Associations between maternal adverse childhood experiences and mind-mindedness: an analysis of mother-infant interaction

Abstract

Adverse childhood experiences (ACEs) can influence parenting behaviours and have lasting effects on child development. This study investigated how maternal ACEs affect mind-mindedness during interactions between 69 Brazilian mothers and their six-month-old infants, and the role of stress in this relationship. Maternal vocalisations were coded for appropriate and non-attuned mind-related comments according to the Mind-Mindedness Coding Manual. The number of maternal ACEs was assessed using the CDC-Kaiser ACE Study Questionnaire. Current stress, postnatal depression, depression, anxiety, adverse life events, and resilience were assessed using the Perceived Stress Scale, Edinburgh Postnatal Depression Scale, Depression Scale Patient Health Questionnaire, Generalized Anxiety Disorder, the Inventory of Major Life Events, and the Resilience Scale, respectively. General Linear Models analysed associations between variables. The results showed no direct link between ACEs and mind-mindedness. However, stress moderated the relationship between maternal ACEs and appropriate mind-related comments. These findings suggest that maternal exposure to ACEs affects how mothers interpret their infants’ mental states, particularly under stress. Understanding these relationships can guide interventions to support maternal mental health and foster positive parent-infant interactions, potentially preventing the intergenerational transmission of trauma.

Introduction

Adverse childhood experiences (ACEs) often result in various negative developmental outcomes that can have a lifelong impact (de Cock et al., 2017). Moreover, a motheŕs ACEs can not only impact their own developmental course, but also those of her children (Leite Ongilio et al., 2023). The interconnectedness of maternal childhood and the subsequent development of her child has emerged as a critical research area as it can have substantial implications for both life trajectories and attachment patterns (Leite Ongilio et al., 2023).

Adverse childhood experiences

Due to the high plasticity of the developing brain, childhood and adolescence are marked by greater vulnerability to environmental influences (Bale et al. 2010; Colovati, 2020). The consequences of exposure to Adverse Childhood Experiences (ACEs) can be manifested as lasting structural and functional changes (ÓConnor et al., 2021; Teicher et al, 2012), as epigenetic modifications (Andersen et al., 2008; Nöthling et al, 2020; Teicher et al., 2006) and can alter the typical trajectory of neurodevelopment (ÓConnor et al., 2021; Stults-Kolehmainen et al., 2014; Sullivan et al., 2019).

The literature has consistently shown that the effects of ACEs can extend to subsequent generations (Moog et al., 2016, 2023; Yehuda & Lerner, 2018). ACEs exposure, with its wide-ranging consequences, disrupts the hormonal regulation necessary for effective stress management, leading to hypervigilant responses to external stressors. These disruptions can negatively impact parenting behaviors, such as emotional availability, biobehavioral regulation, emotional numbing, and sensitivity to children’s needs (Leite Ongilio et al., 2023; Martinez-Torteya et al., 2014; VanBronkhorst et al., 2024). Early trauma exposure can also reduce maternal sensitivity in recognizing and responding to emotional cues or needs, particularly in high-stress contexts, hindering the ability to protect children (Lieberman et al., 2020). This diminished sensitivity has been linked to adverse developmental outcomes in children (Cook et al., 2021; van Ee et al., 2016) and is identified as a key mechanism in the intergenerational transmission of trauma (Dollberg & Hanetz-Gamliel, 2022; Isobel et al., 2019).

The prevalence of exposure to ACEs is significant worldwide. A World Health Organization (WHO) study analysed ACE exposure in 21 countries comprising high income countries, high-middle income countries (including Brazil), and low-middle income countries (n = 51,945). In total, 38.8% of participants reported at least one ACE, with 61.5% of this group reporting only one ACE and 7% reporting four or more (Kessler et al., 2010). In a more recent study, specifically with the Brazilian population (n = 546), it was found that 74.4% of the participants reported at least one ACE, with 13.1% of them reporting four or more (Risso et al., 2023), showing a considerably higher frequency of ACEs in the Brazilian population. The research also revealed a correlation between lower family income and a higher likelihood of experiencing four or more ACEs.

Given the potential effects of maternal exposure to ACEs on offspring development, there is an urgent need to monitor maternal mental health and emotional responsiveness to the infant. One important construct in this regard, which is key to maternal sensitivity and has been shown to be a precursor to positive outcomes in child development, is mind-mindedness (Bernier et al., 2017).

