Mother-Child Neurophysiological Synchrony Moderates the Relation Between Maternal Affect Experiences and Child Emotion Dysregulation

Respiratory sinus arrhythmia (RSA) is conceptualized as a measure of respiratory modulation of vagal control of the heart (i.e., cardiac vagal tone; Grossman & Taylor, 2007; Perry et al., 2018). RSA measures the intersection between the cardiovascular and respiratory systems and is considered a proximal measure of one’s parasympathetic nervous system functioning with implications for individual self-regulation (Grossman & Taylor, 2007). RSA synchrony is an indicator of the degree of parent-child mutual responsiveness to a stressful and/or social interaction task (Davis et al., 2018). However, unlike positive behavioral synchrony (e.g., shared parent-child positive affect), which is uniformly associated with positive child outcomes, relations between RSA synchrony and child outcomes are context-dependent (Hale et al., 2023; Lunkenheimer et al., 2018; Suveg et al., 2019). For example, in the context of positive synchrony, firm parental control related to increased youth externalizing problems in a sample of parent-adolescent dyads. Conversely, the relation between positive parenting and decreased behavioral problems and the association between negative parenting and increased behavioral problems were strengthened in the context of parent-child synchrony (Oshri et al., 2023). In a study of early adolescents and their mothers, maternal depressive symptoms related to elevations in youth psychopathology symptoms in the context of positive RSA synchrony. However, when RSA synchrony was negative, maternal depressive symptoms were associated with fewer symptoms of psychopathology in youth (West et al., 2020). Further, albeit not moderation analyses, one study reported low levels of maternal depressive and child internalizing symptoms related to positive RSA synchrony whereas in the context of high symptoms, mother-child RSA synchrony was observed to be negative (Suveg et al., 2019). Additionally, albeit not moderation directly, studies have reported links between RSA synchrony with broader contextual factors (i.e., income, child gender). Specifically, in a recent meta-analysis, authors reported significantly weaker RSA synchrony when parent-child dyads reported lower income (Miller et al., 2023). Further, a myriad of studies have controlled for child gender due to theorized differences in RSA synchrony by child gender (Feldman, 2021; Lunkenheimer et al., 2021a, b; Oshri et al., 2021). The present study expands empirical work by examining the moderating role of RSA synchrony on the relation between two transdiagnostic indicators of psychopathology (i.e., maternal affect experiences and child emotion dysregulation), while controlling for income and child gender.

Similar to RSA, activation in the dorsolateral prefrontal cortex (dlPFC) is critical for regulating emotions as well as carrying out cognitive functions and engaging in adaptive behavioral exchanges (Lee & Williams, 2024). Activation in the dlPFC has been associated with increased self-monitoring and adaptive emotional and behavioral regulation across development (Nguyen et al., 2021a, b). Research also reports differences in dlPFC activation with socioemotional functioning depending on the hemisphere: activation in the left dlPFC is associated with enhanced positive mood and decreased negative mood (Maeoka et al., 2012; Pena-Gomez et al., 2011) whereas the right dlPFC has been related to emotional processing (Sanchez-Lopez et al., 2018).

Given the implications of dlPFC activation for adaptive emotion regulation and healthy social interactions, work has begun to investigate parent-child dlPFC synchrony (Alonso et al., 2024; Nguyen et al., 2020; Ratliff et al., 2022; Zhang et al., 2017). Miller and colleagues (2019) found then when comparing dlPFC synchrony between tasks, more positive dlPFC synchrony was found between mothers and their 8- to 13-year-old children during cooperative tasks when compared to independent tasks. The same pattern of effects was found in a sample of parents and their 5- and 6-year-old children (Nguyen et al., 2020, 2024). Research has also examined direct associations between neural synchrony and child adjustment, noting significant, positive relations between positive mother-child neural synchrony with adaptive child emotion regulation, shared positive affect expression, and parental warmth across cooperative, competitive, and individual tasks (Liu et al., 2024; Miller et al., 2019; Nguyen et al., 2020; Reindl et al., 2018).

