Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2021 Sep 1.
Published in final edited form as: J Abnorm Child Psychol. 2020 Sep;48(9):1223–1237. doi: 10.1007/s10802-020-00669-w

The interactive effects of child maltreatment and adolescent pregnancy on late-adolescent depressive symptoms

Justin Russotti 1, Elizabeth D Handley 1, Fred A Rogosch 1, Sheree L Toth 1, Dante Cicchetti 1,2
PMCID: PMC7395875  NIHMSID: NIHMS1607657  PMID: 32594294

Abstract

Adolescent females are disproportionately at risk for depression, which is expected to represent the leading cause of disability in 2030 (National Institute of Mental Health [NIMH], 2019). Although prior research has suggested that both child maltreatment and adolescent pregnancy increase the risk for depressive symptoms, less is known about how these two interact to influence depression in late adolescence. The present study tested the unique and interactive effects of adolescent pregnancy and child maltreatment on late-adolescent depressive symptomatology (N=186) with a prospective, longitudinal design that utilized documented records of maltreatment and included demographically comparable (i.e., economically disadvantaged), nonmaltreated and non-pregnant comparisons. Participants were assessed at ages 10-12 and 18-21. Structural equation modeling was used to test whether adolescent pregnancy amplified the effect of child maltreatment on late-adolescent depressive symptoms. In the context of economic disadvantage, results indicated that the effect of child maltreatment on late-adolescent depressive symptoms was significantly enhanced for those who experienced an adolescent pregnancy. This effect remained after controlling for prior depressive symptoms, peer and maternal relationship quality, and romantic relationship violence. The findings are translated to preliminary guidance for practitioners regarding precision depression screening and tailored preventive interventions.

Keywords: child maltreatment, adolescent pregnancy, internalizing psychopathology, depression

Introduction

Depression is a leading cause of world-wide disability and is expected to represent the primary contributor to the global burden of disease by 2030 (World Health Organization [WHO], 2019). The negative consequences of depression are not relegated to psychological distress, as this disorder imposes serious physical health problems and increases the risk for death by suicide (Malhi & Mann, 2018). The extensive impact of depression may be partly due to the particularly high incidence in adolescence that results in unrelenting and cumulative suffering across the life course (Kwong et al., 2019). Indeed, although depression inflicts suffering with indiscriminate impunity, adolescents are a particularly vulnerable group (Kieling et al., 2019).

More specifically, adolescent females are disproportionately at risk, with approximately 20% of adolescent females experiencing a major depressive episode in the past year, compared to 8.7% of adult females or 6.8% of adolescent males (National Institute of Mental Health [NIMH], 2019). Despite significant efforts to prevent or mitigate depression in adolescent females, rates of depression continue to rise in this subgroup, especially in low-income populations (Hasin et al., 2018). Therefore, additional studies that identify subgroups of adolescent females at greatest risk for depression may improve the precision of existing prevention efforts and inform selective prevention.

However, even within the subgroup of adolescent females, depression etiology is complex and multifactorial (Kwong et al., 2019). Research has assembled a substantial list of environmental risk factors that may induce the onset of depression, including social disadvantage, family history of depression, childhood adversity, and exposure to life stressors (Malhi & Mann, 2018). Thus, it is important to examine specific synergistic combinations of depressogenic risk factors so that highly vulnerable subgroups can be detected, and targeted interventions can be designed and implemented with the most vulnerable individuals.

The synergistic combination of two robust depressogenic risk factors, adolescent pregnancy and child maltreatment, are considered in this study. Both adolescent pregnancy and child maltreatment represent significant public health priorities in the United States (Mollborn, 2017; Trickett et al., 2011) and these potent life stressors can have a pervasive influence on later depressive symptomatology (Hodgkinson et al., 2014; Li et al., 2016). Although prior research has suggested that both child maltreatment and adolescent pregnancy increase the risk for depressive symptoms, less is known about how these two interact, or if their influence remains over and above the continuity of preexisting depression.

Child maltreatment and depression

Child maltreatment has been documented as a potent depressogenic risk factor (Cicchetti & Toth, 2016). Maltreated children are often deprived of salubrious environments that promote adaptive functioning; instead, their pathogenic environment is known to result in diverse maladaptations that likely account for the robust association with depression, including hindered resolution of stage-salient developmental tasks, insecure attachments, difficulties with emotion recognition and regulation, negative emotional patterns, and interpersonal challenges (Cicchetti & Toth, 2016; Trickett et al., 2011). One meta-analytic review suggested that over half of global depression can be plausibly attributed to child maltreatment (Li et al., 2016). Further, child maltreatment is associated with greater depression chronicity, severity, and duration (Humphreys et al., 2020).

Though it is established that child maltreatment potentiates depression, additional research is needed to understand how the adverse effects of child maltreatment may differ under certain conditions or contexts (Widom, 2020). Child maltreatment can create a depressogenic state of vulnerability that can pervade across the life course and become amplified or magnified in the face of more proximal stressors (Cicchetti & Toth, 2016; Lippard & Nemeroff, 2020; McLaughlin et al., 2010). Adolescent pregnancy represents one potential proximal stressor that may exacerbate maltreatment-related vulnerabilities for depression.

Adolescent pregnancy and depression

When experienced during adolescence—a sensitive period of emotional, social, and biological change—pregnancy can have a pervasive influence on later depressive psychopathology. Depression occurs in adolescent mothers at more than twice the rate of non-parenting adolescents and older mothers (Estrin et al., 2019; Hodgkinson et al., 2014). Moreover, depressive symptoms are likely to persist and endure throughout adulthood long after the initial pregnancy (Hodgkinson et al., 2014; Schmidt et al., 2006).

The pathophysiology of depression following adolescent pregnancy may be explained by the atypical or ill-timed transition to parenting1. Research suggests that pregnancy itself ushers in biological changes and increased life stress that can exacerbate the risk for depression, regardless of age (Hodgkinson et al., 2014). These effects may be compounded when the pregnancy occurs during the biologically-sensitive periods of adolescent development. Further, adolescent pregnancy can enact a cascade of stress as the individual attempts to navigate the competing demands of adolescent development and the possible responsibilities of parenthood. Indeed, adolescent pregnancy is associated with several depression-related stressors, including interpersonal stress, relationship strain, social isolation, a reduction in social networks, diminished social support, financial disadvantage, and educational barriers (Estrin et al., 2019; Mollborn, 2017).

Combined risk

Studies typically examine these two etiological risks for depression separately, and less is known about how the effect of one may enhance the vulnerability of the other. This is a notable omission, as there are several plausible, theoretical reasons why the context of pregnancy may specifically make the deleterious effects of maltreatment more salient. For instance, due to the interpersonal nature of child maltreatment, there are likely unique mechanisms operating during or after pregnancy for victims that may not be salient for non-victimized adolescents who become pregnant. Pregnancy represents an intimate, emotional period that can initiate a deep psychological re-organization in some women (Narayan et al., 2016). For individuals with maltreatment histories, this transformation may be negatively shaped by previous interpersonal adversities and/or pose a high risk for re-experiencing past abuse (Narayan et al., 2016). For example, the maternal attachment system is triggered during pregnancy, magnifying the quality of current and past relationships in the adolescent’s life, and these processes may awaken the emotional scars of maltreatment that were defensively kept out of conscious awareness, ultimately (re)activating or amplifying depression vulnerability (Narayan et al., 2016). Indeed, caregiving history has been shown to moderate depressive symptoms in pregnant adolescents, such that pregnant adolescents who recall harsh, inconsistent, and/or unavailable caregiving experiences report greater depressive symptoms (Milan et al., 2007), while favorable memories of caregiving are known to reduce depression in pregnant women (Chung et al., 2008). Further, pregnant adolescents with maltreatment histories are more likely to have unresolved states of mind related to their abuse (i.e., unassimilated or disorganized narratives around the abuse; Madigan et al., 2012), which may lie dormant prior to pregnancy and then resurface, with enhanced salience, during pregnancy, ultimately leading to depression (McMahon et al., 2008).

