Functional connectivity of the sensory system and executive functions during a story listening task is related to parent-child interaction during joint reading: a functional MRI-diffusion map study

doi:10.21203/rs.3.rs-5003291/v1

The quality of parent-child interaction during shared reading may influence the activation and synchronization of reading-related brain networks. But could differences in brain activity while a child is listening to stories predict parent-child interaction level during reading? For this study, functional MRI including a stories listening task was performed with 22 4-year-old girls and behavioral measurement scores reflecting parent-child interaction as well as maternal depression levels affecting these interactions were collected using video observation data of a shared reading task of these children with their mothers. The study aim was to apply the fMRI stories-listening data to create a diffusion maps algorithm and then attempt to classify the level of parent-child interaction during a shared reading task outside of the scanner. The diffusion maps algorithm successfully clustered children in this manner, with higher parent-child engagement scores related to diffusion patterns in regions of the brain known to support reading. This study demonstrates that applying this diffusion maps algorithm to brain functional connectivity data can predict parent-child interaction during shared book reading. This algorithmic approach is a potential, novel, data-driven means to quantify parent-child interaction in different contexts (e.g., reading, play) and populations.

Neurobiological correlates for stories listening

Reading, the ability to identify words and understand their meaning, is a crucial skill in modern life. However, since this skill is evolutionary new, brain areas that evolved to support more basic functions (e.g., language, vision) must be stimulated and then integrated, ideally during early childhood when the brain is highly plastic, to achieve reading proficiency[1]. This synchronization process has been termed “emergent literacy” and typically begins in infancy and extends through school age[2]. Brain regions known to support emergent literacy and reading have been well described and include the superior temporal gyrus and the angular gyrus (linguistic abilities) [3], frontal regions (executive functions, comprehension) [2] [4], and visual areas including the fusiform gyrus (letter and word recognition) [5].

Interestingly, the same neural circuits supporting reading abilities are also involved in pre-reading activities, such as during stories listening [6]. More nurturing home literacy environment has been linked to stronger activation of visual processing-related regions and executive functions during stories listening in 3–5 years old children, using magnetic resonance imaging (MRI) [7]. Activation of these regions has been attributed to imagination and comprehension, and likely helps shape the future reading network and abilities [6, 8]. It has also been suggested that stories listening in an engaged manner (i.e. interactive “shared” reading) engages similar brain regions in young children at a pre-reading age [1]. These studies suggest that story listening supports emergent literacy and subsequent reading skills by stimulating and even synchronizing brain regions and networks at this formative age[6].

Stories listening: an important facilitator for parent-child interaction

Brain development is affected by a child’s interactions with their environment beginning in infancy [9]. Although children have many types of interactions with their environment, the most meaningful is with their parents [1]. Behavioral and neurobiological evidence claims that shared book reading, especially involving a child and their parent, is beneficial for developing and synchronizing brain regions supporting reading and learning [7]. It has been shown that high levels of parent-child interaction during this activity may be especially impactful on neurobiological and cognitive levels, leading to an improvement in reading abilities and outcomes [1].

An evidence-based construct for interactive reading has been termed dialogic reading, which involves specific types of prompts and responses during a story [10]. Higher levels of interactivity (dialogic-ness) has been linked to improved language and cognitive abilities in children [1]. This has been recently supported by MRI studies showing associations between interactivity during storytelling and the engagement of neural circuits supporting language and executive functions [1]. Other studies examining the effect of dialogic, interactive reading with children at that age, echoed these findings using EEG measures[11].

Maternal depression and storytelling

Maternal depression is characterized by low mood, loss of interest and reduced activity, which has negative impacts on parenting and child adjustment [12]. It has been shown that mothers with maternal depression tend to read less to their children [13]. Additionally, Mothers with depression tend to read less interactively, with fewer facial expressions and high tones that trigger a child's attention [8]. Maternal depression is related to a range of adverse child health outcomes, including lower linguistic abilities[8]. Impacts on neurobiological measures have also been described, as in lower functional connectivity in brain regions related to executive functions and visual processing during stories listening in 4 year-olds with depressed mothers [8]. This was attributed to lower stimulation during storytelling and consequently lower engagement of functional connections between these brain regions needed to imagine and attend to the stories. Although it is well known maternal depression has a negative effect on the interaction with the child and linguistic outcomes, it is not clear whether children of mothers with maternal depression share differences in brain organization related to future reading readiness. One approach that can be used to test this is by clustering characteristics of individuals using MRI diffusion maps.

