Linking Brain, Heart, and Behavior in Children: Frontal Electroencephalogram Activity, Respiratory Sinus Arrhythmia, and Internalizing Behaviors Authors: Youre Yang, Kristie L. Poole, Louis A. Schmidt Published: 2026 ABSTRACT ABSTRACTInternalizing behaviors are characterized by inward‐focused emotional distress, which often manifests as symptoms such as depression, anxiety, and negative affect. Previous research has found that physiological correlates of heightened emotional arousal, such as phasic changes in overall frontal electroencephalogram (EEG) alpha activity (FAA) and respiratory sinus arrhythmia (RSA), are independently associated with emotion dysregulation and internalizing behaviors in children. The present study examined whether the interaction of changes in overall FAA and changes in RSA from baseline to a stress induction predicted internalizing behaviors in 152 typically developing children (M age = 7.82 years, 73 girls, 81.6% White). Results revealed a significant interaction between changes in overall FAA and changes in RSA from baseline to a stressor in predicting children's internalizing behaviors. Specifically, increases in overall FAA from baseline to stress induction were associated with greater internalizing behaviors in children who also exhibited a decrease in RSA from baseline to stress induction. When children exhibited an increase in RSA from baseline to stress induction, increases in overall FAA were associated with relatively fewer internalizing behaviors. Findings highlight that measures indexing central and peripheral nervous system activity may jointly influence stress regulation and contribute to vulnerability to internalizing behaviors in middle childhood. Introduction Internalizing behaviors are characterized by inward‐focused emotional distress, manifesting as symptoms such as depression, anxiety, and negative affect (Achenbach and Edelbrock 1978). Internalizing behaviors can be observed as early as toddlerhood and often persist throughout development (Battaglia et al. 2016; Hansen and Jordan 2017; Roza et al. 2003; Smith and Bell 2010). The experience of internalizing behaviors in school‐aged children is common, with a prevalence rate of approximately 9%–15% in the United States (Danielson et al. 2021). One hypothesis is that internalizing behaviors are rooted in maladaptive emotion regulation abilities (Eisenberg et al. 1996; Stifter and Augustine 2019). Emotion regulation refers to the processes involved in modulating emotional reactivity during emotionally relevant situations (Eisenberg et al. 1996; Thompson 1994). Individual differences in children's emotion reactivity and regulation can be measured using different methods, including physiological responses to stress such as frontal brain electrical activity and parasympathetic nervous system (PNS) activity (Beauchaine 2015; Calkins and Fox 2002). Frontal cortex activation, specifically in the alpha frequency (8–13 Hz) measured using electroencephalogram (EEG), has been shown to track the intensity of emotional arousal across developmental periods (Davidson 2002; Dawson 1994; see also Fox 1994, for alpha meaning in infant and child samples). For example, infants who displayed greater distress in a maternal separation task also showed increases in overall frontal EEG alpha activity (FAA) from baseline to task (Dawson 1994). Moreover, children with relatively higher overall FAA at baseline (1 week before surgery) reported higher preoperative anxiety right before surgery (Xu et al. 2023). In young adulthood, relatively higher overall FAA has been shown in response to processing intense emotional stimuli (Schmidt and Trainor 2001). Taken together, changes in overall FAA may reflect individual differences in emotional reactivity during stressors and serve as a noninvasive marker of emotional arousal. In addition to overall FAA, the autonomic nervous system facilitates regulation of cardiac activity to support adaptive behavioral responses (Berntson et al. 2007; Hastings and Kahle 2019; Porges 1995; Thayer et al. 1996). The polyvagal theory (Porges 1995, 2007, 2021) suggests that the sympathetic nervous system (SNS) and the PNS coordinate each other through the vagus nerve to maintain homeostasis while engaging in emotionally demanding situations. Respiratory sinus arrhythmia (RSA) measures the high‐frequency heart rate variability that is theorized to originate from PNS activity (Porges 2007). RSA reactivity, conceptualized as the change in RSA from baseline to a stressful condition, is theorized to indicate an individual's threshold of physiological arousal by modulating the activity in the parasympathetic system (Porges 2007). Typically, a decrease in RSA (i.e., RSA withdrawal) under stress facilitates the preparation of the fight‐or‐flight response, allowing increases in heart rate (Hastings and Kahle 2019; Thayer et al. 1996). Conversely, an increase in RSA (i.e., RSA