A limited number of studies have tested whether becoming pregnant during adolescence is more strongly predictive of depression for youth with a history of child maltreatment, compared to those without this early adversity (Hillis et al. 2004; Milan et al., 2004). Studies have demonstrated that maternal maltreatment history may intensify the risk for depression during the antenatal and postnatal period for adult pregnancies (Alvarez-Segura et al., 2014; Chamberlain et al., 2019; McDonnel & Valentino, 2016; Tung et al., 2019). Additionally, multiple studies have confirmed that abuse history is associated with greater depressive symptoms within groups of pregnant adolescents (Easterbrooks et al., 2016; Meltzer-Brody et al., 2013; Romano et al., 2006) and that this effect is independent of confounding factors, such as demographics, previous substance use, and depression (Tzilos et al., 2012). Moreover, Madigan et al (2014) found that abuse history attenuated a decline in depressive symptoms for pregnant adolescents up to 12 months postpartum, indicating that abuse history may act to sustain common increases in depressive symptoms after pregnancy. Though these studies demonstrate the effect of abuse history on depression for pregnant adolescents, they were unable to simultaneously examine the effect of adolescent pregnancy on later depression due to the lack of a comparison group consisting of adolescents who never experienced an adolescent pregnancy.

In one of the few studies that tested the effect of child maltreatment on depression between pregnant and non-pregnant adolescents, Milan et al (2004) found that physical abuse history was related to emotional distress for pregnant adolescents. In line with this finding, Estrin et al (2019), using a cross-sectional design, found that adolescent pregnancy was associated with higher rates of depression and abuse history. However, neither study specifically tested the synergistic role of child maltreatment. Similarly, another study found that becoming pregnant as an adolescent resulted in psychosocial consequences (not depression, per se) only for individuals with a history of adverse childhood experiences (Hillis et al., 2004). However, all of these studies relied on retrospective self-reports of abuse, which are known to have methodological limitations (Baldwin et al., 2019).

Timing and confounds

Previous studies examining the effects of child maltreatment and adolescent pregnancy on later depression have been unable to account for the competing predictive influence of depression continuity and the confounding role of economic disadvantage. For example, the association between depression and adolescent pregnancy is complex and bidirectional (Mollborn & Morningstar, 2009). Whether pregnancy exerts a unique effect on the development of adult depression, over and above pre-existing levels of symptoms, remains unclear. Similarly, adolescent pregnancy and child maltreatment often occur in the same environmental milieu as other depressogenic risk factors, such as economic disadvantage (Boden et al., 2008), low social support (e.g., peer and familial) and relationship violence (Rudolph, 2009). Thus, it is essential that studies examining the effects of adolescent pregnancy and child maltreatment on later depression do so by: employing a sample of individuals with similar economic backgrounds; using longitudinal, prospective designs to control for depression continuity; and controlling for aspects of the individual’s interpersonal milieu that may concurrently influence depression.

Present study

To our knowledge, no prospective, comparative studies exist that isolate the unique and interactive effects of both adolescent pregnancy and a history of child maltreatment and examine whether adolescent pregnancy amplifies child maltreatment vulnerabilities for depression. Much of the research on the interactive effects of child maltreatment and adolescent pregnancy on depression suffers from at least one of the following four methodological limitations: 1) the use of selected samples of pregnant adolescents without non-pregnant adolescent comparisons; 2) the use of cross-sectional designs that are unable to control for depression continuity from childhood to late-adolescence; 3) the use of retrospective self-reports of child maltreatment; 4) failure to account for the confounding roles of economic disadvantage and interpersonal risk factors. Considering the prevalence of both adolescent pregnancy and child maltreatment in the United States and the preponderance of evidence demonstrating considerable risk both stressors pose for the development of depression (see Li et al., 2016; Mollborn, 2017), a more precise understanding of how adolescent pregnancy may enhance vulnerability to later depressive symptoms for adolescents with child maltreatment histories is warranted.

Therefore, the aim of the present study was to examine the presence of an enhancing interaction (Cohen et al., 2003), whereby both child maltreatment and adolescent pregnancy affect the development of depressive symptoms and do so synergistically, rather than additively. We investigated the unique and interactive effects of adolescent pregnancy and child maltreatment on late-adolescent depression with a prospective, longitudinal design that relied on documented records of maltreatment and included demographically comparable (i.e., economically disadvantaged), non-maltreated and non-pregnant comparisons. By first methodologically controlling for economic disadvantage and statistically controlling for interpersonal risks (peer and maternal relationship quality and relationship violence), and then comparing the magnitude of each stressor (i.e., child maltreatment and adolescent pregnancy) in a demographically-comparable sample, more precise inferences can be made regarding the independent attributions of each depressogenic risk factor.

To examine these effects, adolescent pregnancy, child maltreatment, and the interaction between the two, were all included as predictors in our prospective model of the development of later depression. Additionally, given the importance of examining whether adolescent pregnancy and the interaction with child maltreatment exert a unique effect on depressive symptoms, over and above continuity, we included child depression as a predictor. Similarly, we controlled for the influence of peer and maternal relationship quality and romantic relationship violence on depression given their well-documented association with adolescent depression. We hypothesized that children who experienced a pregnancy in adolescence would evidence increased depression in late adolescence. We also hypothesized that children who experienced maltreatment would evidence increased depression in late adolescence. Finally, consistent with prior literature, we hypothesized that child maltreatment would increase the risk of depressive symptoms. However, we anticipated that the relationship between child maltreatment and depressive symptoms may become significantly more robust in the presence of a contextual variable known to enhance the salience of maltreatment experiences (i.e., adolescent pregnancy). Also, we hypothesized this interactive effect would remain over and above any main effects of child maltreatment, adolescent pregnancy, child depression, peer and maternal relationship quality, and romantic relationship violence.

Methods

Participants

The participants for the current investigation included 186 economically-disadvantaged females who were assessed across two longitudinal waves of data collection (childhood ages 10-12 and young adult ages 18-21). Wave 1 of data collection included 334 maltreated (n= 173) and non-maltreated (n =161) female children who participated in a research summer camp (see Cicchetti & Manly, 1990 for a detailed description of the research camp setting) as 10-12-year-olds (M = 11.28, SD = .97). At Wave 2, 186 of the original female participants were followed up at 18-21-years old (M = 19.54, SD= 1.07) and asked to complete a series of individual interviews and research assessments. The participants were racially and ethnically diverse (68.3% African American, 10.8% Caucasian, 7.5% Hispanic, 13.5% biracial/other race) and the families of participants were primarily headed by a single parent (68.7%) and had histories of receiving public assistance (96.1%). At Wave 2, 71.4% of late adolescents were living with their family of origin, 24.3% were in their own residence, and 4.3% were living with friends. There were no maltreatment differences (χ2(1, 333) = .09, p = .76) or significant differences in childhood depressive symptoms t(335) = 1.3 p = .20, in those who completed this study and those who did not complete the follow-up.

Recruitment

Participants were initially recruited based on documented records of child abuse and neglect reports through the Department of Human Services (DHS). A DHS liaison reviewed Child Protective Services (CPS) records and identified children who had been maltreated. Children in foster care were not recruited. A random sample of eligible families was selected for recruitment. The DHS liaison then contacted eligible families and explained the study to parents who were free to either agree to participate or to decline to have their information released to project staff. Interested participants provided project staff with informed consent for both their child’s participation in the summer camp program and for full access to any DHS records pertaining to the family.

A comprehensive and extensive review of DHS records was performed to code for maltreatment information using the Maltreatment Classification System (MCS; Barnett et al., 1993), a reliable and valid nosological maltreatment classification system (Manly, 2005) with operationalized determinations of child maltreatment parameters.

Maltreated children are predominantly from low-income families (Fourth National Incidence Study of Child Abuse and Neglect; Sedlak et al, 2010). Therefore, the DHS liaison identified demographically comparable families (i.e., families receiving Temporary Assistance for Needy Families) without histories of CPS involvement for recruitment for the non-maltreated comparison group. As with the maltreated group, the DHS liaison contacted a random sample of eligible non-maltreated participants to discuss study details. If participants expressed interest, their information was passed to project staff who were provided consent to search family DHS records and further verify the absence of maltreatment information using the MCS. Further, trained research staff completed the Maternal Child Maltreatment Interview (Cicchetti et al., 2003) with all mothers to confirm the lack of maltreatment. If any conflicting information was provided that suggested the non-maltreated participants may have experienced maltreatment, they were excluded from the comparison group.

Procedures

During Wave 1 of data collection, child participants attended a weeklong, full-day summer camp where they engaged in both traditional recreational camp activities and research assessments. At camp, participants were assigned to groups of eight (four non-maltreated and four maltreated) same-age and same-sex peers, with three camp counselors, who also completed behavioral observations. Children provided assent to participate in several individual research sessions at the camp in which they completed research assessments administered by trained research assistants. At Wave 2, after providing informed consent, the late-adolescent participants who were followed completed a set of individual, private, center-based research sessions. Research sessions were conducted by trained research assistants who were unaware of the participants’ maltreatment status and study hypotheses. The research assistants administered an extensive battery of assessments, including self-report measures on psychopathology symptoms and diagnostic clinical interviews.

