Childhood Abuse and Neglect and Self-Reported Symptoms of Psychopathology through Midlife

Abstract

Although abuse and neglect in the early years of life have been reliably linked to poor mental health outcomes in childhood, only a few studies have examined whether the predictive significance of childhood abuse and neglect endures for symptoms of psychopathology into adulthood. Here we examined to what extent prospectively assessed child abuse and neglect is associated with self-reported symptoms of psychopathology measured from ages 23 through 39 years, controlling for early demographic covariates and self-reported symptoms of psychopathology measured at age 16 years. The sample included 140 participants from the Minnesota Longitudinal Study of Risk and Adaptation. Participants were 49% female and 69% White/non-Hispanic. At the time of their child’s birth, 48% of the mothers were teenagers (M = 20.5 years, SD = 3.74), 65% were single, and 42% had completed less than a high school education. Results indicated that childhood abuse and neglect was robustly associated with symptoms of psychopathology in adulthood. Exploratory analyses focusing on specific parametrizations of abuse/neglect suggested that abuse perpetrated by maternal figures (rather than paternal or non-parental figures) was uniquely associated with high levels of self-reported symptoms of psychopathology in adulthood. We found no evidence that any subtype of abuse and/or neglect or abuse/neglect during any particular phase of development uniquely predicted symptoms of psychopathology after controlling for relevant covariates. These results highlight the long-lasting significance of childhood abuse and neglect for reports of mental health in adulthood.

Researchers in various health-related fields have established links between child abuse and neglect and a range of negative mental and physical health outcomes, including anxiety and depression (Lindert et al., 2014; Maniglio, 2013), antisocial behavior (Pollock et al., 1990; Smith, Ireland, & Thornberry, 2005; Sousa et al., 2011), substance use (Herrenkohl et al,. 2013; Lansford et al., 2010), diabetes (Huang et al., 2015; Rich-Edwards et al., 2010), and cardiovascular disease (Batten et al., 2004; Fuller-Thomson, Brennenstuhl, & Frank, 2010). Abuse is present when an act results in non-accidental injury or emotional or physical harm to the child (Child Welfare Information Gateway, 2019). This includes physical, sexual, or emotional abuse. The Center for Disease Control (CDC) identifies neglect as present when a caregiver fails to provide adequate hygiene, shelter, clothing, medical care, supervision, or education (Leeb et al., 2008). Although there is a large body of work in the area of child abuse/neglect and psychopathology, much of this research has utilized retrospective assessments of maltreatment and cross-sectional data. The aim of the current study was to establish a link between child abuse and neglect and symptoms of psychopathology through middle adulthood using prospectively assessed, longitudinal data.

Child Abuse and Neglect and Psychopathology in Childhood and Adolescence

Researchers have been examining the link between child abuse and neglect and psychopathology in childhood and adolescence for over half a century (Briere & Ellio, 1994; Herrenkohl & Herrenkohl, 1979; Kempe et al., 1962; Parton, 1981). Such studies have examined many aspects of how child abuse and neglect relate to psychopathology, and results from laboratory-based studies (Jovanovic et al., 2011; Rodriguez, 2003), meta-analyses (Lindert et al., 2014; Li, D’arcy, & Meng, 2016), and narrative reviews (Nurcombe, 2000; Maniglio, 2013) converge in the conclusion that child abuse and neglect is reliably associated with higher levels of psychopathology in both childhood and adolescence. Moreover, scholars have theorized that abuse and neglect in childhood can lead to childhood behavioral problems by impacting emotion processing and regulation and contributing to non-normative neurological development (Cicchetti & Toth, 2005; Eisenberg, Cumberland, & Spinrad, 1998).

Indeed, experiences of childhood neglect may have severe short- and long-term effects on children’s psychopathology outcomes (Dubowitz & Bennett, 2007; Hildyard & Wolfe, 2002). For example, researchers have found that neglected children are more likely to be withdrawn (Camras & Rappaport, 1993; Crittenden 1992) and experience internalizing, but not externalizing problems compared to physically abused children (Haskett & Kistner, 1991). Bolger and Patterson (2001) similarly found that neglect, compared with physical or sexual abuse, was associated with internalizing problems in late childhood and early adolescence. The authors proposed that this distinction could be explained by perceived external control. Children in their study who experienced neglect reported higher levels of external control compared to sexually or physically abused children; in other words, they believed that others’, rather than their own attributes and actions accounted for their outcomes. The authors found that perceived external control was associated with more internalizing problems and mediated the relationship between neglect and internalizing problems in their sample.

Similar to child neglect, child physical abuse has also been associated with higher levels of psychopathology; however, findings suggest the association between physical abuse and behavioral problems may present as either internalizing or externalizing behaviors in childhood and adolescence (Manly, Cicchetti, & Barnett, 1994; Williamson, Borduin, & Howe, 1991). Shackman and colleagues (2010) found that physically abused children pay more attention to threat, and this mediated the relation between physical abuse and child-reported anxiety. This finding suggests that physical abuse may influence the development of attention regulation which may then contribute to the development of anxiety. Furthermore, Keiley and colleagues (2001) examined the developmental timing of onset of physical abuse and found that children experiencing the earliest onset of abuse (before age 5) were more likely to experience adjustment problems (i.e., internalizing and externalizing symptoms) in adolescence compared to children who experienced later onset of abuse (after age 5) or no abuse. Lansford and colleagues (2002) found that physical abuse can have lasting effects into adolescence. Using prospectively assessed, longitudinal data (from kindergarten to grade 11), these authors found that early physical abuse predicted adolescent psychopathology, including aggression, anxiety, and depression.

Sexual abuse has been consistently linked to psychopathological symptoms, and oftentimes severe psychopathology in childhood. Children who have experienced sexual abuse are more likely to display aggressive and/or hyperactive behaviors (Herrera & McCloskey, 2003; Sonnby et al., 2010; Swanston et al., 2003) as well as experience more emotional distress such as anxiety and depression (Williamson, Borduin, & Howe, 1991; Kim & Cicchetti, 2001; Mcleer et al., 1998) compared to children who have experienced other types of maltreatment or no maltreatment. Using prospective reports of abuse and neglect between ages 4 and 12 years, Lewis and colleagues (2016) found that children who experienced sexual abuse display more extreme levels of internalizing and externalizing symptoms compared to children who had experienced physical abuse or neglect.

