Prenatal Identification of Risk for Later Disrupted Parenting Behavior Using Latent Profiles of Childhood Maltreatment

Abstract

A history of maltreatment during childhood (e.g., physical and sexual abuse, neglect) can threaten the fundamental human need to form and maintain relationships across development, which ensure safety and security. Furthermore, parental maltreatment history presents considerable risk for the emergence of disrupted parenting behaviors (i.e., contradictory communication, sexualized/role-reversed behavior, disorientation, intrusiveness/negativity, and severe withdrawal), which in turn are associated with children’s social-emotional development. The purpose of the present study was to examine whether experiences of childhood maltreatment during pregnancy can predict risk for disrupted parenting behavior before the birth of the child. Given the inherent variability in parenting behaviors, we were interested in how different types or combinations of experiences of maltreatment during childhood are associated with later parenting behaviors. Data were drawn from 120 women from a longitudinal study that spanned from the third trimester of pregnancy through 3-year postpartum. In the current study, mothers’ experiences of childhood maltreatment were assessed during pregnancy, and disrupted parenting behaviors were coded from videotaped mother–infant interactions 1-year postpartum. Four profiles of childhood maltreatment were identified using latent profile analysis: low exposure, high exposure, high sexual maltreatment, and high physical and emotional maltreatment. Results revealed that high exposure to multiple types of childhood maltreatment most strongly predicted later disrupted parenting behavior. Women with multiple exposures to different types of maltreatment during childhood may require more intense intervention during pregnancy to prevent risk for the development of disrupted parenting behavior.

Keywords

neglect physical abuse sexual abuse attachment

Traumatic experiences are events that involve “exposure to actual or threatened death, serious injury, or sexual violence” that produce intense emotional reactions, including shock, horror, anger, and fear (American Psychiatric Association, 2013, p. 271). Traumatic experiences that are interpersonal in nature (i.e., those perpetrated toward human beings by human beings in close proximity), such as child maltreatment, are especially distressing because they threaten fundamental human needs to belong and form protective relationships that offer safety and security (Baumeister & Leary, 1995; Charuvastra & Cloitre, 2008). One of the first, and most important, relationships is often between a mother and her child, and experiences of trauma within this relationship can threaten healthy development of this relationship (Slade et al., 2009). The perinatal period (i.e., pregnancy through the first year postpartum) is a time during which women experience considerable physical and psychological changes associated with the transition to parenthood and may, therefore, be of particular concern for women who experienced maltreatment during childhood (Slade et al., 2009). Therefore, the purpose of the present study was to examine whether specific types or combinations of maltreatment experiences during a mother’s childhood (i.e., physical, sexual, and psychological abuse as well as neglect before the age of 18 years) present greater risk for disrupted parenting behaviors. Childhood maltreatment was assessed during pregnancy as an indicator of parenting risk before the child was born.

There is substantial evidence that experiences of maltreatment during childhood negatively affect future parenting behavior by affecting a mother’s ability to respond effectively and sensitively to her child’s needs. For example, women who report experiencing maltreatment during childhood exhibit decreased maternal sensitivity and engagement and increased anger, control, and intrusiveness during interactions with their young children (e.g., Gustafsson et al., 2012; Hughes & Cossar, 2016; Moehler et al., 2007; Muzik et al., 2013; Schechter et al., 2014). In one study, a significant association between maternal report of childhood maltreatment (i.e., emotional abuse and neglect) and greater observed hostility was shown even after controlling for trauma experienced during adulthood (Bailey et al., 2012), highlighting the unique association between experiences of maltreatment during childhood and later parenting behavior. More recently, a meta-analysis revealed a small but significant association between reports of maltreatment during childhood more broadly (i.e., not specific to type of maltreatment experienced) and problematic forms of parenting behavior (Savage et al., 2019).

It is unclear whether some types of experiences of maltreatment during childhood are more problematic for the parenting of young children than others, as most research focuses on cumulative indices of trauma. However, there is some work suggesting that different types, combinations, and levels of severity of traumatic experiences can lead to different presentations of trauma-related symptoms (Frost et al., 2020; Jowett et al., 2020; Minihan et al., 2018). Thus, it is plausible that different profiles of traumatic experiences might also have different implications for parenting behavior, though this area of research needs further investigation. One study reported that physical and sexual abuse (but not neglect) alone, as well as in combination, were associated with decreased maternal involvement during mother–child interactions (Lyons-Ruth & Block, 1996). Furthermore, maternal experiences during childhood of harsh parenting, witnessing violence, physical abuse, and the combination of physical and sexual abuse (but not sexual abuse alone) were associated with increased hostile and intrusive maternal behavior (Lyons-Ruth & Block, 1996). Other work has also noted associations between a history of childhood sexual abuse and harsh discipline and role reversal (DiLillo & Damashek, 2003). More work is needed to further elucidate individual differences in exposure to maltreatment during childhood and associations with parenting behaviors, particularly parenting behaviors known to be generally associated with a history of maltreatment and trauma during childhood.

Importantly, a body of research has begun to explore how and why experiences of maltreatment during childhood can be harmful to the developing mother–child relationship; this research has identified a group of more fleeting but potentially problematic parenting behaviors known as “disrupted” parenting (see Bronfman et al., 2014). Disrupted parenting behaviors are characterized by anomalous behaviors and communications during mother–child interactions, including maternal role reversal, hostility, disorientation, apprehension, and/or withdrawal from the child. Theoretically, these behaviors arise from a history of childhood maltreatment, particularly unresolved experiences of trauma and loss during childhood (Ballen et al., 2010; Goldberg et al., 2003; Lyons-Ruth et al., 2013; Madigan et al., 2006, 2007). Past work on disrupted maternal behavior has revealed consistent associations between unresolved experiences of loss and other traumatic experiences as assessed on the Adult Attachment Interview (AAI; George, Kaplan, & Main, 1996) and disrupted parenting behavior, as well as the subsequent development of disorganized mother–infant attachment (Lyons-Ruth et al., 1999; Madigan et al., 2006, 2007). However, future work is needed to better understand the unique role of different types or combinations of childhood experiences of maltreatment on disrupted parenting behavior. Understanding how individual differences in childhood maltreatment exposure affect parenting behavior could enrich interventions by allowing for greater individualization of clinical services.

