joliot M, Tzourio-Mazoyer N. Large-scale plurimodal networks common to listening to, producing and reading word lists: an fMRI study combining task-induced activation and intrinsic connectivity in 144 right-handers. Brain Struct & function. 2019;224(9):3075–94. 10.1007/s00429-019-01951-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
96.Zhao Z, Liu Y, Zhang J, Lu J, Wu J. Where is the speech production area? Evidence from direct cortical electrical stimulation mapping. Brain: J neurology. 2021;17(7):e61. 10.1093/brain/awab178. [DOI] [PMC free article] [PubMed] [Google Scholar]
97.Zhang N, Yuan B, Yan J, Cheng J, Lu J, Wu J. Multivariate machine learning-based language mapping in glioma patients based on lesion topography. Brain imaging behavio. 2021;15(5):2552–62. 10.1007/s11682-021-00457-0. [DOI] [PubMed] [Google Scholar]
98.Yuan B, Xie H, Wang Z, et al. The domain-separation language network dynamics in resting state support its flexible functional segregation and integration during language and speech processing. NeuroImage. 2023;274:120132. 10.1016/j.neuroimage.2023.120132. [DOI] [PubMed] [Google Scholar]
99.Lu J, Zhao Z, Zhang J, et al. Functional maps of direct electrical stimulation-induced speech arrest and anomia: a multicentre retrospective study. Brain: J neurology. 2021;4(8):2541–53. 10.1093/brain/awab125. [DOI] [PMC free article] [PubMed] [Google Scholar]
100.Sefcikova V, Sporrer JK, Juvekar P, Golby A, Samandouras G. Converting sounds to meaning with ventral semantic language networks: integration of interdisciplinary data on brain connectivity, direct electrical stimulation and clinical disconnection syndromes. Brain Struct & function. 2022;227(5):1545–64. 10.1007/s00429-021-02438-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
101.Hodgson VJ, Lambon Ralph MA, Jackson RL. Multiple dimensions underlying the functional organization of the language network. NeuroImage. 2021;241:118444. 10.1016/j.neuroimage.2021.118444. [DOI] [PMC free article] [PubMed] [Google Scholar]
102.Adank P, Davis MH, Hagoort P. Neural dissociation in processing noise and accent in spoken language comprehension. Neuropsychologia. 2012;50(1):77–84. 10.1016/j.neuropsychologia.2011.10.024. [DOI] [PubMed] [Google Scholar]
103.Michalek AMP, Ash I, Schwartz K. The independence of working memory capacity and audiovisual cues when listening in noise. Scandinavian J psychology. 2018;59(6):578–85. 10.1111/sjop.12480. [DOI] [PubMed] [Google Scholar]
104.Bronkhorst AW, Plomp R. Effect of multiple speechlike maskers on binaural speech recognition in normal and impaired hearing. The J Acoust Soc America. 1992;92(6):3132–9. 10.1121/1.404209. [DOI] [PubMed] [Google Scholar]
105.Narayan R, Best V, Ozmeral E, et al. Cortical interference effects in the cocktail party problem. Nature neuroscience. 2007;10(12):1601–7. 10.1038/nn2009. [DOI] [PubMed] [Google Scholar]
106.Moore DR, Edmondson-Jones M, Dawes P, et al. Relation between speech-in-noise threshold, hearing loss and cognition from 40–69 years of age. PLoS ONE. 2014;9(9):e107720. 10.1371/journal.pone.0107720. [DOI] [PMC free article] [PubMed] [Google Scholar]
107.Song F, Zhan Y, Ford JC, et al. Increased Right Frontal Brain Activity During the Mandarin Hearing-in-Noise Test. Front NeuroSci. 2020;14:614012. 10.3389/fnins.2020.614012. [DOI] [PMC free article] [PubMed] [Google Scholar]
108.Badcock NA, Bishop DV, Hardiman MJ, Barry JG, Watkins KE. Co-localisation of abnormal brain structure and function in specific language impairment. Brain language. 2012;120(3):310–20. 10.1016/j.bandl.2011.10.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
109.Hancock AS, Warren CM, Barrett TS, Bolton DAE, Gillam RB. Functional near-infrared spectroscopy measures of neural activity in children with and without developmental language disorder during a working memory task. Brain behavior. 2023;13(2):e2895. 10.1002/brb3.2895. [DOI] [PMC free article] [PubMed] [Google Scholar]
110.Lamichhane B, Westbrook A, Cole MW, Braver TS. Exploring brain-behavior relationships in the N-back task. NeuroImage. 2020;15:212:116683. 10.1016/j.neuroimage.2020.116683. [DOI] [PMC free article] [PubMed] [Google Scholar]
111.Rowe JB, Toni I, Josephs O, Frackowiak RS, Passingham RE. The prefrontal cortex: response selection or maintenance within working memory? Science (New York, NY). 2000;288(5471):1656-60. 10.1126/science.288.5471.1656. [DOI] [PubMed]
112.Hagoort P. On Broca, brain, and binding: a new framework. Trends Cogn sciences. 2005;9(9):416–23. 10.1016/j.tics.2005.07.004. [DOI] [PubMed] [Google Scholar]
113.Xiang HD, Fonteijn HM, Norris DG, Hagoort P. Topographical functional connectivity pattern in the perisylvian language networks. Cerebral cortex (New York, NY: 1991). 2010;20(3):549 – 60. 10.1093/cercor/bhp119. [DOI] [PubMed]
114.Price CJ. A review and synthesis of the first 20 years of PET and fMRI studies of heard speech, spoken language and reading. NeuroImage. 2012;15(2):816–47. 10.1016/j.neuroimage.2012.04.062. [DOI] [PMC free article] [PubMed] [Google Scholar]
115.Rowland SC, Hartley DEH, Wiggins IM. Listening in Naturalistic Scenes: What Can Functional Near-Infrared Spectroscopy and Intersubject Correlation Analysis Tell Us About the Underlying Brain Activity? Trends in hearing. 2018;22:2331216518804116. 10.1177/2331216518804116. [DOI] [PMC free article] [PubMed]
116.Pichora-Fuller MK, Kramer SE, Eckert MA et al. Hearing Impairment and Cognitive Energy: The Framework for Understanding Effortful Listening (FUEL). Ear and hearing. 2016;37 Suppl 1:5s-27s. 10.1097/aud.0000000000000312. [DOI] [PubMed]
117.Benasich AA, Tallal P. Infant discrimination of rapid auditory cues predicts later language impairment. Behavioural brain research. 2002;17(1):31–49. 10.1016/s0166-4328(02)00098-0. [DOI] [PubMed] [Google Scholar]
118.Maggu AR, Sharma B. Relationship Between Cognitive Abilities and Basic Auditory Processing in Young Adults. American J audiology. 2024;4(2):422–32. 10.1044/2024_aja-23-00261. [DOI] [PubMed] [Google Scholar]
119.Brenneman L, Cash E, Chermak GD, et al. The Relationship between Central Auditory Processing, Language, and Cognition in Children Being Evaluated for Central Auditory Processing Disorder. Journal Am Acad Audiology. 2017;28(8):758–69. 10.3766/jaaa.16119. [DOI] [PubMed] [Google Scholar]
120.Back NCF, Crippa ACS, Riechi T, Pereira LD. Central Auditory Processing and Cognitive Functions in Children. International archives otorhinolaryngology. 2022;26(1):e020–31. 10.1055/s-0040-1722158. [DOI] [PMC free article] [PubMed] [Google Scholar]
121.Zekveld AA, Heslenfeld DJ, Festen JM, Schoonhoven R. Top-down and bottom-up processes in speech comprehension. NeuroImage. 2006;32(4):1826-36. 10.1016/j.neuroimage.2006.04.199. [DOI] [PubMed]
122. Tsumura K, Shintaki R, Takeda M, Chikazoe J, Nakahara K, Jimura K. Perceptual Uncertainty Alternates Top-down and Bottom-up Fronto-Temporal Network Signaling during Response Inhibition. The J neuroscience: official J Soc Neuroscience. 2022;1(22):4567–79. 10.1523/jneurosci.2537-21.2022. [DOI] [PMC free article] [PubMed] [Google Scholar]
123.McCreery RW, Spratford M, Kirby B, Brennan M. Individual differences in language and working memory affect children’s speech recognition in noise. International J audiology. 2017;56(5):306–15. 10.1080/14992027.2016.1266703. [DOI] [PMC free article] [PubMed] [Google Scholar]
124.Niu T, Wang S, Ma J, Zeng X, Xue R. Executive functions in children with developmental language disorder: a systematic review and meta-analysis. Front NeuroSci. 2024;18:1390987. 10.3389/fnins.2024.1390987. [DOI] [PMC free article] [PubMed] [Google Scholar]
125.Abbott N, Love T. Bridging the Divide: Brain and Behavior in Developmental Language Disorder. Brain sciences. 2023;19(11). 10.3390/brainsci13111606. [DOI] [PMC free article] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data and materials supporting the findings of this study are available from the corresponding author upon reasonable request.