Although research clearly indicates that childhood abuse and/or neglect is associated with psychopathology more generally in childhood, it is not clear whether the different subtypes of abuse are differentially associated with distinct behavioral symptoms. Neglect has consistently been linked to internalizing, but not externalizing symptoms (Bolger and Patterson, 2001; Haskett & Kistner, 1991); however, physical abuse (Keiley et al., 2001; Lansford et al., 2002) and sexual abuse (Lewis et al., 2016; Sonnby et al., 2010) have been linked to both internalizing and externalizing symptoms. Previous studies highlight the need to examine subtypes of abuse and neglect separately to clarify if one type is uniquely associated with more negative outcomes above and beyond other subtypes of abuse and/or neglect. Furthermore, recent investigations have proposed that deprivation (e.g., neglect) and threat (e.g., physical and sexual abuse) may differentially impact psychopathology by influencing the brain and biological systems in unique ways (Busso et al., 2017; McLaughlin et al., 2014; Sheridan & McLaughlin, 2014). Busso and colleagues (2017) found that whereas deprivation was associated with general psychopathology, threat was specifically associated with externalizing, but not internalizing symptoms in adolescents. Therefore, an important aspect of the current study is to disentangle the effect of neglect, physical abuse, and sexual abuse on psychopathology in adulthood. Follow-up analyses also examine if these effects differ for internalizing versus externalizing symptoms in adulthood.

Child Abuse and Neglect and Psychopathology in Adulthood

There is less research examining potential effects of child abuse and neglect on psychopathology in adulthood, and much of the work that does exist leverages retrospective reports of abuse and neglect. It has been theorized that chronic stress, such as childhood experiences of abuse/neglect, may negatively impact physiological mechanisms of stress response that could increase risk for developing psychopathology (Keyes et al., 2012). Using cross-sectional data, researchers have found that sexual abuse in childhood is associated with depression and anxiety (Brown et al., 1999; Musliner & Singer, 2014), suicidal behavior (Brown et al., 1999; Devries et al., 2014), delinquency (Pithers et al., 1998), and substance use (Bulik, Prescott, & Kendler, 2001). Fergusson and colleagues (1996) found that retrospective reports of child sexual abuse were associated with contemporaneous reports of psychopathology above and beyond neglect and physical abuse. Lange and colleagues (1999) found sexual abuse to predict psychopathology in a group of female adult victims, particularly for individuals whose abuse was perpetrated by a primary caregiver.

Reports of childhood physical abuse have also been associated with increased risk of psychopathology in adulthood (Norman et al., 2012; Sugaya et al, 2012), including depressive symptoms (Rhode et al., 2008; Sugaya et al., 2012) and behavioral problems (Teisl & Cicchetti, 2008). Keyes and colleagues (2012) found that childhood physical abuse was associated with increased risk for developing internalizing and externalizing symptoms in adulthood, while childhood neglect was not associated with these symptoms in adulthood. However, other researchers have found reports of childhood neglect to be associated with adult internalizing (Brown et al., 1999; Sachs-Ericsson et al., 2010) and externalizing (Lang, Klinteberg, & Alm, 2002) symptoms.

As noted, there have been a limited number of studies using longitudinal data and prospective reports of abuse and/or neglect when examining how childhood abuse and/or neglect impact adult psychopathology. Horwitz and colleagues (2001) examined prospective reports of abuse and neglect and psychopathology in adulthood (mean age of 28.7 years). They initially found that individuals who had experienced childhood abuse and neglect had higher rates of psychopathology than the comparison group. In a follow-up study with the same cohort about 12 years later (the sample had mean age of 41.2 years), Sperry and Widom (2013) found that child abuse and neglect was related to increased risk of developing anxiety and depression, and this relation was mediated by social support provided to the abused/neglected individual. Individuals who had good social support had a lower risk of developing anxiety and depression compared to individuals who did not have good social support.

Spataro and colleagues (2004) found that individuals who had experienced childhood sexual abuse (prospectively measured) had higher rates of psychiatric treatment in adulthood (mean age 28.4) than the general population. Cutajar and colleagues (2010) followed up with this sample approximately 5 years later (mean age of 33.8 years) and found that individuals who had experienced sexual abuse were at increased risk for psychopathology, including anxiety and psychosis. They also found that older age of onset of abuse was associated with greater risk for psychopathology in adulthood. Though there have been only a handful of studies that have examined the influence of childhood abuse and/or neglect on adult psychopathology using a prospective, longitudinal design, they consistently indicate that childhood abuse and/or neglect is associated with adult psychopathology. The current study attempts to replicate this key result while also building on these findings to investigate whether various aspects of abuse and/or neglect, such as subtype, perpetrator, and developmental timing, influence psychopathology in adulthood.

Methodological Limitations of the Literature

Much of the extant literature examining psychopathology following abuse and/or neglect has focused on outcomes in childhood or, using retrospective reports of abuse and/or neglect, psychopathology in adulthood. Researchers have found that incidents of abuse and/or neglect that were officially documented may not be retrospectively reported in adulthood (Widom, 1998; Widom, Raphael, & DuMont, 2004; Williams, 1995). The use of retrospective assessment of child abuse and/or neglect may result in recall biases in perceptions of abusive experiences (Clark et al., 2010; Shaffer, Huston, & Egeland, 2008) and thus may result in imprecise assessments of abuse/neglect. In contrast, prospective reports may allow for specificity in details about the abuse and/or neglect, such as the timing of onset of abuse. Although retrospective reports can still provide valuable information, especially incidents that prospective assessments may not have been able to obtain (Kendall-Tackett & Becker-Blease, 2004), and are significantly easier to acquire, prospective reports have been shown to provide a more complete and arguably more objective perspective of childhood abuse and/or neglect.

The Current Study

Though previous literature has shown that child abuse and neglect is associated with psychopathology in both children and adults, the majority of studies utilize cross-sectional, retrospective data to examine this association in adulthood. The current study addressed this gap by examining the association between experiences of abuse and/or neglect occurring in the first 17.5 years of life and self-reported symptoms of psychopathology in adulthood from ages 23 through 39 years, using prospective longitudinal data. The study uses data from the Minnesota Longitudinal Study of Risk and Adaptation (MLSRA), an at-risk cohort of participants, all of whom were born into poverty and have been followed prenatally through age 39 years. This report builds on a programmatic set of papers based on the MLSRA examining how experiences of abuse and/or neglect in childhood shape adult attachment states of mind (Martin et al., 2017; Nivison et al., 2020; Raby et al., 2017), romantic relationships (Labella et al., 2018), social and cognitive skills (Raby et al., 2018), parenting (Labella et al., 2018), and physical health (Johnson et al., 2017) in the adult years. The current report addresses the extent to which experiences of abuse and/or neglect in childhood predict adult psychopathology controlling for symptoms of psychopathology assessed at 16 years old and demographic covariates. On an exploratory basis, we examined which aspects of abuse and/or neglect (type, developmental timing, and perpetrator) were uniquely associated with psychopathology in adulthood. Furthermore, given the inconsistent findings regarding the impact of the type of abuse on internalizing versus externalizing symptoms, we follow-up our initial analyses examining total behavior problems with finer grained analyses in order to differentially investigate the predictive significance of the type of abuse for internalizing versus externalizing behavior problems.