From a methodological standpoint, research to date on this topic has been variable centered. Variable-centered techniques, such as correlation and regression, rely on sample-level totals and yield general conclusions about an entire sample. Such techniques assume homogeneity within a sample, meaning that important individual differences go undetected and unappreciated (Nesselroade, 1991). This methodological challenge can be addressed using a person-centered approach, such as latent profile analysis (LPA). LPA utilizes underlying heterogeneity and individual differences within a sample to make conclusions about meaningful, homogeneous groups within a heterogeneous sample (Wang & Bodner, 2007). The use of person-centered techniques to examine individual differences in symptoms and experiences has become more common in trauma literature in recent years (e.g., Cavanaugh et al., 2012; Gaska & Kimerling, 2018; Sutter et al., 2019).

The Present Study

The present study aims to expand current knowledge on predictors of disrupted parenting behavior by examining whether specific types or combinations of experiences of maltreatment during childhood (i.e., physical, sexual, and psychological abuse as well as neglect before the age of 18 years) present greater risk for the development of disrupted parenting behaviors. We assessed experiences of maltreatment during childhood among a group of expectant women, many of whom were living in poverty, in the third trimester of pregnancy using a person-centered technique (LPA). The aim was to explore whether risk for later disrupted parenting behavior could be identified before the child is born. To our knowledge, this is the first known empirical investigation of prenatal predictors of later disrupted maternal parenting behavior in a high-risk sample of women.

To examine associations between different types and combinations of experiences of childhood maltreatment and disrupted maternal behavior, we used a data-driven approach to identify latent profiles of childhood maltreatment experiences. We hypothesized the following: (a) greater severity of total maltreatment experiences during childhood would be associated with greater severity of disrupted maternal parenting behavior 1-year postpartum, (b) different profiles of maltreatment experiences during childhood would exist, and (c) prenatal profiles of childhood maltreatment experiences would have differential relationships with later disrupted parenting behaviors. We did not make specific a priori predictions about possible profiles; however, we predicted that we would find heterogeneity of experiences of childhood maltreatment.

Method

Participants

Participants included a community sample of 120 mother–child dyads who participated in a larger prospective, longitudinal study that began during pregnancy and extended through the child’s third birthday. Demographic characteristics of the sample are displayed in Table 1. In brief, the sample was racially/ethnically diverse with a majority of participants identifying as African American. The sample was predominantly economically disadvantaged with a majority of participants living below the federal poverty line with high rates of social service usage. However, our sample was relatively well educated with a majority reporting having attended at least some post–high school education.

Table 1.

Demographic Information.

Variable	M	SD
Age in years	26.21	5.70
Monthly income in dollars	2,143.14	2,109.69
	n	Percent
Race
African American	56	47
Caucasian	43	36
Biracial	15	13
Other	6	5
Education
High school diploma or less	24	20
Some college	53	44
College degree or higher	43	36
Marital status
Single/never married	76	63
Married	33	28
Divorced	5	4
Separated	6	5
First pregnancy	36	30

Eligibility requirements included the following: (a) being pregnant, (b) being at least 18 years of age, and (c) having the ability to speak fluent English, as bilingual translators were not available. Participants were recruited through the posting of fliers in public locations, as well as local community organizations and agencies serving low-income families. The strategic distribution of fliers allowed for the recruitment of economically disadvantaged pregnant women, which was a specific focus of the overall larger, longitudinal study. Data from the first (T1; third trimester) and the third (T3; 12-month postpartum) collection periods were used herein. The present study was approved by the institutional review board of the institution at which the research occurred.

Procedures

The pregnancy (T1) interviews were primarily conducted in the participant’s home (22% conducted in a research office). Participants reported on their history of childhood maltreatment as well as a broad range of other experiences. Consent was obtained from parents after thoroughly describing the purpose of the study and all tasks that would be completed. Families were given a US$25 gift card for their participation. The 12-month postpartum (T3) visit was also mainly conducted in the participant’s home (7% conducted in a research office). In addition to questionnaires about the health and psychological wellbeing of the mother and child, participants engaged in a 10-min free play and 2-min cleanup interaction task with their child. Consent for participation was obtained from parents and on behalf of their children after a thorough review of the study purpose and tasks that would be completed. A standard set of developmentally appropriate toys was brought to the interview and used for the interaction task. This interaction was videotaped and later coded for disrupted maternal behavior. Families were given US$50 in cash and a small baby gift for their participation.

Both interviews were conducted by teams consisting of graduate and undergraduate students in psychology. All interviewers underwent extensive training and received both in vivo and weekly didactic group supervision with the principal investigator and larger study team, where home visits were discussed at length.

Measures

Childhood trauma

The Childhood Trauma Questionnaire—Short Form (CTQ-SF) was administered in the third trimester of pregnancy to measure mothers’ experiences of maltreatment during childhood (i.e., birth to age 18 years; Bernstein et al., 2003; Bernstein & Fink, 1998). The CTQ-SF is a 28-item self-report questionnaire consisting of five subscales: emotional abuse, physical abuse, sexual abuse, emotional neglect, and physical neglect, each with five items capturing the frequency of each type of abuse. Higher scores on all calculated subscales indicate greater experiences of each type of maltreatment. Items are typically scored from 1 = never true to 5 = very often; however, in the present study, responses were rescaled from 0 to 4, such that scores of 0 indicate an absence of experiences of maltreatment to aid in interpretation of the latent profiles. Combining all five subscales, a total maltreatment score was generated. The psychometric properties of the CTQ-SF were reported by Bernstein et al. (2003) across a wide range of different samples (i.e., substance abusers, inpatient psychosis patients, and a community sample) and revealed good criterion-related validity (compared with therapist ratings of abuse and neglect) and construct validity. Furthermore, internal consistency reliability of the CTQ-SF in the current study is excellent for all subscales: emotional abuse alpha = .91, physical abuse alpha = .90, sexual abuse alpha = .96, emotional neglect alpha = .92, and physical neglect alpha = .84. The alpha for the total score in the present study is .96.

Disrupted maternal parenting behavior

Disrupted parenting behavior was coded using the Disrupted Maternal Behavior Instrument for Assessment and Classification (AMBIANCE; Bronfman et al., 2014). The AMBIANCE consists of five related, but distinct, dimensions of disrupted parenting behavior: affective communication errors, role/boundary confusion, fearful/disoriented, intrusive/negativity, and withdrawal. Empirical investigations of the factor structure of the AMBIANCE revealed that each dimension is a unique factor underlying the larger disrupted parenting behavior construct (Lyons-Ruth et al., 1999). Detailed information about the behaviors coded within each dimension is available elsewhere (Lyons-Ruth et al., 1999). Each dimension is rated on a 1 to 7 scale based on the frequency and severity of behaviors observed within each dimension, with higher values indicating more disrupted behaviors of each type. Based on the five dimensional ratings, a 1-7 total score is also assigned that captures the overall quality of caregiving behavior. Using the total score, caregivers are classified as not disrupted (a score of 1 to 4) or disrupted (a score of 5 to 7).