Measures

Child maltreatment.

Child maltreatment was operationalized with the Maltreatment Classification System (MCS; Barnett et al., 1993) during Wave 1. The MCS is a multidimensional classification system that is applied to official, documented CPS records. This objective measure was developed to assess child maltreatment independently from parent-report and potential bias in reporting (Manly, 2005). The MCS is a reliable and validated measure of maltreatment (Manly, 2005). In this study, children were categorized by the MCS via a binary variable indicating if the child was maltreated (1) or nonmaltreated (0). Fifty-four percent of the sample was maltreated. The participant’s mother was the maltreatment perpetrator in 89% of the cases.

Childhood depressive symptoms

Child Depression Inventory (CDI; Kovacs, 2004).

The CDI was utilized to assess self-reported child depressive symptoms at Wave 1. The CDI is a 27-item questionnaire designed to assess depressive symptoms in school-aged children. Children were asked to respond to a series of items (e.g., feelings of worthlessness) by selecting the response that best described their feelings within the past two weeks. The CDI is a widely-used measure with strong psychometric properties (Kovacs, 2004). The current study relied on a summary score of the 27 items as an indicator of childhood depression (α=.86). In this sample, 19.4% of the participants exhibited symptoms that met the clinical threshold (M= 7.6, SD=6.6).

Child Behavior Checklist Teacher Report Form (TRF; Achenbach, 1991).

Children’s depressive symptomatology was assessed by camp counselors using the TRF at Wave 1. Camp counselors relied on weeklong observations of the children (~35 hours) to provide an assessment of behavioral disturbance from a non-related adult. The TRF is a well-validated and reliable measure (Achenbach, 1991; Achenbach, Dimenci, & Rescorla, 2003) that assesses internalizing and externalizing symptom domains across 118 items. The TRF can yield validated syndrome subscales and DSM-oriented subscales (Achenbach, 1991; Achenbach et al., 2003). Children were independently rated by two separate camp counselors and the counselor scores were averaged to create one T-score for each subscale (e.g., depressive problems) and counselors had an average interrater reliability of (k) 0.68. The T-score for the depressive problems subscale was used as an other-informant indicator of childhood depressive symptoms (α=. 85) and 2.2% of children met clinical threshold (M=52.0, SD=3.2).

Adolescent pregnancy.

During Wave 2 of data collection, late-adolescent participants were asked if they had ever been pregnant. If the participant endorsed a pregnancy, they were asked at what age they experienced the pregnancy (if multiple pregnancies were endorsed, the age for each pregnancy was recorded). Participants also reported on whether an adolescent pregnancy resulted in a birth and if the child was in their care. Adolescent pregnancy was then coded into a binary variable based on the Centers for Disease Control and Prevention (CDC, 2019) definition of adolescent pregnancy. Participants were coded as 1 (pregnancy before age 20) or a 0 (the individual never experienced a pregnancy before age 20). There were 71 participants (38.1%) who reported an adolescent pregnancy (M=1.8; range = 1-4). Ten participants reported a current pregnancy. Fifty-three individuals reported that an adolescent pregnancy resulted in birth (74.6%). Females who experienced adolescent pregnancy were, on average, 16.87-years-old (SD 1.56) when they became pregnant. Individuals who experienced an adolescent pregnancy did not significantly differ from individuals who did not experience an adolescent pregnancy on race/ethnicity, Wave 2 income, or whether or not they were in a relationship in the past six months.

Late-adolescent depressive symptoms (Wave 2)

Adult Self-Report (ASR; Achenbach, 1991).

The ASR is a self-report measure of behavioral symptoms for adult participants (18-35-years). The questionnaire asks participants to rate their symptoms over the last six months across 120-items on a 3-point scale (0 = not at all; 1 = sometimes; 2 = always). The ASR produces normed scales for adaptive functioning, psychological syndromes, internalizing problems, externalizing problems, and DSM-oriented scales. All scales are normed for gender and age based on nationally representative samples. The ASR has strong psychometric properties, including strong test-retest reliability (r = .80 or greater on all scales; Achenbach, 1991). The DSM-oriented subscale for depressive disorders was used as an indicator of a late-adolescent depressive symptoms latent factor (α=.82) and 24.2% of the sample met clinical threshold for depression (M=56.9, SD=7.3).

Beck Depression Inventory-II (BDI-II; Beck et al., 1996).

The BDI is a commonly used 21-item self-report measure of depressive symptoms. Participants were asked to respond to each item (e.g., “I feel lonely”) by choosing one of four statements (0 = never, 1 = sometimes, 2 = often, 3 = always) that best represented their symptomatology over the past two weeks. Clinical-range symptoms are indicated by sum scores of 20 or greater (Beck et al., 1996). In the current sample, 20.1% of the young adults reported sum scores in the clinical range. The BDI-II has good psychometric properties (Wang & Gorenstein, 2013), including test-retest reliability scores ranging from 0.73-0.96 and strong convergent validity with other depression measures (Wang & Gorenstein, 2013). A sum score of all 21 items was used as an indicator of a late-adolescent depression latent factor in the current study (α = .88) and 27% of the sample met clinical threshold for depression (M = 14.4, SD = 9.4).

Diagnostic Interview Schedule (DIS-IV: Robins et al., 1995).

The DIS-IV is a structured clinical interview designed to be administered by lay interviewers, via computer software, to provide clinical psychiatric diagnoses based on Diagnostic and Statistical Manual of Mental Disorders-IV criteria (DSM-IV; APA, 1994). Administrators received training from an established trainer who approved interviewers for validity and reliability. The total count of Major Depressive Disorder (MDD) symptoms in the past year was used as an indicator of a late-adolescent depression latent factor (α=.94) and 29.7% of the sample received a diagnosis.

Interpersonal Covariates (Wave 2)

Peer Relationships.

Participants reported on the quantity and quality, of current peer relationships, as well as amount of peer engagement, using the previously described Adult Self-Report (ASR; Achenbach, 1991). Specifically, this study used the T-score for the Adaptive Functioning: Friends scale, which is a composite of four items (α = .72) related to quantity of friends, time spent with friends, frequency of involvement with friends, and how well participants get along with their friends (rated from 0 = “Not as well as I’d Like” to 3 = “Far Above Average”).

Maternal Support.

Participants self-reported on maternal support using the Network of Relationships Inventory: Behavioral Systems Version (NRI-BSV; Furman & Buhrmester, 2009). The NRI-BSV is a 24-item questionnaire that assesses three features of social support within relationship systems in the participant’s life (i.e., attachment, caregiving, and affiliation). We queried the presence of maternal support by using the established relationship support factor (Furman & Buhrmester, 2009), which incorporates three different scales assessing how the individual uses that relationship for a secure base, safe haven, and companionship. Each scale is comprised of three items that determine the presence of support from the respondent’s mother figure. For example, participants reported on how often they used their mother for support (e.g., “how much do you seek out this person for comfort and support when you are troubled about something?”) via a five-point Likert scale (1 = “Little or None” to 5 = “The Most”). The NRI-BSV is a reliable and valid measure (Furman, 1996; Furman & Burhmester, 2009). Within this sample, reliability for each scale ranged from α = .83-.93, with an alpha coefficient of .84 for the composite factor.

Romantic Relationship Violence.

Participants reported on “whether they had experienced domestic violence in the past 12 months.” A dichotomous variable was created based on the presence or absence of relationship violence (28% of the current sample endorsed the presence of violence).

Data analytic plan

Descriptive data analyses were performed using SPSS 25 and structural equation models (SEMs) were conducted using Mplus Version 8.2 (Muthén & Muthén, 2017) with maximum likelihood estimator. Structural equation modeling was used to determine whether adolescent pregnancy moderated the effect of child maltreatment on late adolescent depressive symptoms. The SEM was specified as follows: child maltreatment, adolescent pregnancy, childhood depression, peer relationship quality, maternal support, and romantic relationship violence were entered as exogenous variables; late-adolescent depression was modeled as an endogenous latent factor with three proposed indicators (BDI sum score, ASR: DSM-oriented subscale for depression, and the DIS symptom count for MDD). Late adolescent depression was predicted by child maltreatment, adolescent pregnancy, the cross-product interaction term of child maltreatment by adolescent pregnancy, and the covariate predictors (childhood depression, peer relationship quality, maternal support, and romantic relationship violence) (see Figure 1). The effect of child maltreatment on depression was probed at each level of the adolescent pregnancy variable (i.e., adolescent pregnancy or no adolescent pregnancy; Cohen et al., 2003).

Figure 1.

Figure 1.