Method

Participants

The current sample consisted of 140 participants drawn from the original poverty sample of the Minnesota Longitudinal Study of Risk and Adaptation (MLSRA; Sroufe et al., 2005). The MLSRA is an on-going longitudinal study that has followed the children of the original mothers from birth to age 39 years. The original sample included 267 pregnant women living below the poverty line seeking free prenatal services from clinics in Minneapolis, Minnesota. At the time of their child’s birth, 48% of the mothers were teenagers (M = 20.5 years, SD = 3.74) 65% were single, and 42% had completed less than a high school education.

The current sample included individuals who had completed at least one measure of psychopathology in adulthood and the Youth Self Report at age 16 and for whom prospectively acquired data on experiences of abuse and/or neglect from birth to 17.5 years were available. Participants were 49% female and 69% White/non-Hispanic. The subsample did not differ significantly from those who attrited from the original sample on biological sex, ethnicity, or childhood socioeconomic status. However, compared to the analytic sample (n = 140, M = 12.32, SD = 1.60), those excluded (n = 126, M = 11.81, SD = 1.94) had significantly lower levels of maternal education (t [243.33] = 2.33, p = .02, r = .14, equal variances not assumed). Nonetheless, average levels of maternal education in the current subsample were still approximately equal to a high school education, consistent with this being a high-risk cohort. This study,, titled Early Life Stress, Developmental Processes, and Adult Health, received IRB approval from the University of Minnesota’s IRB under protocol number 1104S98312.

Measures

Psychopathology.

Self-reported symptoms of psychopathology were measured four times in adulthood with the Achenbach System of Empirically Based Assessment (ASEBA; Achenbach & Rescorla, 2003) at ages 23, 26, 32, and 39 years. At ages 23 and 26 years, participants completed the Young Adult Self Report (YASR; Achenbach, 1997). The YASR consists of 119 items assessing a wide range of problems and socially desirable characteristics. Each item on the behavior checklist was rated by the participant on a 3-point scale: 0=not true, 1=somewhat or sometimes true, and 2=very true or often true. T scores were calculated from the Behavior Checklist at age 23 years for Internalizing (α = .90), Externalizing (α = .88), and Total Psychopathology (a combination of internalizing and externalizing items; α = .95) and at age 26 years for Internalizing (α = .90), Externalizing (α = .85), and Total Psychopathology (α = .94). At ages 32 and 39 years, the Adult Self Report was administered (ASR; Achenbach & Rescorla, 2003). The ASR was administered to participants using the same 3-point scale as the YASR; T Scores were calculated at age 32 years (Total Psychopathology, α = .94; Internalizing, α = .96; and Externalizing, α = .90) and 39 years (Total Psychopathology, α = .94; Internalizing, α = .94; Externalizing, α = .90). T scores at each time point (age 23, 26, 32, and 39 years) were averaged to create composite measures of total psychopathology, internalizing symptoms, and externalizing symptoms in adulthood. Composite measures were created because each measure was significantly correlated over the 4 time points. For Total Psychopathology, the average correlation over the 4 time points was r = .63 with a range of .50-.73 and standardized α = .86. For Internalizing symptoms, the average correlation over the 4 time points was r = .47 with a range of .45-.66 and standardized α = .84. For Externalizing symptoms, the average correlation over the 4 time points was r = .63 with a range of .53-.70 and standardized α = .86.

Adverse caregiving: Abuse and neglect.

The MLSRA uses the rubric childhood experiences of adverse caregiving as an umbrella term to refer to a variety of atypical parent-child experiences that were prospectively measured in the MLSRA cohort and are believed to be harmful to children’s development. The present study focused exclusively on information collected about MLSRA participants’ adverse caregiving experiences of physical abuse, sexual abuse, and neglect. All available data collected from birth to 17.5 years (up to 25 assessments) were reviewed for information regarding caregiving quality, physical discipline, supervision, home environment, physical and sexual assault, child protective service involvement, and foster care history. Information was obtained from parent-child observations, caregiver interviews, reviews of available child protection and medical records, adolescent reports, and teacher interviews.

This information was re-coded to apply contemporaneous definitions of abuse and neglect, to identify the specific perpetrator and ages of the abuse and neglect experiences, and to assess the reliability of those coding decisions. Coding criteria were based on definitions developed by the Centers for Disease Control and Prevention (CDC) in order to “promote consistent terminology and data collection related to child maltreatment” (Leeb, Paulozzi, Melanson, Simon, & Arias, 2008, p. 4). The coding included: (a) neglect of a child’s basic physical or cognitive needs, defined as a caregiver’s failure to provide adequate hygiene, shelter, clothing, medical care, supervision, or education; (b) physical abuse, defined as a caregiver’s “intentional use of physical force against a child that results, or has the potential to result in, physical injury” (Leeb at al., 2008, p. 14); and (c) sexual abuse, defined as sexual contact (e.g., molestation, rape) or noncontact exploitation (e.g., intentional exposure of child to pornography) by a custodial caregiver or by a perpetrator five or more years older than the target child. Although the CDC criteria only addresses sexual abuse perpetrated by a caregiver, the inclusion of non-caregiving perpetrators and the use of a five-year cutoff is consistent with other research in this area (e.g., Stoltenborgh, Van IJzendoorn, Euser & Bakermans-Kranenburg, 2011).

These CDC definitions were supplemented by a set of more specific coding guidelines that distinguished clear indicators of physical abuse, sexual abuse, and physical/cognitive neglect from ambiguous indicators that were not sufficient for classification in isolation of other evidence. Our scoring of abuse and neglect does not necessarily mean that these children or their families were involved with child protective services.

Judgments regarding abuse and neglect experiences were made for participants whose records had been previously flagged as potentially ever abused or neglected (n = 139, 52% of the original sample). Cases were flagged if there was any indication that the participant had been maltreated, even if ambiguous, based on previous maltreatment coding projects, any information on whether child protection had been involved with the family, and identification of periodic corporal punishment. Although many participants shared information about child protection services (CPS) involvement with interviewers, the exact number of participants involved with CPS is unknown.

Coding focused on the presence or absence of physical abuse, sexual abuse, and/or neglect in each of four developmental periods (Infancy: birth to 24 months; Early Childhood: 25 months to five years; Middle Childhood: 6–12 years; and Adolescence: 13–17.5 years). For incidents of physical and sexual abuse, coders additionally specified the perpetrator. Perpetrators included maternal caregivers (biological mothers, stepmothers, grandmothers), paternal or father figures (biological fathers, stepfathers, adoptive fathers, and mothers’ live-in boyfriends), and non-parental figures (relatives, neighbors, babysitters, and family friends). Two coders reviewed each case and demonstrated good to excellent reliability for all parameters: kappas were between .80 and .98 for presence or absence of physical abuse, sexual abuse, and/or neglect, .80 and .84 for presence or absence of each type during each development period; and .80 and .98 for incidents of physical or sexual abuse by each category of perpetrator. All discrepancies were resolved by consensus.