The vast majority of dlPFC synchrony work has centered around direct associations between synchrony and child adjustment. For instance, fathers with more positive perceptions of their parenting role displayed stronger dlPFC synchrony during a cooperation task with their children when compared to fathers with less positive perceptions (Nguyen et al., 2021b). Conversely, mother-preschool dyads experiencing higher levels of adversity, characterized by low income and high familial risk (e.g., household chaos), demonstrated weaker PFC synchrony during a stress task compared to dyads with lower levels of adversity (Hoyniak et al., 2021). Though work has primarily investigated direct links between neural synchrony and parent and child adjustment indicators, biobehavioral frameworks of child development suggest that parent-child synchrony processes can exacerbate or buffer risk (Ratliff et al., 2022). Further, in recent reviews of neural synchrony, Alonso and colleagues (2024) as well as Nguyen et al. (2024) argue that a gap in current research is the examination of neural synchrony as a moderator between contextual factors (e.g., maternal behaviors) and child psychological outcomes. Thus, the present study builds on foundational neural synchrony work and assesses neural synchrony as a moderator of the association between mother’s affect experiences and child emotion dysregulation.

Participants included 80 mothers (M_age = 35.97, SD = 5.46) and children (M_age = 5.88, SD = 0.80; 54% girl; 46% boys). Children were required to be between the ages of 5- and 7-years-old and families were required to speak English and be free from a developmental concern that could prevent the completion of study procedures. Research examining synchrony among caregiver-child dyads have noted differences in male and female caregiver emotional and neurophysiological responding (Kochanska et al., 2008; Liu et al., 2024; Lunkenheimer et al., 2021a, b). Therefore, to reduce variability attributed by caregiver gender, only female caregivers were recruited for the study. Approximately two months into active data collection, inclusion criteria were adjusted to only include additional participants that identified members of an underrepresented racial or ethnic group to increase racial and ethnic diversity within the sample. Regarding child race, mothers reported that 40% of children identified as White (Non-Hispanic), 27.5% Black/African American, 15.0% Other (predominantly reported Biracial), and 2.5% Asian. 15.0% of mothers reported that their child was of Hispanic, Latinx, or Spanish heritage. The majority of mothers endorsed being married (65.0%) with the remaining mothers reporting being single (28.7%), divorced (3.8%), or separated (2.5%). Regarding education level, mothers reported completing a bachelor’s degree (32.5%), graduate degree (37.5%), some college (23.80%), a high school diploma (3.8%), and 2.6% reported having less than a high school diploma. Approximately 11.3% of mothers reported a family income of less than $30,000, 31.3% of mothers reported earning between $30,000 and $59,999, 23.7% reported earning between $60,000 and $89,999, and 33.8% of mothers endorsed a familial income of greater than $90,000.

Study procedures were approved and in accordance with the University of Georgia’s Institutional Review Board (IRB). Participants were recruited from the local community in the southeastern United States through flyers, social media postings, in-person community events, and participant referrals. Interested families were screened over the phone for eligibility and if they met criteria, were scheduled for a two-hour assessment in the laboratory. Mothers and children provided written consent and assent, respectively, prior to completing study procedures. Mothers then completed a battery of questionnaires in relation to their affect experiences and child’s socioemotional functioning. Electrodes and fNIRS caps were then placed on mothers and their children (see below for data acquisition, cleaning, and processing procedures). Dyads were instructed to watch a 2-minute video of calming nature scenery while baseline estimates of resting heart rate and blood-oxygenation levels were collected continuously using BIOPAC MP160 System for physiological data and NIRx for neural activity. Dyads then engaged in a 5-minute mild stress task where children were asked to build a developmentally advanced LEGO figure. Mothers were informed that they could provide verbal assistance to the child but could not physically touch the LEGOs. Participants were told they had 5 min to complete the figure, and a visual timer was displayed with frequent audio cues to remind the dyad of the time remaining on the task. See Supplemental Fig. 1 for a picture of the task. Video recordings for behavioral observations, RSA, and fNIRS were collected continuously throughout the task. Timing was standardized across each assessment such that each dyad had the same time increments between tasks although visits occurred during different times of the day to accommodate participant availability. At the conclusion of the study, participants were informed that the child did not have all of the pieces needed to complete the LEGO figure. Families were compensated $100 in cash for their participation in the study an additional $20 for every family they referred who enrolled in study procedures.

During processing, RSA was sampled at 1 Hz. However, to align with prior work recommending RSA be analyzed in 30-second epochs to best capture physiological reactivity to stress (Berntson et al., 1997; Fisher & Woodward, 2014; Xu et al., 2024a), RSA were aggregated and averaged every 29 samples. This resulted in 10 timepoints for each person as this was the closest approximate to 30-second epochs. Mother-child RSA synchrony was computed using Bayesian Structural Equation Modelling (Asparouhov & Muthén, 2019) following Xu et al. (2024b)’s procedure in MPlus 8.0. The model also used the average RSA of the last 29 timepoints in the baseline as the reference point and controlled for the auto-regression of RSA. Six dyads had missing data due to an inability to clean/derive reliable physiological values, resulting in a total of 74 dyads with usable RSA synchrony data.

