Exploring the Links between the Quality of Early Mother-Infant Interactions and Child Physiological, Behavioural, and Developmental Outcomes. Do we Understand their Complexity?

Twenty-three articles had a longitudinal study design (76,7%), two were randomized control trials (6,7%) and five represented a cross-sectional approach (16,7%).

Among the papers, sample size of children and parent dyads ranged from 24 to 650).

Three studies (10%) included both mother and father interactions with an infant, but the majority (90%) presented observations of mother-infant dyads.

The findings of 13 papers (43,3%) related to healthy infants born full-term. Seven studies (23,3%) concerned preterm children, of which six included groups of very low birth-weight infants (VLBW < 1500 g) or very preterm infants (VPI < 32 weeks of GA). Five papers (16,7%) included both full-term and preterm groups. In five articles (16,7%), gestational age of infants was not clearly indicated.

In four studies, the direct observation and evaluation of the interaction quality was performed in the NICU prior to the hospital discharge. In one study, interactional measurements were taken when infants were 1 month old, and one when infants were 2 months old. Seven of the studies made their interaction quality assessment when infants were 3 months old. Three articles included the interactional measures performed when infants were 4 months old, and two when infants were 5 months old. Thirteen studies evaluated infant interactions when they were 6 months old and one when they were 7 months old. In two of the studies dyadic interactions were assessed by the end of the eighth month postpartum. The period of interactional assessment in two studies was defined as between 6 and 8 months), and in another one between 6 and 12 months).

As far as child developmental outcomes are concerned, 25 studies (83,3%) included developmental assessment in infancy (up to 12 months); 16 (53,3%) in toddlerhood (13–36 months) ; seven (23,3%) included preschool years (4–5 years); in five studies (16,7%) developmental assessment was implemented in school years (6–12 years) (; and just three of the studies (10%) used data from subjects’ adolescence (older than 12 years) in the analyses.

Maternal response constitutes a major part of infant early interactional experience [48]. Various features of maternal signals were evaluated as a measure of interactional quality in 23 papers (76,7%) finally included in this review.

While the majority of researchers examined selected features of parental behaviour in interaction (listed below) and correlated behaviour with infant developmental outcome, a few took an overall perspective on measuring the caregiving quality.

Warnock and colleagues [35] used a microanalytical behavioural tool, called Maternal Behavior Coding System (MBCS) in order to distinguish two general domains of caregiving behaviour: typical behaviour (sensitive and responsive caregiving actions and interactions) from atypical behavious (“being less or overly responsive to their infant cues, intrusive, or they withdraw from interacting with their infants”). Mothers’ reactions were recorded during an infant pain event at NICU, coded and categorized to those two clusters. The results showed that atypical maternal caregiving behaviour was related to atypical infant self-regulation behaviour during and after a stressful event.

Tu and colleagues [26], on the other hand, distinguished “interactive maternal behavior” as a single factor score, calculated on the basis of Crnic’s coding criteria [49], which assessed gratification, affect, sensitivity, and organization in mothers’ behaviour during the session of free play with their infant. They found that the interactive maternal behaviour is correlated with focused attention and infant basal cortisol level (more details on the outcomes in the section Findings: Developmental outcome).

Feldman with colleagues (2013) used her original construct of “early parental care”, which summarized the average of the typical parental care behaviours, such as gaze to infant, positive affect, “motherese” vocalizations, and affectionate touch, measured during free interactions at 1 and 6 months of the infant’s age. The above-mentioned composite index of early parental care describes the proportion of time out of the entire session the parent engaged in caregiving behaviours, and it reflects the quality of early parenting. Child developmental outcome measured in the study was social reciprocity level. Researchers proved that early caregiving experience, together with parents’ oxytocin level (which was also controlled in the study) were factors that jointly explained the child’s social reciprocity in multiple relationships.

Shah and colleagues [21] evaluated maternal caregiving quality on the continuum from more positive to more negative, according to the general sum of scores in the Parent-Child Early Relational Assessment (PCERA) coding system [50].. Points were awarded in three parenting domains: (1) positive affect, involvement, and verbalizations; (2) negative affect and behaviour; and (3) intrusiveness, insensitivity, and inconsistency, and were added together to create the total parenting quality score. After correlating this variable with child behavioural problems and cognitive skills, the authors showed that more negative parenting was a predictor of externalizing behaviour problems and lower IQ.

Likewise, Zeegers and colleagues [33] calculated the general parenting quality index, based on four scales: responsiveness, intrusiveness (reversed), warmth/affectivity, and negativity (reversed). They applied the Meso Behavioural Rating System for Families (MeBRF) with young children [51] to code parental behaviour and affect during parent-infant free play. No statistically significant correlation of parenting quality with physiological regulation as an infant developmental outcome was found.