Within the current sample of those who had at least one measure of adult psychopathology and had completed the YSR at 16 years (N = 156), 76 individuals were classified as having ever experienced physical abuse, sexual abuse, and/or neglect. Among participants with histories of abuse and/or neglect, 61% had experienced neglect, 37% had experienced sexual abuse, and 64% had experienced physical abuse (not mutually exclusive). Within the abused/neglected group, 37% experienced abuse and/or neglect in infancy, 62% during early childhood, 70% during middle childhood, and 25% during adolescence (not mutually exclusive). In terms of chronicity, 37% of this group experienced abuse and/or neglect during one developmental period, 29% during two periods, 25% during three periods, and 3% during all four developmental periods; 6% had insufficient data to determine the number of developmental periods during which abuse and/or neglect occurred. Among participants with histories of abuse and/or neglect, 47% experienced one type of abuse and/or neglect, 37% experienced two types, and 9% experienced all three types; 7% had insufficient data to determine the number of abuse/neglect types experienced. With respect to perpetrator, 42% of participants who experienced abusive acts of commission were abused by a maternal perpetrator, 43% by a paternal perpetrator, and 24% by a non-parental perpetrator (not mutually exclusive). Sixty-nine individuals comprised the non-abused/non-neglected group; the number of missing assessments for this group did not differ from the group of individuals who were classified as having experienced abuse and/or neglect (t [127.71)] = −1.09, p = .28, equal variances not assumed).

Consistent with previous work in the MLSRA (e.g., Nivison et al., 2020) an omnibus measure of total experiences of abuse and/or neglect variable was used. This variable was calculated from summing the number of types of abuse (physical abuse, sexual abuse, neglect) in each developmental period (infancy, early childhood, middle childhood, adolescence). Because each of these subtypes was coded on a dichotomous basis for each developmental period, the total experiences of abuse and/or neglect exposure measure had a theoretical minimum of zero (i.e., the participant did not experience any type of abuse or neglect from infancy to adolescence) and a theoretical maximum of 12 (i.e., a participant experienced all three subtypes in each developmental period from infancy to adolescence). In order to calculate the severity variable, complete information was necessary for each subtype and timing variables. This resulted in an analytic frame of 140 participants. A complete description of the adverse caregiving coding system and descriptive information about abuse and neglect for the entire MLSRA sample can be found in the supplemental materials.

Covariates.

In order to test whether experiences of abuse and/or neglect were associated with psychopathology in adulthood above and beyond covariates, four key control variables used in recent analyses using the MLSRA (e.g., Raby et al., 2017; Waters et al., 2017; Waters et. al., 2018; Nivison et al., in press) were included in the regression analyses. The four control variables included were: child biological sex, child ethnicity, maternal education in childhood, and caregiver socioeconomic status. Child biological sex was coded as 1= female, 0 = male. Child ethnicity was coded as a binary variable 1 = White/non-Hispanic, 0 = non-White, given the predominantly white, non-Hispanic sample. Maternal education was operationalized as the number of years of education the mother had completed. Maternal education was assessed at seven time-points throughout the MLSRA (3 months before the child’s birth, 42 months, Grades 1–3, Grade 6, and age 16 years). The information from each timepoint was averaged to create a composite measure of maternal education. Childhood socioeconomic status was based on the primary caregiver at the time of each assessment (which was often mothers, but in rare cases fathers or grandparents) and was examined at seven time points from birth to 16 years (42 and 54 months, Grades 1–3, Grade 6, and age 16 years). At each time point the primary caregiver reported their occupation. Using the Duncan’s Socioeconomic Index (Stevens & Featherman, 1981), a widely used indicator of occupational ranking, each occupation was assigned a prestige score. Scores from each time point were then averaged to create a composite measure of childhood socioeconomic status.

In addition to the four commonly used control variables, this analysis also controlled for psychopathology in adolescence. At age 16 participants were administered the Youth Self Report (YSR; Achenbach, 1991). Participants completed the Behavior Checklist, a 119-item measure; participants rated each item on a 3-point scale: 0=not true, 1=somewhat or sometimes true, and 2=very true or often true. A Total Problem Scale T Score was calculated from the Behavior Checklist at age 16 (α = .91) and controlled for in subsequent regression analyses.

Results

Correlations among the study variables and descriptive data are presented in Table 1. Total abuse/neglect exposure was associated with higher levels of reported symptoms of total psychopathology (r = .31), internalizing (r = .26) and externalizing (r = .35) in adulthood. Additionally, various parameters of abuse and/or neglect were significantly associated with total psychopathology symptoms in adulthood. Specifically, higher levels of total adult psychopathology symptoms were significantly predicted by the experience of childhood neglect (r = .23), physical abuse (r = .22), and sexual abuse (r = .17); abuse perpetrated by mother (r = .22) and non-caregivers (r = .23); and abuse and/or neglect experienced in infancy (r = .22), early childhood (r = .20 ), and middle childhood (r = .29). Self-reported symptoms of total psychopathology at 16 years was also associated with higher levels of self-reported symptoms of psychopathology in adulthood (r = .40), suggesting that it was an important covariate for analyses presented below.

Table 1.

Correlation matrix of study variables using analytic sampling frame.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
1. Total Psychopathology	-
2. Internalizing Symptoms	.91**	-
3. Externalizing Symptoms	.86**	.65**	-
4. Total Abuse/Neglect Exposure	.31**	.26**	.35**	-
5. Neglect	.23**	.18*	.25**	.76**	-
6. Physical Abuse	.22**	.18*	.27**	.72**	.43**	-
7. Sexual Abuse	.17*	.15	.20*	.48**	.12	.14	-
8. Abuse by Mother	.22**	.15	.31**	.60**	.34**	.74**	.07	-
9. Abuse by Father	.10	.13	.08	.53**	.28**	.57**	.33**	.10	-
10. Abuse by Non-Caregiver	.23**	.17*	.28**	.31**	.08	.11	.68**	.13	−.04	-
11. Abuse/Neglect in Infancy	.22*	.14	.29**	.72**	.66**	.44**	.20*	.49**	.22**	.14	-
12. Abuse/Neglect in Early Childhood	.20*	.14	.24**	.79**	.58**	.65**	.23**	.67**	.36**	.17	.50**	-
13. Abuse/Neglect in Middle Childhood	.29*	.25**	.30**	.78**	.60**	.53**	.56**	.23**	.55**	.35**	.39**	.42**	-
14. Abuse/Neglect in Adolescence	.11	.16	.08	.37**	.18*	.28**	.25**	.29**	.23**	.14	.12	.11	.10	-
15. Child Biological Sex	−.13	−.10	−.14	−.05	−.14	−.14	.18*	−.13	.03	.01	−.08	−.14	.02	.10	-
16. Child Ethnicity	.02	.03	.02	.01	−.03	−.04	.06	.03	.00	−.02	.05	.04	−.03	.00	−.02	-
17. Maternal Education	−.04	.02	−.11	−.33**	−.31**	−.20*	−.10	−.11	−.21*	−.02	−.32**	−.20*	−.27**	−.10	−.08	−.06	-
18. Caregiver SES	−.04	−.07	−.04	−.27**	−.27**	−.20*	−.01	−.18*	−.06	−.09	−.22*	−.23**	−.19*	−.09	.09	.08	.48**	-
19. Psychopathology at 16 years	.40**	.35**	.38**	.17*	.03	.12	.23**	.04	.12	.24**	.04	.05	.22**	.16	.12	.05	.00	.02	-
Mean / %	49.71	48.89	51.11	1.26	.30	.31	.19	.20	.21	.13	.21	.42	.49	.14	48.6%	68.6%	12.32	22.74	54.87
SD	8.84	9.06	8.69	1.60	.46	.47	.40	.40	.41	.34	.48	.68	.74	.38			1.60	8.49	8.71