In the present study, disrupted parenting behaviors were coded from videotaped mother–infant interactions. The first 5 min of the interaction were not coded to allow the dyad to “warm up”; only the remaining 7 min of the interaction was viewed and coded. Interrater reliability (IRR) was established on 33% of interactions in the present sample. IRR (using kappa for classification and intraclass correlation for continuous ratings) was acceptable to excellent for all ratings: overall dichotomous rating (k = .94), affective communication errors (r = .84), role/boundary confusion (r = .78), fearful/disoriented (r = .70), intrusive/negativity (r = .91), and withdrawal (r = .75).

Data Analysis

Hypothesis 1, that greater severity of total maltreatment experiences during childhood would be associated with greater severity of disrupted maternal parenting behavior 1-year postpartum, was examined using scatter plots and bivariate correlations. Childhood maltreatment severity was operationalized as the total severity score on the CTQ and disrupted maternal parenting behavior as the total AMBIANCE score. Additional exploratory correlations examined associations between total childhood maltreatment experiences, the frequency of different types of child maltreatment (emotional, physical, sexual abuse, and emotional and physical neglect), and ratings of different types of disrupted parenting behavior (affective communication errors, role/boundary confusion, fearful/disoriented, intrusive/negativity, and withdrawal).

Hypothesis 2 stated that different prenatal profiles of childhood maltreatment experiences exist. Prenatal profiles of experiences of maltreatment during the mother’s childhood were explored using LPA. Five variables determined profiles of maltreatment experiences: physical abuse, sexual abuse, emotional abuse, physical neglect, and emotional neglect. Unconditional models (i.e., without covariates and AMBIANCE variables) were examined first to ensure heterogeneity of the data. For the conditional models, the maternal level of education was included as a covariate due to a significant correlation with overall (total) maternal trauma history (r = −.25, p<.01).

LPA is a data-driven approach, meaning that the number of profiles within the data is not predetermined, but rather, guided by fit indices, which identify the optimal number of profiles that exist within the data (i.e., the number of profiles that best represent the data). There are many different indices to choose from, and it is important to consider more than one index of fit. Although there are no firm guidelines for choosing which fit indices to examine, the Bayesian information criterion (BIC) as well as the Lo–Mendell–Rubin adjusted likelihood ratio test (LMR-A) and bootstrap likelihood ratio test (BLRT) have shown to be consistent indicators of the adequate number of profiles (Nylund et al., 2007). Furthermore, although fit statistics are important, theoretical interpretability of the profiles is also meaningful in choosing the most appropriate model.

Hypothesis 3 stated that prenatal profiles of childhood maltreatment experiences would have differential relationships with later disrupted parenting behaviors. Differences between maltreatment profiles and the AMBIANCE dimensions were explored by including the AMBIANCE dimensions as auxiliary variables in the conditional LPA models. When using multiple imputation in Mplus, only descriptive information is provided for auxiliary variables (e.g., M and SE, no p values); therefore, the SE was converted to SD that was used to calculate Cohen’s d values to evaluate the size of the difference in AMBIANCE dimensions between the maltreatment profiles.

Data preparation and handling of missing data

Missing data were minimal and missing at random. All 120 participants provided information on childhood trauma experiences during pregnancy (T1). Multiple imputation was used for the final conditional LPA models, with AMBIANCE entered as auxiliary outcome variables, to account for missingness on the AMBIANCE scales; the maximum likelihood ratio (MLR) for missing data cannot address missingness on auxiliary variables in Mplus. A total of 12% (n = 14) of women were missing AMBIANCE scores for various reasons: three were unable to be located at the time of the 1-year interview, two withdrew from the study, one moved out of the country, one did not have consistent contact with her child at the time of the interview, two did not have custody of their children at the time of the interview, four interviews were completed over the phone, and one video was unable to be coded due to various interruptions and distractions during the free play and cleanup. Using MLR and multiple imputation allowed for all analyses (except for descriptive statistics) to be based on the full sample size of N = 120.

Results

Descriptive Statistics

Women reported a wide range of experiences of childhood maltreatment. A total of 64% of women reported at least one experience of childhood maltreatment; 50% reported emotional abuse, 32.5% physical abuse, 29.2% sexual abuse, 42.5% emotional neglect, and 29.2% physical neglect.

The overall level of disrupted parenting was moderately high, with a mean of 4.25 (of 7). That is, across all five AMBIANCE dimensions, on average, women in the present study demonstrated nonoptimal to disrupted parenting behaviors with their 1-year-old infants. In regard to the dichotomous disrupted versus not-disrupted classification, nearly half of the women (49%) in the present study were classified as disrupted.

Associations Between Childhood Maltreatment and Disrupted Parenting Behavior

A scatter plot between childhood maltreatment severity (CTQ total) and the overall total AMBIANCE score revealed a very weak relationship. In fact, the bivariate correlation between the CTQ total and overall AMBIANCE score was not statistically significant (r = .10, ns; see Table 2).

Table 2.

Correlations Between Disrupted Parenting and Experiences of Childhood Maltreatment.

AMBIANCE Dimension	Emotional Abuse	Physical Abuse	Sexual Abuse	Emotional Neglect	Physical Neglect	Total Child Maltreatment
Affective communication errors	.06	.14	.21	.12	.13	.20
Role/boundary confusion	.01	.07	.09	.03	.21	.12
Fearful/disoriented	.04	.05	.16	.05	−.01	.06
Intrusiveness/negativity	−.02	.08	.04	.02	.12	.08
Withdrawal	−.04	−.05	.02	−.03	−.17	−.05
Overall level of disrupted behavior	.04	.14	.13	.04	.06	.10

Exploratory associations were conducted to examine relationships between the frequency of different types of child maltreatment (emotional, physical, and sexual abuse and emotional and physical neglect) and different types of disrupted parenting behavior (affective communication errors, role/boundary confusion, fearful/disoriented, intrusive negativity, and withdrawal). Given the large number of associations explored (five childhood maltreatment by five AMBIANCE dimensions), we applied a Bonferroni adjustment, such that the significance value was set at the p < .002 level. Using this correction, none of the bivariate associations between types of childhood maltreatment and dimensions of disrupted parenting behavior were statistically significant. However, without the correction, it is noteworthy that a history of childhood sexual abuse was positively associated with affective communication errors (r = .21, p < .05), and childhood physical neglect was positively associated with role/boundary confusion (r = .21, p < .05).