Maternal sensitivity concept derives from Ainsworth’s works, and it stands for the capability to receive infant signals, to process, interpret them correctly, and respond adequately and in a timely manner [52]., which creates the proper infant relational experiences. In order to assess the level of maternal sensitivity two research groups [17,19] (applied a gold standard measure, i.e., the macro-analytic Ainsworth Maternal Sensitivity Scales [52]. The tool consists of four subscales of maternal characteristics: acceptance, availability, cooperation, and sensitivity. One group [19] proved that less sensitive maternal behaviours can predict a disorganized type of attachment [53]; however, another one [17] did not find any significant correlation between maternal sensitivity and child’s security of attachment. In addition, Wolke and colleagues in their paper [27] reported significantly higher levels of attachment disorganization in 1.5-year-toddlers of mothers with lower sensitivity measured at the hospital unit, close after delivery, using the Boston City Hospital Assessment of Parental Sensitivity (BCHAPS) [54]. The tool was completed by nurses on the basis of their observation of maternal behaviours, and consisted of 12 items, which evaluated mothers’ caring, interacting, understanding, and enjoyment with their infant. At 3 months of infants’ age, the same team of researchers applied two different measures of maternal sensitivity, i.e., the Mother-Infant Structured Play Assessment, adapted from the Emotional Availability Scales [55] and Infant and Caregiver Engagement Phases [56]. Interestingly, they were not significantly correlated with attachment patterns, and only BCHAPS remained significantly related.

Another approach to measuring maternal sensitivity was presented by Milgrom with colleagues [29] and Newnham with his research group [30] in their randomized controlled trials. They implemented the Mother–Infant Transaction Program (MITP) intervention [57], which teaches mothers to recognize and minimize stress responses in preterm infants. After intervention, mothers in the experimental group [30] were more responsive to their infants than control mothers when assessed with the Synchrony Coding Scales [58]. The Maternal Respond Item measured the level of sensitivity in mothers’ responses to infant cues. Milgrom and colleagues [29] applied the same intervention and proved that it can successfully enhance the level of maternal sensitivity in the experimental group. They used the Preterm Mother–Infant Interaction Scale (PREMIIS), based on Synchrony Scales [58] to assess it. Maternal sensitive behaviour in this tool represents better awareness of infant negative cues, lower level of stressful impact on infants, and more appropriate caregiving. Both Milgrom’s and Newnham’s research groups found a positive correlation between maternal sensitivity and infant communication abilities.

Kivijärvi and colleagues [24] explored the construct of maternal sensitivity behaviour and its possible longitudinal consequences, namely, infant temperament. They implemented the PCERA Scale developed by Clark [50] to assess seven maternal interactional features: (1) enjoyment/pleasure, (2) contingent responsivity to child’s positive or age-appropriate behaviour, (3) contingent responsivity to child’s negative or unresponsive behaviour, (4) structures and mediates environment, (5) parent reads child’s cues and responds sensitively and appropriately, (6) connectedness, and (7) mirroring. The mothers who had the so-called “concern scores” on any of the above-mentioned items were categorized as the Less Sensitive (LS) mothers and those without any “concern score” were categorized as the More Sensitive (MS) mothers. The results point to interrelation between maternal sensitivity behaviour and infant temperament with potentially bidirectional links. Apparently, 12-month-old infants of LS mothers belonged mostly to the intermediate temperamental cluster, while this was true for only 35% of MS mothers’ infants. In addition, there were more infants of LS mothers with temperamental “concern scores” in mood than those of MS mothers.

Also, Sethna and colleagues, in their research [25], considered maternal sensitivity as an important factor related to infant development; they found a positive correlation of moderate effect size between maternal sensitivity and infant’s subcortical grey matter volume. The maternal sensitivity here was understood as warmth, acceptance, non-demanding, and non-intrusive behaviour; it was measured using the Global Rating Scales (GRS) [59].

Another closely related concept is maternal responsiveness, which is sometimes used interchangeably with maternal sensitivity. Although some researchers distinguish the two as separate constructs, others point out to the fact that it might be difficult to theoretically differentiate sensitivity from responsiveness [28]. Therefore, in this section, we present studies that also include maternal responsiveness defined as “a sensitivity in responding to infants’ cues” [34]. The more responsive a mother is, the more attentively she will notice and react to the subtle, minimal verbal or non-verbal cues. Milgrom and colleagues rated maternal responsiveness by frequency of its occurrence (0: < 30% of the time; 1: 30–60% of the time; and 2: > 60% of the time) during free interactions comparing two groups of depressed mothers and a control group of healthy mothers. Afterwards, they correlated maternal responsiveness with children’s cognitive competences and their temperamental profiles. The statistical modelling that authors used revealed that cognitive deficits in children of depressed mothers can be explained by lowered maternal responsiveness. No such mediating effect of maternal responsiveness was found, as far as children’s temperamental changes were concerned.

Erickson and Lowe [60] examined a level of maternal responsiveness using Haley and Stansbury’s coding system [61], including variables such as: (a) watching; (b) attention-seeking; (c) Mirror A (i.e., mother mimics infant’s behaviour in an exaggerated fashion); and (d) Mirror B (i.e., mother mimics infant’s behaviour; in return, the infant responds to the mother’s response in a positive manner). In their research, maternal responsiveness predicted the child’s baseline positive affect, and positive affect during and after an interactive stressor (more details on these developmental outcomes in the section Findings: Developmental outcome).