Maternal mind-mindedness

The term mind-mindedness (MM) was coined by Elizabeth Meins (2001) and is defined as a caregiver’s propensity to identify their child as an individual with distinct thoughts, feelings, and desires. (Meins et al., 2003). It represents a mental and cognitive ability to accurately interpret an infantś inner mental states as expressed in vocalisations addressed to or about the child (Mein & Fernyhough, 2015). Despite being widely understood as a type of mentalization, some studies conceptualise the construct as a relational quality (Barreto et al., 2015; Meins et al., 2014).

According to a meta-analysis conducted by Zeegers and collaborators (2017), MM may predict secure attachment in children and parental sensitivity, with associations between appropriate mind-related comments (AMRC), secure attachment, and maternal sensitivity. It is considered as a predictor of parental caregiving quality (McMahon & Bernier, 2017) and increased parental emotional availability (Mcmahon & Newey, 2018; Planalp et al., 2019).

Regarding outcomes in children’s development, Bernier and colleagues (2017) suggested that promoting MM in parents is a path that helps trigger a cascade of positive long-term consequences. Parental MM has been linked to enhanced physiological emotional regulation (Zeegers et al., 2018), better inhibitory control (Cheng et al., 2018), theory of mind (Goffin et al., 2020; Licata et al., 2016), and indirect long-term effects on the child’s future moral consciousness, including in respect of discomfort in following transgressions and pro-social judgments (Goffin et al., 2020). MM was also a predictor of greater expressive vocabulary (Laranjo & Bernier, 2013; Meins et al., 2013) and better reading performance (Bernier et al., 2017).

Meins and collaborators (2001) suggested that MM is a potential mediator between early attachment quality and lifelong child development. Thus, it is reasonable to consider MM as a potential cognitive framework for understanding the intergenerational transmission of trauma.

Maternal ACEs may result in the mother being less understanding of what the child wants or in recognizing the infant’s mental states and intentions, and thus less likely to be appropriately mind-minded in interactions with the infant (Easterbrooks et al., 2017), although this has rarely been studied. Easterbrooks et al. (2017) were the first to explore the relationship between ACEs and MM in their study of adolescent mothers and post-traumatic stress disorder (PTSD). Their findings revealed a complex association. Among mothers exposed to trauma but without current PTSD, higher MM was linked to fewer child behavior problems. However, this relationship did not hold for mothers with full PTSD (based on three symptom clusters as per DSM-IV). Additionally, the study showed that for children of mothers without trauma exposure, higher MM was associated with greater social competence. In contrast, for children of mothers with partial PTSD (two symptom clusters according to DSM-IV), higher MM was linked to lower social competence. The authors suggested that mothers with partial PTSD may be actively processing their traumatic experiences, leading to changes in their interaction dynamics. They may also exhibit hypervigilance or hyperarousal regarding their children’s mental states, potentially contributing to their children’s dysregulation. Camisasca and colleagues (2017) also studied the relationship between MM and trauma, but only those related to the birth of the offspring. They found associations between traumatic symptoms and less AMRC when mothers were interacting with their 17-month-old offspring and reported greater parental stress. They also found that MM mediated the association between traumatic symptoms, specifically hyperexcitation, and parental stress. Furthermore, results from the study by Brown et al. (2023) indicate that MM at 27.51 months was associated with fewer behavioural difficulties at 39.31 months in children from families with a history of maltreatment, suggesting that MM may serve as a protective factor in higher-risk populations.

Given the scarcity of research on the impact of maternal ACEs on MM, there is a need for further studies in this area to develop a better understanding of these relations. This knowledge can support the development of interventions to promote positive outcomes for children, including in respect of the potential intergenerational consequences of past maternal experience, as an individual’s developmental roots are deeply anchored in these experiences.

The current study

The current study examined the associations between maternal exposure to ACEs and mothers’ mind-mindedness during mother-infant interactions. This study aims to identify whether maternal ACEs have an impact on appropriate mind-related comments and non-attuned mind-related comments during interactions with their offspring at 6 months of age in a Brazilian population. Additionally, the study explored whether stress moderates this relationship.

Materials and method

Recruitment and participants

This research protocol was approved by the IRB of Duke University (Pro00110664) and Columbia University/New York State Psychiatric Institute (7927). In Brazil, it was sanctioned by the National Research Ethics Committee (CONEP; 78018417.2.0000.5505) and the Research Ethics Committee of Universidade Federal de São Paulo (CEP; 1200/2017). All procedures were conducted in compliance with applicable guidelines and regulations. Written consent was obtained from all mothers participating in the study.