Note. Child Biological Sex was coded as 1 = female, 0 = male. Child ethnicity was coded as 1 = White/non-Hispanic, 0 = non-White. SES = socioeconomic status. N= 140.

^*p < .05

^**p < .01

Do experiences of abuse and/or neglect predict higher levels of adult symptoms of total psychopathology above and beyond symptoms of psychopathology at 16 years old and demographic covariates?

A hierarchical linear regression analysis was conducted to address this primary question. Total experiences of abuse and/or neglect was entered in the initial step. The second and final step included the covariates: symptoms of psychopathology at 16 years old, child biological sex, child ethnicity, maternal education, and caregiver socioeconomic status. Results are reported in Table 2. Total experiences of abuse and/or neglect were uniquely predictive of total psychopathology symptoms in adulthood (step 2), though inclusion of covariates also explained additional variance in symptoms of psychopathology in adulthood (from step 1 to step 2, ΔR² = .15, p = < .01). Additionally, self-reported symptoms of psychopathology at age 16 years and child sex were uniquely predictive of total psychopathology symptoms in adulthood.

Table 2.

Hierarchical regression of total child abuse/neglect exposure predicting adulthood total psychopathology, internalizing behaviors, and externalizing behaviors.

	Total Psychopathology						Internalizing Behaviors						Externalizing Behaviors
Variable	B	SE	β	t	p	R2	B	SE	β	t	p	R2	B	SE	β	t	p	R2
Step 1
Total Abuse/Neglect	1.70	0.45	0.31	3.80	<0.01	0.10	1.44	0.47	0.26	3.10	<0.01	0.07	1.87	0.43	0.35	4.32	< 0.01	0.12
Step 2
Total Abuse/Neglect	1.36	0.45	0.25	3.00	<0.01	0.25	1.21	0.48	0.21	2.51	0.01	0.19	1.49	0.44	0.27	3.37	< 0.01	0.26
Psychopathology at 16	0.39	0.08	0.38	4.94	<0.00		0.34	0.08	0.33	4.10	<0.00		0.36	0.08	0.36	4.64	< 0.01
Child Biological Sex	−2.95	1.36	−0.17	−2.17	0.03		−1.95	1.45	−0.11	−1.35	0.18		−3.14	1.33	−0.18	−2.37	0.02
Child Ethnicity	−0.14	1.44	0.01	−0.10	0.92		0.42	1.54	0.02	0.27	0.79		−0.38	1.41	−0.02	−0.27	0.79
Maternal Education	0.05	0.50	0.01	0.10	0.92		0.64	0.53	0.11	1.22	0.22		−0.41	0.48	−0.08	−0.85	0.39
Caregiver SES	0.03	0.09	0.03	0.31	0.76		−0.07	0.10	−0.06	−0.70	0.49		0.09	0.09	0.09	0.98	0.33

Note: N = 140. Dependent variable = mean composite of psychopathology, internalizing, and externalizing behaviors at 23, 26, 32, and 39 years. Total Abuse/Neglect Exposure = sum of abuse and neglect experiences from birth to 17.5 years. Child Biological Sex = child’s biological sex at birth coded as 1 = female, 0= male; Child Ethnicity = child’s ethnicity coded as 1 = White/non-Hispanic, 0 = non-White; Maternal Education = maternal education from 3 months before birth to 16 years; Caregiver SES = caregiver occupational prestige from 42 months to 16 years.

Which aspects of abuse and/or neglect are uniquely associated with psychopathology in adulthood?

To address this exploratory question, three parameterizations of abuse and neglect were examined. Specifically, type (physical abuse, sexual abuse, or neglect), perpetrator (abused by mother, father, or non-parental figure), and developmental timing of abuse and/or neglect (abused and/or neglected in infancy, early childhood, middle childhood, or adolescence) were examined. Paralleling the initial analysis, three hierarchical linear regression analyses were conducted. In each regression, the abuse/neglect parameters were entered as the initial step and symptoms of psychopathology at age 16 and the other covariates were entered as the second step.

Type.

As reported in Table 3, neglect (but not sexual or physical abuse) was marginally uniquely predictive of symptoms of total psychopathology in adulthood in step 1. The inclusion of covariates did explain significantly more variance in psychopathology symptoms in adulthood (from step 1 to step 2, ΔR² = .15, p < .01) with symptoms of psychopathology at 16 years and child biological sex uniquely predictive of psychopathology in adulthood. With the addition of covariates in the model, neglect remained marginally predictive of adult self-reported symptoms of total psychopathology.

Table 3.

Hierarchical regression of type of abuse/neglect exposure predicting adulthood total psychopathology internalizing behaviors, and externalizing behaviors.

	Total Psychopathology						Internalizing Behaviors						Externalizing Behaviors
Variable	B	SE	β	t	p	R2	B	SE	β	t	p	R2	B	SE	β	t	p	R2
Step 1
Neglect	3.03	1.75	0.16	1.74	0.09	0.09	2.36	1.82	0.12	1.30	0.20	0.06	2.83	1.69	0.15	1.68	0.10	0.12
Physical Abuse	2.61	1.73	0.14	1.51	0.13		2.25	1.80	0.12	1.25	0.21		3.46	1.67	0.19	2.07	0.04
Sexual Abuse	2.88	1.85	0.13	1.56	0.12		2.72	1.92	0.12	1.41	0.16		3.43	1.79	0.16	1.92	0.06
Step 2
Neglect	3.14	1.70	0.16	1.85	0.07	0.24	2.73	1.81	0.14	1.51	0.13	0.19	2.64	1.65	0.14	1.60	0.11	0.26
Physical Abuse	1.50	1.64	0.08	0.92	0.36		1.34	1.74	0.07	0.77	0.44		2.39	1.60	0.13	1.50	0.14
Sexual Abuse	1.59	1.80	0.07	0.89	0.39		1.53	1.91	0.07	0.80	0.42		2.31	1.75	0.11	1.32	0.19
Psychopathology at 16	0.40	0.08	0.39	4.92	< 0.01		0.35	0.09	0.34	4.08	< 0.01		0.36	0.08	0.36	4.53	< 0.01
Child Biological Sex	−2.82	1.42	−0.16	−1.99	0.05		−1.85	1.51	−0.10	−1.23	0.22		−3.07	1.38	−0.18	−2.22	0.03
Child Ethnicity	−0.02	1.46	−0.00	−0.01	0.99		0.52	1.56	0.03	0.33	0.74		−0.26	1.43	−0.01	−0.18	0.86
Maternal Education	0.03	0.50	0.01	0.06	0.95		0.62	0.53	0.11	1.17	0.25		−0.45	0.49	−0.08	−0.92	0.36
Caregiver SES	0.02	0.09	0.02	0.26	0.79		−0.07	0.10	−0.07	−0.73	0.47		0.08	0.09	0.08	0.92	0.36