Crawford, in her research [28], assessed macroscopically maternal sensitivity/responsiveness coding parent-child free interactions according to a 7-point Likert scale (1 = extremely non-responsive/non-sensitive; 4 = moderately responsive/sensitive; 7 = extremely responsive/sensitive). It correlated with infant attentional skills. (More details on these developmental outcomes are included in the section Findings: Developmental outcome).

Other researchers [32] tried to measure differently sensitive responsiveness by using the ELO scales (Dutch Experimental Longitudinal Project; ELO) [62], and they connected results with the speed of development in infant executive functioning; however, they were not able to find significant effects.

Another two qualities of parenting appear to be seemingly close to each other but require a clear distinction because of their opposite effects on infant-caregiver interactions. The first one is positive parental directiveness, namely, “the degree to which a parent selects topics of conversation or play, uses imperatives and prompts to control or regulate the child’s attention or behavior” [32] that enriches the baby’s experience and supports their developmental outcomes. Van de Weijer-Bergsma with colleagues [32] used the ELO scales [62] to assess the level of maternal directiveness in play sessions and proved that directive mothers’ behaviour was positively related to faster developmental change of infants’ executive functions over time.

The second mentioned parenting quality is maternal intrusiveness, which in contrast to directiveness, seems to impair infant developmental outcome. It may include different kinds of behaviours: not only intrusive physical interactions, such as forcing toys or self on infant, but also overcontrolling reactions, failing to modulate own behaviour when the infant withdraws or turns away from interaction or higher focus on own agenda, while ignoring the infant’s cues [22]. Bernier and colleagues [22] scored the prevalence of those behaviours in early mother-infant interactions using a particular coding scheme [64] and correlated it with neurodevelopmental outcomes of infants, but the results showed no interrelation. However, another group of researchers [18] found a positive correlation between maternal intrusiveness and infants’ longitudinal outcome at 12 months of age, i.e., feeding difficulties. Maternal intrusiveness was observed during both feeding and non-feeding mother-child interactions and coded using the Coding of Interactive Behavior-Newborn (CIB) scale [65].

A mother can display positive affection to her child in early interactions by cuddling and cooing or present negative affect towards an infant such as expressing anger, frustration, or disgust. There is evidence of a long-term intercorrelation between the valence of maternal affect and child developmental outcome [20,22]. Lorber and colleagues [36], in their longitudinal study, rated the mother’s expression of positive and negative regard during feeding in infants’ first months of life and found out that negative parenting variables predicted conduct problems in children during preschool age (5–6 years).

Bernier [22] proved that maternal positive affect, here coded as positive emotions expressed through the tone of voice or facial expressions, are positively related to infants’ higher frontal resting EEG power, which reflects brain development (further details below). However, Sethna’s results [25] showed that maternal affective state, here characterized by the level of maternal enjoyment, effort, and vitality, degree of self-consciousness, and the extent of anxiety in the interaction and expressions of negative affect (i.e., depressive-like statement), measured with the GRS [59], was significantly related to infants’ total grey and white matter volume, and CSF volume (more details on these developmental outcomes in the section Findings: Developmental outcome).

Silberstein and colleagues [18] microcoded maternal behaviour during feeding and non-feeding interaction with their infants using the CIB scale [65], which distinguishes the following behavioural attributes related to mother’s touch: (1) affectionate (kissing, caressing, hugging, gently stroking, or mother touching infant with clear positive affect); (2) functional (wiping mouth, arranging clothes or blanket); and (3) no touch (mother just holds baby), as well as related to mother’s gaze: (1) to the infant; (2) avert to bottle (i.e., mother averts her gaze from infant to look at the bottle); and (3) gaze aversion (no eye contact with infant or bottle). Subsequently, researchers correlated mothers’ touch and gaze in the neonatal period with the occurrence of feeding problems at the infant age of 12 months. The results showed that feeding difficulties were associated with less maternal affectionate touch and gaze during non-feeding interactions and more gaze aversion during feeding interactions.

This construct [63] refers to the caregiver’s tendency of interpreting an infant’s behaviour in terms of the underlying mental processes, such as desires, thoughts, preferences and feelings, rather than an individual with needs that must be satisfied. Mind-mindedness (MM) is measured by observation of a parent’s verbal expressions towards their infants. High scores of MM reflect a parent’s capacity to make appropriate mind-related comments that accurately describe an infant’s internal states. Zeegers in 2018 [33] and Giovanelli with his colleagues [34] in 2020 used this theoretical concept to differentiate parental interactional competences and correlate them with infants’ developmental outcome. In their research, they assessed MM according to Meins and Fernyhough’s coding instruction [66], and found a positive correlation between mothers’ MM and average length and level of infants’ symbolic play in toddlerhood [34]. Zeegers and colleagues [33] found an independent association between mothers’ and fathers’ appropriate mind-related comments and infants’ higher HRV baseline index, which reflects better physiological regulation. In addition, mothers’ and fathers’ non-attuned comments independently contributed to worse physiological regulation.