The data processing procedure for fNIRs was based on steps outlined by Nguyen et al. (2020). The dataset was evaluated for missing values. A visual inspection of all Channels was conducted to identify and exclude those with significant issues. Additionally, an automated Channel pruning process was employed to detect and remove poor-quality channels using criteria such as signal integrity. Once the data were cleaned, they were converted into optical density values to measure the absorption of light by brain tissue. Motion artifacts caused by participant movement were addressed through spline interpolation and wavelet-based correction methods. Then, filters were applied to eliminate irrelevant fluctuations. The cleaned data was then converted into concentrations of HbO, HbR, and HbT, and downsampled from 7.81 Hz to 1 Hz. Finally, the fNIRS data were temporally aligned with the baseline and LEGO tasks. Of the 80 dyads, 65 were useable for further analysis. Each participant of the 65 dyads used 8 channels (4 for the left and right dorsolateral prefrontal cortex) out of the total 18-channel montage (8 out of 18 channels used). Out of the 8 Channels used for each dyad, there were 1,040 total Channels analyzed across 65 dyads (65 dyads x 2 participants x 8 channels) which resulted in 81.25% useable data from the total possible dataset. The 15 (18.75%) subjects that were excluded were removed due to poor signal-to-noise ratio, excessive motion artifacts, or study interruptions.

For analysis, brain activity measures from the left and right dorsolateral prefrontal cortex were aggregated across fNIRS channels for both mothers and children. The resulting columns represent these aggregated activity measures. HbR is typically less susceptible to artifacts from extracerebral sources, provides greater spatial specificity, and is less influenced by autonomic nervous system activity (Kayhan et al., 2022; Kirilina et al., 2013; Piazza et al., 2020; Tachtsidis & Scholkmann, 2016). Conversely, the HbO signal is more sensitive to changes in regional cerebral blood flow and exhibits a higher signal-to-noise ratio (Hoshi, 2007; Kinder et al., 2022; Pan et al., 2017). Given signal-type has specific advantages, the present study reported both HbR and HbO results as recommended by the literature (Boas et al., 2014; Yücel et al., 2021; see Tables 2 and 4).

Wavelet transform coherence (WTC) of fNIRS time-series data was used as an estimate of brain-to-brain synchrony for each mother-child dyad (Grinsted, et al., 2004). More specifically, WTC was used to compute the cross-correlation between each dyad’s aggregated fNIRS time series, as a function of time and frequency. This computation evaluates overall coherence patterns in the dyad’s brain activity, while considering both phase-lagged and in-phase correlations (Nguyen et al., 2021a, b). Higher coherence values indicate a higher correlation between mother and child brain activity across task duration and a chosen frequency range. However, it is worth noting that while WTC is useful in identifying similar patterns of change in brain activity, it is unable to detect directionality, such that coherence values alone may not be used to determine whether synchrony is being driven by in-phase, anti-phase, or lagged synchronization. WTC was calculated with the WTC function from the “biwavelet” package in R (v0.20.22; Gouhier et al., 2024), using the default Morlet wavelet. Coherence values were averaged across a frequency band of interest (FOI) of 0.02–0.1 Hz (corresponding to periods of 10–50 s) for the duration of the 5-minute stress task for each of the aggregated dlPFC signals, resulting in 4 coherence values for each dyad (right and left dlPFC HbO and HbR). This FOI was chosen due to its identification as most relevant to parent-child problem solving tasks and reduction of artifacts from physiological noise produced by respiration (~ 0.2–0.3 Hz), cardiac pulsation (~ 1 Hz), and Mayer waves (~ 0.1 Hz; Nguyen et al., 2021a, b).

To test for the significance of neural synchrony analyses above spurious correlations, we conducted random pair analyses in the same region (right and left dlPFC) and signal type (HbR and HbO) with a permutation approach. To that end, each mother was paired with a random child from another dyad 100 times, and coherence was averaged for each region and signal type across all random pairings. The random pair coherence values were entered in the same analyses as the true dyad coherence values to validate that results were due to synchrony between true dyads and not random chance or signal noise. Values for true mother-child dyads are expected to be higher (i.e., more strongly related) when compared to values observed in random pairings.