augmentation) under stress may signal more parasympathetic influence in regulatory response by lowering heart rate and demonstrating higher variability in heartbeat interval (Hastings and Kahle 2019). RSA reactivity has been associated with emotion regulation and children's internalizing behaviors. However, the results are mixed. Although a moderate decrease in RSA is often observed as a normative regulatory response to stress, excessive RSA withdrawal has been correlated with relatively poorer emotion regulation and internalizing behaviors (Hastings and Kahle 2019). For example, greater decreases in RSA during stressful experimental tasks have been associated with higher internalizing behaviors reported by mothers and teachers (Fortunato et al. 2013; Hinnant and El‐Sheikh 2009; Shanahan et al. 2014) and temperamental styles related to fear (Hassan and Schmidt 2024; Kiel and Aaron 2024; Poole and Schmidt 2021a). On the other hand, some research has found the opposite relation where greater RSA withdrawal is associated with less problem behaviors (Fletcher et al. 2017; Gentzler et al. 2009; Graziano and Derefinko 2013). There is a possibility that the mixed results could be explained by methodological reasons such as different sample compositions, different experimental tasks, and heterogeneity in internalizing symptomology. However, it is also important to consider the complex nature of physiological regulation and further investigate the relation between RSA reactivity and internalizing behaviors within a broader regulatory system. A well‐established line of research suggests that the central and autonomic nervous systems operate in a coordinated way to facilitate adaptive emotion regulation (R. Smith et al. 2017; Thayer et al. 2009). Indeed, brain and cardiac measures of emotional arousal have been shown to interact to predict individual differences in emotional reactivity and negative affect in adults (Davidson et al. 2000; Hagemann et al. 2003; Kop et al. 2011; Waldstein et al. 2000). However, existing work in child samples has focused largely on the independent relations between overall frontal brain activity1 (Dawson 1994; Xu et al. 2023) and RSA (Beauchaine 2015; Hastings and Kahle 2019) with indices of emotion regulation and internalizing behaviors. There is, however, relatively little work examining the interaction between overall FAA and RSA. Specifically, some other studies with children have examined the interaction between brain and autonomic systems in relation to social adjustment, using other indicators of regulation such as baseline frontal alpha asymmetry (Gatzke‐Kopp et al. 2014) or error‐related negativity (ERN) activity (Peoples et al. 2024). Investigating the joint role of different physiological systems (i.e., the brain and the heart) during a stress‐inducing context may provide a more fulsome understanding of their role in children's internalizing behaviors than examining these systems independently (Calkins and Fox 2002; Hastings and Kahle 2019; Miskovic and Schmidt 2012). In the present study, we examined the interactive role of overall FAA and RSA during an anticipation of a stressor (i.e., self‐presentation speech task) in relation to internalizing behaviors in typically developing children aged 7–8 years. We focused on middle childhood because during this developmental period, children develop self‐conscious emotions and can interpret social‐evaluative cues in their environment (Lagattuta and Thompson 2007). Children aged 7–8 years may also face diverse contexts requiring cognitive, emotional, and behavioral regulation, such as interaction with peers or performing a self‐presentation task in formal schooling. Further, children in middle childhood can comprehend the concept of anticipating a stressor, and previous work has shown that social stressors can elicit individual differences in physiological responses during this age (Schmidt et al. 1999). While prior research has linked each physiological measure independently to emotion dysregulation (e.g., Shanahan et al. 2014; Xu et al. 2023), to our knowledge, the interaction between baseline‐to‐task changes in overall FAA and changes in RSA has not been examined in relation to children's internalizing behaviors. Based on previous work examining these measures separately (Fortunato et al. 2013; Xu et al. 2023), we predicted that children who demonstrated a combination of greater increases in overall FAA and greater decreases in RSA from baseline to a stress induction task would have the highest level of internalizing behaviors. Method Participants Participants were 152 typically developing children (73 girls; M age = 7.82 years, SD = 0.44) aged 7–8 years and their primary caregivers (M age = 39.95 years, SD = 4.27, 90% mothers, 10% fathers). The sample was predominantly Caucasian (81.6%), followed by Mixed (9.9%), Asian (3.9%), African American (2.6%), and Latin American (2%). The