The participants of this study were derived from the Maternal Influences on Child Neurobehavioral Development Study (Healthy MINDs) cohort study. The Healthy MINDs study aims to examine the mechanisms by which maternal ACEs affect the neurodevelopment of the offspring. To this end, pregnant women with and without a history of ACEs were recruited from the Brazilian Universal Health Care System (SUS), in order to follow their children during the first two years of life, with a focus on the neurodevelopment of cognitive control. The recruitment took place in primary health care units in Guarulhos, São Paulo and in a maternity hospital in the city of São Paulo, Brazil.

This study included a sub-sample of 69 mother-infant dyads who completed the 6-month follow-up. Initially, 95 mothers were selected, but 23 were excluded due to incomplete responses on the required scales, and 3 were removed due to poor video recording quality. As a result, the final sample consisted of 69 mother-infant dyads. Of the infants, 38 were girls. The average age of the mothers was 27.32 years (SD ± 4.96), while the infants had an average age of 185.64 days (SD ± 18.06).

The maternal inclusion criteria were: a) low-risk pregnancy; b) being able to understand, consent to, and sign the Informed Consent Form; c) being between 18–38 years-old.

The maternal exclusion criteria were: a) report of a severe psychiatric (e.g., schizophrenia, persistent delusional disorder, bipolar disorder, obsessive-compulsive disorder, or suicidal ideation) or neurological disease (e.g., dementia, brain injury or epilepsy under treatment); b) a history of head trauma with diagnosed brain injury; c) diagnosis of epilepsy; d) a history of neurosurgery; e) significant intellectual deficits; f) decompensated clinical diseases that required intensive treatment; g) illicit drug use, other than cannabis; and h) the presence of toxoplasmosis, other infections, rubella, cytomegalovirus, and herpes (also known as TORCH infections).

The infant inclusion criteria were: a) 6 months of age; and b) whose mother was part of the Healthy MiNDS study. The exclusion criteria for infants were: a) premature (<37 weeks); b) classified with 5-minute Apgar below 7; c) born with low weight (below 2.5 kg); d) admitted to the neonatal intensive care unit; and e) presenting kernicterus or inborn errors of metabolism.

Procedures

The data were collected at three different times. Information regarding maternal socio-economic status, and level of education were collected between the 25–39 gestational weeks in the primary health care unit. Information regarding the exposure to adversities, resilience, and postnatal depression were obtained by means of specific scales in meetings within 4 weeks of the birth of the infant. The mother-infant interaction was videotaped, and additional information relating to maternal mental health (current stress, anxiety and depression) was collected when the infants were 6 months old during a home visit.

Socioeconomic and clinical assessment

Sociodemographic (maternal age, social-economic status (SES), marital status, race and education level) and clinical data (current and past six months hospitalizations and medical and/or psychological treatments, current use of medications, and occurrence of physical or mental disability) were collected with the mother during pregnancy. To assess SES, the Brazilian Economic Classification Criteria (CCEB/2019) by salary range, developed by the Brazilian Association of Opinion and Media Researchers (ABEP), was used. Using monthly family income (Brazilian currency - Reais) as a measure of socioeconomic class five socioeconomic strata were assigned: D/E =≤719.81; C2=719.82 – 1748.59; C1=1748.60 – 3085.48; B2=3085.49 – 5641.64; B1=5641.65 – 11279.14.

Maternal exposure to adversities

ACEs were assessed through the CDC-Kaiser ACEs Questionnaire Full (Maternal), developed by the ACE Study (Felitti et al., 1998). The questionnaire covers 10 different ACEs, and the cumulative score ranges from 0 to 10, depending on the number of ACEs experienced before the age of 18. ACEs are categorised into three primary categories: abuse, neglect, and household challenges. Each category is further divided into multiple subcategories. A higher score indicates more different types of adverse childhood experiences.

The Inventory of Major Life Events (MLE) based on Turner and Lloyd (2004) was used to assess potentially traumatic events experienced through life. It consists of 47 items, with a higher score indicating more different types of adverse experiences.

Stress

Current stress was assessed using the Perceived Stress Scale (PSS), developed by Cohen et al. (1983) and validated for use with the Brazilian population by Luft et al. (2007). The scale was designed to investigate how individuals perceive situations considered to be stressful, focusing on aspects such as unpredictability, lack of control, and overload of activities that occurred during the last month.