As a test of the interaction effects between maternal positive affect experiences and dlPFC synchrony in the right and left hemispheres, all moderations were computed a second time with random pairings to ensure significant effects for both HbR and HbO signals were due to dlPFC synchrony between mothers and their children, rather than a product of noise and/or randomness. The interaction between maternal positive affect experiences and dlPFC synchrony with random pairings using the HbR signal was nonsignificant in the right (B = −0.40, SE = 0.50, β = −0.24, p =.437) and left (B = −0.73, SE = 0.39, β = −0.24, p =.0.067) hemispheres. The same nonsignificant pattern emerged for the interaction between maternal negative affect experiences with right (B = −0.88, SE = 0.44, β = −0.24, p =.053) and left (B = −0.68, SE = 0.41, β = −0.24, p =.106) dlPFC synchrony using the HbR signal.

The interaction between maternal positive affect experiences and dlPFC synchrony when measured using the HbO signal with random pairings was nonsignificant in the right (B = 0.12, SE = 0.60, β = 0.03, p =.839) and left (B = 0.99, SE = 0.54, β = 0.23, p =.070) hemispheres. The same nonsignificant pattern emerged for the interaction between maternal negative affect experiences with right (B = 0.41, SE = 0.46, β = 0.11, p =.376) and left (B = −0.60, SE = 0.54, β = −0.14, p =.274) dlPFC synchrony using the HbO signal. Results from random pairings bolster confidence that the significant interactions noted above with dlPFC synchrony are products of mother-child neural synchrony and not chance.

In support of hypotheses, there was also a negative association in the context of positive RSA synchrony. This negative association was also found for right and left dlPFC synchrony. Positive RSA synchrony has been positively associated with both mother and child individual and shared positive affect expression in a sample of 8-to-10-year-old children and their mothers (Zehra Capraz et al., 2023). Similarly, Xu and colleagues (2024b) found that the link between parent-child RSA synchrony and child emotion regulation was positive in the context of high parental emotional support (e.g., likely also indicative of maternal positive affect experiences) in a sample of Chinese school-age children. In short, prior work has shown the benefits of mother positive affect experiences and RSA synchrony for maximizing positive child functioning. Interestingly, and contrary to hypotheses, there was also a positive relation between maternal positive affect experiences and child emotion dysregulation at low values of RSA synchrony. Regarding this perplexing association, perhaps at low levels of positive RSA synchrony, children do not benefit from the positive affect experiences of mothers or were unable to experience these positive effects, resulting in increased emotion dysregulation. Findings from the current study expand on this work by showing the benefits of mother positive affect for minimizing child emotion dysregulation, an effect that is strengthened in the context of high levels of positive RSA synchrony. It may be especially beneficial for children of well-regulated mothers to synchronize physiologically with their caregiver. Physiological synchrony at high levels in the context of maternal modeling of positive emotion may buffer against the development of child emotion dysregulation as these processes scaffold healthy self-regulation. Longitudinal designs can examine the potential long-term protective benefits of an emotionally-positive caregiving environment in combination with physiological synchrony for child psychopathology symptoms (McLaughlin & Gabard-Durnam, 2022).

Findings provide direct empirical support for the extension of the parent-child emotion regulation dynamics model (Ratliff et al., 2022), suggesting that the relation between maternal affect experiences and child emotion dysregulation specifically varies depending on the neural synchronization of mothers with their children. With respect to positive dlPFC synchrony, in line with earlier research, there was a positive link between positive dlPFC synchrony with HbR signals and child emotion dysregulation (Miller et al., 2019; Nguyen et al., 2020). While preliminary work has established the value in dlPFC synchrony across cooperative, competitive, and individual tasks for adaptive child socioemotional functioning (Miller et al., 2019; Nguyen et al., 2020; Reindl et al., 2018), we extend this literature by demonstrating the value of positive dlPFC synchrony during a mildly stressful and emotionally laden parent-child interaction task. Mirroring the results for physiological synchrony, current findings suggest that when children are neurally-attuned to their mothers during a mildly stressful task, their symptoms of emotion dysregulation are significantly reduced. DlPFC activation represents neural underpinnings in relation to one’s self-regulation (Lee & Williams, 2024; Monsell, 2003). Thus, when a mother-child dyad is able to synchronously move with one another during a stressful time, this has implications for individual child self-regulation. Although extent literature has linked maternal adaptive self-regulation to positive child adjustment (Are & Shaffer, 2016; Morris et al., 2018; Ratliff et al., 2022; Tan et al., 2019), present findings suggest that dlPFC synchrony and maternal positive affect experiences are synergistically related to a robust transdiagnostic indicator of child psychopathology.