Next, we examined the relationship between different levels (none, low, moderate, severe) of each type of childhood maltreatment (e.g., emotional abuse and physical abuse) and disrupted classification (i.e., a score of 5–7). The percentage of women classified as disrupted increased among those who reported severe levels of childhood maltreatment. More specifically, of the women who reported severe emotional abuse, 63% (n = 12) were classified as disrupted. Similarly, 60% (n = 9) of women with severe physical abuse and 60% (n = 6) of women with severe physical neglect were classified as disrupted. The highest percentages of women classified as disrupted were among women who reported severe sexual abuse (83% disrupted, n = 10) and severe emotional neglect (80% disrupted, n = 8).

Latent Profiles of Childhood Maltreatment Experiences and Later Disrupted Parenting Behavior

Five variables determined unconditional (i.e., without maternal education) latent profiles of childhood maltreatment experiences: emotional abuse, physical abuse, sexual abuse, physical neglect, and emotional neglect. A two-profile solution demonstrated improved fit compared with a single profile, indicating the presence of latent profiles (see Supplemental Table 1). In the conditional models, controlling for maternal education, BIC values continued to decrease for each additional profile added and bottomed out at the five-profile model, which demonstrated good fit based on the BIC and BLRT statistics (see Table 3). However, the model had specification issues and, therefore, was not a suitable model. Therefore, the four-profile model was chosen as the best fitting model; it was superior to the three-profile model in terms of BIC and the BLRT statistics and produced four theoretically interesting profiles (see Figure 1). Classification quality for the four-profile model was excellent (entropy = .99). The four-profile model consisted of a low exposure group with less severe childhood maltreatment compared with other groups (n = 78, 65%), a high exposure group with high reported frequency of childhood maltreatment across all five types (n = 15, 12.5%), a relatively low maltreatment group except for high frequency of sexual abuse, labeled high sexual abuse (n = 12, 10%), and a relatively high physical and emotional maltreatment exposure group with few incidences of sexual abuse (n = 15, 12.5%).

Table 3.

Conditional Latent Profile Models for Childhood Trauma Experiences.

Models	BIC	LMR-A	p Value	BLRT	p Value
Two profile	1,884.47	432.56	<.001	−1,119.49	<.001
Three profile	1,827.65	87.71	.30	−896.75	<.001
Four profile	1,785.55	73.42	.34	−851.59	<.001
Five profile	1,759.46	57.88	.23	−813.78	<.001
Six profile	1,770.03	27.21	.33	−786.52	<.001

Note. BIC = Bayesian information criterion; LMR-A = Lo–Mendell–Rubin adjusted likelihood ratio test; BLRT = bootstrap likelihood ratio test.

Figure 1.

Latent profiles of mothers’ childhood trauma experiences.

To explore differences in AMBIANCE scores between profiles, the conditional, four-profile model was examined with all five AMBIANCE dimensions entered as auxiliary variables. Moderate to large effect sizes were observed for the difference in affective communication errors between the high exposure profile and all other profiles (Cohen’s d ranged from .46 to .68; see Table 4). For role/boundary confusion, only the difference between the high exposure and low exposure profiles was notable, with the high exposure profile demonstrating more role/boundary confusion (moderate effect, d = .49). Meaningful differences between profiles were not observed for fearful/disoriented, intrusive/negativity, or withdrawal parenting dimensions.

Table 4.

Childhood Trauma Profile Differences on AMBIANCE Dimensions.

Variable	Profile 1Low Exposure		Profile 2High Sexual Abuse		Profile 3High Physical and Emotional Maltreatment		Profile 4High Exposure
Variable	M	SD	M	SD	M	SD	M	SD
Affective communication errors	3.21^a	1.68	3.57^a	1.70	3.30^a	2.05	4.33^a	1.63
Role/boundary confusion	2.63^a	1.41	2.45	1.39	2.96	1.63	3.33	1.47
Fearful/disoriented	3.29	1.50	3.44	1.49	3.36	1.63	3.60	1.51
Intrusiveness/negativity	4.18	1.68	4.05	1.70	4.12	2.17	4.66	1.63
Withdrawal	2.79	1.41	2.59	1.32	2.71	1.67	2.68	1.43
Education	3.77	1.50	3.19	1.14	3.26	1.32	2.74	1.20

Note. Mplus only provides two decimal places for means and standard deviations of distal outcomes.

Profile differs from profile 4 (Cohen’s d greater than .4).

Discussion

The current study examined associations between experiences of maltreatment during childhood as reported prenatally and disrupted maternal parenting behavior 1-year following the birth of a child among an economically disadvantaged sample that was diverse in regard to race/ethnicity. Specifically, using LPA, we examined whether different types (e.g., sexual vs. physical vs. psychological) and combinations of childhood maltreatment were examined. The best fitting model included four groups: low exposure to maltreatment (n = 78, 65%), high exposure to all forms of maltreatment (n = 15, 12.5%), high sexual abuse only (n = 12, 10%), and high physical and emotional maltreatment and low sexual abuse (n = 15, 12.5%). The intended aim of the present study was to examine associations between specific childhood maltreatment profiles identified during pregnancy and later disrupted parenting behaviors, which are known to be associated with infants’ social-emotional development (Lyons-Ruth et al., 1999; Madigan et al., 2007; Shi et al., 2012).

Results revealed that women in the group characterized by high exposure to all maltreatment types displayed more affective communication errors with their infants compared with women in all other profiles. Affective communication errors include behaviors such as contradictory signaling (e.g., frowning while speaking in a sweet voice), failure to initiate proper responses to infant cues, and inappropriate responses to infant cues. Experiencing interpersonal trauma, particularly multiple types of childhood trauma, is known to negatively affect interactions with others due to high levels of distrust in others, a strong desire to avoid social interaction, and alterations in how women view themselves (Cloitre et al., 2005; Herman, 1992). Therefore, mothers who have experienced multiple types of maltreatment may be very uncomfortable with intimacy and closeness with their infant. This type of extensive maltreatment history may also lead to a defensive denial of children’s own needs for closeness and connection. As a result, these mothers may be more prone to miss or respond inappropriately to their infant’s cues for closeness and exploration. For example, a mother may experience an infant’s cries as overwhelming/overpowering or, alternatively, as intimidating and aggressive. If so, they may respond to distress by minimizing the infant’s affective experience by saying something such as, for example, “you don’t need me” or “stop that right now.” Along similar lines, mothers may also respond inappropriately to their child’s cues by, for example, laughing while the infant is crying.