Diffusion maps as a method of data clustering

Diffusion maps (DM) are a method for finding meaningful geometric descriptions of data sets based on diffusion processes [14]. This method changes the representation of data sets with many variables into a low-dimensional description, while preserving quantities of interest, in turn enabling classification of the data sets in a more clear and representative manner [15]. This method is based on the use of eigenfunctions of Markov matrices that can construct coordinates called diffusion maps in which the data could be represented. As the two points on the diffusion map coordinates are closer, the relationship between those data points is stronger. Moreover, the relationship between sets of data can be represented as a geometric structure on the diffusion map coordinates [15].

The goal of the current study is to determine whether parent-child interaction during an observed shared book reading session could be reflected in the child’s neural functional connectivity during a story listening fMRI task, while focusing on networks known to be involved with narrative processing. The approach was to determine the distance in brain functional connectivity matrices during shared story reading for maternal-child dyads with high vs low levels of interaction. As mothers with depression tend to interact less with their children[16], a secondary aim was to determine if high vs low levels of interaction during this book reading session would also be reflected in maternal depression levels.

We hypothesized that: 1)applying the DM method to fMRI data from a stories-listening task involving children of mothers with high vs low levels of depression would successfully cluster functional connectivity networks. The selected networks in the current study are known to support reading based on the level of parent-child interaction during a shared reading task; 2) dyads with more interactive reading and also lower maternal depression levels would show greater functional connections within and between reading-related networks and that these networks would successfully cluster children based on parent-child interaction.

Participants

The study involved 22 mother-child dyads, recruited from a longitudinal home injury prevention trial serving mothers of low-socioeconomic status (Cincinnati Home Injury Prevention (CHIP) trial). Participating children were girls between 3.0 and 4.5 years (mean age: 4.1 years, SD = 0.2) who met the following criteria: native English speakers from monolingual households, right-handed, full-term gestation, no history of head trauma with loss of consciousness or stimulant use, and no standard contraindications to MRI. All parents provided written consents. The study was approved by the Cincinnati Children’s Institutional Review Board (IRB) in accordance with the Declaration of Helsinki.

Behavioral measurements

Two methods were used to measures parent-child interaction level during storytelling as well as offline (i.e. Without directly measuring parent-child interaction):

Parent-child interaction during storytelling (online interaction measurement): Behavioral measurements of the mother and the mother-child interaction were collected during a shared reading task (also reported in [1]). To measure the quality of interaction, the child's engagement during the shared reading was scored by a scale that was developed and reviewed by experts [1] as follows: 0 – not engaged, 1 – somewhat engaged, 2 – very engaged, and 3 – extremely engaged. Children's engagement scores were calculated as the mean score from the 3 experts' scores. To determine the level of interaction challenges (per [17]), number of times when mothers checked their phones during the shared reading task was measured, and was scored as 1 (i.e. checked the phone) or 0 (did not check the phone), respectively (see also [18] for the scoring method).

Maternal depression (an offline measure of parent-child interaction): Maternal depression levels were assessed using the validated Beck Depression Inventory-Ⅱ (BDI-Ⅱ) [19].

Behavioral data analysis

Independent t-tests were conducted for each behavioral measure and medians were calculated. Moreover, to determine the relations between parent-child interaction measures while listening to stories and the offline parent-child interaction levels, Pearson correlations between maternal depression measures and behavioral measures were conducted.

Neuroimaging measurements

The MRI scans were acquired in Cincinnati Children’s Hospital Medical Center, Ohio using a 3T Phillips Achieva MRI system. For fMRI, a time series of 165 blood-oxygen-level-dependent (BOLD) weighted scans covering the entire brain with 38 slices in the axial plane were continuously acquired with voxel size 3.75x3.75x5 mm at 2-second intervals (TR = 2) during the story listening task. In addition, a 3D anatomical (T1) brain image was acquired for the co-registration to the functional scans. All children were awake and non-sedated during the scan.

Neuroimaging task: stories listening

The stories’ listening task was composed of 11 alternating blocks of control and active condition (6 and 5 each, respectively), with a duration of 30 seconds for each block. The active condition was listening to stories which was presented by a female voice and included 9,10 or 11 sentences with varying syntactic structures and vocabulary which was matched for the preschool level (see [20] for the stories transcript and description). For the control condition, a backward speech condition was used with voices with a range of frequencies of 200–400 Hz.

