Neurobiological Implications of Parent–Child Emotional Availability: A Review

Parental influences are important for a child’s behavior, overall adjustment, as well as cognitive/language development. New research is exploring how relationships with parents can influence a child’s neurobiological functioning and development. In this systematic review, our first aim is to describe how the caregiving environment influences these aspects of child development. The second and main aim is to review and recommend that the concept (and measurement) of “emotional availability” may provide a new window in this continued exploration. Emotional availability (EA) refers to the capacity of a dyad to share an emotionally healthy relationship. The EA Scales assess this construct using a multi-dimensional framework, with a method to measure the affect and behavior of both the child and adult partner (caregiver). In this review, we first provide an overview of child development research, with regards to stress physiology, neuroendocrine system, genetics and epigenetics, and brain mechanisms. We then summarize the results of specific EA research in these areas, and propose a theoretical model integrating these constructs. Finally, we offer areas for future research in this area.


Introduction
There is a great deal of scientific focus on the effect of early experience on a child's neurobiological functioning and development. This line of research began with animal studies uncovering that maternal licking/grooming behavior in rodents is associated with less offspring anxiety and more optimal prefrontal cortical functioning [1]. In humans, evidence has accumulated that institutionally reared children have larger amygdala volumes as compared to those raised in the home setting [2]. The amount of time spent in the institutional environment has also been linked with amygdala volume [3]. Those children raised in the home setting, but under conditions of trauma, have also been a topic of focus. Trauma, such as child maltreatment exposure, can be harmful, in particular, when traumatic experiences happen at a young age, and such exposures may have long-term adverse effects on individuals. Specifically, studies of child maltreatment-exposed mothers have shown that they have a higher risk for adult psychopathological problems and low-quality parenting behaviors when compared to non-child-maltreatment-exposed mothers [4]. Early maltreatment has also been found to be associated with a lesser volume of the child's corpus callosum, as well as lesser hippocampal volume in adulthood [5]. Further, maltreated children have been found to show enhanced neural responsiveness to "angry stimuli" such as facial expressions [6]. Additional work by Neukel, et al. [7] indicates that mothers with a history of early maltreatment are more likely to show effort in processing their own infant's facial expressions of emotion, as evidenced by elevated activation of areas that are associated with the visual processing of faces (e.g., cuneus, middle temporal gyrus). These studies have established the impact of early adverse experiences on brain and physiological functions.
Outside of extreme adverse circumstances, such as child maltreatment, human research on the influence of the environment is only beginning. Learning how "normative" caregiving influences may be associated with children's brain and physiological functioning is of vital importance. As such, Belsky and de Haan [8] wrote that research on the role of the environment on the brain is at the point of "the end of the beginning", and that "work is now needed to determine whether and how variation in parenting in the normal range affects the brain development of children not exposed to extreme adversity" (p. 423).
While the evidence is scant, research indicates that parental warmth during interactions with one's child at 18 and 24 months, as well as at 10 and 11 years, was associated with the child's prefrontal cortical responding during a reward or loss, with maternal warmth serving an especially protective role in the case of boys who were exposed to maternal depression. Thus, even in the context of maternal depression, children (especially the boys) benefitted from the buffering role of maternal warmth [9].
Another important line of work is on attachment security in infancy. While a great deal of research has informed us of the importance of attachment in predicting behavioral and affective development, attention is only now turning towards the effect of attachment on brain and neural function. For example, Leblanc, et al. [10] reported that attachment security during infancy predicted whole-brain gray matter volume (assessed through structural magnetic resonance imaging) in the superior temporal sulcus dan gyrus, temporo-parietal junction and precentral gyrus, which are involved in social, cognitive, and emotional functioning in late childhood, at 10-11 years of age. However, there was no association with cortical thickness [10].
Maternal sensitivity has also been investigated to some extent, in relation to brain development in typically developing children. Higher maternal sensitivity has been found to be associated with greater subcortical volume in infants [11], but also to smaller hippocampal volume and smaller amygdala volume in infants [12]. In contrast, Kok, et al. [13] reported that early parental sensitivity (maternal and paternal combined) was actually not linked with hippocampal or amygdala volume, revealing some inconsistencies in specific brain regions in relation to this parenting construct. However, as would be expected, parental sensitivity was in fact associated with total brain volume in school-age children [12]. Moreover, as would be expected, the mother's self-reported hostility, as well as observed aggression, were related to the child's smaller total volume and cortical thinning, which are structures that are important for children's social emotional/cognitive development, as well as emotion regulation [14]. Given the findings by Kok, et al. [15] that indicate a clearer link between brain volume and behavior for girls, at least in relation to prosocial behavior, additional research is needed to better understand the link between brain morphology and how this translates into behavior for each gender.
In addition to brain structures, research has also explored the link between the caregiving environment and neural activity, as measured by electroencephalography (EEG) or magnetoencephalography (or MEG). The EEG is a noninvasive means to assess the brain's neural activity via the recording of electrical signals at the scalp. For an MEG, a Dewar containing multiple sensor coils are near, but not touching, the participant's head, with recording of the brain's magnetic activity. Perhaps the clearest example of this is a study by Hane and Fox [16], where they measured the observed caregiving environment (using a composite of the different Ainsworth scales, including sensitivity, acceptance, interference, degree of availability, pacing during feeding), and referred to this as maternal caregiving behavior (MCB). They found that low-quality MCB was associated with greater right frontal electroencephalographic asymmetry and fewer positive behaviors (e.g., fear, less joint attention). Further, they also found that lower MCB was associated with more negative observed infant affect. Levy, et al. [17] found a link between the caregiving environment (using a molecular index of mother-child synchronous interactions) and MEG recordings for that child many years later.
In addition to early deprivation and maternal behaviors, chronic stress has also been of emotional availability (child responsiveness to the adult and child involvement of the adult). The child's side has been virtually ignored in the field of affective neuroscience, except for some minimal investigation of the child's attachment security. A system that includes both the parental and child sides, as well as attachment security/bonding, would appear to be useful.
In recent decades, emotional availability (EA) has been a topic of interest for developmental researchers, and has been widely studied in the scientific and scholarly literature [30]. The current conceptualization of emotional availability encompasses the following two aspects: (1) how parental EA influences the child's development and responses, and (2) how child EA influences the parent's feelings of value and ability to provide appropriate care for their child [30]. In 2014, Biringen, Derscheid, Vliegen, Closson and Easterbrooks [30] conducted a comprehensive review of the published EA literature that emphasized the need to include at-risk populations with mental health disorders and/or disabilities, and to increase EA measures in intervention work [30]. Since the 2014 review, EA researchers have indeed increased their focus on at-risk populations, and have provided valuable information on a variety of disorders (e.g., depression, substance use disorders, personality disorders, and schizophrenia), as well as a new measure of prenatal EA [32,33]), which can be assessed both as self-report as well as observation [34]. Moreover, past research has suggested that environmental experiences and behavioral practices can influence brain structure and function, and may be shaped by genetic co-evolutionary mechanisms [26]. However, there has been limited research focused on understanding the underlying biological and neural mechanisms of EA on parent-child interactive behaviors and development. Moreover, cultivating secure, emotionally available relationships, along with developing appropriate stress regulation skills, are important aspects to a child's healthy brain development [24,35,36]. As the dyadic nature of EA focuses both on the parent as well as the child, with the goal of improving the parent-child relationship, addressing physiological cascades that are influenced by improved EA would dramatically add to our understanding of affective neuroscience, as well as the intergenerational genetic transmission of emotional connections. Thus, a review of stress physiology and the neuroendocrine system, genetic and epigenetic influences, and brain mechanisms associated with EA, is warranted.
Our first aim of this review is to explore and describe how the caregiving environment influences a child's neurobiological functioning and development. Our second, and primary aim, is to review and recommend the concept (and measurement) of EA as an innovative methodology that will add important information into furthering this field's research. Our review unfolds as follows. First, we provide an overview of the EA framework. Second, we review neurobiological factors that are related to EA, with regards to stress physiology and the neuroendocrine system, genetics and epigenetics, and brain mechanisms. We present a theoretical model beginning with prenatal EA and postnatal EA. We then discuss implications of improving the health and wellbeing of parent-child relationships, as well as the potential for informing future interventions.