Note. N = 140. Dependent variable = mean composite of psychopathology, internalizing, and externalizing behaviors at 23, 26, 32, and 39 years. Neglect = ever neglected by any perpetrator from birth to 17.5 years; Physical Abuse = ever physically abused by any perpetrator from birth to 17.5 years; Sexual Abuse = ever sexually abused by any perpetrator from birth to 17.5 years. Child Biological Sex = child’s biological sex at birth coded as 1 = female, 0 = male; Child Ethnicity = child’s ethnicity coded as 1 = White/non-Hispanic, 0 = non-White; Maternal Education = maternal education from 3 months before birth to 16 years; Caregiver SES = caregiver occupational prestige from 42 months to 16 years.

Perpetrator.

As reported in Table 4, abuse perpetrated by the mother and non-caregiver were uniquely predictive of symptoms of total psychopathology in adulthood. The inclusion of covariates did explain significantly more variance in symptoms of total psychopathology in adulthood (from step 1 to step 2, ΔR² = .15, p < .01). Once covariates were added into the model, abuse perpetrated by the non-caregivers was no longer uniquely predictive of symptoms of total psychopathology in adulthood; however, abuse perpetrated by mother remained uniquely predictive of total psychopathology symptoms in adulthood. As in the previous analysis, psychopathology at age 16 years and child biological sex were uniquely predictive of symptoms of psychopathology in adulthood.

Table 4.

Hierarchical regression of abuse exposure by perpetrator predicting adulthood total psychopathology internalizing behaviors, and externalizing behaviors.

	Total Psychopathology						Internalizing Behaviors						Externalizing Behaviors
Variable	B	SE	β	t	p	R2	B	SE	β	t	p	R2	B	SE	β	t	p	R2
Step 1
Abuse by Mother	4.06	1.82	0.18	2.24	0.03	0.10	2.59	1.90	0.12	1.36	0.18	0.06	5.98	1.73	0.28	3.47	<0.01	0.16
Abuse by Father	2.04	1.78	0.09	1.15	0.25		2.79	1.86	0.13	1.50	0.14		1.29	1.69	0.06	0.76	0.45
Abuse by Non-Caregiver	5.63	2.16	0.21	2.60	0.01		4.42	2.26	0.16	1.96	0.05		6.35	2.05	0.25	3.09	<0.01
Step 2
Abuse by Mother	3.63	1.73	0.17	2.10	0.04	0.25	2.16	1.85	0.10	1.16	0.25	0.18	5.63	1.64	0.26	3.42	<0.01	0.30
Abuse by Father	0.99	1.71	0.05	0.58	0.57		2.25	1.83	0.10	1.23	0.22		−0.02	1.63	0.00	−0.01	0.99
Abuse by Non-Caregiver	3.36	2.09	0.13	1.61	0.11		2.26	2.23	0.08	1.01	0.31		4.39	1.98	0.17	2.22	0.03
Psychopathology at 16	0.39	0.08	0.38	4.80	<0.01		0.34	0.09	0.33	3.98	<0.01		0.35	0.08	0.35	4.57	<0.01
Child Biological Sex	−2.92	1.38	−0.17	−2.12	0.04		−2.03	1.48	−0.11	−1.37	0.17		−2.94	1.31	−0.17	−2.25	0.03
Child Ethnicity	−0.20	1.46	−0.01	−0.14	0.89		0.44	1.56	0.02	0.26	0.80		−0.51	1.39	−0.03	−0.36	0.72
Maternal Education	−0.23	0.50	−0.04	−0.46	0.64		0.47	0.53	0.08	0.88	0.38		−0.77	0.47	−0.14	−1.63	0.11
Caregiver SES	0.03	0.09	0.03	0.32	0.74		−0.08	0.10	−0.08	−0.81	0.42		0.11	0.09	0.10	1.21	0.23

Note: N = 140. Dependent variable = mean composite of psychopathology, internalizing, and externalizing behaviors at 23, 26, 32, and 39 years. Abuse by Mother = ever physically/sexually abused by mother from birth to 17.5 years; Abuse by Father = ever physically/sexually abused by father from birth to 17.5 years; Abuse by Non-Caregiver = ever physically/sexually abused by non-parent figure from birth to 17.5 years; Child Biological Sex = child’s biological sex at birth coded as 1 = female, 0= male; Child Ethnicity = child’s ethnicity coded as 1 = White/non-Hispanic, 0 = non-White; Maternal Education = maternal education from 3 months before birth to 16 years; Caregiver SES = caregiver occupational prestige from 42 months to 16 years.

Developmental timing.

As reported in Table 5, abuse and/or neglect in middle childhood, but not infancy, early childhood, or adolescence, was uniquely predictive of symptoms of total psychopathology in adulthood (step 1). The inclusion of covariates did explain significantly more variance in symptoms of psychopathology in adulthood (from step 1 to step 2, ΔRR² = .16, p = < .01). However, once covariates were added to the model, no developmental time period was uniquely predictive of total psychopathology symptoms. Again, psychopathology at 16 years and child biological sex were uniquely predictive of symptoms of psychopathology in adulthood.

Table 5.

Hierarchical regression of abuse/neglect exposure in each developmental period predicting adulthood total psychopathology internalizing behaviors, and externalizing behaviors.