Maternal mental health

Postnatal depression was measured using the Edinburgh Postnatal Depression Scale (EPDS) (Cox et al., 1987), validated by Santos et al. (2007) for the Brazilian population. To assess depression, the Brazilian version of the Patient Health Questionnaire (PHQ-9) was used (Santos et al. 2013). It was developed by Spitzer (1999) and is a self-administered questionnaire based on the depression module of the Diagnostic and Statistical Manual Fourth Edition (American Psychiatric Association et al., 2000). Anxiety was self-assessed using the Generalised Anxiety Disorder (GAD-7) questionnaire. The questionnaire, designed by Spitzer et al. (2006) and validated for the Brazilian population by Moreno et al. (2016), consists of seven items aimed at gauging the patient’s health status during the preceding two weeks.

Maternal resilience was assessed with the 25 item Resilience Scale (RS) (Wagnild & Young, 1993). It is considered the “gold standard” for resilience assessments (Damasio et al., 2011) and it was validated for the Brazilian population by Pesce et al. (2005).

Mind-mindedness assessment

Maternal mind-mindedness was assessed through the Mind-mindedness Coding Manual (Mein & Fernyhough, 2015). A 8-minute video recording of an infant-mother free play session was captured during the home visit by the project team. The MM codings were carried out in a home environment, which gives the interaction a more ecological and less laboratory factor. Mothers were instructed to play with their infants as they would if they had free time together. The video recording was required to clearly show the faces of the mother and infant, with all audio being distinctly audible. Six minutes of each recording, from minute 1 to 7, were used for vocalization coding, excluding the initial and final minutes to avoid noise interference. All maternal speech and sounds throughout the session were meticulously transcribed into individual comments. Raters identified each comment as mind-related or not. Mind-related comments are those that: (a) refer to the infant’s current internal state (e.g., “Are you frightened?”), (b) refer to cognitions (e.g., “You know how to play with the frog”), (c) refer to wishes or desires (e.g., “Do you want this book?”), (d) suggest activities the infant might enjoy (e.g., “Do you want to play with the giraffe now?”), or (e) express what the infant might say if able to talk (e.g., “This is a teddy bear, mummy”). Each mind-related comment was further classified as either an appropriate mind-related comment (AMRC) or a non-attuned mind-related comment (NAMRC). AMRCs are those where: (a) the researcher agrees with the mother’s interpretation of the infant’s current internal state, (b) the comment links the current activity with similar events from the past or future, or (c) the comment provides guidance on how to proceed after a pause in the interaction. In contrast, NAMRCs are those where: (a) the rater disagrees with the mother’s reading of the infant’s current internal state, (b) the comment refers to a past or future event unrelated to the infant’s present activity, (c) the caregiver suggests a new activity or asks what the child wants to do while the child is already engaged in something else, or (d) the caregiver attributes internal states that are not implied by the infant’s behavior, possibly reflecting projections of the caregiver’s own internal states.

The coding process involved three blind raters. The raters were psychologists trained in the Mind-mindedness Coding Manual. To ensure reliability, an intraclass correlation coefficient (ICC) of 0.85 was used as the threshold for 31.8% of the videos. Videos that did not meet this reliability standard were thoroughly discussed, and all discrepancies in coding between assessors were resolved through mutual agreement, ultimately resulting in an ICC of 0.97 across all the videos. The AMRC and NAMRC scores were based on the proportion of the total number of comments made during the interaction, allowing for control of maternal verbosity.

Statistical analysis

The software used to perform statistical analysis was SPSS for Windows (version 21.0; SPSS, Armonk, NY). Descriptive analyses were conducted to characterise the sociodemographic profiles, the quantification of occurrences and types of reported ACEs (See Table 1 and Table 2) and the frequency of the maternal vocalisations related to mind-mindedness.

Table 1.