Apart from established quantitative methods centred on measuring the valence of maternal characteristics in interaction, there are studies that searched for more overall patterns capturing mothers’ behaviours. Davis and colleagues [15] looked for sequences of maternal caregiving by coding the process of transitioning between auditory, visual, or tactile sensory stimuli. Researchers estimated the degree of probability of mothers’ behaviour by counting the number of possible combinations of subsequent sensory signals expressed to the child. They compared this entropy rate between mothers and correlated it with self-regulation of infants in their first postnatal months. Results showed that the more unpredictable maternal signals, the poorer self-regulation in infancy.

The observed maternal behaviour can be coded according to duration, form, and intensity of specific sensory signals and can relate to parental interactional qualities, such as sharing attention and vocalization. Wazana and colleagues [19] used the microanalytic Behavioral Evaluation Strategies and Taxonomies coding system to measure maternal sensory events, namely, frequency and duration of looking away from the infant and characteristics of vocalization to the infant (quiet talk, motherese, singing, and adult talk). It occurred that less attentive maternal care measured with higher frequency of looking away from infant behaviours was related to a disorganized style of attachment.

Britton and Gronwaldt [38] used the Nursing Child Assessment Satellite Training Feeding Scale [67] to probe the quality of mutual mother-infant interactions during the ongoing process of feeding infants by their mothers at 6 months postpartum. This tool is based on Barnard’s Parent/Caregiver-Child Interaction theoretical model, which describes the interactive system as a contingency of signals from a child to caregiver, and from caregiver to the child, since they constantly influence and modify each other’s behaviours, which makes the whole system acquire new properties [68]. In the above-mentioned interactional scales, the following factors are assessed: child’s clarity of clues and responsiveness, caregiver’s alleviation of distress and sensitivity to cues, and also environmental characteristics referring to both animate and inanimate elements. Britton and coworkers [38] found that a higher quality of bidirectional interactions, scored on previously mentioned scales, predicts an infant’s secure attachment at 12 months of age.

Stolt et al. [40] used the PCERA method to assess the quality of mother-child interaction at the same age (6 months). PCERA [50] is based on clinical observations and developmental theories and was meant to capture and analyse a child’s experience of the mother, and vice versa. The scales reflected three maternal, three infant, and three dyadic features, namely, dyadic mutuality, flatness, disorganization, and tension. These dyadic factors correlated positively with later linguistic competences of children measured at 12 and 24 months of age.

Poehlmann and Fiese [39] assessed quality of parent-infant interaction macro-analytically, using an observational rating of early parent-child interactions, called the Pediatric Infant Parent Exam (PIPE) [69]. This tool is designed to assess the degree of early interactional reciprocity and positive affect during interactional games. As a potentially related outcome, infant mental abilities were measured and the PIPE scores significantly correlated with infants’ mental development (further details below).

In addition, the already mentioned CIB manual, elaborated by Feldman (1998) [65], measures several aspects of mother-infant communication during free play, including dyadic reciprocity [16,22,37]. It addresses “the formal features of the interaction and describes an exchange where the dyad is moving in harmony, each is engaged and contributes to the mutual exchange, the interaction is the end product of the input of both partners, and the atmosphere is one of collaboration and joint activity, whether the activity is verbal or non-verbal, focuses on social give-and-receive or on object manipulation” [42], and it is calculated as an average of selected codes related to verbal and non-verbal signals, expressions, exchanges in gaze, smiles, gestures, breathing, dyadic tension, and mutual adaptation [23,37,42]. In 2010, Feldman correlated dyadic reciprocity in early mother-infant interactions with the data on the participant’s psychological adaptation across development until 13 years of age [22]. Her findings show that CIB’s dyadic reciprocity construct is a meaningful independent predictor of adolescents’ adaptation. In her later study [42], Feldman and colleagues explored both father-infant and mother-infant dyadic reciprocity and correlated it with children’s social competence and aggression. They found strong evidence that reciprocity contributes to lower aggression and better social skills (more details in the developmental outcome section). In 2015, Feldman [37] also discovered that dyadic reciprocity is a meaningful factor for emotion regulation and again, empathy.

Costa and Figueiredo [44] applied in their research the GRS [70], which probe the nature of engagement between mother and infant during their interaction. The total score on five assessed items (smooth/easy versus difficult; fun versus serious; satisfying versus unsatisfying; much engagement versus no engagement; excited engagement versus quiet engagement) results in two categories: good interaction cluster and poor interaction cluster. Authors Costa and Figueiredo [44] found that mother-infant interactional quality, measured this way, was associated with the infant’s temperamental profile (details below).

The above-mentioned scales reflect some interactional dyadic factors that can be meaningful for the quality of interaction; however, some researchers tried to capture the dynamics of communication sequences over time.