In regards to the lateralization of the dlPFC, prior literature has linked individual left dlPFC activation to experiences of positive and negative mood whereas right dlPFC has been linked to broader emotional processing (Maeoka et al., 2012; Pena-Gomez et al., 2011; Sanchez-Lopez et al., 2018). Although lateralization of dlPFC activation has been important in understanding individual processes, the current findings regarding the link between maternal positive affect experiences and child emotion dysregulation were consistent for left and right dlPFC synchrony. However, the interaction between maternal positive affect experiences and left dlPFC synchrony was stronger when compared to that of right dlPFC synchrony. Individual-level left dlPFC activation has been related to efficient emotion regulation and affect experiences, which is consistent with the extended parent-child emotion regulation dynamics model and is echoed in the current findings (De Raedt et al., 2015; Ratliff et al., 2022). Prior work suggests that mothers’ intrapersonal affect experiences shape how they interact with their children and, in turn, influence children’s abilities to regulate their own emotions (Deater-Deckard et al., 2016; Feng et al., 2008; Halberstadt et al., 1993). Results build upon prior literature which suggests that maternal affect experiences directly relate to children’s emotion dysregulation and that this association varies depending on dyadic indicators. While this relation has been theorized to vary by neurophysiological markers, the present study provides empirical support for this proposed association.

HbO synchrony did not show a comparable pattern of results as HbR synchrony. This discrepancy may be attributed to differential signal sensitivity. HbR has a lower amplitude than HbO. Nonetheless, HbR has been shown to be less sensitive to extracerebral hemodynamics and systemic physiological noise (e.g., scalp blood flow, respiration), potentially making it a more specific marker of localized cortical activity in naturalistic interaction contexts (Tachtsidis & Scholkmann, 2016; Kirilina et al., 2013). In contrast, HbO typically exhibits a higher signal-to-noise ratio but is also more susceptible to these confounds, which can obscure task-related neural signals. That said, the absence of HbO findings should be interpreted with caution, and future work is needed to determine under what conditions HbO and HbR may diverge in parent-child neural synchrony research.

Contrary to hypotheses, the interactions between maternal negative affect experiences and RSA synchrony as well as dlPFC synchrony (both left and right) were nonsignificant. However, the interaction between maternal negative affect experiences and RSA synchrony neared significance (p =.059), so the moderation was probed exploratorily. As expected, as RSA synchrony became more positive (raw score ≥ −0.03), the relation between maternal negative affect experiences and child emotion dysregulation was strengthened, suggesting that children show increasing emotion dysregulation as mothers’ self-reported negative affect experiences increase. During the early school-age period, children are highly attuned to the affect experiences and regulatory patterns of their parents and these observations can influence their own emotional development (Davis et al., 2015). For children with mothers reporting high degrees of negative affect experiences, children are likely observing more frequent emotional dysregulation when compared to children with mothers reporting lower levels. Domains of self-regulation (e.g., behavioral, physiological, emotional) do not operate in isolation but rather, are theorized to work as a coordinated system (Blair & Ku, 2022). Thus, if a mother is experiencing high degrees of negative affect, and the child becomes synchronized physiologically with her, this concordance likely exacerbates the child’s emotion dysregulation. Albeit only marginally significant, current findings provide empirical support for the notion that maternal negative affect experiences coupled with positive RSA synchrony confers increased risk for child maladjustment.

Contrary to hypotheses, there was no significant interaction between maternal negative affect experiences and neural synchrony in relation to child emotion dysregulation. This finding is perplexing particularly in light of other significant associations noted above. Perhaps the lack of effects could be due to the lower reported rates of maternal negative affect experiences (M = 19.54) when compared to positive affect (M = 36.21) in combination with a lower sample size for dlPFC synchrony. Decreased variance coupled with fewer data points could result in decreased power to detect statistically significant effects (Preacher et al., 2007). When looking at Pearson’s r values in Table 1, although not significant, all indicators of neural synchrony were negatively related to maternal positive affect experiences. The exact opposite appears to have emerged regarding the relation between maternal negative affect experiences and neural synchrony indices. This pattern of effects may suggest that a significant interaction may emerge with a larger sample size, though additional studies are needed to test that claim.