majority of the children were from families of middle and upper socioeconomic status as indicated by total household income in Canadian dollars (<$60,000 = 14.5%, $60,001–$100,000 = 19.7%, >$100,000 = 65.8%). Children were recruited from a database containing contact information of caregivers who consented in the hospital after delivery before being discharged to be contacted for future developmental studies in the Department of Psychology, Neuroscience & Behaviour at McMaster University. All children were full‐term and typically developing without prenatal or postnatal health complications. Procedure Children and their caregivers visited the Child Emotion Laboratory at McMaster University. After obtaining written consent from both the parent and the child, the child was outfitted with an EEG net and an ambulatory electrocardiogram (ECG) monitor while the parent was present. The experimenter and the child then moved to a psychophysiology recording room while their parent completed an online questionnaire battery that asked questions about their child's socioemotional development. The child's EEG and ECG were recorded simultaneously during a baseline condition and a stress induction task (described below). Following the collection of the psychophysiology measures, the experimenter and the child moved to an adjacent room and completed additional tasks as part of the larger study on children's socioemotional development (see Poole and Schmidt 2021b). Because these tasks and measures were not part of the present study, they are not discussed further. The lab visit took approximately 75–90 min. The child received a Junior Scientist Certificate, and the family was compensated with a $20 CAD gift card as a token of appreciation for their participation. All procedures were approved by the McMaster University Research Ethics Board. Study Conditions Baseline Baseline EEG and ECG were concurrently collected for 2 min (1 min with eyes open and 1 min with eyes closed). For consistency with the stress induction task (described below), baseline psychophysiological measures were collected with the child standing, as postural differences are known to influence autonomic activity (Bush et al. 2011). All children were asked to stay still and relaxed during this period. Stress Induction Following the baseline psychophysiological condition, the experimenter informed the child that they would next have to deliver a speech. For more detailed information on the scripts and procedures for the stress induction task, see the Supporting Information. Briefly, children were told that they have 2 min to prepare a speech about their most recent birthday, which would be videotaped for other children to watch. During this speech anticipation period, EEG and ECG were recorded in the same standing position as the baseline condition. As a manipulation check, children reported on their own subjective nervousness after the baseline and anticipation period in the larger study. Subjective nervousness showed a significant increase from baseline to anticipation. These data are presented elsewhere (see Poole et al. 2021). Psychophysiological Data Collection and Reduction EEG Data Collection and Reduction EEG activity was recorded using a 128‐channel geodesic dense sensor net (Electrical Geodesics Inc., Eugene, OR). Data from each channel were digitized at a 250‐Hz sampling rate and filtered at 0.1 Hz (high pass) and 30 Hz (low pass) with a 60‐Hz notch filter. EEG channels were collected with reference to Cz and re‐referenced offline to the average of all electrodes (Junghöfer et al. 1999). All data were prepared and processed offline using Brain Vision Analyzer (Brain Products GmbH, Gilching, Germany). Electrodes with artifacts (e.g., high impedance or noisy signal) were removed by visually identifying them and were interpolated by surrounding channels. Ocular artifacts from eye blinks and horizontal eye movements were corrected using independent component analysis. Data were edited into 1‐s epochs using 50% overlap and baseline corrected using the entire segment. Epochs exceeding 120 µV, a voltage step of more than 75 µV between sample points, or a maximum voltage difference of less than 0.50 µV within any 100‐ms interval were marked as artifacts and automatically removed. Alpha frequency (8–13 Hz) was chosen given its established associations with individual differences in affective responses in child samples of this age (Chow et al. 2023) and adult samples (Schmidt and Trainor 2001; see also Dawson 1994; Fox 1994, for alpha meaning in child samples). For both the baseline and stress induction conditions, we created an overall alpha power score by averaging alpha power across three frontal sites (left frontal: F3; mid frontal: Fz; right frontal: F4). Regional absolute EEG power (in µV2) was derived in the alpha (8–13 Hz) frequency bands for each electrode of interest. Locations of the electrode sites