Feldman’s approach [41] was to microcode mother’s and infant’s affective states in 1 s frames and assess synchrony of reactions with a time-series analysis (CCF; cross-correlation function). Feldman was searching for its contribution to specific child competences across 13 years of life, i.e., verbal IQ, child self-regulated compliance, behaviour problems, moral cognition, and dialogical empathy. Mother-infant synchrony across the first year appeared to be a significant predictor of some developmental outcomes, such as IQ, behaviour adaptation, and empathy (details below).

Crawford, in her doctoral dissertation [28], measured synchrony/reciprocity in mother-infant interactions by using a 7-point Likert scale, based on the work of Bemieri, Reznick, and Rosenthal (1988). A global rating was assigned to parent-child interactional behaviour by assessing three aspects of synchrony: simultaneous movement, tempo similarity, and coordination/smoothness. The author tried to relate the synchrony scores to follow up on child attentional skills; however, no significant interrelations were found.

Evans and Porter [43] observed the communication structure in mother-infant interactions, regardless of emotional content of the communicated message, looking for co-regulated sequences. They used Fogel’s Global Relational Coding System [71] in order to code co-regulation patterns of four dimensions: symmetrical, asymmetrical, unilateral, and disruptive, or a dimension of non-regulation or unengaged behaviours. Subsequently, they tried to correlate these dimensions with mental and psychomotor developmental scores and attachment style. Findings showed that a symmetrical structure of communication is positively related to infant mental and psychomotor development and secure attachment status. Asymmetrical and unilateral co-regulation patterns and unengaged behaviours correlated negatively with infant development; a unilateral pattern was also related to an insecure style of attachment.

In this section, we summarize the evidence for the relationship between early interactional quality and biological functions/markers of development on the basis of neuroimaging research, studies on hormonal fluctuations, autonomic nervous system activity changes, or eating problems.

Some authors tried to find a way to measure neurological outcomes relevant to the quality of early mother-infant interactions. Bernier and colleagues [22] measured infants 1-minute resting frontal EEG power, which, according to the literature, may be a strong indicator of brain development during infancy. Researchers explored the associations between quality of maternal behaviour during early interactions, infant’s resting EEG power, and increase of resting EEG power at three points: at the infant’s age of 5, 10, and 24 months. Apparently, higher quality maternal care predicted higher frontal resting EEG power at 10 and 24 months, as well as increases in power between 5 and 10 months, and between 10 and 24 months.

Sethna [25] examined infants’ dimensions and brain volumes at the age of 4 months using MRI neuroimaging and correlated it with several maternal and infant behaviours during interaction. Their results showed that infants’ total grey and white matter volume was positively correlated with maternal affect in early interactions (r = 0.33, p = 0.042), and cerebral-spinal fluid (CSF) volume correlated negatively with it (r =−0.33, p = 0.039), whereas in the case of maternal negative affect (i.e., depressive-like expressions) opposite correlations were observed. However, these associations did not survive correction for multiple testing in further analyses. In addition, positive correlations were found between subcortical infants’ grey volumes and maternal sensitivity in early interactions (r = 0.54, p < 0.001), which this time survived correction for multiple comparisons.

In the articles we collected, there were also examples of exploring correlation between quality of parent-infant early interactions and physiological outcomes. Tu and colleagues [26] in their research examined infant levels of basal cortisol. According to current knowledge, high levels of this hormone can contribute to the risk of impaired neurodevelopment. The authors succeeded in proving that in the high-stressed group of mothers, higher quality of interactional behaviour correlates with lower infant basal cortisol level, both measured at 8 months CA. The same association was not found in the group of low-stress mothers.

Zeegers and colleagues [33] registered infants’ autonomic, physiological regulation capabilities controlled by parasympathetic nervous system (PNS) activity by measuring high-frequency (HF) heart rate variability (HRV). The HRV baseline index (mean value over time), assessed in a patient’s resting state, provides information about an individual’s general capacity for emotion regulation. A decrease of the HRV level during and after situational external stimulation can be regarded as an index of the affective arousal regulation. In their research, the stranger approach protocol was used in order to explore HRV decrease after stressful situations. In this task, a male stranger interrupted parental interaction by coming in, talking, and picking up the infant for 30 seconds. Authors correlated both variables (baseline HRV and HRV decline) with maternal and paternal MM and caregiving quality. They found strong evidence that a higher baseline HRV at 12 months of an infant’s age was associated with higher levels of mother’s and father’s MM, and a lower baseline HRV with parents’ poorer MM, independently. Moreover, infants’ larger HRV decline after distress could be predicted by mothers’ appropriate mind-related comments at 4 months and fathers’ non-attuned mind-related comments at 12 months.

Different infants’ physiological functioning aspects were observed in a study by Silberstein and colleagues [18]. They explored children’s feeding difficulties in the late infancy, as a physiological outcome related to the early interactional qualities measured in the postnatal period, prior to hospital discharge. At 12 months of age, researchers performed direct observations of feeding interactions, coupled with interviews with participating mothers. Feeding difficulties were associated with maternal interactional features in the neonatal period: higher intrusiveness, less affectionate touch and gaze, more gaze aversion, lower adaptability, and with the infant’s early interactional behaviours, i.e., lower involvement and greater withdrawal.