Results also showed that women in the high exposure profile demonstrated the highest average level of role/boundary confusion with their infants, with meaningfully higher levels compared with the low exposure and high sexual abuse profiles. Furthermore, women in the high physical and emotional maltreatment profile demonstrated higher levels of role/boundary confusion compared with women in the high sexual abuse group. Role/boundary confusion within the AMBIANCE system includes behaviors that suggest clear difficulty in the caregiver’s ability to prioritize the infant’s needs over her own including repeatedly asking for affection, asking the child for permission, and in extreme cases, treating the infant like a spouse or romantic partner. It is possible, although speculative at this stage, that having experienced multiple types of childhood maltreatment reduces a mother’s sense of agency and effectiveness in relationships including in her own parental role. For instance, the pervasive experience of maltreatment may result in a sense of having no control in relationships, an internalized sense of helplessness, and excessive dependency on others for care and support (including one’s own children). This presentation is similar to symptoms observed in borderline personality disorder and broader conceptualizations of post-traumatic stress disorder (PTSD; for example, complex PTSD), which have each been heavily associated with multiple experiences of trauma and victimization across development (Frost et al., 2020; Jowett et al., 2020; MacIntosh et al., 2015).

Interestingly, meaningful differences between childhood maltreatment profiles were not observed for the fearful/disoriented, intrusive/negativity, or withdrawal parenting dimensions. Women across all childhood maltreatment profiles demonstrated similar levels of fearful/disoriented behavior, which approached the nonoptimal range for all groups. A score of 4 indicates nonoptimal (though not disrupted) levels, and women across all profiles averaged 3.29 to 3.60 on this scale. Similarly, women across all profiles demonstrated, on average, nonoptimal levels of intrusive/negativity behaviors (means ranged 4.05–4.66). Levels of withdrawal were also similar across profiles, though not particularly elevated (2.59–2.79). A lack of findings for the withdrawal dimension, in particular, could be due to the fact that severe withdrawal was not frequently observed in the present study; therefore, there was not enough range of maternal withdrawal behaviors to detect differences between profiles.

In sum, LPA results suggest that different combinations of experiences of childhood maltreatment have implications for the display of maternal affective communication errors and role/boundary confusion during mother–infant interactions at 1-year postpartum. In particular, experiencing multiple, different types of childhood maltreatment appears to affect a mother’s ability to communicate with her child in sensitive and predictable ways, including her ability to read, understand, and respond to her child’s needs for safety and emotional support without ignoring or mocking her child’s experience. Experiencing multiple types of childhood maltreatment was also associated with the ability to form appropriate roles within the parent–child relationship. A large body of past research suggests that exposure to disrupted parenting behavior (more generally) contributes to the development of child disorganized attachment (e.g., Lyons-Ruth et al., 1999; Madigan et al., 2007; Shi et al., 2012); thus, it will be important for future research to further elucidate how different dimensions of disrupted parenting behavior, resulting from extensive maternal childhood trauma, affect the parent–child relationship and child social-emotional development.

Importantly, this is the first known study to examine latent profiles of maternal childhood trauma experiences as they relate to different dimensions of disrupted parenting behavior. Identifying latent profiles of trauma experiences is an innovative approach to examining predictors of parenting behavior and highlights the importance of examining individuals’ unique experiences, compared with relying on trends within an entire sample. Furthermore, this is one of the few studies to examine predictors of the different dimensions of disrupted parenting behavior coded with the AMBIANCE system; most of the past research on disrupted parenting behavior has only utilized the total score and/or the disrupted versus not-disrupted classification.

Although the LPA revealed meaningful differences in disrupted parenting behavior between profiles of childhood maltreatment, statistically significant bivariate associations between experiences of childhood maltreatment and disrupted parenting behavior were not found. In fact, some women in the present study who reported few experiences of childhood maltreatment were still observed to display a wide range of disrupted parenting behaviors. There was also considerable variability in AMBIANCE scores among those who reported numerous experiences of childhood maltreatment. These correlational findings are inconsistent with past findings. For instance, previous work by Lyons-Ruth et al. (1999) demonstrated a link between psychosocial risk (i.e., history of physical or sexual childhood maltreatment and/or inpatient psychiatric care) and disrupted parenting behavior; however, their entire sample reported one or more operationally defined risk experiences. In contrast, the women in the present study made up a nontreatment seeking, community sample who reported a wider range of maltreatment experiences including 36% of women who did not report any experiences of childhood maltreatment. It is possible, therefore, that the association between childhood trauma and disrupted parenting was weaker in the present study due to the presence of women without experiences of maltreatment during childhood. However, among women in the present study with severe experiences of childhood maltreatment, particularly sexual abuse and emotional neglect, rates of being classified as disrupted were considerably higher compared with the entire sample (49% vs. 83% and 80%, respectively), highlighting the negative effect of severe experiences of childhood maltreatment on parenting behavior.

The weak correlational associations between mothers’ reports of their own experiences of childhood maltreatment and observed disrupted parenting behavior suggest the classification of childhood trauma experiences as “unresolved” may be a critical condition to detect an association between childhood maltreatment experiences and disrupted parenting behavior. The classification of trauma or loss as unresolved indicates that the traumatic experience or traumatic grief has not yet been psychologically processed (i.e., not yet been made conscious and/or coherent), which can lead to ongoing impairments in many domains of functioning, including parenting (Lyons-Ruth et al., 1999; Main & Hesse, 1990). Because the AAI was not administered in the present study, the degree to which participants’ childhood trauma was (un)resolved could not be determined. Even so, the present study contributes to the field’s understanding of how experiences of trauma may (or may not) affect disrupted parenting behavior. Having an experience of trauma in childhood may not, by itself, indicate risk for serious parenting difficulties; furthermore, unresolved traumatic experiences might not be the only source of risk for disrupted parenting. For example, in addition to findings with more severe childhood trauma experiences noted earlier, LPA results in the present study suggest that experiencing multiple types of trauma puts women at a greater risk of displaying disrupted parenting behaviors. Furthermore, in the present study, women with severe experiences of childhood maltreatment, particularly sexual abuse and emotional maltreatment, had exceptionally high rates of disrupted parenting (>80%) compared with the rate of disrupted parenting in the sample as a whole (49%).