The EA Framework
The EA framework comprises the following three operationalized measures that allow researchers to evaluate how each member of the parent-child dyad affects the other member: (1) observational EA Scales, (2) observational EA zones, and (3) EA Self-report [30]. The observational EA measures are currently on their 4th edition [30].

The EA Scales
The EA Scales specifically differentiate adult and child experiences and perspectives for use in research and clinical practice. Specifically, there are the following four dyadic EA Scales for the parent: (1) sensitivity, (2) structuring, (3) nonintrusiveness, and (4) nonhostility. Further, there are the following two dyadic EA Scales for the child: (1) responsiveness and (2) involvement [30].

Adult EA Scales
Adult sensitivity refers to the caregiver's ability to attend to the child's emotional needs and behavioral cues; this scale is dyadic, and is dependent on the relationship with, and the response of, the child. Adult structuring is the ability of the caregiver to support the child's activities and autonomy through guidance, scaffolding, or mentorship. Adult nonintrusiveness refers to the caregiver's lack of interference with the child's behaviors through over-direction, over-stimulation, interference, or over-protection. Adult nonhostility is the absence of hostile acts that parents may intentionally or unintentionally directly target towards their children. A parent's behavior is rated and viewed in a manner that is dependent on the way the child responds [30].

Child EA Scales
Child responsiveness refers to the responsiveness of the child to the parent's bids. Child involvement refers to the child's ability and desire to involve the caregiver in their play or activities [30].

EA Zones
The EA System now also includes EA zones for the parent and the child, which align with the four attachment styles of secure, insecure-avoidant, insecure-resistant, and disorganized [37]. These zones provide a dimensional view of "emotional attachment" (on a 100-point scale), in addition to providing information on the four attachment categories. Moreover, the EA zones have been validated in terms of secure/insecure attachment in different studies, using the strange situation procedure, the attachment Q-set, and the adult attachment interview [38]. Specifically, the EA zones include the following four different zones of emotional attachment of the parent and child: (1) emotionally available, (2) complicated, (3) detached, and (4) problematic/disturbed [29].
The zone placement for the parent is predominantly dependent on the parent's sensitivity score, and the zone placement for the child relies on the child's responsiveness score, with higher scores in their respective scale being associated with being in the "emotionally available" zone. The mid-range scores of child responsiveness refer to a "complicated" emotional attachment, and low scores, below the mid-range of child responsiveness, are referred to as being "detached". Finally, "problematic/disturbed/traumatized or traumatizing" is the lowest zone [29]. The additional EA dimensions (structuring, nonintrusiveness, nonhostility, and child involvement) are further important when assessing the relationship, and may be used to place an individual into the lowest zone, but sensitivity and child responsiveness scales "kickstart" the EA zones placement decision [29]. Most importantly, the EA zones provide an attachment perspective from both the parental side as well as the child side, which extends traditional attachment research in which observing only one perspective is a common practice, with the parent and child having their own distinct zone, which may or may not be the same, suggesting that the attachment of each may be different. Additionally, within each zone, there is the option to rate some higher than others, with the inclusion of a dimensional way of looking at a zone, as for example, low on emotional availability, but nonetheless in the emotionally available zone, or high on complicated, but still in the complicated zone rather than one zone up, which would be the emotionally available zone. Finally, the EA zones include a description of emotions as they impact attachment; thus, they are zones of "emotional attachment".
1.6. EA Self-Report (EA-SR) The EA Self-Report (EA-SR) is a parent-reported measure that is inspired by the observational EA Scales and correlates with the observational EA dimensions, to a moderate extent [39,40]. The EA-SR consists of 36 statements that are rated on a 5-point Likert scale, from "1" (Almost Never) to "5" (Always), which assesses the parent's perception of the parent-child relationship, and the EA between themselves and their child. Parents are asked how characteristic each statement is about their relationship with their child, with questions such as, "My child clearly enjoys being together with me". At this time, and with only one study that conducted a factor analysis [41], the EA-SR items load on to the following five factors: (1) mutual attunement, (2) affect quality, (3) dyad interactions, (4) intrusiveness, and (5) hostility. The EA-SR has demonstrated good internal reliability and construct validity compared with the observational EA Scales [40,41]; however, it is important to note that the link between the observational EA Scales and the EA-SR is moderate at best [41]. The EA-SR is now included in the overall EA System, but it is still recommended to use the EA Scales or EA zones rather than the EA-SR, if the situation permits. The EA-SR is referred to as "self-reported EA" rather than EA, since the latter refers to the observational system.