Temperament, defined as “individual differences in emotional, motor, and attentional reactivity (...) and self-regulation processes (…)” [72,73], is often considered an important infant contribution to the quality of their interactions with an adult. At the same time, temperament is, to a certain degree and via complex mechanisms, shaped by an infant’s interactions with the environment, and as such, can also be considered in terms of potential outcomes.

Across studies collected in this review, several well-established tools were used to measure infants’ and children’s temperament as a developmental outcome, starting with a gold standard, the Infant Behavior Questionnaire-Revised (IBQ-R) [74]. Van de Weijer-Bergsma and colleagues [32] used this questionnaire in order to explore the correlation between infant self-regulation and maternal parenting behaviours in early interactions. They did not find significant interdependence; however, other researchers were more successful.

In Costa’s and Figueiredo findings [44], IBQ-R scores representing mothers’ perception of infant high pleasure (IC95% [1.54, 2.05]; p = .042) and infants’ smile (IC95% [−2.04, −.56]; p = .038) were lower when the quality of early mother-infant interaction was categorized as a poor one. Again, when mother-infant interaction was rated as poor, mothers’ perceptions of infant activity level was higher (IC95% [−1.56, .−54]; p = .043).

Davis and colleagues (2018) [15] showed that scoring on the orienting/regulation dimension of IBQ-R can be related to the maternal behavioural pattern, i.e., a degree of predictability. Infants at the end of their first year of life showed poorer regulation capacities if their mothers were more unpredictable in early interactions (there were two cohorts in the study: the Turku Cohort: r = −.28, p < .01; the Irvine Cohort: r = −.16, p < .05). In this study, the unpredictable maternal behaviours presaged a poorer value of self-control over time in the follow-up covering three time points (5, 6.5, and 9 years of age). To reduce the possibility of maternal bias in reporting infants’ self-regulation abilities, Davis and colleagues also implemented the Children’s Behavior Questionnaire (CBQ) [75] and the Temperament in Early Childhood Questionnaire (TMCQ) [15] and performed direct self-control evaluation via Ericksen’s Flanker task [76], which measures the child’s response inhibition. Those diverse child self-control measures correlated with each other over time.

Another researcher who applied IBQ-R in order to assess infant temperament and correlate it with the quality of maternal interaction was Crawford [28], who explored infants’ attentional capabilities. She found that perceptual sensitivity, which relates to the infant’s ability to detect slight, low intensity stimuli in the environment, was positively related to the maternal sensitivity/responsivity (original terminology of the author) in early interactions.

Milgrom [31] used other parent-completed questionnaires on infant temperament: the Short Temperament Scale for Infants (STSI) and the Short Temperament Scale for Toddlers (STST) [77], trying to explore interdependence between temperament and maternal responsiveness in early interactions, but this relationship was not significant.

Kivijärvi and colleagues [24] discovered an interrelation between maternal sensitivity behaviour and temperamental profile, asking mothers to complete the Revised Infant Temperament Questionnaire (R-ITQ); [78], at 6 months of age and the Toddler Temperament Questionnaire (TTQ) [79] at 12 months of age. Lower levels of maternal sensitivity were found more frequently in the intermediate temperamental clusters of infants, characterized by moderate levels of temperamental features, between the high or low ends of specific temperament dimensions.

Nevertheless, parent-completed questionnaires were not the only method used to assess self-regulation capacities of infants. Some researchers [35] applied a microanalytic approach that included recording infant motor movement, posture, respiration responsiveness, and cry behaviour during and after a painful procedure in NICU and using the Neonatal Distress Pain Related Behavioral Coding Schema (ND-BSC) [80] to code infant reactions. They correlated infants’ pain behaviour self-regulation level with caregivers’ behaviour, clustered in two domains–typical and atypical. Presented findings prove that more atypical infants’ reactions to pain relate to atypical ways of caregiving.

Kivijärvi [24] measured infant temperament, i.e., mood, sociability, and activity, by implementing infant items of the PCERA scales [50]. Mood scores reached concern levels more frequently in the case of less sensitive mothers.

Feldman [37] explored emotion regulation (ER) in children between 3 months and 5 years of age by using several age-specific ER interactional paradigms based on presenting intrusive stimuli, delayed gratification, separation-reunion with caregiver episodes, and other interactional situations that revealed child’s regulatory behaviours and compliance. Infants and children’s ER at five different time points was next correlated with previously measured mother-infant reciprocity in early interactions. Cross-time correlations proved that higher child ER capacities appeared when mother-infant early reciprocity increased.

The concept of attachment refers to the individual’s internal organization regarding the tie from an infant to a caregiver [81], which can be established as a secure or insecure pattern, depending on early dyadic relationship with the principal attachment figure, i.e., a parent. If the mutual interaction is satisfactory and provides a secure base for exploring the environment, it is likely that a child will develop a secure attachment [82]. Disorganized attachment is characterized by the lack of consistent strategy to organize emotions and behaviours toward the caregiver.