Sociodemographic characterization of participants (n=69)

		n	%
SES	B1	1	1.4
	B2	8	11.6
	C1	14	20.3
	C2	30	43.5
	D/E	16	23.2
Race	Mixed-race	35	50.7
	White	19	27.5
	Black	12	17.4
	Asian	3	4.3
Educational level	SE Incomplete	1	1.4
	SE II Complete	3	4.3
	HSE Incomplete	5	7.2
	HSE Complete	28	40.6
	UE Incomplete	13	18.8
	UE Complete	19	27.5
Marital Status	Single	27	39.1
	Married	23	33.3
	Stable Union	16	23.2
	Divorce	3	4.3
Infant sex	Female	38	55.1
	Male	31	44.9
	Total	69	100

Note: SES = Social-economic status (SES) defined by monthly salary range class by family in Brazilian reais: D/E = ≤ 719.81; C2 = 719.82 – 1748.59; C1= 1748.60 – 3085.48; B2 = 3085.49 – 5641.64; B1 = 5641.65 – 11279.14. SE = Secondary Education (9 years of schooling); HSE = Higher Secondary Education (12 years of schooling); UE = University Education

Table 2.

Sample distribution according to the number of ACEs

Number of ACEs	n	%
0	6	8.7
1	14	20.3
2	11	15.9
3	9	13
4	9	13
5	8	11.6
6	6	8.7
7	3	4.3
8	1	1.4
9	0	0
10	2	2.9
Total	69	100

Note: According to data obtained from the CDC-Kaiser ACEs questionnaire. ACE = Adverse Childhood Experiences

Five bivariate Pearson correlation analyses were conducted: (a) between the number of ACEs and stress; (b) between AMRC and stress; (c) between the number of ACEs and the AMRC; (d) between NAMRC and stress and; (e) between NAMRC and the number of ACEs.

A multivariate generalised linear model (GLM) was performed to verify the effect of ACEs on two types of mind-related comments: AMRC and NAMRC (unconditional model). Model was adjusted for major life events, current depression, postnatal depression, anxiety and resilience. The other covariates were the age and sex of the infant, SES, race, and maternal education. Although the infants were approximately six months old (M = 185.64 days; SD ± 18.06), variations in days can still impact developmental outcomes. To investigate whether the effect of ACEs on the mothers’ mind-related comments were moderated by stress, a condition (moderation) model was tested, as represented in Figure 1.

Figure 1. Conditional model to verify if the effect of ACEs on AMRC and NAMRC.

Conditional model to verify if the effect of ACEs on AMRC and NAMRC is moderated by stress. Infants’ sex, mothers’ educational level, socioeconomic strata (SES) and race were categorical variables. Infants’ age (in days) and the covariants anxiety and resilience were continuous variables. ACE = Adverse Childhood Experiences; Stress = Perceived Stress (PSS).

The unconditional model revealed no significant effect of depression, postpartum depression, or maternal life events (MLE) on the dependent variables, leading to the exclusion of these three covariates in subsequent analyses. The model was adjusted for the infant’s age and sex, socioeconomic status (SES), race, and maternal education. Anxiety and resilience were included as covariates based on their relevance in influencing both parenting behaviors and child development. Anxiety was included as Crugnola, Tagini, and Ierargi (2022) demonstrated that maternal anxiety is a significant mental health risk factor, strongly correlating with non-attuned parenting behaviors. Resilience, on the other hand, was considered as a potential protective factor, given its capacity to buffer the negative effects of ACEs and promote more adaptive maternal responses (Armans et al., 2020). Tests to verify the absence of multicollinearity between the independent variables were conducted using Jamovi software (version 2.3.38).

The level of significance was set at 0.025 (i.e., 0.05/2, also known as Bonferroni correction). To conduct these analyses, the equality of covariance matrices was examined using Box’s M test, alongside Levene’s test to assess variance homogeneity. Both tests were executed with a confidence interval of α = 0.05, yielding p-values of 0.03 and > 0.05, respectively. This outcome indicates homogeneity among the variance-covariance matrices, suggesting that the variance error of the dependent variable is consistent across all groups.

Given that this was a convenience sample, a post hoc sample size calculation was performed to assess its statistical power. The effect size was set at r = 0.15, and based on the sample size of n = 69, the statistical power was estimated as 0.88.

Results

The study included a sub-sample of 69 mother-infant dyads from the Healthy Minds Study. The sociodemographic characterization of the sample is presented in Table 1.

Table 1 shows that our sample was predominantly composed of mothers with a monthly family income of up to 1748.59 Brazilian reais (approximately USD 318.48; C2 and D/E), indicating a low-income population profile, representing the most vulnerable socioeconomic strata. Approximately half of the participants self-identified as mixed race. The majority of women completed higher secondary education (86.9%), and more than half were in a stable relationship, either married or in a stable union (56.5%).

The participants reported an average of 3.25 ACEs (SD = 2.37) as depicted in Table 2.