Collecting information on experiences of abuse and maltreatment using a trauma-informed approach (see Racine et al., 2020) when women are pregnant may help identify those at risk for displaying disrupted parenting behavior after the birth of a child, which is known to negatively affect the security of the parent–infant relationship (Lyons-Ruth et al., 1999; Madigan et al., 2006, 2007). Furthermore, knowing whether pregnant women have experienced multiple types of childhood maltreatment matters. In particular, women who report experiencing multiple, different types of past trauma may display higher levels of certain forms of disrupted parenting behavior compared with women who experience one type of trauma. Therefore, early prevention of disrupted parenting behaviors requires an individualized approach that considers the varied trauma experiences women may have had during childhood. In addition, our sample was predominately economically disadvantaged and racially/ethnically diverse, capturing an underrepresented group in the larger society that is disproportionally represented in regard to experiences of trauma. Furthermore, disrupted parenting behavior has been largely examined in low-risk samples comprising predominately Caucasian women. Thus, it is important to understand associations between trauma during childhood and later disrupted parenting behaviors in this vulnerable and understudied group of women.

Clinical Implications

Overall, the results of the present study show that indicators of risk for disrupted parenting can be identified during the prenatal period. Yet, many of our current parent–infant interventions focus on the postpartum period, after the baby is born and after the mother has begun to display disrupted parenting behaviors while interacting with her infant. There is a critical need to develop more focused prenatal interventions that address how experiences of trauma might affect specific forms of parenting. In addition, interventions with mothers who have certain profiles of trauma experiences may target decreasing certain forms of disorganized parenting. In fact, recent work highlights that reducing disruptive forms of parenting behaviors is an essential, explanatory component in the relationship between attachment-based interventions and more optimal attachment outcomes for young children (Tereno et al., 2017; Yarger et al., 2019).

Given also that unresolved trauma is known, based on past research, to be associated with disrupted parenting, interventions should include techniques designed to help mothers process and meaningfully make sense of their traumatic experiences; this may prove more challenging and time-consuming for those who have disavowed or repressed such experiences. Finally, it will be important for researchers to continue exploring protective factors that might buffer women from the development of disrupted parenting behaviors and serve as feasible intervention targets as opposed to unchangeable factors such as age, race, and trauma history.

Strengths and Limitations

The use of both variable- and person-centered analyses in the present study was a significant strength. Few, if any, meaningful findings were revealed when using traditional, variable-centered analyses (i.e., bivariate correlations). However, the use of LPA provided meaningful information that adds to our understanding about the role of maternal childhood trauma on later disrupted parenting behavior. The lack of findings in the present study when using variable-centered statistics compared with the significant findings using person-centered statistics, many of which demonstrated moderate to large effects, underscores the importance of utilizing techniques like LPA.

In addition, the use of the AMBIANCE system to measure disrupted parenting behavior and the strong IRR across all AMBIANCE dimensions is another significant strength of the current study. Most of the research thus far on disrupted parenting behaviors has focused on the disrupted versus not-disrupted classification or the overall AMBIANCE score. The use of the five disrupted parenting behavior dimensions in the present study allowed for a more in-depth and nuanced examination of how maternal trauma may adversely affect specific types of problematic parenting behavior. Knowing what types of behaviors might emerge among women with different experiences of trauma allows for more focused intervention aimed at reducing or preventing specific and serious parenting problems (e.g., affective communication errors or role/boundary confusion).

One limitation is the relatively small sample size, which may have negatively affected the ability to detect statistically significant differences between trauma profiles and later disrupted parenting behavior. Furthermore, although past research has employed LPA with similar sample sizes in other areas of trauma research (e.g., Au et al., 2013, n = 119), the ability to detect meaningful profiles can be affected by sample size. Future research with larger samples is needed to replicate and expand on the findings in the present study.

Another limitation is that unresolved status with respect to maternal experiences of trauma (as seen in studies utilizing the AAI) was not measured in the present study. Instead, latent profiles of self-reported experiences of childhood trauma were explored as predictors of disrupted parenting behavior. Future studies should explore whether latent profiles of self-reported accounts of experiences of trauma provide incremental predictive validity above unresolved status when predicting disrupted parenting behavior.

Conclusion

In sum, experiences of childhood maltreatment assessed during pregnancy are not rare, especially within diverse, low-income community samples, and present significant risk for disrupted parenting behavior following the birth of a child. In particular, experiencing multiple, different types, and more severe forms of childhood maltreatment may confer unique risks to the mother–infant relationship via disrupted parenting. Given that maltreatment history can be identified during pregnancy, before the child is born, the development of preventive interventions, guided by both attachment and trauma theories, during the prenatal period is imperative. Future research is needed to continue identifying profiles of prenatal risks, as well as possible protective factors, for disrupted parenting that can be integrated into the ongoing development of such interventions.

Supplemental Material

Childhood_maltreatment_and_parenting_Supplemental_Table_1 – Supplemental material for Prenatal Identification of Risk for Later Disrupted Parenting Behavior Using Latent Profiles of Childhood Maltreatment

Supplemental material, Childhood_maltreatment_and_parenting_Supplemental_Table_1 for Prenatal Identification of Risk for Later Disrupted Parenting Behavior Using Latent Profiles of Childhood Maltreatment by Katherine L. Guyon-Harris, Sheri Madigan, Elisa Bronfman, Gloria Romero and Alissa C. Huth-Bocks in Journal of Interpersonal Violence

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by the American Psychoanalytic Association (Huth-Bocks), Psi Chi International Honor Society in Psychology (Huth-Bocks), International Psychoanalytical Association (Huth-Bocks), Eastern Michigan University Office of Research Development (Huth-Bocks), the Blue Cross Blue Shield of Michigan Foundation Student Award Program (Guyon-Harris), and the International Society for Traumatic Stress Studies (ISTSS) Frank W. Putnam Trauma Research Scholars Award (Guyon-Harris). T32HP22240 (Guyon-Harris).

ORCID iD

Katherine L. Guyon-Harris

Supplemental Material

Supplemental material for this article is available online.

Author Biographies

Katherine L. Guyon-Harris, PhD, is a postdoctoral fellow in the University of Pittsburgh School of Medicine, Department of Pediatrics. Her research focuses on the impact of trauma and substance use on parenting and parent–child relationships. She is also interested in the integration of parenting interventions and behavioral health services into primary care settings serving families with young children, particularly underserved families.

Sheri Madigan, PhD, holds a Canada Research Chair in Determinants of Child Development and is an associate professor in the Department of Psychology at the University of Calgary. Her research focuses on enhancing understanding of how social stressors and adversities influence child socio-emotional development and why some mothers and children show resilience to these stressors, whereas others remain vulnerable.