Prenatal EA
In addition, a new measure of EA has been developed-assessing prenatal emotional availability (pre-EA), which has been validated in Finland [34]. In the pre-EA, the mother is videotaped while performing activities with her fetus. Unlike the EA (postpartum to 14 years, soon to be extended to 17-19 years) that is scored on six scales (described above), the pre-EA only includes the following two maternal scales: maternal sensitivity and nonhostility. Assessment of the pre-EA focuses on the affective and behavioral cues of the mother. The pre-EA measure has similarities to previous work on prenatal attachment; however, the pre-EA is an observational measure about the mother in interaction with her unborn baby, in contrast to prenatal attachment measures that are typically self-reported [32,33]. As numerous studies have examined postpartum or early postnatal EA to two months of age, it would also be a beneficial addition to incorporate further study of EA during pregnancy [42][43][44].

Methods
The preferred reporting items for systematic reviews and meta-analyses (PRISMA) statement was designed to transparently report the reasoning, methods and results discovered by the authors, and improve the quality of reporting in systematic reviews through an evidence-based 27-item checklist [45]. This systematic review was conducted in adherence to the guidelines and checklist. A visual representation of this review's search and selection criteria is depicted in Figure 1.

Literature Search
The journal articles included in this review were compiled through online searches, utilizing the entirety of the Colorado State University (CSU) online library as well as Google Scholar to access 374 journal databases. A general search of the key terms scoured all 374 databases in the online library, which included databases such as PsychInfo and Medline. Searching for studies in this way assured that all studies using an EA system, and those that were published in a database made available through the CSU online library, were included and found. Thus, the literature search was conducted by searching for terms in the general online library, rather than searching each database individually.

Study Selection
Studies were selected a priori using the following inclusion criteria: (1) utilized at least one of the measures in the EA System (EA Scales, EA-SR, or EA zones) as a measure; (2) directly included the measure in the reported results; and (3) related the measure to the neurobiological sciences. The exclusion criteria was as follows: (1) studies not using the EA System; (2) did not report on one of the EA measures in the results; (3) did not relate to neurobiological sciences; and (4) were theses or dissertations, and therefore, not peer reviewed. Thus, the authors chose to include any study using any part of the EA System and focusing on neurobiology, and this was conducted on an a prior basis, given that we wanted to focus on the literature on EA specifically.

Data Extraction
We extracted and condensed the following information: (1) target population, including age, sample size, and sample characteristics; (2) the EA measure used in the study and how it was scored; (3) any applicable additional variables (e.g., cortisol sampling or brain imaging); (4) EA reliability scoring, where the observational system is used; and (5) key results. The tables which show this in detail are at the end of the document.

Quality Evaluation
The quality of the included articles was rated in line with guidelines by the Cochrane Handbook for Systematic Reviews of Interventions [46].

Results
The literature search yielded 7633 articles. After excluding duplicates and studies that did not meet the inclusion criteria, 26 articles in total were included in this systematic review.
3.1. Stress Physiology, the Neuroendocrine System, and EA 3.1.1. Stress Physiology Throughout the existing literature on EA, multiple studies have found a connection between this aspect of parent-child relationships and child stress physiology. In one study using the EA Scales to measure parent sensitivity, structuring, nonintrusiveness, and nonhostility, Kertes, et al. [47] used a principal component analysis to combine these four measures of observed parental EA into a factor score that was labeled "Parenting Quality". Their results suggested that sensitive, supportive parenting that was indicated by a higher score of parenting quality, may act as a buffer of the HPA axis stress response in infants, toddlers, and preschoolers [47].
This result is also consistent in samples of children with an immigrant background, in which those who experience higher EA scores on sensitivity, structuring, nonintrusiveness, child responsiveness, and child involvement, demonstrated more effective stress regulation at kindergarten entry [48]. Rickmeyer, Lebiger-Vogel and Leuzinger-Bohleber [48] reported that in immigrant households (immigrants from some 17 different cultures in Germany), the children's cortisol levels were higher after kindergarten entry than before; lower EA before kindergarten entry was associated with a rise in hair cortisol concentrations (HCC) (negative correlation); the children with lower mother intrusiveness and higher child responsiveness showed lower cortisol increases. Similar to the above, but using the EA zones, which is a measure of "emotional attachment", Senehi, et al. [49] investigated HCCs in relation to adverse childhood experiences (ACEs) and EA. They found that only in the low EA zones (which was defined as anything that is not in the emotionally available zone), there was a link between ACEs and HCCs. Thus, for children with parental history of multiple ACEs, those children of parents with higher EA had lower HCC compared to those children of parents with lower EA, indicating the buffering role of EA [49].
Similarly, Philbrook and Teti [50] examined EA at bedtime in relation to the stress reactivity of the infant. Infants with mothers who were scored as more emotionally available (measured through creating a composite maternal EA score with the EA Scales 3rd edition) at bedtime, showed less infant distress and more sleep (coded through videosomonography, where the infant's sleep-wake states were objectively assessed), as well as lower infant salivary cortisol levels, as compared to those with lower EA mothers [50].
Another study by Ruttle, et al. [51] used the EA Scales from the EA 2nd edition to create a variable capturing dyadic behavioral sensitivity through a principal component factor analysis with sensitivity, structuring, nonhostility, child responsiveness, and child involvement, and examined the synchronization (or "attunement") of the hypothalamicpituitary-adrenal (HPA) axis between the mother and child. The results of this study indicated that such cortisol attunement only occurred during significant periods of challenge, and only in behaviorally sensitive mother-child dyads [51].
Several studies have linked EA and stress, but few have looked at the relationship between EA, stress, and stress mitigating factors. One well-researched stress contributing factor is the impact of low socioeconomic status (SES). Tarullo, et al. [52] examined the link between SES and cortisol functioning in early childhood, using a multidimensional approach to assess SES and cortisol concentration. Measuring income-to-need, parent education, occupational prestige, neighborhood risk, food insecurity, and household chaos allowed the researchers to look at the underlying SES factors that play a role in cortisol functioning. Additionally, researchers used both hair cortisol concentrations (HCC) and salivary cortisol concentrations (SCC) to measure how SES risks may contribute to parenting. The EA sensitivity scale was used to measure parenting. A relation between higher child HCC and higher parent HCC, as well as lower SES (indicated by parent education, occupational prestige, and greater food insecurity), was established; however, lower parental sensitivity was only related to higher child HCC in the 3.5-year-old group compared to the 12-month-old group. Parental sensitivity was unrelated to the infant cortisol measures. Despite finding a connection between parental sensitivity and HCC, parental sensitivity was not found to mediate links between SES risks and HCC in the 12-month-old group or 3.5-year-old group [52]. These results suggest that the interactions between parent sensitivity and child hair cortisol are likely to be complex, and it would therefore be beneficial for researchers to utilize the entirety of the EA Scales in order to more accurately observe how SES risks may play a role in parenting and child cortisol functioning.