Notably, more than two thirds experienced up to four events (70.9%), 8.6% reported seven or more adverse events, while 8.7% of the mothers indicated no adverse childhood experiences. The frequencies of the types of experience lived are depicted in Table 3.

Table 3.

Distribution of types of ACEs reported by mothers (n=69)

	n	%
Physical abuse	19	27.5
Emotional abuse	29	42
Sexual abuse	17	24.6
Physical neglect	9	13
Emotional neglect	25	36.2
Parental separation or divorce	39	56.5
Mother treated violently	14	20.3
Substance abuse in the household	27	39.1
Mental illness in the household	25	36.2
Incarcerated household member	20	29

Note: According to data obtained from the CDC-Kaiser ACEs questionnaire. ACE = Adverse Childhood Experiences

Important variability was observed in the types of ACEs experienced, with more than half of the participants reporting parental separation (56.5 %), and 42% emotional abuse as a child.

The last descriptive analysis was conducted to examine the percentage of the maternal vocalisations related to mind-mindedness. It was observed that in the interaction with their infants the mothers made on average 7% of mind-related comments, with 5.8% (SD = 5.4, range: 0–26.3) categorized as appropriate and 1.2% (SD = 1.8, range: 0–7.52) classified as non-attuned.

Five bivariate correlation analyses were performed. The correlation analysis between AMRC and stress was not significant (p = 0.68), as was the correlation between AMRC and total ACEs (p = 0.14). The correlation analysis between NAMRC and total ACEs was also not significant (p = .47), as was the correlation between NAMRC and stress (p = .72). Nonetheless, the correlation between total ACEs and stress was significant (r = .38; p = .001), with a moderate positive correlation.

An unconditioned GLM was performed to verify whether there was an association between the mothers’ ACEs and the two types of comments related to mind-mindedness: AMRC and NAMRC. No statistical significance was observed for any of the considered variables regarding the independent variables AMRC and NAMRC (p > 0.05).

The results of the analysis verifying the absence of multicollinearity for each independent variable in the conditional model (moderation model) are detailed in Table 4.

Table 4.

Multicollinearity analysis of variables

Variable	VIF	Tolerance
PSS	1.70	0.588
ACE	1.31	0.766
Infant Age	1.24	0.809
Infant Sex	1.16	0.866
SES	1.12	0.892
Race	1.12	0.892
Educational level	1.17	0.857
Anxiety	1.63	0.614
Resilience	1.26	0.797

Note: PSS = Perceived stress; ACE = Adverse Childhood Experiences; SES = Social-economic Status. VIF = Variance Inflation Factor. Tolerance = proportion of variability that is not shared with other variables.

The multicollinearity analysis showed low levels of multicollinearity. Overall, the results indicate that the model is robust, with the independent variables contributing independently and without redundancy.

The conditional analyses revealed a strong trend toward statistical significance for the ACEs × PSS (perceived stress) interaction term on the outcome variables (Pillai’s Trace = 0.113; F = 3.048; df = (2), p = 0.057, η² = 0.113). Following Bonferroni correction, the effect of this interaction between ACEs and perceived stress remained significant only on the dependent variable AMRC (B coefficient = −0.112; p = 0.020; η² = 0.106). Furthermore, the moderation model showed a tendency for an effect of ACEs (Pillai’s Trace = 0.115; F = 3.128; df = (2); p = 0.053, η²= 0.115) on the two dependent variables (AMRC and NAMRC). The tendency remained after the Bonferroni correction only for the AMRC variable (p = 0.018; η² = 0.108). Despite the lack of association between the ACEs variable and AMRC in the unconditional model, the relationship is significant at the level of perceived stress. In other words, it is considered a relation moderated by stress (Hayes, 2013).

A significant effect of infants’ age on AMRC and NAMRC was also observed (Pillai’s Trace = 0.162; F = 4.627; df = (2), p = 0.015, η² = 0.162). After Bonferroni correction, this effect remained only for AMRC (p = 0.013, η²= 0.119).

We can analyse through the results of the parameter estimates how much the ACEs and stress interaction contributed to the observed effect (Table 5).

Table 5.

Parameter estimates on AMRC percentage

	B	p	η2
ACE*Stress	−.112	.020	.106
Infant’s age	−.110	.013	.119

Note: ACE = Adverse Childhood Experiences

AMRC = Appropriate Mind-related Comments

The analysis of the interaction between ACEs and stress indicated that if the AMRC increased in one unity the value of the interaction between ACEs and stress would decrease by the coefficient .112 (β value). Such results suggest that a higher number of reported ACEs and higher current stress are associated with lower percentages of AMRC. The analysis of parameter estimates also suggests that if the AMRC value increases in one unity the infant’s age value decreases by the coefficient .110 (β value). Thus, the results suggest that the older the offspring is, the lower the percentage of AMRC from the mother.