of interest corresponded to the traditional 10–20 system of the International Federation (Jasper 1958). Because we were interested in children's change in overall FAA from baseline to stress induction, we created an overall FAA change score (i.e., baseline overall frontal alpha power minus stress induction frontal alpha power). Because of the inverse relation between EEG alpha power and activity (Davidson and Tomarken 1989), increases in FAA (i.e., decreases in frontal alpha power) are thought to reflect an anxious responseshown through the change score. Accordingly, on the overall FAA change score, a lower change score (i.e., negative score) reflected decreases in overall frontal brain activity from baseline to the stress induction task, while a higher change score (i.e., positive score) reflected increases in overall frontal brain activity from baseline to the stress induction task. Thus, a higher change score ostensibly reflects greater anxious arousal. ECG Data Collection and Reduction Cardiac and respiration data were continuously collected using MindWare Mobile Impedance Cardiograph, Model 50‐2303‐00. The unit acquired the cardiac R‐waves with a sampling rate of 500 Hz and 24‐bit ADC digitization. One ECG electrode was placed on the child's upper right back, and the other two were placed on the lower left and right sides in a triangle shape. A respiration belt was placed around the child's chest. The ECG unit was placed in a small backpack that the child wore for the duration of the study. Cardiac data were reduced and analyzed through MindWare software version 3.1.1. The software allowed visualization of the cardiac R‐waves. Missing or erroneous peaks were manually edited according to the guidelines from Berntson and Stowell (1998). We used the respiratory frequency band of 0.15–0.40 Hz, which is commonly applied to data collected on children in middle childhood (Erath and El‐Sheikh 2015; Keller et al. 2014; Schmitz et al. 2011). Mean RSA value (lnms2) was derived for each 1‐min epoch and averaged across the epochs of experimental tasks, which were baseline (2 min) and stress induction (2 min). Children were required to have at least 30s of usable data. We were interested in baseline‐to‐task changes in RSA, and therefore we created an RSA change score by subtracting baseline RSA from stress induction RSA (stress induction RSA minus baseline RSA). On this metric, a higher number (i.e., positive score) reflected an increase in RSA from baseline to a stress induction task. A lower number (i.e., negative score) reflected a decrease in RSA referring to RSA withdrawal, which has been associated with anxious arousal (Porges 2007) and emotion dysregulation (Hastings and Kahle 2019). Parent Report of Children's Internalizing Behaviors Parents completed the five‐item internalizing behavior subscale from the Social Skills Rating System (SSRS; Gresham and Elliot 1990). Parents reported on how often the child displays the behavior on a 3‐point Likert scale ranging from 0 = Never to 2 = Very often. Sample items on the internalizing subscale include “My child acts sad or depressed” and “My child shows anxiety about being with a group of children.” The sample demonstrated acceptable internal reliability (α = 0.68). Data Loss and Missing Data For the baseline EEG measure, there were missing data due to thick hair (n = 4), child refusal to wear the EEG cap (n = 1), equipment failure (n = 2), excessive artifact (<20 segments of usable data; n = 3), and outliers (3 SD above the mean; n = 3), resulting in 139 children with usable baseline EEG data. For the stress induction task EEG measure, there were missing data from thick hair (n = 4), child refusal to wear the EEG cap (n = 1), equipment failure (n = 2), excessive artifact (<20 segments of usable data; n = 3), and outliers (3 SD above the mean; n = 2), resulting in 140 children with usable stress induction EEG data. For the ECG measure, three children were missing baseline and stress induction data due to equipment failure (n = 1), child refusal to wear electrodes (n = 1), and an outlier (3 SD above the mean; n = 1), resulting in 149 children with usable ECG baseline and stress induction data. Children who were missing either baseline or task EEG data were significantly different from children who were missing either baseline or task ECG data (baseline: t = −3.779, p < 0.001; task: t = −3.983, p < 0.001). Only two children were missing both EEG and ECG data. Children without complete physiological data were not included in the analysis. There were 133 children who had complete EEG and ECG data. Children with missing EEG or ECG data did not significantly differ from children with complete data on child age, familial income, and parent‐reported internalizing behaviors (ts > 0.059, ps > 0.107). However, girls were more likely than boys to be missing baseline EEG (χ 2 = 4.756, p = 0.041) and stress induction EEG data (χ 2 = 9.941, p = 0.002). Statistical Analyses We first performed a series of Pearson correlations to examine relations among study measures. We next performed a linear regression to examine if changes in RSA from baseline to stress induction moderated the relation between changes in overall FAA and internalizing behavior scores. In the first step, we entered continuous scores of RSA change and overall FAA change as main effects. In the second step, we entered the interaction term between RSA change and overall FAA change scores. The interaction term was computed with centered RSA change and overall FAA change scores. A significant interaction was probed using the Johnson–Neyman method to investigate the influence of overall FAA change on children's internalizing behaviors for children with varying levels of RSA change. Analyses were performed in SPSS version 30.0.0, JASP version 0.19.1, and RStudio 2024.12.0. Results Descriptive Statistics Descriptive statistics and Pearson's correlations for key study variables are presented in Table 1. All variables were normally distributed as indicated by skewness and kurtosis statistics. Overall FAA change and RSA change from baseline to stress induction were not significantly correlated, suggesting they were independent of one another. Further, internalizing behavior score was not significantly correlated with overall FAA change or RSA change across conditions. Child age and income were not significantly correlated with any study variables (all ps > 0.08). There were no sex differences in RSA change scores or parent report of internalizing behaviors, but girls demonstrated greater increase in overall FAA from baseline to the stress induction task than boys (t = −2.23, p = 0.027).2 TABLE 1Descriptive statistics and correlations of main study variables.Variables12345 M (SD)RangeSkewnessKurtosis1. Child age—7.82 (0.43)7.08 to 8.75−0.17−1.272. Income0.06—6.14 (1.45)2.00 to 7.00−1.641.603. Overall FAA change score−0.02−0.06—0.26 (0.57)−1.01 to 2.000.581.034. RSA change score0.040.01−0.09—0.11 (0.72)−1.94 to 1.91−0.300.605. Internalizing behavior0.05−0.100.060.06−0.010.49 (0.37)0 to 1.600.64−0.15 Note: M = mean; Overall FAA change score = change in overall frontal alpha activity from baseline to stress induction (higher score = increase in overall frontal alpha activity); RSA change score = change in RSA from baseline to stress induction (lower score = decrease in RSA, higher score = increase in RSA).Abbreviations: FAA, frontal alpha activity; RSA, respiratory sinus arrhythmia; SD, standard deviation.*p < 0.05.John Wiley & Sons, Ltd. Descriptive statistics and correlations of main study variables. Note: M = mean; Overall FAA change score = change in overall frontal alpha activity from baseline to stress induction (higher score = increase in overall frontal alpha activity); RSA change score = change in RSA from baseline to stress induction (lower score = decrease in RSA, higher score = increase in RSA). Abbreviations: FAA, frontal alpha activity; RSA, respiratory sinus arrhythmia; SD, standard deviation. *p < 0.05. Main Analyses Full regression parameters are presented in Table 2. In the first step, there were no significant main effects of either overall FAA change or RSA change predicting internalizing behaviors. In the second step, there was a significant interaction between overall FAA change and RSA change predicting parent‐reported internalizing behavior (β = 0.278, p = 0.006). As illustrated in Figure 1, the Johnson–Neyman analysis revealed that in children with decreases in RSA from baseline to stress induction (RSA change score below −0.54), overall FAA change was positively related to internalizing behaviors. For children with increases in RSA from baseline to stress induction (RSA change score above 0.89), overall FAA change was negatively related to internalizing behaviors. There were 23.31% of children with RSA change below −0.54, and there were 10.53% of children with RSA change above 0.89. The relation between overall FAA change and internalizing behaviors was not significant for children with RSA change scores between −0.53 and 0.88. TABLE 2Results of the linear regression for overall FAA change and RSA change predicting children's internalizing behavior.Predictors β SE p‐value R 2 ΔR 2 Step 10.0030.003Overall FAA change0.0510.0570.567RSA change0.0100.0450.908Step 20.0600.057** Overall FAA change0.0110.0560.897RSA change0.1400.0500.146Overall FAA change × RSA change0.2780.0810.006 Note: Overall FAA change = change in overall frontal alpha activity from baseline to stress induction; RSA change = change in RSA from baseline to stress induction.