Elisa Bronfman, PhD, is an assistant professor at Harvard Medical School and staff psychologist at Boston Children’s Hospital. She is the first author of the AMBIANCE scale, which was first developed as part of her dissertation in the laboratory of Dr. Karlen Lyons-Ruth. She trains extensively on the use of AMBIANCE with collaborators, Dr. Lyons-Ruth and Dr. Sheri Madigan.

Gloria Romero, MS, is a doctoral student in the Clinical Psychology PhD program at the University of Tennessee, Knoxville. Her research focuses on the intergenerational transmission of violence. Specifically, she is interested in the association between witnessing and experiencing violence within one’s family of origin and the perpetration and victimization of intimate partner violence in adulthood.

Alissa C. Huth-Bocks, PhD, is a professor in pediatrics, psychiatry, and psychological sciences at University Hospitals Cleveland Medical Center, Case Western Reserve University School of Medicine. She has conducted multiple longitudinal studies examining the impact of traumatic stress exposure and adversity on mothers and their young children. Other special interests are in early childhood mental health, parent–child attachment, and early social-emotional difficulties.

References

American Psychiatric Association. (2013). Diagnostic and statistical manual of mental disorders (5th ed.). American Psychiatric Publishing.

T. M.

Dickstein

B. D.

Comer

J. S.

Salters-Pedneault

Litz

B. T.

(2013). Co-occurring posttraumatic stress and depression symptoms after sexual assault: A latent profile analysis. Journal of Affective Disorders, 149, 209–216. https://doi.org/10.1016/j.jad.2013.01.026

Bailey

H. N.

DeOliveira

C. A.

Wolfe

V. V.

Evans

E. M.

Hartwick

(2012). The impact of childhood maltreatment history on parenting: A comparison of maltreatment types and assessment methods. Child Abuse and Neglect, 36, 236–246. https://doi.org/10.1016/j.chiabu.2011.11.005

Ballen

Bernier

Moss

Tarabulsy

G. M.

St-Laurent

(2010). Insecure attachment states of mind and atypical caregiving behavior among foster mothers. Journal of Applied Developmental Psychology, 31, 118–125. https://doi.org/10.1016/j.appdev.2009.10.001

Baumeister

R. F.

Leary

M. R.

(1995). The need to belong: Desire for interpersonal attachments as a fundamental human motivation. Psychological Bulletin, 117(3), 497–529. https://doi.org/10.1037/0033-2909.117.3.497

Bernstein

D. P.

Fink

(1998). Manual for the childhood trauma questionnaire. The Psychological Corporation.

Bernstein

D. P.

Stein

J. A.

Newcomb

M. D.

Walker

Pogge

Ahluvalia

… Zule

(2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Abuse and Neglect, 27, 169–190. https://doi.org/10.1016/S0145-2134(02)00541-0

Bronfman

Madigan

Lyons-Ruth

(2014). Atypical maternal behavior instrument for assessment and classification (AMBIANCE): Manual for coding disrupted affective communication. Advance online publication.

Cavanaugh

C. E.

Messing

J. T.

Petras

Fowler

La Flair

Kub

… Campbell

J. C.

(2012). Patterns of violence against women: A latent class analysis. Psychological Trauma: Theory, Research, Practice, and Policy, 4(2), 169–176. https://doi.org/10.1037/a0023314

10.

Charuvastra

Cloitre

(2008). Social bonds and posttraumatic stress disorder. Annual Review of Psychology, 59(1), 301–328. https://doi.org/10.1146/annurev.psych.58.110405.085650

11.

Cloitre

Miranda

Stovall-McClough

K. C.

Han

(2005). Beyond PTSD: Emotion regulation and interpersonal problems as predictors of functional impairment in survivors of childhood abuse. Behavior Therapy, 36, 119–124. https://doi.org/10.1016/S0005-7894(05)80060-7

12.

DiLillo

Damashek

(2003). Parenting characteristics of women reporting a history of childhood sexual abuse. Child Maltreatment, 8(4), 319–333. https://doi.org/10.1177/1077559503257104

13.

Frost

Hyland

Shevlin

Murphy

(2020). Distinguishing complex PTSD from Borderline Personality Disorder among individuals with a history of sexual trauma: A latent class analysis. European Journal of Trauma & Dissociation, 4, Article 100080. https://doi.org/10.1016/J.EJTD.2018.08.004

14.

Gaska

K. A.

Kimerling

(2018). Patterns of adverse experiences and health outcomes among women veterans. American Journal of Preventive Medicine, 55(6), 803–811. https://doi.org/10.1016/j.amepre.2018.06.029

15.

George

Kaplan

Main

(1996). Adult Attachment Interview. Unpublished manuscript, Department of Psychology, University of California, Berkeley (third edition).

16.

Goldberg

Benoit

Blokland

Madigan

(2003). Atypical maternal behavior, maternal representations, and infant disorganized attachment. Development and Psychopathology, 15, 239–257. https://doi.org/10.1017/S0954579403000130

17.

Gustafsson

H. C.

Cox

M. J.

Blair

(2012). Maternal parenting as a mediator of the relationship between intimate partner violence and effortful control. Journal of Family Psychology, 26, 115–123. https://doi.org/10.1037/a0026283

18.

Herman

J. L.

(1992). Complex PTSD: A syndrome in survivors of prolonged and repeated trauma. Journal of Traumatic Stress, 5, 377–391. https://doi.org/10.1007/BF00977235

19.

Hughes

Cossar

(2016). The relationship between maternal childhood emotional abuse/neglect and parenting outcomes: A systematic review. Child Abuse Review, 25(1), 31–45. https://doi.org/10.1002/car.2393

20.

Jowett

Karatzias

Shevlin

Albert

(2020). Differentiating symptom profiles of ICD-11 PTSD, complex PTSD, and borderline personality disorder: A latent class analysis in a multiply traumatized sample. Personality Disorders: Theory, Research, and Treatment, 11, 36–45. https://doi.org/10.1037/per0000346

21.

Lyons-Ruth

Block

(1996). The disturbed caregiving system: Relations among childhood trauma, maternal caregiving, and infant affect and attachment. Infant Mental Health Journal, 17(3), 257–275. https://doi.org/10.1002/(SICI)1097-0355(199623)17:3<257::AID-IMHJ5>3.0.CO;2-L

22.