Discussion

The present study investigated the role of current maternal stress as a potential moderator of the relationship between adverse maternal childhood experiences (ACEs) and mind-mindedness (MM), a construct that involves a caregiver’s ability to understand and respond to their child as an individual with their own thoughts, feelings, and intentions. It is reflected in how the caregiver talks to or about the child, often by attributing mental states to the child in a way that shows an understanding of the child’s internal world. The assessment took place when the infants were 6 months of age. After controlling for the age and sex of the infants, the anxiety of the mothers, and socioeconomic variables, our results showed a strong trend toward statistical significance effect of the ACEs in interaction with stress on MM. Interestingly, this effect remained significant after Bonferroni correction only on the AMRCs, with the interaction between ACEs and stress being negatively related to AMRC, suggesting that the greater the interaction, the lower the AMRC. Given that the unconditional analyses showed no effect of ACEs on the outcome variables, this result clearly suggests the moderating role of stress on the relationship between ACEs and MM construct, and this leads us to infer that the accumulation of different types of adverse experiences alone is not enough to modify mothers’ MM behaviour towards their children. However, considering that higher percentages of AMRC were correlated with a lower number of reported adverse experiences and a lower perception of stress, it seems likely that exposure to stress in an interaction with ACEs is the most relevant risk factor for impaired expression of MM.

Our results also showed that MM was not influenced by socioeconomic factors, as reported in some samples from other countries (Meins et al., 2011; McMahon & Bernier, 2017). Although this outcome may be partially attributed to the homogeneity of our low-income sample, it suggests a degree of stability in the expected frequency of MM comments during mother-infant interactions, regardless of demographic variables. This seems to confer validity on MM for cross-cultural studies, more specifically in the Brazilian context. Although no studies were found in Latin America for comparison, the frequency of comments related to mental states in maternal vocalizations is in agreement with those observed in previous studies in Western populations, which generally varies between 5% and 9% (Bozicevic et al., 2023; Dai et al., 2019; Meins, 2011; Suttora et al., 2020).

Moreover, the positively correlated moderate correlation between ACEs and stress validates the moderation analysis results, indicating that mothers with more types of childhood adversities reported higher current stress levels. This association suggests a direct influence of ACEs on maternal mental well-being. Life stressors may lead to difficulty in finding coping strategies for new stressful situations, such as motherhood. Stress intensifies the effects of ACEs, becoming a risk factor for caregiving and making it more challenging for mothers to maintain a mentally attuned approach while interacting with their infants (Dollberg & Hanetz-Gamliel, 2022).

Our results suggest that the negative impact of adversities in childhood may be mitigated in environments with lower exposure to stress, highlighting a complex interplay between individual and environmental factors in the expression of MM. The study also underscores the importance of avoiding the stigmatisation of women who have experienced childhood traumas. In a systematic review, Vaillancourt, Pawby, and Fearon (2017) presented evidence showing that not all women with a history of childhood abuse face challenges in motherhood. They emphasise the need for a better understanding of the neurobiological and psychosocial mechanisms that differentiate mothers with a history of ACEs who can adequately care for their infants from those who struggle. Our findings suggest that the combined impact of ACEs and current stress on maternal mentalization may trigger changes in maternal sensitivity that impact infant attachment patterns, and child development outcomes, as the use of AMRC promotes various future child skills, such as cognitive, language and socio-emotional skills (McMahon & Bernier, 2017; van IJzendoorn, 1995). These results reinforce the importance of effective support for mothers with a history of ACEs.

The absence of significant associations between number of types of ACEs and the frequency of NAMRC suggest that some individual or socio-environmental characteristics could have been protective factors for quality of interactions. For example, socio-environmental factors such as the availability of adequate social support and public policies that promote care for pregnant and postpartum women may also help reduce the impact of ACEs - factors not captured by the resilience scale.