**p < 0.01.John Wiley & Sons, Ltd. Results of the linear regression for overall FAA change and RSA change predicting children's internalizing behavior. Note: Overall FAA change = change in overall frontal alpha activity from baseline to stress induction; RSA change = change in RSA from baseline to stress induction. **p < 0.01. Johnson–Neyman plot identifying the region of significance for the moderating effect of RSA change score on the relation of overall FAA change and internalizing behavior in middle childhood. Overall FAA change = change in overall frontal alpha activity from baseline to stress induction; RSA change = change in RSA from baseline to stress induction (lower score = decrease in RSA, higher score = increase in RSA). The association between overall FAA change and internalizing behavior is significant for children with RSA change scores below −0.54 or above 0.89. Discussion In the present study, we examined whether the interaction between two well‐established psychophysiological indices of emotional arousal, change in overall FAA and change in RSA, was associated with children's internalizing behaviors. We found that children who showed increases in overall FAA had a relatively higher level of internalizing behaviors when they also exhibited decreases in RSA from baseline to a stress induction task. This finding highlights the interactive role of brain and heart measures on children's internalizing behaviors, which has been largely underexplored in samples of typically developing children. Our finding is consistent with prior work that has separately examined overall FAA and RSA reactivity in emotionally evocative contexts. Changes in overall FAA under stressful situations are thought to reflect intensity of emotional arousal in both children and adults (Davidson 2002; Fox 1994; Schmidt and Trainor 2001). In this study, increases in overall FAA from baseline to a stress induction may be interpreted as heightened physiological reactivity and stress from anticipating a self‐presentation task, a pattern that has been linked to arousal and anxiety (Dawson 1994; Xu et al. 2023). Changes in RSA under stress also have been associated with emotional reactivity and regulatory ability (Porges 2021). Specifically, a decrease in RSA may reflect a decrease in PNS activity in preparation for a fight‐or‐flight response to perceived threat (Hastings and Kahle 2019; Porges 2021; Seery 2013), and the magnitude of this withdrawal may have implications for social adjustment. Moderate decreases in RSA may facilitate active engagement in emotional situations (Hastings and Kahle 2019; Porges 2021); however, extreme levels of cardiac withdrawal have been linked to hypervigilance and internalizing behaviors (Fortunato et al. 2013; Graziano and Derefinko 2013; Thayer et al. 1996). Taken together, a combination of increases in overall FAA and decreases in RSA during a stress‐inducing task was associated with elevated internalizing symptoms. Different from our hypothesis, we also found that when children showed increases in RSA from baseline to a stressful task, an increase in overall FAA was associated with a relatively lower level of internalizing behaviors. Increases in RSA reflect higher parasympathetic influence on heart rate compared to baseline and under stress (i.e., the speech task in the present study), and this may allow the cardiac system to react more flexibly to the environment (Porges 2021). In line with this theory, an increase in RSA has been associated with lower internalizing behaviors in children (Hastings et al. 2008). In our study, increases in RSA from a baseline condition to a stress induction may allow children with increased reactivity in overall FAA to regulate stress through potentially more adaptive cardiac processes, contributing to lower internalizing behaviors. Of note, we did not find evidence of main effects such that overall FAA change or RSA change individually did not predict internalizing behavior in our sample, which contrasts with some previous work (e.g., Fortunato et al. 2013; Xu et al. 2023). The discrepancy between our findings and those of previous work may be due to low average internalizing behavior in our typically developing, low‐risk sample and/or differences in recording procedures, as our participants were standing rather than seated during psychophysiological recording, as is the case in the majority of other studies. Instead, it was the interaction of the two psychophysiological markers of arousal that was associated with higher levels of children's internalizing behaviors. To our knowledge, there is relatively little research that has examined the interaction between measures of state‐related changes in overall FAA and RSA in the same study in typically developing children. However, our results are in line with other research using complementary cortical measures. For example, one study by Gatzke‐Kopp et al. (2014) found that right frontal EEG asymmetry at baseline significantly interacted with increased electrodermal activity (EDA; reflects greater sympathetic response under stress) during viewing a sad film clip to predict greater internalizing symptoms in preschoolers, although the