Neuroimaging data analysis

Functional MRI data and the 3D anatomical image of each child were pre-processed using the CONN toolbox (http://www.nitrc.org/projects/conn/; [21]). These steps included realigning and unwarping the data; slice-timing correction, coregistration, segmentation, normalization and smoothing. We used the Power’s atlas [22] to divide the brain to different regions of interest (ROIs) for calculating at each one the changes in BOLD over time during the stories listening task. Due to a high motion level of several children (threshold of 80% invalid volumes after the pre-processing step) only 17 functional scans were included in the analysis. Using the processed data of the functional MRI scans, we conducted a functional connectivity analysis. For each child, a connectivity matrix was created, containing the correlation of the changes in BOLD over time (during the stories listening task) between all the different ROIs as divided by the Power atlas. Since this atlas divides the brain into 264 ROIs [22], the connectivity matrix of each child was created as a 2D matrix with a size of 264×264.

Diffusion Maps analysis

Following a previously published method[23], connectivity matrices either for the whole brain or for selected networks supporting storytelling (i.e. visual processing, auditory processing and executive functions networks, i.e. Cingulo-opercular and fronto-parietal, see more in the next sections) were turned into a united three-dimensional connectivity matrix with a size of 17×264×264, which served as the input for the DM algorithm. This algorithm enabled lowering the high-dimension data of each child (connectivity matrix with a size of 264×264) and representing it as a single data point on a set of 3D diffusion map coordinates.

For each behavioral measurement (i.e. online assessment of parent-child interaction during storytelling and offline measure for parent -child interaction), the median was calculated, and the scores were divided into two groups – above the median and under/equal to the median. Then, each data point was colored according to values greater than or lower than the median values of a chosen behavioral measurement. For each of the two groups, the centroid on the DM coordinates was calculated. Then, the distance between the two centroids was calculated and was mathematically characterized to allow a separation of the group by the behavioral measurement.

To determine whether the DM better cluster the groups based on whole-brain ROIs or based on the selected networks associated with storytelling, both approaches were tested, including a whole brain analysis and network-based analysis with the following combinations: 1) visual, auditory, and EF networks (cingulo-opercular, fronto-parietal); 2) visual, and EF networks (cingulo-opercular, fronto-parietal); and 3) auditory, and EF networks (cingulo-opercular, fronto-parietal). These networks were previously uses in our studies[24, 25], and the ROIs were defined in [22].

Correlations between the distance of the points from the centroids and behavioral measurements

After applying the between-networks connectivity information to the DM algorithm, the distance between the centroids of each group (above the median or less/equal than the median) was calculated applying each behavioral measure (parent-child interaction during storytelling, maternal depression).

Behavioral measures

Demographic characteristics for the study’s sample are described in Table 1.

---Insert Table 1 about here---

Table 1. Demographic characteristics

		N	%
Gender	Female	17	100
Annual household income ($)	Under 5,000	7	41
	5,000-10,000	4	23
	10,000-30,000	3	18
	30,000-50,000	3	18
Maternal Education level	High school graduate or less	8	47
	Some college	8	47
	College graduate	1	6

Characterizing the study population by parent-child interaction online and offline Parent-child interaction (online) median calculation: Of the 17 children who participated in the both the MRI session and in the shared reading task with their mothers, eight had engagement scores less than the median (1.67) and one child got a score that was equivalent to the median (a total of nine children who had under/equal the median score), whereas eight had greater scores for their engagement in the shared reading task. Interaction challenges were also assessed, whereases nine mothers checked their phones and eight mothers did not. See Table 2.

Parent-child interaction (offline) median calculation: Of the 17 mothers, seven demonstrated BDI scores less than the median (median: 8.67) and two had BDI scores which are equivalent to the median (a total of nine mothers had under/equal the median scores) whereas eight mothers had BDI scores greater than the median. See Table 2.

----Insert Table 2 about here---

Table 2. Behavioral measurements of the mothers and the parent-child interaction during the shared reading task

	Mean (SD)	Median	Min-max	Number of mothers below or equal the median	Number of mothers above median
Child engagement level (online measure)	1.57 (0.97)	1.67	0-3	9	8
Engagement challenges (online measure)	0.47 (0.51)	0	0-1	9	8
Maternal depression; BDI (offline measure)	12.73 (7.98)	8.67	2.67-29.27	9	8

Table 2. The engagement challenges were assessed using the CONNECT measure (see [18]) focusing on the number of times the parent checked their phones during the storytelling activity.