	Total Psychopathology						Internalizing Behaviors						Externalizing Behaviors
Variable	B	SE	β	t	p	R2	B	SE	β	t	p	R2	B	SE	β	t	p	R2
Step 1
Abuse/Neglect in Infancy	1.78	1.80	0.10	0.99	0.32	0.10	0.55	1.86	0.03	0.30	0.77	0.08	3.18	1.74	0.17	1.82	0.07	0.13
Abuse/Neglect in Early Childhood	0.59	1.27	0.05	0.47	0.64		0.31	1.32	0.02	0.24	0.81		0.81	1.24	0.06	0.66	0.51
Abuse/Neglect in Middle Childhood	2.69	1.10	0.23	2.45	0.02		2.64	1.14	0.22	2.32	0.02		2.37	1.06	0.20	2.23	0.03
Abuse/Neglect in Adolescence	1.62	1.90	0.07	0.86	0.39		3.03	1.97	0.13	1.54	0.13		0.65	1.84	0.03	0.36	0.72
Step 2
Abuse/Neglect in Infancy	2.19	1.72	0.12	1.28	0.20	0.26	1.20	1.83	0.06	0.66	0.51	0.20	3.33	1.67	0.18	1.99	0.05	0.27
Abuse/Neglect in Early Childhood	0.43	1.20	0.03	0.36	0.72		0.05	1.28	0.00	0.04	0.97		0.74	1.17	0.06	0.64	0.53
Abuse/Neglect in Middle Childhood	1.88	1.06	0.16	1.77	0.08		2.06	1.13	0.17	1.82	0.07		1.51	1.03	0.13	1.46	0.15
Abuse/Neglect in Adolescence	0.89	1.80	0.04	0.49	0.62		2.38	1.91	0.10	1.25	0.21		−0.05	1.75	0.00	−0.03	0.98
Psychopathology at 16	0.38	0.80	0.38	4.76	<0.01		0.33	0.09	0.31	3.80	<0.01		0.36	0.08	0.36	4.65	<0.01
Child Biological Sex	−3.01	1.39	−0.17	−2.16	0.03		−2.23	1.48	−0.12	−1.51	0.13		−3.01	1.35	−0.17	−2.22	0.03
Child Ethnicity	−0.08	1.46	0.00	−0.05	0.96		0.55	1.55	0.03	0.35	0.72		−0.40	1.42	−0.02	−0.28	0.78
Maternal Education	0.12	0.51	0.02	0.24	0.81		0.68	0.54	0.12	1.27	0.21		−0.31	0.49	−0.06	−0.64	0.53
Caregiver SES	0.02	0.09	0.02	0.24	0.81		−0.07	0.10	−0.07	−0.74	0.46		0.08	0.09	0.08	0.91	0.37

Note. N = 140. Dependent variable = mean composite of psychopathology, internalizing, and externalizing behaviors at 23, 26, 32, and 39 years. Abuse/Neglect in Infancy = ever physically/sexually abused or neglected by any perpetrator from 0–2 years; Abuse/Neglect in Early Childhood = ever physically/sexually abused or neglected by any perpetrator from 3–5 years; Abuse/Neglect in Middle Childhood = ever physically/sexually abused or neglected by any perpetrator from 6–12 years; Abuse/Neglect in Adolescence = ever physically/sexually abused or neglected by any perpetrator from 13–17.5 years. Child Biological Sex = child’s biological sex at birth coded as 1 = female, 0 = male; Child Ethnicity = child’s ethnicity coded as 1 = White/non-Hispanic, 0 = non-White; Maternal Education = maternal education from 3 months before birth to 16 years; Caregiver SES = caregiver occupational prestige from 42 months to age 16 years.

Does overall abuse/neglect or aspects of abuse/neglect differentially predict reported internalizing and/or externalizing symptoms in adulthood?

As in the analyses above, hierarchical linear regressions were used to determine whether measures of abuse and/or neglect were uniquely associated with internalizing or externalizing symptoms in adulthood. Each model included the abuse/neglect parameter as the first step and added covariates in the second step. As reported in Table 2, total experiences of abuse and/or neglect were predictive of both internalizing and externalizing symptoms before and after covariates were added to the model. When examining subtypes of abuse, differential effects emerged. Though no subtype was uniquely predictive of adulthood internalizing symptoms, physical and sexual abuse were uniquely predictive of externalizing symptoms before covariates were added to the model. However, after the covariates were added, no subtype was uniquely predictive of externalizing symptoms in adulthood. These results are reported in Table 3.

Furthermore, there were some differences in unique associations between internalizing and externalizing symptoms in adulthood when examining perpetrators (see Table 4). Abuse by non-caregiver was uniquely predictive of adulthood internalizing symptoms, though this effect disappeared after covariates were added to the model. On the other hand, abuse by mother and non-caregiver was uniquely predictive of adulthood externalizing symptoms before and after covariates were added to the model. Finally, there were differential associations between developmental timing of abuse and internalizing and externalizing symptoms in adulthood. Abuse/neglect in infancy was uniquely predictive of externalizing, but not internalizing, symptoms in adulthood after controlling for all covariates. Abuse/neglect in middle childhood was uniquely predictive of both internalizing and externalizing symptoms in adulthood before covariates were added to the model. After covariates were added, however, abuse/neglect in middle childhood only marginally predicted internalizing symptoms. No other developmental period was uniquely predictive of these symptoms (see Table 5).

Discussion

Although there is substantial research examining the associations between child abuse and neglect and psychopathology over the lifespan, much of this work focuses on psychopathology in childhood or uses cross-sectional, retrospective data. The current study utilized prospective, longitudinal data to examine to what extent child abuse and/or neglect was associated with self-reported symptoms of psychopathology in adulthood and if certain aspects of abuse and/or neglect were uniquely associated with psychopathology symptoms in adulthood. Results indicated that overall, child abuse and neglect was robustly associated with symptoms of total psychopathology in adulthood, controlling for symptoms at age 16 and demographic covariates (i.e., child biological sex, child ethnicity, maternal education, and caregiver SES). Furthermore, neglect (but not physical or sexual abuse) was marginally uniquely predictive of adult symptoms of total psychopathology above and beyond covariates. Likewise, abuse perpetrated by mothers (but not fathers or non-caregivers) uniquely predicted adult symptoms of total psychopathology above and beyond covariates. Finally, after controlling for covariates, no developmental time period from birth to 17.5 years uniquely predicted adult symptoms of total psychopathology.

Although there have been few prospective, longitudinal studies examining this question, it was expected that overall abuse and/or neglect would predict adult symptoms of total psychopathology given the abundant literature discussing relations between these two variables in childhood, adolescence, and adulthood (Cicchetti & Toth, 2005; Lansford et al., 2002; Norman et al., 2012). Furthermore, these findings are consistent with results from other prospective longitudinal studies (Cutajar et al., 2010; Spataro et al., 2004; Sperry & Widom, 2013).