Open in a new tab

SEM model.

Note. The covariances among all predictor variables are modeled, though not depicted in the figure.

Child Maltreatment is coded (0 = nonmaltreated, 1 = maltreated). Adolescent pregnancy is coded (0 = no, 1 = yes). Child depx (child) = the sum score on the Child Depression Inventory. Child depx (counselor) = the TRF Depressive sx T-score. Relx Violence is coded as (0 = no romantic relationship violence, 1 = presence of any romantic relationship violence in last year). Maternal Support = The score on the NRI-BSV maternal support factor. Peer Relx = Adaptive Functioning: Friends T-score from the ASR. BDI = sum score on the Beck Depression Inventory. ASR:Depx = the T score for the depression subscale on the Adult Self Report. DIS.MDD = the number of Major Depressive Disorder symptoms endorsed on the Diagnostic Interview Schedule.

*p < .05, *** p < .01, *** p < .001.

Missing data for endogenous variables were estimated as a function of exogenous variables based on the missing data at random assumption (Schafer & Graham, 2002). Specifically, both the BDI-II and DIS-IV scales were missing data from one respondent, respectively (0.5% of the sample). Structural relationships were tested using maximum likelihood estimation. Model fit for the SEM was determined using the following criteria: Comparative fit index (CFI), root mean square error of approximation (RMSEA), and standardized root mean square residual (SRMR; Yu & Muthén, 2002). Acceptable model fit was determined by RMSEA values less than 0.06, CFI greater than 0.95, and SRMR less than 0.08 (Kline, 2011; Yu & Muthén, 2002).

Results

Preliminary analysis

Table 1 provides the zero-order correlations among study variables. Child maltreatment was significantly related to adolescent pregnancy as well as greater late-adolescent depressive symptoms on the BDI and ASR. Child maltreatment was not significantly associated with greater childhood depressive symptoms on the TRF or CDI, but it was significantly associated with lower adaptive peer functioning. Additionally, adolescent pregnancy was related to two indicators of late-adolescent depression, as well as greater romantic relationship violence and lower peer functioning and maternal support. However, child depressive symptoms (TRF and CDI) were not significantly associated with adolescent pregnancy. Finally, child depression, relationship violence, and lower peer functioning were all significantly related to late-adolescent depressive symptoms on the BDI and ASR.

Table 1.

Zero-order correlations among study variables.

1. 2. 3. 4. 5. 6. 7. 8. 9.
1. Maltreatment -
2. Adolescent Preg .20** -
3. BDI .17* .25** -
4. ASR:Depx .18* .23** .72*** -
5. DIS:MDD .01 .09 .29*** .33***
6. CDI. .07 .11 .24** .21** .10 -
7. TRF:Depx .09 −.07 .24** .08 .03 .05 -
8. Relx Violence .12 .20** .12 .29*** .14 .16* −.07 -
9. Peer Functioning −.23** −.20** −.26*** −.18* .05 −.29*** −.08 −.07 -
10. Maternal Support −.12 −.15* −.13 −.16* .04 −.15* −.06 −.06 .16
Open in a new tab

Note. Maltreatment is coded (0 = nonmaltreated, 1 = maltreated). Adolescent pregnancy is coded (0 = no, 1 = yes). BDI = sum score on the Beck Depression Inventory. ASR:Depx = the T score for the depression subscale on the Adult Self Report. DIS:MDD = the number of Major Depressive Disorder symptoms endorsed on the Diagnostic Interview Schedule. CDI = the sum score on the Child Depression Inventory. TRF: Depx = T score for depressive problems on TRF. Relx Violence is coded as (0 = no romantic relationship violence, 1 = presence of any romantic relationship violence in last year). Peer Functioning = T score for Adaptive Functioning: Friends via ASR. Maternal Support = Composite of maternal support scales via the NRI.

*

p < .05,

**

p < .01,

***

p < .001.

Structural equation models

The SEM model was specified as described above and this model fit the data well χ2 (16) = 12.94, p = .68, RMSEA=.00 (90% CI: .00-.06), CFI=1.00, SRMR=.04. An examination of the standardized path coefficients revealed that, with the inclusion of the interaction term, child maltreatment did not significantly predict greater late-adolescent depressive symptoms (β = −0.03, p > .05). Similarly, in the context of the interaction term, adolescent pregnancy did not significantly predict greater late-adolescent depressive symptoms (β = −0.01, p > .05). Greater self-reported depressive symptoms in childhood predicted greater depressive symptoms in late adolescence (β = 0.18 p < .05) but the effect for counselor-rated child depressive symptoms was not significant. Romantic relationship violence also predicted greater late-adolescent depressive symptoms (β = 0.22 p < .01), but peer relationship quality and maternal support did not. Finally, the interaction term (child maltreatment x adolescent pregnancy) significantly predicted greater late adolescent depressive symptoms (β = 0.28 p < .05). See Figure 1 for graphical representation of the results.2

Next, we probed the interaction of child maltreatment and adolescent pregnancy. Among those who did not experience an adolescent pregnancy, child maltreatment was not related to later depressive symptoms (β = −0.02). Next, we reverse scored the adolescent pregnancy variable, recalculated a new interaction term (reversed adolescent pregnancy x child maltreatment), and estimated the model again. Results indicated that among individuals who experienced an adolescent pregnancy, child maltreatment predicted significantly greater late adolescent depressive symptoms (β = 0.30 p = .05). Thus, results indicated that the effect of child maltreatment on late-adolescent depressive symptoms was significantly enhanced for individuals who experienced adolescent pregnancy. To visualize the interaction, we tested an ANCOVA using the depressive symptoms factor score as an outcome and child maltreatment and adolescent pregnancy as predictors (retaining all the model covariates) F(12, 173) = 5.20, p <.05 (See Figure 2). The factor score (M = 0.00; SD = 6.9) was generated from the original CFA model using the regression method in Mplus (also known as the maximum a posteriori method; Muthén & Muthén, 2017).

Figure 2.

Figure 2.

Open in a new tab

Effect of child maltreatment on late-adolescent depressive symptoms varies by adolescent pregnancy

Sensitivity analyses.

Additional post hoc analyses were conducted to examine whether the interactive effect of child maltreatment history (yes or no) and adolescent pregnancy on later depressive symptoms was dependent on the outcome of the adolescent pregnancy. Of the 71 participants who reported an adolescent pregnancy (38.1% of the sample), 53 of these pregnancies (74.6%) resulted in birth. We created a three-level adolescent pregnancy variable (0 = no adolescent pregnancy, 1 = adolescent pregnancy that did not result in birth, 2 = adolescent pregnancy resulting in birth) and then performed a 2x3 ANCOVA, using the depression factor score as an outcome, to determine if the interaction remained when considering pregnancy outcome (while continuing to control for model covariates). Results of pairwise multiple comparisons (F (14, 171) = 6.80, p < .01) revealed that among those without a history of maltreatment, there were no mean differences on late adolescent depressive symptoms (factor score) between the three levels of adolescent pregnancy. However, among those with a history of childhood maltreatment, individuals with adolescent pregnancies (both birth and no birth outcomes) demonstrated significantly greater depressive symptoms than maltreated individuals who did not experience an adolescent pregnancy (LSD, p < .01). There were no significant depression differences between the two categories of adolescent pregnancy (birth or no birth) among those with maltreatment histories.

Discussion

The present study used a prospective design to investigate the interactive impact of both a documented history of child maltreatment and the experience of adolescent pregnancy on late-adolescent depression. In addition to examining the effect of each predictor alone, we investigated whether adolescent pregnancy enhanced the risk of child maltreatment history on the development of late-adolescent depression. Results indicated that the effect of child maltreatment on late-adolescent depressive symptoms may become significantly more robust for those who experienced an adolescent pregnancy. This enhancing effect remained even after controlling for prior depressive symptoms, peer and maternal relationship quality, and romantic relationship violence and considering the context of economic disadvantage, which have all been identified as relevant risk factors for depression (Boden et al., 2008; Mollborn et al., 2009).

Child maltreatment and depressive symptoms.

Notably, at the bivariate level, child maltreatment was not significantly associated with greater depressive symptoms in childhood but was significantly associated with greater depressive symptoms in later adolescence. It may be that maltreatment initiates a cascade of negative developmental consequences that begin in childhood and eventually result in depressive symptoms in late adolescence (Cicchetti & Toth, 2016).