Another important factor, that was highlighted in the study by Harris et al. (2021), is the relevance of the type of ACE suffered (e.g., sexual abuse) in predicting negative outcomes. While not analysed separately in this study due to the small sample size, the type of adversity could indeed have influenced outcomes. Moreover, it is well-known that a higher number of experiences negatively impacts coping strategies and the stress becomes more likely to be toxic, triggering a cascade of adverse effects (Agorastos et al., 2019). In our results, 71% of the mothers reported more than one type of ACE, and over half (54.9%) reported three or more types of ACEs. This buildup of stressors, justifying our initial examination of how this accumulation impacts the relationship between women and their infants, irrespective of the type of adversity. This understanding helps identify at-risk groups and mitigate the deleterious effects of intergenerational transmission.

Despite the small age difference in our sample, the infant age also had a significant effect on AMRC in the moderation analysis, suggesting that AMRC decreases as age increases. Previous findings on the effects of age on MM are inconsistent. Our results differed from Meins and colleagues (2011), who found mothers engaging in more AMRC as infants grew (3 to 7 months). Other results indicated lower AMRC as children grew, showing negative associations with caregivers attributing appropriate mental states to their older children (Dai et al., 2019; Silletti et al., 2022); however, these studies, unlike ours, tracked this effect longitudinally. As infant communication develops over time as a part of natural child development, older infants may require less maternal input as they become increasingly capable of communicating their needs, even if primitively and non-verbally. For those infants who are older and can talk, mothers no longer need to interpret their infants’ mental states, which may explain the lower MM observed.

To the best of our knowledge, there are no published studies that directly evaluate the impacts of ACEs on maternal MM. This study not only provides new results and perspectives on the MM construct, but also considers an underexplored population, as MM studies predominantly focus on developed countries with predominantly white participants (Keller, 2012; McMahon & Bernier, 2017).

Clinical Implications and Limitations

Our findings have clear practical implications, particularly in regard to the care of mother-child dyads. First of all, a better understanding of the potential long-term effects of ACEs and the effect of current stress may help health professionals to identify risk groups in prenatal checks, and promote early interventions that aim to mitigate the deleterious effects of ACEs on parenting and potential intergenerational transmission. For instance, training can help to improve MM and increase maternal sensitivity to infant mental states. As caregivers become more proficient in MM, we can expect more positive outcomes in respect of child development. Moreover, strategies devoted to stress management could also enhance a readiness to engage emotionally with the infant. Interventions like this may be particularly relevant for developing countries such as Brazil because of the increased risk of exposure to childhood adversities due to socioeconomic inequalities.

This study has some limitations that should be considered. The small sample size did not allow a more extensive analysis of the effects of different adversities. Future studies involving larger cohorts could contribute to developing a better understanding of these effects. Moreover, longitudinal studies would be useful in assessing changes over time, particularly in children’s development. Furthermore, the use of a single variable that comprises multiple traumas did not allow us to identify which type of trauma or its frequency were most associated with behavioral changes involving the mother-child relationship.

It is also important to consider that, since MM is a complex ability influenced by various factors, the intrinsic complexity can make it challenging to detect the specific effects of a single factor, despite careful control of the variables. Future studies that explore additional modulating factors will further enrich the understanding of these complex dynamics, including the influence of protective factors and other family variables. Additionally, a posteriori qualitative analysis of the dyads interaction can be a great resource for better understanding the MM construct and its implications for the mother-child bond.

Conclusion

In conclusion, this study revealed a negative interaction between the number of types of adverse events in childhood and perceived current stress on mothers’ appropriate mind-related comments to their six-month-old infants. Therefore, more adverse types of childhood events combined with current stress seem associated with lower Mind-Mindedness skills. The lack of evidence of any direct effect of reported events or stress on the outcome variables in the unconditional analysis suggests that the accumulation of different types of adverse experiences alone is not enough to modify mothers’ expression of Mind-Mindedness behaviour towards their children, and that stress plays a primary role in this relationship. This study contributes to expanding mind-mindedness research to Latin America. Our results indicate that the percentage of MM comments aligns with findings from Western studies, supporting the validity of MM for cross-cultural research, particularly within the Brazilian context. Future studies should further explore cross-cultural differences by directly comparing Brazilian samples with other populations.

Figure 2.

Analysis of the moderation effect of perceived stress on the relationship between ACEs (adverse childhood experiences) and AMRC (appropriate mind-related comments). Maternal educational levels, socioeconomic status (SES) and race, as well as infant’ sex, were categorical variables. The infant’s age (in days) and the covariates anxiety and resilience were continuous variables. ACE = Adverse Childhood Experiences; Stress = Perceived Stress (PSS).