In an experimental setup, infant attachment pattern can be revealed by implementing the protocol of the Strange Situation Procedure (SSP) [83], in which a baby is exposed to a series of separations and reunions with the caregiver and the infant’s reactions can be observed when being left alone, left in the presence of a stranger, or being reunited with a parent. O’Donnell and colleagues [17] implemented this paradigm to assess the security of attachment in preschool-age children and tried to find the interrelation with early maternal behaviour and maternal sensitivity; however, they did not succeed.

Wolke and others [27] applied Main and Solomon’s continuous scale [84] to assess attachment disorganization during the SSP and related it to the maternal sensitivity scores; they found an interrelation between lower maternal sensitivity and infant attachment disorganization (B = −0.57, SE = 0.25, β = −0.21, p = 0.019). Evans and Porter [43] administered SSP and explored the interrelation of attachment patterns with early mother-infant co-regulation patterns. Specifically, they found associations between infants’ secure attachment status and symmetrical structure of co-regulation and also between insecure attachment style and higher levels of unilateral mother-infant co-regulation structure. Also, Britton’s findings [38] show that the quality of early dyadic interactions can be predictive of further infant attachment security. The scores on interaction quality scales were higher among dyads with securely attached infants assessed by the Ainsworth Strange Situation [83] (OR = 1.05, 95% CI: 1.01–1.10; p < .05).

Ainsworth’s approach to attachment classification was also applied to the observation of preschool age children during a separation-reunion procedure with their mothers [85]. Wazana and colleagues [19] explored the interrelation between early maternal care and the type of attachment at 3 years of the children’s age.

Researchers used Moss’s coding criteria [53] and found that children of mothers who presented less attentive and sensitive early maternal care were those who revealed a disorganized pattern of attachment at the time of mother-child reunion.

Positive and negative affect can be measured either at a baseline or after an exposure to some external stressors. The example of the experimental procedure that can probe the infant affect in response to a standardized interpersonal stressor is the Still Face (SF) procedure [86], which reveals emotional reactions of a child before, during, and after mothers express indifference to their child’s attention-seeking by intentionally maintaining a neutral facial expression, avoiding eye contact, touch, and any interaction with their infant, remaining still, and looking slightly above the infants’ head. Erickson and Lowe [60] coded changes of affect when applying the SF paradigm and related it to the maternal responsiveness level measured before and after the SF procedure. It appeared that higher maternal responsiveness was highly correlated with infant positive affect (r = .70, p < 0.001 before SF; r =.64, p < .001 after SF). However, no significant correlation with infant negative affect could be found.

Parent-child relationships may have a crucial role in building social competences and prosocial behaviours, or may contribute to the early onset of conduct problems (CPs), such as aggression or oppositionality [86]. Feldman [41] found a positive interrelation between a child’s behavioural adaptation measured with the Child Behavior Checklist (CBCL) [88] and mother-infant affective states’ synchrony during early interactions. Feldman also revealed that this kind of synchrony contributes to the development of dialogical empathy in adolescents, assessed during laboratory-driven observation of conflict discussions with their mothers. Dialogical empathy reflects the child’s ability to see other people’s perspective and adapt their view accordingly, discuss their issues in a dialogical manner, and raise adaptive solutions and concern about others’ feelings. The effect of early parent-child reciprocity on empathy was also confirmed in another Feldman study [37] in a group of 10-year-olds In 2010, Feldman [23] proved that a lower level of adolescents’ adaptation measured with the CBCL and the Child Depression Inventory (CDI) [89] was correlated with poorer dyadic reciprocity in early mother-infant interactions. Moreover, Feldman and colleagues [16] assessed a child’s social behaviour at 3 and 13 years of age during interactive events with different partners: mother, father, and a best friend. They correlated it with the indicators of social reciprocity measured at 5 months of age during early caregiver-infant interactions, using the Social Reciprocity composite of CIB manual [65]. The CIB scores represented types of behaviours when partners were engaged in give-and-receive interactions, shared activity, communication, requests, when they were sensitive to each other’s verbal and non-verbal cues, adapted to each other’s needs, or just reflected moments when interaction was fluent and rhythmic. In the preschool period, children’s social competences, aggression, and prosocial behaviour were assessed via the protocol of structured play. In adolescence, dialogical skills were rated during conflict discussion with a partner. Statistical analysis revealed that social reciprocity in early interactions was related to child’s social competences, aggression, and prosocial behaviour at preschool age, as well as to childrens’ dialogical skills in adolescence. It also showed that social competence and lower aggression in preschool age, which shaped dialogical skills in adolescence, were also predicted independently by early maternal and paternal reciprocity.

Other research groups point out that negative parenting in infancy [36], which involves parental expression of anger, frustration, and disgust, can predict conduct problems measured in 5–6-year-olds, using the CBCL [90] and the Teacher Report Form [91]. Shah and colleagues [21] proved that experiencing more negative parenting during the first two years of life predicts externalizing behaviour problems, measured at the age of 3 years using CBCL 1½–5.