Synergistic effects of child maltreatment and adolescent pregnancy

Although previous studies have demonstrated effects of both adolescent pregnancy and child maltreatment on later depression (Hodgkinson et al., 2014; Li et al., 2016), these studies focused on either child maltreatment or adolescent pregnancy, not both, and none have drilled down to examine whether adolescent pregnancy enhances the vulnerability associated with documented history of child maltreatment in the development of later depression. The lack of main effects for either variable in our model is qualified by the presence of the robust interaction effect (i.e., they should not be interpreted in isolation). This finding underscores the need to go beyond examining independent predictors of depression by exploring the synergistic effects of multiple predictors and the way in which one risk may enhance the risk of another vulnerability factor to further clarify those most susceptible to the development of depression.

Our findings suggest that the effect of child maltreatment on late-adolescent depression is significantly enhanced for those who become pregnant in adolescence. This finding accentuates the need to elucidate for whom the depressogenic vulnerabilities of child maltreatment may be activated or amplified (Cicchetti & Toth, 2016; Widom, 2020). Indeed, although increased depressive symptomatology is often a resultant consequence of child maltreatment (Li et al., 2016), this association is not an ineluctable outcome. The findings presented herein indicate that the depressogenic vulnerabilities of child maltreatment may be activated or amplified by the experience of adolescent pregnancy.

However, child maltreatment was associated with greater late-adolescent depressive symptoms at the bivariate level; thus, the effect of maltreatment on subsequent depressive symptoms is not wholly dependent on adolescent pregnancy. Further, due to the high-risk nature of the sample, all participants, even those who did not experience maltreatment or adolescent pregnancy, are living under conditions of stress and economic disadvantage, which can potentiate depression (Inaba et al., 2005), likely contributing to the high rates of depression within this sample. As such, group differences on isolated risk factors (e.g., maltreatment) may be masked by the sample-wide prevalence of adversity and depression without examining amplifying factors, as done in this study.

Similarly, the finding that late adolescent depressive symptoms were enhanced for individuals who experienced both child maltreatment and adolescent pregnancy is consistent with previous studies suggesting that adolescent pregnancy, in isolation, does not represent a heightened period of risk for depression (Hipwell et al., 2016; Mollbom & Morningstar, 2009). For example, some studies have contended that adolescent pregnancy is merely an indicator of a depressogenic trajectory that originated with preexisting depression and/or socioeconomic disadvantage (Mollborn & Morningstar, 2009). Notably, there was not a significant bivariate association between childhood depression and adolescent pregnancy, indicating that adolescents who become pregnant were not more likely to have been depressed prior to the pregnancy. Further, the current study was able to account for both prior depression and economic disadvantage and demonstrate that, over and above those factors, it may be the combined context of child maltreatment and adolescent pregnancy that increases subsequent depression risk.

Moreover, this interaction effect remained even when accounting for the several well-documented depressogenic risk factors, such as the quality of peer and maternal relationships and romantic relationship violence. Demonstrating the unique combinative effects of child maltreatment and adolescent pregnancy within the context of these other depressogenic risks is critical to advancing this line of research. There is evidence for a robust association between one’s interpersonal milieu and depression (Cicchetti & Toth, 2009) and this association is salient for individuals who experience adolescent pregnancy, as the quality of their social support networks can attenuate depression risk (Easterbrooks et al., 2016).

That adolescent pregnancy significantly amplified the effect of child maltreatment on late-adolescent depression is consistent with a substantial body of research linking increased depression to maltreatment histories within groups of pregnant adolescents (Milan et al., 2004; Meltzer-Brody et al., 2013; Romano et al., 2006). However, this study advances those findings by testing the interaction in a sample that included comparison groups for each variable of interest: child maltreatment (maltreated and nonmaltreated) and adolescent pregnancy (individuals who experienced adolescent pregnancy and those who did not).

As demonstrated by the sensitivity analysis conducted, our findings suggest that individuals who experience adolescent pregnancy are more vulnerable to subsequent depression if they have maltreatment histories, and this effect is not dependent on whether or not the pregnancy is carried to term. This finding is not surprising, given that adolescent parenthood, miscarriage, and abortion are associated with depression (though in the case of abortion, the association with depression is often related to shared risk, not the abortion itself; Biggs et al., 2017). Instead, it appears the specific context of pregnancy may allow for the deleterious depressogenic effects of child maltreatment to amplify.

It has been hypothesized that pregnancy represents a period of enhanced susceptibility to psychological material, historical and/or current (Busch & Lieberman, 2010). Thus, it is possible that adolescent pregnancy, regardless of the outcome, creates a window of vulnerability for past traumas (i.e., child maltreatment) to resurface, which may then induce depressive symptoms or exacerbate preexisting symptoms (Busch & Lieberman, 2010). Even if the pregnancy is brief and does not result in a birth, the temporary psychological transformation may evoke or amplify old wounds or betrayals, allowing the insidious effects of child maltreatment to gain enhanced salience. However, in the absence of maltreatment histories, the transformative experiences of pregnancy, even when occurring in adolescence, may remain uncontaminated by maltreatment memories and protected against increased depression (Narayan et al., 2016).

Additionally, partner and maternal supports represent two powerful protective factors that are known to buffer subsequent depression following adolescent pregnancies (Easterbrooks et al., 2016). However, for pregnant adolescents with maltreatment histories, those relationships are more likely to be conflictual (Handley et al., 2019). Among the individuals who experienced both child maltreatment and adolescent pregnancy in the current study, the mother was the indicated perpetrator of the participant’s maltreatment experience in 90% of the cases. Therefore, child maltreatment may enhance the risk for subsequent depression in pregnant adolescents because maltreated individuals may be uniquely thwarted in their attempts to access positive versions of the most effective depression-buffering relationships for pregnant adolescents (namely, parents and partners). Additionally, for maltreated-pregnant adolescents (compared to nonmaltreated pregnant adolescents), the pregnancy may increase reliance on maltreating caregivers and intensify interpersonal conflicts that are known to increase depression (Mufson, 2004).

Child maltreatment and adolescent pregnancy

Adolescent pregnancy is not a random event, and its association with child maltreatment is well documented in meta-analytic studies (Noll et al., 2009), prospective longitudinal studies (Noll et al., 2019), and population-based studies (Putnam-Hornstein et al., 2013). As with previous studies, we found a significant bivariate association between child maltreatment and adolescent pregnancy. The conceptual goal of this study was to examine the synergistic effects of child maltreatment and adolescent pregnancy on depressive symptoms, rather than the sequencing of their effects. Nonetheless, an alternative or competing model may exist that demonstrates a mediational chain, whereby child maltreatment causes adolescent pregnancy, which then causes greater depressive symptoms.

Implications

Depression is a debilitating and potentially deadly condition and, as previously noted, adolescent females represent an especially vulnerable group with disproportionate risk compared to other demographics (NIMH, 2019). This study has identified a particular combination of depressogenic risk factors (i.e., child maltreatment and adolescent pregnancy) that interact to explain some of the striking level of risk for adolescent females. Fortunately, adolescence is an optimum period to identify this vulnerable subgroup of females and deliver targeted prevention.

Screening.

In the recent momentum to prevent depression, there is a tendency to conglomerate depressogenic stressors into a single categorical definition of depression risk (i.e., risk vs. no risk; Kieling et al., 2019; U.S. Preventive Task Force, 2018). That is, if an individual presents with one or more factors determined to be a risk for depression, they are fitted to the “risk” group and referred for preventive intervention. While this type of screening widens the proverbial net and minimizes screening failures, the results reported herein suggest that if we refrain from recognizing the diverse interactive effects of coexisting stressors (i.e., child maltreatment and adolescent pregnancy), we lose critical information required for designing targeted and tailored prevention, and the precision of our interventions will be compromised. For example, imagine a clinician checks for the presence of adolescent pregnancy or child maltreatment history, determines the individual meets criteria for categorical “depression risk” because of the presence of one or both risk factor, and then refers for universal preventive intervention. The practitioner may understandably feel satisfied with their preventive action, and no doubt, this is preferable to no screening at all. However, universal preventive approaches unrelated to a priori risk have been ineffective in preventing depression (Kieling et al., 2019). Rather, it is equally important that the practitioner refine their assessment to consider how the co-occurrence of child maltreatment and depression may represent an especially vulnerable set of risks for young women. In doing so, they can deliver selected prevention (i.e., prevention that targets subgroups at increased risk for depression and aims to attenuate a profile of risk with tailored techniques), which demonstrates greater promise in preventing depression (Kieling et al., 2019). Therefore, when screening for depression risk in adolescent females, practitioners should move beyond “checkbox” screening for stressors such as child maltreatment and adolescent pregnancy and consider how the two may distinctively interact and contribute to amplified risk (see Chamberlain et al., 2019 for a review of screening tools for child maltreatment history).