PERMALINK

Exploring the relationship between auditory processing characteristics and cognition in preschool children with developmental language disorder based on fNIRS

Wenmin Wang

Panting Liu

Bingzi Hao

Jia Zhou

Bilan Chen

Yuying Sun

Xinyue Yu

Tingyu Li

Yaqin Xu

Liping Meng

Lei Zhang

Jun Qian

Xia Chi

Qin Hong

Abstract

Background

Methods

Results

Conclusions

Introduction

Methods

Participants

Materials and procedure

Basic demographic information

Pure Tone Audiometry (PTA)

DREAM-C

Subjective AP assessment: PAPAS

Objective AP assessment: SIN comprehension

Fig. 1.

Acquisition of NIRS data

Fig. 2.

Behavioral data

Cognitive behavioral assessment

WPPSI-IV/WISC-IV

BRIEF-P

Preprocessing of fNIRS data

Statistical analysis

Visualization

Results

Comparison of basic demographic information

Table 1.

Language and related cognitive performance in children with DLD

Fig. 3.

AP characteristics in children with DLD

Behavioral performance of AP in children with DLD

Results of the PAPAS scale

Behavioral performance on the SIN comprehension task

Table 2.

Table 3.

Table 4.

Table 5.

Table 6.

Table 7.

Brain activation characteristics during the SIN comprehension task in children with DLD

Table 8.

Fig. 4.

Table 9.

Fig. 5.

Correlations between behavioral performance on the SIN comprehension task and ROI activation levels

Fig. 7.

Associations between AP characteristics and language and cognitive performance in the two groups

Correlations between PAPAS scores and language and cognitive behavioral performance

Fig. 6.

Correlations between ROI activation levels during the SIN comprehension task and language and cognitive behavioral performance

Regression analysis of AP characteristics and language performance

Table 10.

Table 11.

Associations between AP characteristics and language and cognitive performance in children with DLD

Table 12.

Discussion

Comparison of basic demographic information

Language and related cognitive performance in children with DLD

AP characteristics in children with DLD

Behavioral performance of AP in children with DLD

Results of the PAPAS scale

Behavioral performance on the SIN comprehension task

Brain activation characteristics during the SIN comprehension task in children with DLD

Correlations between behavioral performance on the SIN comprehension task and ROI activation levels