Pearson correlation between parent-child interaction during storytelling measures and offline parent-child interaction measures

A positive trend was found between maternal depression levels and the online measure for parent-child interaction challenges (i.e the number of phone checks by the mother) (r=.389, p=.06). Additionally, a negative trend between maternal depression and storytelling interaction (r=-.363, p=.07).

A Pearson correlation within the parent-child online interaction measures during storytelling revealed a negative correlation between the number of phone checks and the level of child engagement (r=-0.778, p<0.001) indicating more phone checks were related to lower interaction during storytelling. Results suggest that higher maternal depression levels are related to more phone checks and lower interaction with the child during storytelling and that more phone checks are related to lower child engagement.

Neuroimaging results

Whole brain analysis

The whole brain matrices included in the DM algorithm, demonstrated a successful classification of the participants by the “Child Engagement” and “Phone Check” online behavioral measurements as the input. The DM algorithm also successfully clustered the participants by their maternal depression level (BDI measurement). See Figure 1.

----Insert Figure 1 about here---

Figure 1. Whole brain analysis during story listening. The black color represents datasets with values lower than the median; red color: values greater than the median. The green color represents the centroid for datasets with values lower than the median and the blue color represents centroids for datasets with values greater than the median. The upper maps represent the data for the division by parent-child interaction during storytelling (left; child engagement, right; number of times the mother checked the phone). The lower map represents the data for the division by offline mother-child interaction (i.e. maternal depression).

DM containing the combination of the Visual, FP, and CO networks

Importing into the DM the specific functional networks for the ROIs in the selected visual processing and EF networks (FP, CO) during the stories-listening task, and coloring them based on the median values above and below the three selected measurements, the DM algorithm was able to split the full cohort into the two groups of high vs low interaction (i.e. successfully classified) the participants by the “Child Engagement” and “Phone Check” online behavioral measurements. See Figure 2.

----Insert Figure 2 about here---

Figure 2. Visual processing and EF networks-based DMs. The black color represents datasets with values lower than the median; red color: values greater than the median; green color; centroid for datasets with values lower than the median and the blue color represents centroids for datasets with values greater than the median. The upper maps represent the data for the division by parent-child interaction during storytelling (left; child engagement, right; number of times the mother checked the phone). The lower map represents the data for the division by offline mother-child interaction (i.e. maternal depression).

DM containing the combination of the Auditory, FP, and CO networks

Importing into the DM the functional connectivity networks from the auditory processing and EF networks (FP, CO) during the stories-listening task, and coloring them based on the median values above and below the three selected measurements, the DM algorithm successfully classified the participants only by the “Child Engagement” behavioral measurement and was not sensitive for the differences between the groups while focusing on the number of phone checks. See Figure 3.

----Insert Figure 3 about here---

Figure 3. Auditory processing and EF networks-based DMs. The black color represents datasets with values lower than the median; red color: values greater than the median; green color; centroid for datasets with values lower than the median and the blue color represents centroids for datasets with values greater than the median. The upper maps represent the data for the division by parent-child interaction during storytelling (left; child engagement, right; number of times the mother checked the phone). The lower map represents the data for the division by offline mother-child interaction (i.e. maternal depression).

DM containing the combination of the Auditory, Visual, FP, and CO networks

Importing into the DM the functional connectivity networks from the auditory and visual processing and EF networks (FP, CO) during the stories-listening task, and coloring them based on the median values above and below the three selected measurements did not show successful clustering. See Figure 4.

----Insert Figure 4 about here---

Figure 4. Visual and auditory processing and EF networks-based DMs. The black color represents datasets with values lower than the median; red color: values greater than the median; green color; centroid for datasets with values lower than the median and the blue color represents centroids for datasets with values greater than the median. The upper maps represent the data for the division by parent-child interaction during storytelling (left; child engagement, right; number of times the mother checked the phone). The lower map represents the data for the division by offline mother-child interaction (i.e. maternal depression).