The current study extended the literature by examining the unique predictive significance of different aspects of abuse/neglect. Results indicated that no subtype was significantly uniquely predictive of total psychopathology symptomatology in adulthood, though neglect had a marginally significant effect, above and beyond physical and sexual abuse. Other work has found evidence for the unique predictive effects of neglect on psychopathology, above and beyond physical abuse (Haskett & Kistner, 1991) and sexual abuse (Bolger & Patterson, 2001). As discussed earlier, this marginal effect (should it prove more reliable in future studies) might be explained by the theory of external control (Bolger & Patterson, 2001), which claims that children who experience neglect have higher levels of perceived external control compared to sexually or physically abused children and has been found to mediate the relation between neglect and psychopathology. Furthermore, it is possible that parents and youth are more likely to self-report neglect rather than physical or sexual abuse. Thus, it is possible that there are more false negative reports for sexual and physical abuse compared to neglect. It is important to note that in the current study, physical (r = .22) and sexual abuse (r = .17) were also correlated with psychopathology in adulthood, just not uniquely so.

Results also indicated that abuse perpetrated by mothers predicted symptoms of total psychopathology beyond abuse perpetrated by fathers or non-caregivers. This might reflect the special early bonds that children form with their mothers as implied by attachment theory (Bowlby, 1973, 1980), potentially making abuse perpetrated by primary caregivers particularly detrimental to development. Furthermore, previous work has found that abuse perpetrated by a caregiver, rather than a non-caregiver, is more often to be repetitive and children are more likely to assume blame and responsibility for abuse that, in turn, has been related to increased symptoms of psychopathology (Lange et al., 1999). Families in the MLSRA sample were more likely to have a single mother as a caregiver (and thus not have an involved father) than in more normative-risk samples. When the children in the study were born, 65% had single mothers. This could help explain why abuse perpetrated by mothers (r = .22) but not fathers (r = .10) was significantly associated with psychopathology in adulthood in this sample. Although abuse perpetrated by non-caregivers was significantly correlated with psychopathology in adulthood (r = .23), few children in this sample experienced abuse by non-caregivers, and these low frequencies may have impacted power to detect this effect in the regression analyses.

The study also found that developmental timing of abuse/neglect did not predict symptoms of total psychopathology in adulthood controlling for covariates. However, all developmental time periods except adolescence were significantly correlated with psychopathology in adulthood. The results suggest that experiences of abuse across development may be meaningful in thedevelopment of adult symptoms of psychopathology.

In follow-up analyses, few results indicated any differential association between child abuse and/or neglect and internalizing and externalizing symptoms in adulthood. Only physical abuse and sexual abuse uniquely predicted adult externalizing symptoms; however, no subtypes of abuse and/or neglect were associated with internalizing symptoms in adulthood. This is consistent with Busso and colleague’s (2017) conceptual model, discussed above, in which deprivation (such as neglect) and threat (such as physical and sexual abuse) are differentially related to psychopathology. They hypothesize that physiological responses to threat will impact psychopathology in a way that deprivation does not. Busso and colleagues (2017) found that while deprivation was associated with general psychopathology (as was found in the current study), threat was specifically associated with externalizing, but not internalizing, psychopathology in adolescents. The current study extended these findings into adulthood.

Overall abuse and/or neglect was associated with both adulthood internalizing and externalizing symptoms, and abuse and/or neglect in middle childhood also predicted both internalizing and externalizing symptoms in adulthood. Additionally, after controlling for covariates, no perpetrator was uniquely associated with internalizing symptoms in adulthood whereas abuse perpetrated by mother and non-caregiver was associated with externalizing symptoms in adulthood.

It is important to consider that the key findings discussed all involved controls for psychopathology at age 16 years. As such, it is not simply that childhood abuse and neglect leads to childhood psychopathology, that, in turn, is related into adult psychopathology. Rather, these data suggest that childhood abuse and neglect has unique predictive significance for adult psychopathology even after controlling for earlier psychopathology. The current study aimed to understand the unique effects of child abuse and/or neglect on adult mental health outcomes, above and beyond effects on childhood mental health outcomes. Thus, importantly, results should be interpreted specifically as the unique predictive significance of child abuse and/or neglect on adult psychopathology, beyond the effects it had on child or adolescent psychopathology. Furthermore, controlling for psychopathology at 16 years allowed us to more compellingly rule out reverse causation.

Although this study is unique in the extent of time participants have been followed prospectively from prior to birth to mid-adulthood, findings of the present study should be interpreted in light of study limitations. First, the racial and ethnic makeup of the sample is not ideally diverse and mothers were all poverty at the time of the original recruitment. As such, results may not generalize (for example) to the entire population of the United States. A second limitation of the current study is attrition and missing data, which are not unusual in longitudinal cohort studies. Third, the sample size is modest, limiting the extent to which moderators and mediators can be explored.

Fourth and finally, though obtaining prospective reports of abuse is valuable, it is likely that not all children in the MLSRA were willing to share incidents of abuse at the time they were occurring and not all caregivers were knowledgeable of abuse or willing to disclose abuse they had perpetrated or a known other had perpetrated. Therefore, it is possible that some of the children who we considered to not be abused or neglected might have been experiencing abuse or neglect. That said, the coding system implemented in this study was conservative in that we: (a) coded for possible childhood abuse and/or neglect all cases that were ever flagged by staff personnel and (b) removed from this analysis cases without known abuse but missing two or more full assessments within any given developmental period (as it is possible that such assessments would have yielded evidence for abuse and/or neglect). Moreover, over the course of the MLSRA, participants developed relationships with the research staff and often felt comfortable disclosing private and personal information—though, of course, it is impossible to know if every participant disclosed such information.

Despite these limitations, the current study addressed important gaps in research by leveraging prospective, longitudinal data to understand the associations between childhood abuse and/or neglect and symptoms of psychopathology in mid-adulthood. The current study’s key contribution is the unique design that allows for inferences about the long-term associations between child abuse and neglect on adult mental health. The design of this study is unusual in this field, as it is rare to have a prospective longitudinal study that spans the life course through mid-adulthood focusing on adult outcomes and their relation to experiences of child abuse and neglect. Another advantage of the current study is that it allows for a nuanced parameterization of maltreatment. It is important to note few differences based on those parameterizations were found. However, abuse perpetrated by mothers seemed to be a particularly significant driver of longer-term effects on self-report mental health. Furthermore, the study was comprised of an at-risk sample, making comparisons between the abused/neglected group and non-abused/neglected group less attributable to correlated aspects of the environment (e.g., low SES).The current paper also contributes to a set of papers based in MLSRA data examining the deleterious associations between child abuse and/or neglect and a wide range of adult outcomes, including physical health (Johnson et al., 2017), attachment relationships (Martin et al., 2017; Nivison et al., 2020; Raby et al., 2017), social and cognitive skills (Raby et al., 2018), romantic relationships (Labella et al., 2018), and parenting (Labella et al., 2019) in adulthood. Leveraging large sample longitudinal data sets, future research in this area might examine mediators and moderators of links across time to better understand why and for whom abuse and/or neglect impacts adult psychopathology. In particular, future work could examine the role of the quality of interpersonal relationships throughout childhood (siblings, friendships) and adulthood (romantic relationships) in altering the associations between childhood abuse and/or neglect and psychopathology in adulthood we report here.