Secondary and tertiary prevention

Secondary and tertiary depression prevention designed to address the stress of child maltreatment or adolescent pregnancy, independently, may not suffice for pregnant adolescents with maltreatment histories. Thus, the results of this study suggest practitioners should be cognizant of the unique issues presenting for adolescents with histories of maltreatment and pregnancy and deliver trauma-specific support to concurrently address both stressors and mitigate subsequent depression

Insight-oriented approaches.

Insight-oriented approaches that link current emotional states to unresolved childhood traumas and unmet attachment needs (Narayan et al., 2016) represent one promising example to dually treat depression in the context of child maltreatment and adolescent pregnancy. As previously noted, pregnancy represents a transformative period (i.e., emotional and biological changes; intergenerational linking) of change during which a woman references her past caregiving experiences (Narayan et al., 2016), creating an environment wherein the insidious aspects of child maltreatment can thrive. Using interventions such as Perinatal Child-Parent Psychotherapy (P-CPP; Lieberman et al., 2005; Narayan et al. 2016), practitioners can create emotional holding environments that allow clients to acknowledge, process, and reconcile their unconscious representations of previous caregiving betrayals (“ghosts in the nursery”) and augment those memories with recollections of benevolent childhood experiences (“angels in the nursery”), which have been shown to reduce depressive symptoms (Lavi et al., 2015; Narayan et al., 2017) and facilitate healing.

Interpersonal approaches.

Additionally, empirically-informed, family-based approaches, such as Attachment-Based Family Therapy (ABFT; Diamond & Siqueland, 1995), may be particularly effective for individuals with both a maltreatment history and an adolescent pregnancy. AFBT is an integrative, process-oriented, trauma-focused approach that aims to mitigate depressive symptoms in adolescents by mending disrupted attachment bonds, reducing parent-adolescent conflict and parental criticism, and increasing trust and communication within the family system (Diamond et al., 2002). AFBT also has documented efficacy in treating adolescent depression within low-income samples similar to the current study (Diamond et al., 2016). Rather than helping clients gain insight about the impact of previous caregiving failures, this approach aims to directly repair attachment bonds and change current interpersonal patterns (Diamond et al., 2016).

Interpersonal Psychotherapy (IPT Weissman et al., 2000), and its adaptation for depressed adolescents (IPT-A; Mufson, 2004) is another efficacious interpersonal intervention that may address the unique interpersonal challenges operating for adolescents who experience both child maltreatment and adolescent pregnancy. For example, IPT-A (Mufson, 2004) asserts that interpersonal conflicts, problems, and deficits contribute to the onset and continuation of depressive symptoms and individuals who experience both child maltreatment and adolescent pregnancy may experience greater social isolation and interpersonal conflict (Madigan et al., 2014). IPT has been found to be efficacious in treating depression within high-risk samples of economically disadvantaged, depressed mothers (Toth et al., 2013), as well as reducing symptoms of pregnancy-related depression (Stuart, 2012).

Intergenerational implications

The levels of depression among the individuals who experienced both adolescent pregnancy and child maltreatment in this study are troubling given the potential for intergenerational consequences. Perinatal and postnatal depression can have deleterious physical and psychological consequences for the developing offspring (Glover & Capron, 2017). Moreover, adolescent parents with maltreatment histories are more likely to have children who experience maltreatment (Dixon et al., 2005) and are less likely to respond effectively to maltreatment prevention programs when their depression remains unaddressed (Easterbrooks et al., 2013). Further, Easterbrooks et al (2011) found that depression can act as a “cost” for resilient parenting in young mothers, suggesting that efforts to mitigate risk for the child may levy psychopathology risk for the mother. Thus, the offspring of individuals in the high-risk group identified by this study (i.e., pregnant adolescents with maltreatment histories) are at considerable risk for negative outcomes. And, because adolescent mothers are less likely to seek care for depression (Mollborn, 2017), it is critical that when practitioners do refer for preventive interventions, these vulnerable dyads receive effective, tailored care that addresses their trauma histories in the context of the dyad (Narayan et al., 2016).

Strengths and limitations

These study findings represent important advances in understanding the unique vulnerability of young women with maltreatment histories who become pregnant as adolescents. Strengths include the prospective, longitudinal design that controls for potential confounding of depression continuity across development; the rigorously obtained, documented records of child maltreatment; the involvement of maltreated/nonmaltreated and pregnant/non-pregnant comparison groups; the use of a latent construct of depression, informed by multiple empirically-sound depression measurements; and the use of an economically-disadvantaged sample to methodologically control for the confounding impact of poverty on depression. In spite of these strengths and contributions, there are limitations worth considering.

First, recruiting and retaining high-risk samples of individuals is very difficult (Cicchetti & Toth, 2016), especially after a gap of ten years. The degree of attrition in this study may limit generalizability of the findings. Second, the single-item measurement of relationship violence is limited, and it may underestimate the amount of relationship violence and dampen effects. Next, although we partially interpret our findings within attachment theory and mental representations, we did not directly assess these constructs. Future research utilizing measures of caregiving representations among pregnant adolescents with maltreatment histories (see Narayan et al., 2016) will provide important insights.

Finally, adolescent pregnancy is complex, and we were unable to assess several related factors that may also influence depressive symptoms, such as pregnancy desires, intentions, cultural beliefs and expectations, partner involvement, victimization, family approval, or the outcomes of pregnancies not resulting in birth. Similarly, adolescent pregnancy is associated with higher rates of pre-term births and low-birthweight infants, and this association is even stronger in adolescent mothers who also have maltreatment histories (Cederbaum et al., 2013).

For some in our study who became mothers and experienced these outcomes, it is possible that the potential stressors associated with caring for vulnerable infants may influence or exacerbate maternal depressive symptoms via transactional or bidirectional processes (Miles et al. 2007). Thus, future studies may consider pregnancy intentions, wantedness, and outcome of terminated pregnancies when further clarifying the interactive effects of child maltreatment and adolescent pregnancy on depression.

Conclusion

Despite great effort and significant expenditure to alleviate the burden of depression, rates for young women continue to climb to alarmingly high rates (NIMH, 2019). Given this trend, a compelling argument can be made that investing in targeted prevention programs that address individuals with the highest risk would have the highest yield. Sharpening the precision of depression prevention requires the identification of high-risk groups and this study has advanced the understanding of one such vulnerable group in the form of adolescent females who experience child maltreatment and adolescent pregnancy. The results of this study provide preliminary guidance to practitioners regarding precision depression screening and tailored preventive interventions to ameliorate the abhorrent rates of depression plaguing young women.

Acknowledgments:

This research was supported by grants received from the National Institute on Drug Abuse (R01DA17741), National Institute of Child Health and Human Development (P50-HD096698), and the Spunk Fund, Inc.

Footnotes

Publisher's Disclaimer: This Author Accepted Manuscript is a PDF file of a an unedited peer-reviewed manuscript that has been accepted for publication but has not been copyedited or corrected. The official version of record that is published in the journal is kept up to date and so may therefore differ from this version.

1

It should be noted that these findings are culturally dependent and apply primarily to a Western-culture viewpoint where adolescent pregnancy is non-normative.

2

Variables representing Wave 2 income, racial/ethnic minority status, age, and currently pregnant were considered as potential covariates. None were significant predictors of the outcome variable and participants did not differ on these variables based on maltreatment status or adolescent pregnancy status. Their inclusion in the model did not change the pattern of results and were therefore trimmed based on the parsimony principle.