Correlations between the distance from the centroids of each group, per DM condition and behavioral measures

Table 3 demonstrates the distances between centroids of each group (low vs high parent-child interaction), for each networks’ combination and parent-child interaction condition. The largest distance between centroids were found for the combination of the visual, CO, and FP networks, by the parent-child interaction measures during storytelling (i.e. “Child Engagement” and “Phone Check” measurements), which were 0.3067 and 0.3187, respectively. These results indicate that the DM algorithm best classified the neural matrices by these behavioral measurements, into two different clusters. The distances between the centroids, separating the neural data by the offline parent-child interaction (BDI measurement), were nearly close, for all the between-networks analyses, which indicates that the DM algorithm could not cluster more successfully between the different between-networks analyses.

----Insert Table 3 about here---

Table 3. Distances between the centroid

Whole brain

Visual-EF

Auditory-EF

Visual-Auditory-EF

Offline

Online

Offline

Online

Offline

Online

Offline

Online

BDI

Child Engagement

Phone Check

BDI

Child Engagement

Phone Check

BDI

Child Engagement

Phone Check

BDI

Child Engagement

Phone Check

Distance between centroids

0.206

0.185

0.264

0.196

0.306

0.318

0.168

0.188

0.279

0.192

0.158

0.214

Table 3. The distances between centroids for each parent-child interaction measure (o line and offline) for each networks combinations is presented in the table.

This study aimed to determine the feasibility of DM to cluster children in terms of more vs less interactive maternal-child shared reading, based on brain connectivity patterns captured using fMRI during a stories listening task. This approach allows an objective clustering of groups with shared characteristics, without the need to down-sample the data or reduce its dimensionality.

In line with our hypothesis, the DM algorithm successfully classified the children in terms of parent-child interaction level as measured using both “Child Engagement” and “Phone Check” behavioral measurements during the shared-reading task, based on the fMRI connectivity data involving ROIs of interest during the stories-listening task. The best classification (compared to a whole brain approach or other networks combinations) was obtained when importing the connectivity matrix of the visual processing and EF networks to the DM algorithm, quantified via larger distances between centroids of each group. However, contrary to our hypothesis, the DM algorithm did not successfully classify the children by maternal depression level. We attribute this to the moderate correlation between higher maternal depression level and behavioral measures (phone check, reading engagement) , which may not be sensitive enough to differentiate children of mothers with maternal depression. Additional behavioral measures might be needed for this purpose (i.e. the level of eye contact, touch, etc)

The importance of sensory processing and EF in stories engaging

The results of the current study are in line with previous findings highlighting the involvement of neural networks related to sensory processing and EF while listening to studies to children’s engagement in listening to the story[18, 26, 27]. These studies, as well as ours, strengthen the evidence that stories listening is not a passive process. Indeed, robust engagement in shared reading involves both imagery (i.e. visual processing) and cognitive control (i.e. EF), neural networks that are engaged via higher parental interactivity. Critical questions are raised regarding the importance of parent-child engagement during storytelling in populations that engage these networks differently than in typical populations. These include children with reading difficulties who share challenges in engaging the visual cortices[28] or children with attention difficulties who share challenges in engaging neural networks related to EF[29]. Moreover, it is unclear whether interventions such as Dialogic Reading that are intended to enhance interactivity during storytelling[11][30] also can fuel higher engagement and connectivity in these networks. Additional studies should address this question.

Study’s limitations

This study has several limitations that should be noted. The results are based on a relatively small sample size (n = 17), limiting statistical power. Conducting a study with more participants would possibly improve the DM algorithm's ability to cluster the fMRI data and would likely better differentiate connectivity matrices of participants. Additionally, in the current study, no strong correlation was found between the depression measure (BDI) and behavioral measures of children during the shared reading task. A future study should reassess the correlation of those measurements and determine the ability of the DM algorithm to cluster participants by this measure or others. Moreover, the main (online) measure for child engagement involved direct observation of maternal-child interaction during storytelling, while the offline measure for engagement was the BDI questionnaire, which may not be optimally aligned. Lastly, the focus of this study was clustering participants by the DM algorithm based on several between-networks information. Although the DM algorithm successfully clustered participants here by applying the neural information of the Vis-CO-FP networks, there are additional networks that may more efficiently cluster participants, which is worthy of further study.