Skin Conductance and Heart Rate Variability
Other methods to assess stress physiology have been heart rate variability and skin conductance. An interesting study by Gilissen, et al. [53] reported that temperamentally fearful children showed less stress reactivity to fear-evoking film clips if they were being raised by higher EA mothers, but this link was not found for those who were not temperamentally fearful. These findings support the importance of the rearing environment, especially for this group of more susceptible children [53].
EA research, to this point, has consistently found a connection between EA and stress; therefore, it is important to further investigate rigorous, appropriate measures of stress (e.g., hair and saliva cortisol), along with the potential adverse effects of stress, in order to best inform prevention and intervention research. Moreover, it is recommended to use the entire EA System (i.e., sensitivity, structuring, nonintrusiveness, nonhostility, child responsiveness, and child involvement) when assessing for EA, rather than a select number of dimensions, such as only sensitivity or parenting scales, as each dimension has the potential to add unique information to the research.

Testosterone
Of course, father-child interactions are the place to study the link between basal testosterone levels and EA. van der Pol, et al. [54] studied this link in the context of selfregulation, and found that higher testosterone levels in the evening were associated with less respect for the child's autonomy, but only if the father also had problems with selfcontrol. For fathers who were generally well regulated, higher testosterone in the evening was associated with greater sensitivity [54].

Oxytocin
Oxytocin is another important neurotransmitter to include when discussing the neuroendocrine system and caregiving. MacKinnon, et al. [55] examined the influences of oxytocin during the end of pregnancy and postpartum on maternal behaviors, using the EA scales. Researchers found that oxytocin that was measured at 32-34 weeks gestation was indirectly associated with greater maternal structuring and fewer intrusive maternal behaviors at 2-3 years postpartum, through the mothers' theory of mind (one's ability to understand and take into account another individual's mental state) [56]. The results of this follow-up study indicated that one's theory of mind ability may represent a social cognitive mechanism linking endogenous maternal oxytocin and maternal caregiving [55].
In a study by Naber, et al. [57], the influences of oxytocin on father-child interactions were investigated using intranasal oxytocin administration and the EA Scales. Specifically, fathers who received the intranasal oxytocin were able to stimulate their child's exploration and autonomy more effectively compared to the placebo condition. Moreover, fathers in the oxytocin condition tended to demonstrate less hostility compared to the placebo condition. However, the oxytocin condition did not lead to increases in sensitive interactions. This is an intriguing example of the transactional nature of parenting and the neuroendocrine system, with changes in oxytocin levels inducing changes in caregiving, and specific parenting behaviors leading to changes in oxytocin levels. These findings suggest that oxytocin may be an important mechanism in parenting, by which socially relevant cues in the dyad activate dopaminergic pathways and can positively reinforce responsive parenting behaviors in a positive feedback loop [57].

Genetics
In a study by Reichl, et al. [58], researchers examined three variants of the OXTR gene on parenting behavior as well as plasma oxytocin. The EA Scales were used to measure one dimension of EA, which was maternal sensitivity. One of the OXTR gene variants, rs53576, did demonstrate a direct effect on maternal sensitivity; however, the other two gene variants, rs2254298 and rs104778, did not have this direct effect. Moreover, another of the OXTR variants (a rare variant A of the rs2254298) did moderate the relation between mothers' experience of childhood abuse and her sensitivity, suggesting that this variant may act as a protective factor between mothers' childhood abuse and their parenting sensitivity. It was also found that mothers' childhood abuse was related to lower maternal sensitivity and to a higher child abuse potential, but only among the mothers who were homozygous for the common allele (GG) compared to those who were not homozygous. Finally, the G allele of the OXTR variant rs2254298 was associated with elevated plasma oxytocin levels [58]. Overall, developing a more comprehensive understanding of the interconnections between genetic level variations and full assessments of parent and child emotional variability would provide the field with information that could be used to provide predictions of maternal behavior.
As the EA System assesses parent-child relationship quality from both parent and child perspectives, the use of EA in assessing qualities would provide essential information about potentially modifiable aspects of the relationship that may influence the oxytocinergic system. Moreover, the incorporation of EA in genetics research would further provide information needed for targeted, preventative interventions, designed to promote healthy parent-child relationships.