References

  1. Achenbach TM (1991). Manual for the Teacher’s Report Form and 1991 profile. Vermont: University of Vermont Department Psychiatry. [Google Scholar]
  2. Alvarez-Segura M, Garcia-Esteve L, Torres A, Plaza A, Imaz ML, Hermida-Barros L, … & Burtchen N (2014). Are women with a history of abuse more vulnerable to perinatal depressive symptoms? A systematic review. Archives of Women’s Mental Health, 77(5), 343–357. [DOI] [PubMed] [Google Scholar]
  3. American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disorders (4th Edition-Revised). Washington, DC: American Psychiatric Association. [Google Scholar]
  4. Baldwin JR, Reuben A, Newbury JB, & Danese A (2019). Agreement between prospective and retrospective measures of childhood maltreatment: A systematic review and meta-analysis. JAMA Psychiatry, 76(6), 584–593. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Beck AT, Steer RA, & Brown GK (1996). Manual for Beck Depression Inventory—II. San Antonio, TX: Psychological Corporation. [Google Scholar]
  6. Biggs MA, Upadhyay UD, McCulloch CE, & Foster DG (2017). Women’s mentalhealth and well-being 5 years after receiving or being denied an abortion: A prospective, longitudinal cohort study. JAMA Psychiatry, 74(2), 169–178. [DOI] [PubMed] [Google Scholar]
  7. Boden J, Fergusson D, & Horwood J (2008). Early motherhood and subsequent life outcomes. Journal of Child Psychology & Psychiatry & Allied Disciplines, 49: 151–160. [DOI] [PubMed] [Google Scholar]
  8. Cederbaum JA, Putnam-Hornstein E, King B, Gilbert K, & Needell B (2013). Infant birth weight and maltreatment of adolescent mothers. American Journal of Preventive Medicine, 45(2), 197–201. [DOI] [PubMed] [Google Scholar]
  9. Chamberlain C, Gee G, Harfield S, Campbell S, Brennan S, Clark Y, … & Healing the Past by Nurturing the Future group. (2019). Parenting after a history of childhood maltreatment: A scoping review and map of evidence in the perinatal period. PLoS One, 14(3), e0213460. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Chung EK, Mathew L, Elo IT, Coyne JC, & Culhane JF (2008). Depressive symptoms in disadvantaged women receiving prenatal care: the influence of adverse and positive childhood experiences. Ambulatory Pediatrics, 8(2), 109–116. [DOI] [PubMed] [Google Scholar]
  11. Cicchetti D, & Manly JT (1990). A personal perspective on conducting research with maltreating families: Problems and solutions In Brody G & Sigel I (Eds.), Methods of family research: Families at risk (Vol. 2, pp. 87–133). Hillsdale, NJ:Erlbaum. [Google Scholar]
  12. Cicchetti D, & Toth SL (2009). A developmental psychopathology perspective on adolescent depression In Nolen-Hoeksema S, & Hilt L (Eds.), Handbook of Adolescent Depression (pp. 3–31). New York, NY: Taylor & Francis. [Google Scholar]
  13. Cicchetti D, & Toth SL (2016). Child maltreatment and developmental psychopathology: A multilevel perspective. Developmental Psychopathology, 1–56. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Cicchetti D, Toth SL, & Manly JT (2003). Maternal maltreatment classification interview. Unpublished manuscript, Mt. Hope Family Center, Rochester, NY. [Google Scholar]
  15. Cohen P, West SG, & Aiken LS (2003). Applied multiple regression correlation analysis for the behavioral sciences. Psychology Press. [Google Scholar]
  16. Diamond GUY, & Siqueland L (1995). Family therapy for the treatment of depressed adolescents. Psychotherapy: Theory, Research, Practice, Training, 52(1), 77. [Google Scholar]
  17. Diamond GS, Reis BF, Diamond GM, Siqueland L, & Isaacs L (2002). Attachment based family therapy for depressed adolescents: A treatment development study. Journal of the American Academy of Child & Adolescent Psychiatry, 41(10), 1190–1196. [DOI] [PubMed] [Google Scholar]
  18. Diamond G, Russon J, & Levy S (2016). Attachment-based family therapy: A review of the empirical support. Family Process, 55(3), 595–610. [DOI] [PubMed] [Google Scholar]
  19. Dixon L, Browne K, & Hamilton-Giachritsis C (2005). Risk factors of parents abused as children: a mediational analysis of the intergenerational continuity of child maltreatment (Part I). Journal of Child Psychology and Psychiatry, 46{l), 47–57. [DOI] [PubMed] [Google Scholar]
  20. Easterbrooks MA, Chaudhuri JH, Bartlett JD, & Copeman A (2011). Resilience in parenting among young mothers: Family and ecological risks and opportunities. Children and Youth Services Review, 33(1), 42–50. [Google Scholar]
  21. Easterbrooks MA, Bartlett JD, Raskin M, Goldberg J, Contreras MM, Kotake C, … & Jacobs FH. (2013). Limiting home visiting effects: maternal depression as a moderator of child maltreatment. Pediatrics, 732(Supplement 2), S126–S133. [DOI] [PubMed] [Google Scholar]
  22. Easterbrooks M, Kotake C, Raskin M, & Bumgarner E (2016). Patterns of depression among adolescent mothers: Resilience related to father support and home visiting program. American Journal of Orthopsychiatry, 86{ 1), 61. [DOI] [PubMed] [Google Scholar]
  23. Estrin GL, Ryan EG, Trevillion K, Demilew J, Bick D, Pickles A, & Howard LM (2019). Young pregnant women and risk for mental disorders: findings from an early pregnancy cohort. BJPsych Open, 5(2). [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Fergusson DM, Horwood L, & Ridder EM (2006). Abortion in young women and subsequent mental health. Journal of Child Psychology and Psychiatry, 47(1), 16–24. [DOI] [PubMed] [Google Scholar]
  25. Furman W, & Buhrmester D (2009). Methods and measures: The network of relationships inventory: Behavioral systems version. International Journal of behavioral Development, 33(5), 470–478. [DOI] [PMC free article] [PubMed] [Google Scholar]
  26. Glover V, & Capron L (2017). Prenatal parenting. Current Opinion in Psychology, 15, 66–70. [DOI] [PubMed] [Google Scholar]
  27. Handley ED, Russotti J, Rogosch FA, & Cicchetti D (2019). Developmental cascades from child maltreatment to negative friend and romantic interactions in emerging adulthood. Development and Psychopathology, 37(5), 1649–1659. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Hasin DS, Sarvet AL, Meyers JL, Saha TD, Ruan WJ, Stohl M, & Grant BF (2018). Epidemiology of adult DSM-5 major depressive disorder and its specifiers in the United States. JAMA Psychiatry, 75(4), 336–346. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Hillis SD, Anda RF, Dube SR, Felitti VJ, Marchbanks PA, & Marks JS (2004). The association between adverse childhood experiences and adolescent pregnancy, long term psychosocial consequences, and fetal death. Pediatrics, 773(2), 320–327. [DOI] [PubMed] [Google Scholar]
  30. Hipwell AE, Murray J, Xiong S, Stepp SD, & Keenan KE (2016). Effects of adolescent childbearing on maternal depression and problem behaviors: A prospective, population-based study using risk-set propensity scores. PloS One, 77(5), e0155641. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Hodgkinson S, Beers L, Southammakosane C, & Lewin A (2014). Addressing the mental health needs of pregnant and parenting adolescents. Pediatrics, 733(1), 114–122. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Humphreys KL, LeMoult J, Wear JG, Piersiak HA, Lee A, & Gotlib IH (2020). Child maltreatment and depression: A meta-analysis of studies using the Childhood Trauma Questionnaire. Child Abuse & Neglect, 102 10.1016/j.chiabu.2020.104361 [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Inaba A, Thoits PA, Ueno K, Gove WR, Evenson RJ, & Sloan M (2005). Depression in the United States and Japan: Gender, marital status, and SES patterns. Social Science & Medicine, 67(11), 2280–2292. [DOI] [PubMed] [Google Scholar]
  34. Kieling C, Adewuya A, Fisher HL, Karmacharya R, Kohrt BA, Swartz JR, &Mondelli V (2019). Identifying depression early in adolescence. The Lancet Child & Adolescent Health, 3(4), 211–213. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Kline RB (2011). Convergence of structural equation modeling and multilevel modeling In Williams M & Vogt WP (Eds.), Handbook of methodological innovation (pp. 562–589). London, England: Sage. [Google Scholar]
  36. Kovacs M (2004). Children’s depression inventory (CDP). North Tonawanda, NY: Multi Health Systems [Google Scholar]
  37. Kwong AS, Lopez-Lopez JA, Hammerton G, Manley D, Timpson NJ, Leckie G, & Pearson RM (2019). Genetic and environmental risk factors associated with trajectories of depression symptoms from adolescence to young adulthood. JAMA Network Open, 2(6), el96587–el96587. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Lavi T, Gard AM, Hagan M, Van Horn P, & Lieberman AF (2015). Child-parent psychotherapy examined in a perinatal sample: Depression, posttraumatic stress symptoms and child-rearing attitudes. Journal of Social and Clinical Psychology, 34{ 1), 64–82. [Google Scholar]