The current study suggests that the DM algorithm has the potential to predict the quality of parent-child interaction using fMRI connectivity data. By conducting the neural information of the Visual processing, CO and FP networks, the DM algorithm successfully clustered children by two direct behavioral measures (dialogic interactivity, phone checking) during observed shared reading with their mothers. The results suggest that the DM algorithm may be able to represent visually connections between neurological and behavioral information related to shared book reading. Additional, potential applications of this data-driven algorithmic approach could include populations of interest (e.g., behavioral conditions, poverty) and other cognitive domains.

Author Contribution

THK and JH secured funding; AJ- analysis, AJ, THK, JH- wrote the first draft and reviewed, THK- mentorship

Data Availability

Data will be available upon request from the corresponding author (THK)

Funding declaration

This study was supported by the Technion-CCHMC next generation grant (PI: Horowitz-Kraus, Hutton).

Human Ethics and Consent to Participate declarations

All parents provided written consents. The study was approved by the Cincinnati Children’s Institutional Review Board (IRB) in accordance with the Declaration of Helsinki.

J. S. Hutton et al., "Story time turbocharger? Child engagement during shared reading and cerebellar activation and connectivity in preschool-age children listening to stories," Plos one, vol. 12, no. 5, p. e0177398, 2017.
N. U. Dosenbach, D. A. Fair, A. L. Cohen, B. L. Schlaggar, and S. E. Petersen, "A dual-networks architecture of top-down control," Trends in cognitive sciences, vol. 12, no. 3, pp. 99-105, 2008.
J. J. Vannest et al., "Comparison of fMRI data from passive listening and active‐response story processing tasks in children," Journal of Magnetic Resonance Imaging: An Official Journal of the International Society for Magnetic Resonance in Medicine, vol. 29, no. 4, pp. 971-976, 2009.
V. J. Schmithorst, S. K. Holland, and E. Plante, "Cognitive modules utilized for narrative comprehension in children: a functional magnetic resonance imaging study," Neuroimage, vol. 29, no. 1, pp. 254-266, 2006.
M. M. Berl et al., "Functional anatomy of listening and reading comprehension during development," Brain and language, vol. 114, no. 2, pp. 115-125, 2010.
T. Horowitz-Kraus, R. Meri, S. K. Holland, R. Farah, T. Rohana, and N. Haj, "Language First, Cognition Later: Different Trajectories of Subcomponents of the Future-Reading Network in Processing Narratives from Kindergarten to Adolescence," Brain Connectivity, 2024.
J. S. Hutton, T. Horowitz-Kraus, A. L. Mendelsohn, T. DeWitt, S. K. Holland, and C.-M. A. Consortium, "Home reading environment and brain activation in preschool children listening to stories," Pediatrics, vol. 136, no. 3, pp. 466-478, 2015.
R. Farah et al., "Maternal depression is associated with altered functional connectivity between neural circuits related to visual, auditory, and cognitive processing during stories listening in preschoolers," Behavioral and Brain Functions, vol. 16, no. 1, pp. 1-12, 2020.
S. E. Fox, P. Levitt, and C. A. Nelson III, "How the timing and quality of early experiences influence the development of brain architecture," Child development, vol. 81, no. 1, pp. 28-40, 2010.
R. Farah, R. Meri, D. S. Kadis, J. Hutton, T. DeWitt, and T. Horowitz-Kraus, "Hyperconnectivity during screen-based stories listening is associated with lower narrative comprehension in preschool children exposed to screens vs dialogic reading: an EEG study," PLoS One, vol. 14, no. 11, p. e0225445, 2019.
E. Twait, R. Farah, N. Shamir, and T. Horowitz‐Kraus, "Dialogic reading vs screen exposure intervention is related to increased cognitive control in preschool‐age children," Acta Paediatrica, vol. 108, no. 11, pp. 1993-2000, 2019.
R. T. Ammerman, C. E. Shenk, A. R. Teeters, J. G. Noll, F. W. Putnam, and J. B. Van Ginkel, "Impact of depression and childhood trauma in mothers receiving home visitation," Journal of child and family studies, vol. 21, pp. 612-625, 2012.
M. Kavanaugh, J. S. Halterman, G. Montes, M. Epstein, A. D. Hightower, and M. Weitzman, "Maternal depressive symptoms are adversely associated with prevention practices and parenting behaviors for preschool children," Ambulatory pediatrics, vol. 6, no. 1, pp. 32-37, 2006.