Epigenetics
Promoting healthy parenting may help promote the healthy intergenerational transmission of genetic material. Lecompte, et al. [59] examined mother-child interactions, preschool-aged child's controlling-attachment behaviors, and DNA methylation of the oxytocin receptor (OXTR) gene in the child. Mother-child interactions were assessed by a 5-min free play and coded with the EA Scales, child attachment behaviors were measured by the separation-reunion procedure, and children's DNA methylation of the OXTR gene was obtained by buccal swab. They found that lower maternal sensitivity was predictive of a more child controlling-caregiving form of attachment and less structuring was pre-dictive of controlling-punitive attachment. Maternal structuring was associated with the hypomethylation of the OXTR gene (associated with caregiving) [59].
Further, a different study by Lewis, et al. [60] investigated how epigenetic processes may be a mechanism by which early social experiences may shape immune functioning and general health. This particular study investigated the differences in monozygotic twins, to assess how their primary caregiver's emotional availability with each of them, measured by the EA-SR over time, predicted their immune gene methylation at 8 years. They found that maternal self-rated emotional availability at a child age of 1 year was related to the methylation of multiple inflammation genes in the monozygotic twins at 8 years of age. Further, twin pairs who were discordant in health had more differences in their methylation of inflammation genes, and twin pairs who were discordant in health had larger differences in EA with their mothers, compared to twin pairs who were more concordant in health. However, note that the EA findings are based on self-reports, such that mothers may have been rating these relationships based on the child's health. Alternatively, there may have been real differences in EA in these mother-twin relationships. It would be important to investigate EA in an observational context [60].

Brain Mechanisms
EA research has largely focused on behavioral factors. Much less is known about the brain mechanisms that are associated with EA. Existing studies have shown the importance of parent-child interactions in children and parents' brain development. To our best knowledge, there are 13 studies in total linking brain mechanisms to different EA constructs.

Neglectful Mothers
Neglectful parenting is considered an adverse childhood experience (ACE) for children, and can have negative long-term impacts on children's health and development [61]. Moreover, neglectful mothers tend to show lower levels of emotion expression and less reactivity to their child's emotional signals during mother-child interactions [62].
Two recent studies investigated the EA Scales in relation to structural brain healthneglectful and non-neglectful mothers. In the first study, Rodrigo, et al. [63] studied the white and gray matter volumes of 25 neglectful (NM) and 23 non-neglectful or control mothers (CM), and related this to mother-child interactions. They performed structural MRI with voxel-based morphometry to examine brain volumes, and coded a gameplay task of the mother and child with the six adult and child EA Scales. Compared to CM mothers, NM mothers showed smaller gray matter volume in the right insula, anterior/middle cingulate, and right inferior frontal gyrus, as well as less white matter volume in the bilateral frontal regions. Specifically, areas related to empathy, regulatory control, and reactions to infant pain and distress, as well as the mirror neuron system, which enables the parent to attune at an intuitive level, appear to be affected in NM mothers [63].
Additionally, in a structural MRI study by León, et al. [64], cortical thickness and surface area were examined for 45 mothers (24 NM and 21 CM mothers), and examined in relation to EA as well as alexithymia (difficulty identifying and expressing emotions) traits. Compared to the CM mothers, the NM mothers showed thinner cortical thickness in the right rostral middle frontal gyrus and orbitofrontal cortex, and greater surface area in the right lingual and lateral occipital cortices. Further, although the direct path between the right rostral middle frontal gyrus and EA was significant, and remained significant even when the alexithymia traits were subtracted, nonetheless, such alexithymia traits reduced the strength of the relation. Thus, the frontal areas of the brain have important implications in the bidirectional attunement of mother-child interactions. The results of this study suggest the importance of further considering the association of cortical thickness and EA among NM mothers, and providing insights into the understanding of the neural mechanism behind neglectful mothers' behaviors [64].
In sum, these two studies, examining the effect of neglectful experiences on the mother's own brain volume and functionality, may help the design of intervention pro-grams that seek to improve maternal response to infant distress cues and signals, as well as assess interventions that can create the context for modifications of the brain.