  39. Li M, D’arcy C, & Meng X (2016). Maltreatment in childhood substantially increases the risk of adult depression and anxiety in prospective cohort studies: Systematic review, meta-analysis, and proportional attributable fractions. Psychological Medicine, 46(4), 717–730. [DOI] [PubMed] [Google Scholar]
  40. Lippard ET, & Nemeroff CB (2020). The devastating clinical consequences of child abuse and neglect: increased disease vulnerability and poor treatment response in mood disorders. American Journal of Psychiatry, 777(1), 20–36. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Madigan S, Wade M, Plamondon A, Vaillancourt K, Jenkins JM, Shouldice M, &Benoit D (2014). Course of depression and anxiety symptoms during the transition to parenthood for female adolescents with histories of victimization. Child Abuse &Neglect, 38(1), 1160–1170. [DOI] [PubMed] [Google Scholar]
  42. Malhi G, & Mann J (2018). Depression. Lancet, 392 (10161), 2299–2312. [DOI] [PubMed] [Google Scholar]
  43. Manly JT (2005). Advances in research definitions of child maltreatment. Child Abuse & Neglect, 29, 425–439. [DOI] [PubMed] [Google Scholar]
  44. McDonnell CG, & Valentino K (2016). Intergenerational effects of childhood trauma: Evaluating pathways among maternal ACEs, perinatal depressive symptoms, and infant outcomes. Child Maltreatment, 21(4), 317–326. [DOI] [PubMed] [Google Scholar]
  45. McLaughlin KA, Conron KJ, Koenen KC, & Gilman SE (2010). Childhood adversity, adult stressful life events, and risk of past-year psychiatric disorder: A test of the stress sensitization hypothesis in a population-based sample of adults. Psychological Medicine, 40(10), 1647–1658. [DOI] [PMC free article] [PubMed] [Google Scholar]
  46. McMahon C, Trapolini T, & Barnett B (2008). Maternal state of mind regarding attachment predicts persistence of postnatal depression in the preschool years. Journal of Affective Disorders, 707(1–3), 199–203. [DOI] [PubMed] [Google Scholar]
  47. Meltzer-Brody S, Bledsoe-Mansori SE, Johnson N, Killian C, Hamer RM, Jackson C, … & Thorp J (2013). A prospective study of perinatal depression and trauma history in pregnant minority adolescents. American Journal of Obstetrics and Gynecology, 208(3), 211. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Milan S, Ickovics JR, Kershaw T, Lewis J, Meade C, & Ethier K (2004). Prevalence, course, and predictors of emotional distress in pregnant and parenting adolescents. Journal of Consulting and Clinical Psychology, 72(2), 328. [DOI] [PubMed] [Google Scholar]
  49. Milan S, Kershaw TS, Lewis J, Westdahl C, Rising SS, Patrikios M, & Ickovics JR (2007). Caregiving history and prenatal depressive symptoms in low-income adolescent and young adult women: Moderating and mediating effects. Psychology of Women Quarterly, 31(3), 241–251. [Google Scholar]
  50. Miles MS, Holditch-Davis D, Schwartz TA, & Scher M (2007). Depressive symptoms in mothers of prematurely born infants. Journal of Developmental & Behavioral Pediatrics, 25(1), 36–44. [DOI] [PubMed] [Google Scholar]
  51. Mollborn S (2017). Teenage mothers today: what we know and how it matters. Child Development Perspectives, 77(1), 63–69. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Mollborn S, & Morningstar E (2009). Investigating the relationship between teenage childbearing and psychological distress using longitudinal evidence. Journal of Health and Social Behavior, 50(3), 310–326. [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Muthén LK and Muthén BO (1998-2017). Mplus User’s Guide (8th Edition). Los Angeles, CA: Muthén & Muthén. [Google Scholar]
  54. Mufson L (Ed.). (2004). Interpersonal psychotherapy for depressed adolescents. Guilford Press. [Google Scholar]
  55. Narayan AI, Bucio GO, Rivera LM, & Lieberman AF (2016). Making sense of the past creates space for the baby: Perinatal child-parent psychotherapy for pregnant women with childhood trauma. Zero to Three, 36(5), 22–28. [Google Scholar]
  56. Narayan AJ, Ippen CG, Harris WW, & Lieberman AF (2017). Assessing angels in the nursery: A pilot study of childhood memories of benevolent caregiving as protective influences. Infant Mental Health Journal, 35(4), 461–474. [DOI] [PubMed] [Google Scholar]
  57. National Institute of Mental Health (NIMH). (2019). Major depression. Retrieved from: https://www.nimh.nih.gov/health/statistics/major-depression.shtml
  58. Noll JG, Shenk CE, & Putnam KT (2009). Childhood sexual abuse and adolescent pregnancy: A meta-analytic update. Journal of Pediatric Psychology, 34(4), 366–378. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Noll JG, Guastaferro K, Beal SJ, Schreier ΗM, Barnes J, Reader JM, & Font SA (2019). Is sexual abuse a unique predictor of sexual risk behaviors, pregnancy, and motherhood in adolescence? Journal of Research on Adolescence, 29(4), 967–983. [DOI] [PMC free article] [PubMed] [Google Scholar]
  60. Putnam-Hornstein E, Cederbaum JA, King B, Cleveland J, & Needed B (2013). A population-based examination of maltreatment history among adolescent mothers in California. Journal of Adolescent Health, 53(6), 794–797. [DOI] [PubMed] [Google Scholar]
  61. Robins L, Cottier L, Bucholz K, & Compton W (1995). Diagnostic interview schedide for dsm-iv (dis-iv). Washington University School of Medicine: St Louis, MO. [Google Scholar]
  62. Romano E, Zoccolillo M, & Paquette D (2006). Histories of child maltreatment and psychiatric disorder in pregnant adolescents. Journal of the American Academy of Child & Adolescent Psychiatry, 45(3), 329–336. [DOI] [PubMed] [Google Scholar]
  63. Schafer JL, & Graham JW (2002). Missing data: Our view of the state of the art. Psychological Methods, 7(2), 147. [PubMed] [Google Scholar]
  64. Schmidt RM, Wiemann CM, Rickert VI, & Smith EB (2006). Moderate to severe depressive symptoms among adolescent mothers followed four years postpartum. Journal of Adolescent Health, 38(6), 712–718. [DOI] [PubMed] [Google Scholar]
  65. Sedlak AJ, Mettenburg J, Basena M, Petta L, McPherson K, & Greene A (2010). Fourth national incidence study of child abuse and neglect (NIS-4): Report to congress. Washington, DC: US Department of Health and Human Services, Administration for Children, Youth and Families. [Google Scholar]
  66. Stuart S (2012). Interpersonal psychotherapy for postpartum depression. Clinical Psychology & Psychotherapy, 19(2), 134–140. [DOI] [PMC free article] [PubMed] [Google Scholar]
  67. Toth SL, Rogosch FA, Oshri A, Gravener-Davis J, Sturm R, & Morgan-Lopez AA (2013). The efficacy of interpersonal psychotherapy for depression among economically disadvantaged mothers. Development and Psychopathology, 25(4ptl), 1065–1078. [DOI] [PMC free article] [PubMed] [Google Scholar]
  68. Trickett PK, Negriff S, Ji J, & Peckins M (2011). Child maltreatment and adolescent development. Journal of Research on Adolescence, 21(1), 3–20. [Google Scholar]
  69. Tung T, Keenan K, Stepp SD, & Hipwell AE (2019). The moderating effects of traumatic stress on vulnerability to emotional distress during pregnancy. Development and Psychopathology, 1–14. [DOI] [PMC free article] [PubMed] [Google Scholar]
  70. Tzilos GK, Zlotnick C, Raker C, Kuo C, & Phipps MG (2012). Psychosocial factors associated with depression severity in pregnant adolescents. Archives of Women’s Mental Health, 15(5), 397–401. [DOI] [PMC free article] [PubMed] [Google Scholar]
  71. U.S. Preventive Services Task Force. (2018). Perinatal depression: Preventive interventions. Retrieved from: https://www.uspreventiveservicestaskforce.org/Page/Document/draftrecommendation-statement/perinatal-depression-preventive-interventions [Google Scholar]
  72. Yu CY (2002). Evaluating cutoff criteria of model fit indices for latent variable models with binary and continuous outcomes (Vol. 30). Los Angeles: University of California [Google Scholar]
  73. Wang YP, & Gorenstein C (2013). Psychometric properties of the Beck Depression Inventory—II: A comprehensive review. Revista Brasileira de Psiquiatria, 35, 416 431. [DOI] [PubMed] [Google Scholar]
  74. Widom CS (2020). Commentary: A challenge for a higher bar in research on childhood trauma: Reflections on Danese 2019. Journal of Child Psychology and Psychiatry. [DOI] [PMC free article] [PubMed] [Google Scholar]
  75. World Health Organization (WHO). (2019). Depression. Retrieved from: https://www.who.int/news-room/detail/30-03-2017--depression-let-s-talk-savs-who-as-depression-tops-list-of-causes-of-ill-health

RESOURCES