Evans and Porter [43] measured infant mental and psychomotor development using the Bayley Scales of Infant Development (BSID II) [92] and found a positive correlation between MDI (mental developmental index) and PDI (psychomotor developmental index) scores at 9 months of age and symmetrical structure of dyadic interaction at 6 months postpartum (r = .44, p < .001, r = .21, p < .06, respectively). A lower MDI score was related to the asymmetrical (r = −.28, p < .01) and unilateral (r =−.37, p < .001) patterns of interaction. Moreover, the lower PDI results were linked to the unengaged mutual behaviours of interactive partners (r = −.27, p < .05). Poehlmann and Fiese [39] also used BSID II in order to assess infants’ developmental abilities. Their results show direct interrelation between Bayley’s mental scales and the quality of parent-infant interaction. MDI scores, measured at 12 months of age, were higher when infants’ early interaction with mothers (6 months postpartum) was more reciprocal, affectively positive, and engaging.

Milgrom and colleagues [31] applied the Wechsler Preschool Primary Scale of Intelligence (Revised) (WPPSI-R) [93] in two groups of infants, the clinical one, whose mothers were depressed at 6 months postpartum and the control one, whose mothers were healthy. Children of mothers with postpartum depression had significantly lower full-scale IQ measured in toddlerhood, as well as poorer performance on particular Wechsler’s subscales like geometric design and arithmetic. Statistical modelling revealed a large mediating effect of maternal responsiveness level explaining the above-mentioned correlation, since responsiveness is the interactional quality strongly associated with postpartum depression. Another example of the interrelation between maternal parenting and IQ was presented by Shah with co-workers [21]. Their results show a positive correlation between quality of maternal parenting and general IQ score of a very preterm infant (VPI: < 30 weeks), measured with the Abbreviated Battery IQ Scale (ABIQ) from the Stanford-Binet Intelligence Scales, 5th edition [94]. Feldman [41] used the same tool to measure the verbal IQ of preschool and school-aged children and found its interrelation with mother-infant synchrony of affective states across the first year postpartum.

Van de Weijer-Bergsma and colleagues [32] correlated early interactional experiences of infants with the development of executive functions (EFs). EFs are differently defined in the literature, but in general this term covers a number of higher-order cognitive processes that allow one to perform goal-oriented behaviour [95]. In the above-mentioned study, authors measured working memory, inhibitory control, and attention control via performance on so-called reversal tasks, namely, the A-not-B task [96]. The speed of developmental progress in EF measured at three time points via the A-not-B task was positively correlated with maternal level of directiveness assessed in an interactional context (r = 0.54, one-sided p < 0.05). An attempt to correlate developmental change in EF with maternal sensitive responsiveness did not result in statistically significant findings.

Tu and colleagues [26] concentrated on a different infants’ cognitive competence, namely, focused attention. Tu applied Lawson and Ruff’s [97] scoring method to mark out the attentional processes during videotaped interactional mother-infant toy exploration and related it to the interactivity level in maternal behaviour. The results differed according to the index of parenting stress. In the high-stress group of parents, the higher interactive behaviour was associated with lower quality of focused attention, and, in contrast, in the low-stress group of parents, the higher interactive behaviour had a positive correlation with focused attention.

Stolt and colleagues[40 explored the association between early maternal interactions and child early-language development. With children at the age of 6 months, they used the Checklist for the Development of Early Vocalizations (CDEV), and at 12 and 24 months, they tested lexical development using the Finnish version of the MacArthur Communicative Development Inventory (CDI; FinCDI). The results showed a positive correlation between linguistic competences and several determinants of interactional quality at 6 months CA, i.e., dyadic mutuality, flatness, disorganization, and tension. Milgrom and colleagues [29] explored the interrelation between interactional quality in the mother-newborn relationship and the development of infant communication. She measured child early communication abilities at the age of 6 months by asking mothers to fill out the Infant–Toddler Checklist of the Communication and the Symbolic Behavior Scales Developmental Profile (CSBS DP Infant–Toddler Checklist) [98]. Milgrom proved that higher scores in symbolic composite of CSBS were found in infants whose mothers were assigned to the experimental group, which received the MITP parent training program, and in consequence, presented higher maternal sensitivity, more positive affect toward infants, and more synchronic dyadic interactions. Furthermore, Newnham and colleagues [30] found significant differences in communication competences, measured by the Ages and Stages Questionnaire (ASQ) [99], between analogical groups of two-year-old children with diverse levels of maternal responsiveness in early interactions. The correlation was positive. Symbolic behaviour was also measured in free play at 12 and 18 months by Giovanelli [34] using the Coding System of Symbolic Play [100]. The infants’ maturity level of symbolic play at 18 months of age was positively correlated with higher scores of mothers’ MM regarding appropriate comments in the subcategory of emotions. The average level and length of symbolic play at 12 months of age was also positively related to the appropriate mind-related comments in subcategories of desires and cognitions.