Child Maltreatment-Exposed Mothers
Neuroimaging studies of linking child maltreatment-exposed (CME) mothers to EA constructs are sparse. We identified five studies here in this review. First, Rodrigo, et al. [65] studied the white matter alterations in neglectful mothers who had experienced maltreatment in childhood. They used the six EA Scales to access the quality of the emotional exchanges of the parent and child in 22 neglectful and 22 control mothers with children who are younger than 5 years old. They performed functional MRI to estimate six elements of diffusion tensors and associated fractional anisotropy and mean diffusivity, and tracked the white matter connectivity between the frontal, temporal, and occipital regions of the brain. Overall, the two groups of mothers differed by the streamlines of the inferior fronto-temporo-occipital connectivity, in particular, neglectful mothers showed fewer streamlines in this region that is known for the involvement of the face-processing network. Furthermore, they found fewer streamlines, especially in the right inferior longitudinal fasciculus tract (IFL-R), and observed that this was predictive of lower EA scores. Neglectful mothers showed lower sensitivity to the child's cues and demands, as compared to non-neglectful mothers. These white matter alterations that form atypical fronto-temporooccipital patterns, have an impact on the interactions indexed by EA between a mother and her child; thus, greater volume in the ILF-R was also predictive of positive mother-child interactions [65].
In addition, Olsavsky, et al. [66] recruited 26 CME and 19 NE mothers to participate in a mother-infant social communication study, and measured amygdala reactivity from functional MRI during an adult-infant face task. They used a 15 min mother-infant free play to code the four maternal EA scales (sensitivity, structuring, nonhostility, and nonintrusiveness), with a focus on maternal sensitivity. They found that CME mothers showed higher amygdala reactivity to infant faces, and this was associated with greater maternal sensitivity during mother-infant interactions, but such an effect was not observed during adult face tasks. The study results suggest that childhood maltreatment could affect maternal neural processing of social cues of infants [66].
Moreover, Olsavsky, et al. [67] focused on mothers with CME and how such experiences influence their responses to children's cues. Specifically, they collected information on mothers' CME through the risky families questionnaire (using the physical abuse, verbal abuse, and witnessed domestic violence questions), and recorded (by functional MRI) amygdala activation and connectivity to motor planning and empathy regions in the brain in response to their child's cry vs. other stimuli (other infant's cry and white noise). The researchers found that mothers reporting higher levels of CME showed higher amygdala activation to their own baby's distress cries when compared to other infant's distress cries or white noise. Thus, the connectivity between the amygdala and the areas associated with the motor planning and empathy regions of the brain (middle frontal gyrus) are heightened with greater mother-reported CME. Moreover, higher amygdala activation and heightened connectivity with prefrontal areas were positively associated with maternal nonintrusiveness (curiously, with no differences documented for maternal sensitivity and hostility). Those who reported greater CME may have been showing a maternal adaptive response (i.e., an increased sense of protection over the infant) in the context of their infant's distress signals (vs. non-distress signals) [67].
Neukel, et al. [68] investigated the caregiving behavior of mothers with early life maltreatment (ELM) history (physical and/or sexual abuse or neglect), and used the maternal sensitivity scale of EA to investigate the effect of ELM on interactions in the next generation. They invited 47 mothers (22 ELM and 25 control mothers, who had none of the above indicators of maltreatment in their early life history) to interact with their children (7-11 years old), and collected their functional MRI during real-life interactions as well as imagined conflictual and pleasant interactions. The findings indicated that the ELM mothers were less sensitive than the control mothers in their real-life interactions, but while imagining conflictual interactions with their own child, they showed increased activation in the amygdala, insula, and hippocampus, which are known to be involved in emotion-processing, with early experiences of trauma likely sensitizing these mothers to their child's distress communications, but not their pleasant communications. Thus, ELM mothers appear to be highly vigilant and responsive to negative interactions with their own child, but lack proper responsiveness and sensitivity during imagined pleasant or real-life mother-child interactions [68].
Mielke, et al. [69] investigated the relation between maternal sensitivity and "the empathic brain" in 25 mothers with, and 28 mothers without, an ELM history of physical and/or sexual abuse. Mother-child pairs were invited to engage in a 15-min free play and a 6-min problem-solving task, and their interactions were rated by EA Scales (only the maternal sensitivity scale). A voxel-based morphometry method was used to collect the structural MRI of the mothers' brains. Overall, compared to mothers without ELM, mothers with ELM were less sensitive when interacting with their own child, and yet there was a huge range in sensitivity in both the groups. Importantly, for mothers with ELM, a positive relation was found between their maternal sensitivity and their self-reported cognitive empathy, as well as grey matter volume (in the superior temporal sulcus, temporal poles), which are core regions of the "cognitive empathy" network. For mothers in the control group, on the other hand, their maternal sensitivity was related to grey matter volume (in the anterior insula), which is a core region of the "emotional or intuitive empathy" network. The authors concluded that mothers with ELM compensate for their emotional deficits by recruiting more brain regions involved in cognitive empathy when engaging in the gameplay [69].

Normative Samples
Six studies examine brain mechanisms in normative samples. A study by Firk, et al. [70] focused on the amygdala response to infant crying, but in a normative sample. Firk, Dahmen, Lehmann, Herpertz-Dahlmann and Konrad [70] investigated 26 mothers without any genetic syndrome or severe disease with their fullterm infants. The quality of the mother-infant interactions was assessed by a 12-min free-play coded by using EA Scales (maternal sensitivity, maternal instructing, maternal nonhostility, and maternal nonintrusiveness) and a 6-min still-face task. Functional MRI was used to collect the mothers' amygdala activities while presenting sound stimuli (own infant cry, other infant cry, or control sound). The authors reported that maternal sensitivity and maternal nonhostility were negatively associated with amygdala activation during mother-infant interactions. Furthermore, downregulation was shown in the amygdala when mothers practiced self-distraction as an emotion regulation method. Therefore, it was suggested that enhanced activation in the amygdala, in response to infant crying, might be related to less sensitive and more hostile maternal behaviors; fortunately, self-distraction is an effective method in decreasing the emotional response to infant crying [70].
Another study looking at families who are socioeconomically disadvantaged (again in a normative sample), examined amygdala responses, using fMRI, to infant emotional expressions (positive, negative, and neutral faces). Compared to the previous study that took a multidimensional approach to SES, Kim, et al. [71] only looked at the incometo-needs (ITN) ratio, but they did use the entirety of the EA Scales and further only looked at maternal sensitivity and nonintrusiveness. They found that mothers did have an elevated amygdala response to negative infant faces and less neural responses to positive infant faces, and this was associated with the mothers' parenting behaviors and ITN ratio. Their indirect effect finding suggested that a lower ITN ratio was linked to elevated amygdala responses to negative infant faces, and this was further associated with lower nonintrusiveness (meaning the mother was observed to be more intrusive) during the interaction. Similarly to the study by Tarullo, Tuladhar, Kao, Drury and Meyer [52], there was no effect for maternal sensitivity [71]. This finding has huge implications for future interventions for low-income mothers, who may be more vulnerable to altered neural processing of emotional expressions, and calls for more research on the links between SES, stress physiology, neural processing, and all of the EA Scales. For example, it would be interesting to see how household chaos (a SES index) may be related to, or impact, the EA scale structuring, due to unstable routines. Insight into the different EA dimensions, including child responsiveness and involvement, may provide valuable information on the relation between cortisol functioning and low SES, and inform future child intervention and prevention efforts focused on low-income children and youth.
In addition to the aforementioned studies, there are additional studies on brain correlates of EA Scales on the parent-child relationship that are worth noting. First, Schneider-Hassloff, et al. [72] recruited 4-6-year-old children and their mothers, to investigate whether EA is associated with actual behavioral and electrophysiological measures of executive functioning (EF). They administered behavioral EF tasks (head-toes-knees-shoulders task, HTKS, and a delay of gratification task, DoG), and evaluated with the observational EA System. The electrophysiological correlates of EF (go/nogo tasks) were administered and evaluated using event-related potentials (ERP). The authors reported that higher EA nonintrusiveness, as well as higher EA zones scores, were positively associated with the behavioral aspects of EF (that is, HTKS and DoG), and both maternal structuring and nonintrusiveness, and in some cases child responsiveness, were associated with the electrophysiological correlates of EF (that is, in the go/nogo task) [72]. These findings suggest that parenting qualities are associated with the functionality of neural circuits that are involved in the response inhibition component of EFs.
To better understand the child's side of emotional availability in the mother-child relationship, Licata, et al. [73] investigated left frontal electroencephalogram (EEG) activation vs. right frontal activity in 14-month-old children, and measured EA when children were 7 months and then again at 50 months old. They reported that a higher left frontal EEG activation (as compared to right activation) at 14 months was associated with higher child involvement at 50 months, even when controlling for earlier maternal sensitivity, child responsiveness, and child involvement at 7 months, as well as these EA variables at 50 months. This result implies that child involvement may be dependent on biological factors within the child, and suggest that left frontal asymmetry is related to a child taking the initiative and proactive tendencies in the mother-child relationship. There was no correlation between child responsiveness and frontal EEG activation. Instead, child responsiveness was more closely associated with the mother's sensitivity [73].
To fully capture all six EA Scales' relations to different infant emotion cues, Killeen and Teti [74] examined mother's frontal EEG asymmetry at rest and during emotionstimulated (joy, anger/distress, and neutral interest) and free-play videos of their infants. Specifically, the researchers recruited 27 right-handed mothers and their 5-8-month-old infants, and recorded infant emotion and mother-infant free-play videos. The free-play videos were used to code for mother-infant EA Scales (maternal sensitivity, maternal structuring, maternal nonintrusiveness, maternal nonhostility, child responsiveness, and child involvement). It was found that maternal frontal EEG asymmetry at rest and during infant emotion videos were not related to mother-infant EA in response to emotionalstimulated states and infant emotion videos, respectively. However, in response to infant anger/distress videos, maternal sensitivity and structuring were related to a shift towards greater relative right frontal activation [74]. The results imply that maternal in-the-moment empathetic response to the infant was related to mother-infant EA.
Moreover, a final study by Taylor-Colls and Pasco Fearon [75] linked parental sensitivity and infant's neural responses, measured by event-related potentials (ERP), which is an electrophysiological response to a stimulus. Specificity, 40 healthy mothers and their 7-month-old infants were invited to participate, and their interactions were coded using EA Scales (maternal sensitivity, maternal structuring, maternal nonintrusiveness, and maternal nonhostility) during a free play. These maternal dimensions were standardized and summed to form an EA composite. Although the authors seemed to refer to this composite as sensitivity, it is in fact a composite of all the adult scales. Further, given that they used the 4th edition of the system, these are measured on the same metric and contribute equal amounts. The authors found that those with greater maternal EA had increased amplitudes to positive facial expressions, relative to fearful and neutral expressions. The authors argued that maternal responses to the child's cues contribute to the child's brain development, and are crucial for their social development and adaptation [75].