Lyons-Ruth

Bronfman

Parsons

(1999). Atypical attachment in infancy and early childhood among children at developmental risk. IV. Maternal frightened, frightening, or atypical behavior and disorganized infant attachment patterns. Monographs of the Society for Research in Child Development, 64(3), 67–96. https://doi.org/10.1111/1540-5834.00034

23.

Lyons-Ruth

Bureau

J.-F.

Holmes

Easterbrooks

Brooks

N. H.

(2013). Borderline symptoms and suicidality/self-injury in late adolescence: Prospectively observed relationship correlates in infancy and childhood. Psychiatry Research, 206(2–3), 273–281. https://doi.org/10.1016/J.PSYCHRES.2012.09.030

24.

MacIntosh

H. B.

Godbout

Dubash

(2015). Borderline personality disorder: Disorder of trauma or personality, a review of the empirical literature. Canadian Psychology/Psychologie Canadienne, 56(2), 227–241. https://doi.org/10.1037/cap0000028

25.

Madigan

Moran

Pederson

D. R.

(2006). Unresolved states of mind, disorganized attachment relationships, and disrupted interactions of adolescent mothers and their infants. Developmental Psychology, 42(2), 293–304. https://doi.org/10.1037/0012-1649.42.2.293

26.

Madigan

Moran

Schuengel

Pederson

D. R.

Otten

(2007). Unresolved maternal attachment representations, disrupted maternal behavior and disorganized attachment in infancy: Links to toddler behavior problems. Journal of Child Psychology and Psychiatry and Allied Disciplines, 48(10), 1042–1050. https://doi.org/10.1111/j.1469-7610.2007.01805.x

27.

Main

Hesse

(1990). Parents’ unresolved traumatic experiences are related to infant disorganized attachment status: Is frightened and/or frightening parental behavior the linking mechanism? In Greenberg

M. T.

Cicchetti

Cummings

E. M.

(Eds.), Attachment in the preschool years: Theory, research, and intervention (pp. 161–182). University of Chicago Press.

28.

Minihan

Liddell

B. J.

Byrow

Bryant

R. A.

Nickerson

(2018). Patterns and predictors of posttraumatic stress disorder in refugees: A latent class analysis. Journal of Affective Disorders, 232, 252–259. https://doi.org/10.1016/J.JAD.2018.02.010

29.

Moehler

Biringen

Poustka

(2007). Emotional availability in a sample of mothers with a history of abuse. American Journal of Orthopsychiatry, 77, 624–628. https://doi.org/10.1037/0002-9432.77.4.624

30.

Muzik

Bocknek

E. L.

Broderick

Richardson

Rosenblum

K. L.

Thelen

Seng

J. S.

(2013). Mother-infant bonding impairment across the first 6 months postpartum: The primacy of psychopathology in women with childhood abuse and neglect histories. Archives of Women’s Mental Health, 16, 29–38. https://doi.org/10.1007/s00737-012-0312-0

31.

Nesselroade

J. R.

(1991). Interindividual differences in intraindividual change. In Collins

L. M.

Horn

J. L.

(Eds.), Best methods for the analysis of change: Recent advances, unanswered questions, future directions (pp. 92–105). American Psychological Association. https://doi.org/10.1037/10099-006

32.

Nylund

K. L.

Asparouhov

Muthén

B. O.

(2007). Deciding on the number of classes in latent class analysis and growth mixture modeling: A Monte Carlo Simulation Study. Structural Equation Modeling, 14(4), 535–569. https://doi.org/10.1080/10705510701575396

33.

Racine

Killam

Madigan

(2020). Trauma-informed care as a universal precaution: Beyond the Adverse Childhood Experiences Questionnaire. JAMA Pediatrics, 174, 5–6. https://doi.org/10.1001/jamapediatrics.2019.3866

34.

Savage

L. É.

Tarabulsy

G. M.

Pearson

Collin-Vézina

Gagné

L. M.

(2019). Maternal history of childhood maltreatment and later parenting behavior: A meta-analysis. Development and Psychopathology, 31, 9–21. https://doi.org/10.1017/S0954579418001542

35.

Schechter

D. S.

Suardi

Manini

Cordero

M. I.

Rossignol

A. S.

Merminod

… Serpa

S. R.

(2014). How do maternal PTSD and Alexithymia interact to impact maternal behavior? Child Psychiatry and Human Development, 46, 406–417. https://doi.org/10.1007/s10578-014-0480-4

36.

Shi

Bureau

J. F.

Easterbrooks

M. A.

Zhao

Lyons-Ruth

(2012). Childhood maltreatment and prospectively observed quality of early care as predictors of antisocial personality disorder features. Infant Mental Health Journal, 33, 55–69. https://doi.org/10.1002/imhj.20295

37.

Slade

Cohen

L. J.

Sadler

L. S.

Miller

(2009). The psychology and psychopathology of pregnancy: Reorganization and transformation. In Zeanah

C. H.

(Ed.), Handbook of infant mental health (3rd ed., pp. 22–39). The Guilford Press. https://books.google.com/books?hl=en&lr=&id=yLscK_LGW3EC&oi=fnd&pg=PA22&dq=Slade+AND+pregnancy&ots=gzrJU-uguw&sig=uPW3pZ2bAhSj5QQnQqd9UgaETEE#v=onepage&q=SladeANDpregnancy&f=false

38.

Sutter

M. E.

Rabinovitch

A. E.

Trujillo

M. A.

Perrin

P. B.

Goldberg

L. D.

Coston

B. M.

Calton

J. M.

(2019). Patterns of intimate partner violence victimization and perpetration among sexual minority women: A latent class analysis. Violence Against Women, 25(5), 572–592. https://doi.org/10.1177/1077801218794307

39.

Tereno

Madigan

Lyons-Ruth

Plamondon

Atkinson

Guedeney

… Guedeney

(2017). Assessing a change mechanism in a randomized home-visiting trial: Reducing disrupted maternal communication decreases infant disorganization. Development and Psychopathology, 29(2), 637–649. https://doi.org/10.1017/S0954579417000232

40.

Wang

Bodner

T. E.

(2007). Growth mixture modeling: Identifying and predicting unobserved subpopulations with longitudinal data. Organizational Research Methods, 10(4), 635–656. https://doi.org/10.1177/1094428106289397

41.

Yarger

H. A.

Bronfman

Carlson

Dozier

(2019). Intervening with attachment and biobehavioral catch-up to decrease disrupted parenting behavior and attachment disorganization: The role of parental withdrawal. Development and Psychopathology. Advance online publication. https://doi.org/10.1017/s0954579419000786

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.01 MB