authors did not find a significant interaction with RSA change from baseline to task. A recent study by Peoples et al. (2024) found that children who had higher anxiety symptoms at Age 5 also showed more negative coupling (i.e., inverse relation between ERN activity and RSA suppression) measured with emotionally evocative tasks at Age 3. This result is consistent with our finding, which also showed a differential effect of the pattern in RSA change from baseline to emotional/stressful task on the relation between frontal brain activity and children's internalizing behaviors. Moreover, previous studies with adult samples have demonstrated a similar association between frontal brain activity and cardiac reactivity in studies of emotion dysregulation and psychopathology (see Davidson et al. [2000], Thayer and Lane [2000], and Waldstein et al. [2000] for a review). It is important to note that overall FAA change and RSA change were not significantly correlated with each other in our sample. This finding may suggest that we were possibly capturing the interaction of two distinct physiological patterns of children under stress, which have been independently associated with internalizing‐related behaviors in previous literature (Fortunato et al. 2013; Xu et al. 2023). We speculate that demonstrating both neural and cardiac physiological responses associated with emotion regulatory difficulties may have a cumulative effect on children's internalizing behaviors. Taken together, the current findings emphasize the importance of considering a combination of brain and heart regulatory patterns when studying children's internalizing behaviors. The present study has theoretical and practical implications. By using reactivity indices from both brain and heart measures, we captured real‐time psychophysiological changes in children's regulation of social stress, illustrating coordination of brain and heart influences on affective chronometry across different contexts (Lapate and Heller 2020). Compared to prior research linking brain and heart in emotion regulation of adults (Thayer and Lane 2000), our findings highlight the importance of examining how different physiological systems jointly operate under stress and further explain vulnerability to internalizing symptoms in children. Furthermore, our findings may pinpoint effective emotion regulation strategies for children targeting physiological arousal. For example, deep breathing exercises in everyday settings have been found to increase state RSA and decrease heart rate (Obradović et al. 2021) and help children to cope with anxiety in stressful situations such as during a hospital visit (Jyskä et al. 2023). The present study should also be interpreted with the following limitations in mind. First, the design of the present study was cross‐sectional, so inferences about causality cannot be derived. Given that children's physiological regulatory systems may change throughout developmental stages (Hastings and Kahle 2019), future studies should examine the relations reported here using a longitudinal design to investigate how children's physiological regulation shapes future developmental outcomes. Further, children's internalizing behaviors were only reported by their parents. Possible reporting bias or discrepancies between parent and child reports might exist. Moreover, our sample was relatively homogeneous in terms of ethnicity, socioeconomic status, and age. Future work should replicate the current findings in more diverse samples to ensure generalizability. Finally, we recruited a low‐risk, community sample of children, resulting in a relatively lower overall prevalence of internalizing behaviors than observed in clinical samples. Thus, it remains unclear how the current findings might generalize to samples of children with clinically diagnosed internalizing and emotional disorders. Conclusion The interaction between changes in brain and heart measures during a stress induction task was associated with individual differences in children's internalizing behaviors during the early school age years. Specifically, children who exhibited an increase in overall FAA in combination with decreases in RSA had the highest level of internalizing behaviors. Conversely, children who displayed an increase in overall FAA in combination with increases in RSA had the lowest level of internalizing behaviors. These findings highlight the importance of considering the combination of neural and cardiac responses under stress when attempting to understand children's emotion regulation difficulties and behavioral problems. Future work should examine these physiological and behavioral dynamics longitudinally to understand how brain–heart–behavior relations develop, coordinate, and predict later socioemotional adjustment and mental health. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Supplementary Information: dev70192‐sup‐0001‐SuppMat.docx