Theoretical Framework
Pre-EA and postnatal EA, stress physiology, the neuroendocrine system, genetics and epigenetics, brain mechanisms, and behavior bidirectionally interact in ways that are not yet fully understood. For example, past research has shown that changes in parent oxytocin levels are associated with respective changes in emotionally available parenting, and specific parenting emotional availability behaviors lead to changes in oxytocin levels, further emphasizing the transactional nature of the neuroendocrine system [57]. Previous literature is aligned with our theory that specific environmental experiences and parent/child behavioral practices have the potential to influence the brain structure and function in a complex feedback loop, and are thought to be shaped by genetic/epigenetic co-evolutionary processes [26]. With the goal of conceptualizing a theory that integrates these complex concepts [76][77][78], we propose the following theoretical model, presented in Figure 2. In this model, prenatal EA (be it self-reported or observed) is linked with the EA of parent and child postnatally (be it observed or self-reported) and with stress physiology, neuroendocrine system, genetics and epigenetics, and brain mechanisms, and behavior. Each of the included constructs are associated, in multiple ways, with varying strength and direction of associations; however, we have chosen to omit illustrating these interactions for simplicity, and instead have illustrated these dynamic processes through the inclusion of a feedback loop bidirectional arrow. It is important to note that these constructs may influence each other bidirectionally with feedback interactions. In this model, prenatal EA (be it self-reported or observed) is linked with the EA of parent and child postnatally (be it observed or self-reported) and with stress physiology, neuroendocrine system, genetics and epigenetics, and brain mechanisms, and behavior. Each of the included constructs are associated, in multiple ways, with varying strength and direction of associations; however, we have chosen to omit illustrating these interactions for simplicity, and instead have illustrated these dynamic processes through the inclusion of a feedback loop bidirectional arrow. It is important to note that these constructs may influence each other bidirectionally with feedback interactions.

Discussion
After a general overview of parenting and stress physiology and the neuroendocrine system, genetics and epigenetics, as well as brain mechanisms, we have focused on EA as a worthy area of work in affective neuroscience, with projection that it can bring a consistent framework and measurement specificity. Studies appear to use different measures of sensitivity, and yet refer to them all as sensitivity. Even with the use of the EA System, investigators sometimes hone in on only one or two qualities, rather than take advantage of the entire framework. Particularly noteworthy is how many investigators have reduced EA to 'sensitivity' to the neglect of other qualities. Particularly striking is that few studies have focused on the construct of 'structuring', even though this is a construct that is