R. R. Coifman and S. Lafon, "Diffusion maps," Applied and computational harmonic analysis, vol. 21, no. 1, pp. 5-30, 2006.
R. R. Lederman and R. Talmon, "Learning the geometry of common latent variables using alternating-diffusion," Applied and Computational Harmonic Analysis, vol. 44, no. 3, pp. 509-536, 2018.
A. Y. Farmer and S. K. Lee, "The effects of parenting stress, perceived mastery, and maternal depression on parent–child interaction," Journal of Social Service Research, vol. 37, no. 5, pp. 516-525, 2011.
Y. Lederer, H. Artzi, and K. Borodkin, "The effects of maternal smartphone use on mother–child interaction," Child development, vol. 93, no. 2, pp. 556-570, 2022.
J. S. Hutton, Phelan, K., Horowitz-Kraus, T., Dudley, J., Altaye, M., DeWitt, T., & Holland, S. K., "Story time turbocharger? Child engagement during shared reading and cerebellar activation and connectivity in preschool-age children listening to stories," PLoS One, , vol. 12, no. 5, 2017.
A. T. Beck, R. A. Steer, and G. Brown, "Beck depression inventory–II," Psychological assessment, 1996.
J. S. Hutton et al., "Shared reading quality and brain activation during story listening in preschool-age children," The Journal of pediatrics, vol. 191, pp. 204-211. e1, 2017.
S. Whitfield-Gabrieli and A. Nieto-Castanon, "Conn: a functional connectivity toolbox for correlated and anticorrelated brain networks," Brain connectivity, vol. 2, no. 3, pp. 125-141, 2012.
J. D. Power et al., "Functional network organization of the human brain," Neuron, vol. 72, no. 4, pp. 665-678, 2011.
N. Habouba et al., "Parent-child couples display shared neural fingerprints while listening to stories: A functional magnetic resonance study," Scientifc Report, 2023.
N. Habouba et al., "Parent–child couples display shared neural fingerprints while listening to stories," Scientific Reports, vol. 14, no. 1, p. 2883, 2024/02/04 2024, doi: 10.1038/s41598-024-53518-x.
M. Appel, D. Hasin, R. Farah, and T. Horowitz-Kraus, "Greater utilization of executive functions networks when listening to stories with visual stimulation is related to lower reading abilities in children," Brain and Cognition, vol. 177, p. 106161, 2024.
J. S. Hutton, T. Horowitz-Kraus, A. L. Mendelsohn, T. DeWitt, and S. K. Holland, "Home Reading Environment and Brain Activation in Preschool Children Listening to Stories," (in eng), Pediatrics, vol. 136, no. 3, pp. 466-78, Sep 2015, doi: 10.1542/peds.2015-0359.
R. Farah et al., "Maternal depression is associated with altered functional connectivity between neural circuits related to visual, auditory, and cognitive processing during stories listening in preschoolers," Behavioral and Brain Functions, vol. 16, pp. 1-12, 2020.
B. L. Schlaggar and B. D. McCandliss, "Development of neural systems for reading," Annual review of neuroscience, vol. 30, pp. 475-503, 2007, doi: 10.1146/annurev.neuro.28.061604.135645.
S. Cortese, "The neurobiology and genetics of Attention-Deficit/Hyperactivity Disorder (ADHD): what every clinician should know," (in eng), European journal of paediatric neurology : EJPN : official journal of the European Paediatric Neurology Society, vol. 16, no. 5, pp. 422-33, Sep 2012, doi: 10.1016/j.ejpn.2012.01.009.
A. A. Zevenbergen, & Whitehurst, G. J. , "Dialogic reading: A shared picture book reading intervention for preschoolers," On reading books to children: Parents and teachers, pp. 177-200, 2003.

No competing interests reported.

Functional connectivity of the sensory system and executive functions during a story listening task is related to parent-child interaction during joint reading: a functional MRI-diffusion map study

Status:

Version 1

Abstract

Figures

Introduction

Neurobiological correlates for stories listening

Stories listening: an important facilitator for parent-child interaction

Maternal depression and storytelling

Diffusion maps as a method of data clustering

Methods

Participants

Behavioral measurements

Behavioral data analysis

Neuroimaging measurements

Neuroimaging task: stories listening

Neuroimaging data analysis

Diffusion Maps analysis

Correlations between the distance of the points from the centroids and behavioral measurements

Results

Discussion

Study’s limitations

Conclusions and Future directions

Declarations

Author Contribution

Data Availability

References

Additional Declarations

Status:

Version 1