Discussion
After a general overview of parenting and stress physiology and the neuroendocrine system, genetics and epigenetics, as well as brain mechanisms, we have focused on EA as a worthy area of work in affective neuroscience, with projection that it can bring a consistent framework and measurement specificity. Studies appear to use different measures of sensitivity, and yet refer to them all as sensitivity. Even with the use of the EA System, investigators sometimes hone in on only one or two qualities, rather than take advantage of the entire framework. Particularly noteworthy is how many investigators have reduced EA to 'sensitivity' to the neglect of other qualities. Particularly striking is that few studies have focused on the construct of 'structuring', even though this is a construct that is closely aligned with scaffolding and teaching, which are qualities that are likely to enhance the brain development of children. In the larger EA literature, adult structuring appears to be a very important quality, and one that is very amenable to intervention efforts [30].
EA is a particularly beneficial measure in brain sciences, due to its utilization of both the parental and child perspectives. Overall, the EA framework's emphasis on both the parental and the child's side of the relationship, be it with the EA Scales, EA zones, or EA-SR, is especially promising when investigating how stress physiology and the neuroendocrine system, genetics/epigenetics, may be associated with adults' as well as children's brain development, and intergenerational transmission of specific genetic and neurobiological markers. With the exception of Licata and colleagues, few studies have examined the child side of EA with an a priori focus [64]. Including the child's side of EA would be important, given its links with attachment security [30].
Research into prenatal and early life experiences, when the brain is highly malleable and epigenetic changes may occur, is also an important area for future research, in order to better identify potentially modifiable behavioral, genetic, and environmental factors, to promote supportive, quality parent-child relationships during these years. The development of the prenatal EA construct and measurement [34] provides an avenue to investigate parent-child relationships (using an observational lens) from the prenatal period to adolescence [30]. Work on imagined scenarios of conflict and pleasurable interactions [7] with one's child (even before birth) could open up new areas for intervention and prevention.
Not only can the findings from studies such as the one by Neukel, Bertsch, Fuchs, Zietlow, Reck, Moehler, Brunner, Bermpohl and Herpertz [68] open new areas for intervention and prevention, but they may also provide support for already current parent intervention and prevention groups. If a provider knows a mother has experienced early life maltreatment (ELM), and knows how this life experience may impact the ability of the mother to accurately respond to the cues of her child, the mother may be referred to parent groups that foster sensitivity and responsiveness skills for pleasant mother-child interactions, in addition to negative or challenging interactions. It would be interesting to measure the brain mechanisms and EA of an ELM mother and her child pre-and post-participation in parenting groups that are focused on emotion regulation and building secure attachment. The literature emphasis on child maltreatment-exposed mothers, and the impact on their brain functioning or mechanisms, highlights the need to create and provide support to this population through prevention and intervention efforts.
Moreover, although the dyad can include mothers and fathers, developmental research has only recently begun to study the father-child relationship. Increased incorporation of the father figure has rarely been conducted in this area of inquiry [13], and would provide valuable insights into the similarities and differences between parent figures' EA and influences on the body.
It may also be interesting to measure the brain's response to free-play interactions between a mother, father, and child, at the same time. This would provide valuable insights into how the biological response of both parents, who have had similar experiences or relationships to the same child, may differ in the same interactions, and how this may correlate with their EA.
Considering the recent and lasting global impacts of COVID-19, and the heightened stress families have endured, and knowing what we know about the association between stress physiology and EA, it would be interesting and important to study the EA of families who have been most impacted by the virus. Studying children's cortisol levels who have been significantly impacted by COVID-19-linked stressors (parents' loss of job, financial instability, death of a parent or close family member) may inform future prevention and intervention efforts focused on improving the EA and quality of relationship between the parent and child.
Finally, it is important to move beyond the parents, to include teachers and child care professionals. Furthering our understanding of stress physiology (salivary and hair cortisol) in these professionals, particularly those working in low-income contexts, will be important. Offering prevention and intervention programs will be key.

Conclusions and Some Limitations
Utilizing EA to obtain a more comprehensive understanding of the parent-child relationship would provide more specificity in behavioral and affective neuroscience. As the EA field grows in relation to the interconnections of stress and the neuroendocrine system, genetics, epigenetics, and the brain, it is important to utilize all of the EA Scales dimensions (sensitivity, structuring, nonintrusiveness, nonhostility, child responsiveness, and child involvement), as well as EA zones on emotional attachment, to provide the most comprehensive understanding of the strength and direction of the relations between the different constructs. The incorporation of EA into research would provide information that is essential for informing targeted, preventative interventions, to promote healthy parentchild relationships in the present and future. However, more research using a variety of observational measures would also be useful, as for example, molecular indicators, such as synchronous interactions, or frequency counts of behaviors, which the EA System does not offer. Thus, we encourage more observational work including both global as well as molecular measures that can be useful in affective neuroscience, with EA being the global measure that can provide consistency, and with the inclusion of other measures that are more fine-grained rather than holistic. Further, more studies including attachment measures would provide an important focus to this line of work. Tables 1-3 show data extraction in detail. Children's cortisol levels were higher after kindergarten entry than before; lower EA before kindergarten entry associated with a rise in HCC (negative correlation); children with low mother intrusiveness and higher child responsiveness associated with lower cortisol increases [49] Senehi  Higher father basal testosterone levels in the evening were related to less respect for child autonomy (only in fathers with low self-control). Higher father basal testosterone levels in the evening was related to more sensitive parenting (only in fathers with high self-control) In the oxytocin condition fathers were more stimulating of their child's exploration than in the placebo condition, and they tended to show less hostility Lower maternal sensitivity associated with more controlling caregiving behaviors; less maternal structuring associated with more controlling punitive behaviors; hypomethylation of the OXTR gene associated with greater maternal structuring behaviors and with more child controlling caregiving behaviors; no interaction effect found of OXTR gene as a moderator in the association between interactive behaviors and child controlling behaviors  Table 3. Characteristics of studies included in the review with brain mechanisms focus.  Neglectful mothers, compared to the control, had disruptions in the structural organization of connectors between the occipital lobe and the temporal and frontal lobes: ILF-R and bilaterally the inferior fronto-occipital fasciculi (IFO-R and IFO-L). Neglectful mothers, compared to controls, showed reduced volumes in ILF-R and IFO-L. Positive mother-child interactions were predicted by increased volume in the ILF-R; neglectful mothers had a higher likelihood of exposure to early adversity, higher vulnerability to psychopathologies, and lower cognitive integrity compared to the control  Higher maternal sensitivity and higher maternal nonhostility were associated with lower amygdala activation during mother-infant interaction; self-distraction decreased subjective emotional intensity and bilateral activations in the amygdala [71] Kim et al.,