Content uploaded by Adriano Schimmenti
Author content
All content in this area was uploaded by Adriano Schimmenti on Apr 29, 2016
Content may be subject to copyright.
Author’s Accepted Manuscript
Associations between depression and specific
childhood experiences of abuse and neglect: A
meta-analysis
Maria Rita Infurna, Corinna Reichl, Peter Parzer,
Adriano Schimmenti, Antonia Bifulco, Michael
Kaess
PII: S0165-0327(15)30530-9
DOI: http://dx.doi.org/10.1016/j.jad.2015.09.006
Reference: JAD7688
To appear in:
Journal of Affective Disorders
Received date: 2 July 2015
Revised date: 20 August 2015
Accepted date: 5 September 2015
Cite this article as: Maria Rita Infurna, Corinna Reichl, Peter Parzer, Adriano
Schimmenti, Antonia Bifulco and Michael Kaess, Associations between
depression and specific childhood experiences of abuse and neglect: A meta-
a n a l y s i s , Journal of Affective Disorders,
http://dx.doi.org/10.1016/j.jad.2015.09.006
This is a PDF file of an unedited manuscript that has been accepted for
publication. As a service to our customers we are providing this early version of
the manuscript. The manuscript will undergo copyediting, typesetting, and
review of the resulting galley proof before it is published in its final citable form.
Please note that during the production process errors may be discovered which
could affect the content, and all legal disclaimers that apply to the journal pertain.
www.elsevier.com/locate/jad
1
Associations between depression and specific childhood experiences of abuse and neglect: a
meta-analysis
Maria Rita Infurna
a,b,c
, Corinna Reichl
a
, Peter Parzer
a
, Adriano Schimmenti
d
, Antonia Bifulco
e
,
Michael Kaess
a
a
Clinic of Child and Adolescent Psychiatry, Centre of Psychosocial Medicine, University of
Heidelberg, Heidelberg, Germany
b
Department of Psychological and Educational Sciences, University of Palermo, Palermo, Italy
c
Institute of Psychology, Faculty of Behavioural and Cultural Studies, University of Heidelberg,
Heidelberg, Germany
d
Faculty of Human and Social Sciences, UKE – Kore University of Enna, Enna, Italy
e
Centre for Abuse and Trauma Studies, Middlesex University, London, UK
_____________________
Corresponding Author
Associate Professor Michael Kaess, M.D.
Clinic of Child and Adolescent Psychiatry
Centre of Psychosocial Medicine, University of Heidelberg
Blumenstrasse 8
69115, Heidelberg, Germany
Mailto: michael.kaess@med.uni-heidelberg.de
Phone: +49 6221 566914
2
Abstract
Background: Research documents a strong relationship between childhood maltreatment and
depression. However, only few studies have examined the specific effects of various types of
childhood abuse/neglect on depression. This meta-analysis estimated the associations between
depression and different types of childhood maltreatment (antipathy, neglect, physical abuse,
sexual abuse, and psychological abuse) assessed with the same measure, the Childhood
Experience of Care and Abuse (CECA) interview.
Method: A systematic search in scientific databases included use of CECA interview and
strict clinical assessment for major depression as criteria. Our meta-analysis utilized Cohen´s d
and relied on a random-effects model.
Results: The literature search yielded 12 primary studies (reduced from 44), with a total of
4372 participants and 34 coefficients. Separate meta-analyses for each type of maltreatment
revealed that psychological abuse and neglect were most strongly associated with the outcome of
depression. Sexual abuse, although significant, was less strongly related. Furthermore, the effects
of specific types of childhood maltreatment differed across adult and adolescent samples.
Limitations: Our strict criteria for selecting the primary studies resulted in a small numbers of
available studies. It restricted the analyses for various potential moderators.
Conclusion: This meta-analysis addressed the differential effects of type of childhood
maltreatment on major depression, partially explaining between-study variance. The findings
clearly highlight the potential impact of the more “silent” types of childhood maltreatment (other
than physical and sexual abuse) on the development of depression.
Key words: depression; CECA; abuse; neglect; maltreatment; meta-analysis
Introduction
The association between childhood adversity and the development of depression has been
widely studied. Substantial evidence from both cross-sectional and prospective studies indicates
3
that childhood abuse and neglect are strongly associated with the development of clinical
depression in both adolescence and adulthood (Abela and Skitch, 2007; Bifulco et al., 1998;
Gibb et al., 2001; MacMillan et al., 2001; Widom et al., 2007).
Most research has so far focused on physical and sexual abuse (Cutajar et al., 2010; Dube et
al., 2005; Fergusson et al., 2008; Kendler et al., 2000; Molnar et al., 2001); whereas fewer
studies have examined the effects of other types of abuse or neglect (Alloy et al., 2006; Bifulco
et al., 2002; Chen et al., 2014; Liu et al., 2009; Musliner and Singer, 2014). Nonetheless, several
authors have argued that emotional abuse in childhood, which typically includes experiences of
being rejected, degraded, terrorized, isolated, or teased, might be more strongly related to
internalizing symptoms and the development of depression than physical abuse or sexual abuse
(Alloy et al., 2006; Gibb et al., 2003; Lumley and Harkness, 2007; Shapero et al., 2014).
Unfortunately, high levels of heterogeneity can be observed across the published studies,
which limits the comparability of previous research. This could be due to the use of different
sampling procedures and methods of assessment. Indeed, different measurement methods (self-
reports, interviews, hospital records, and official records) have been used to investigate
childhood maltreatment and, more importantly, many different definitions of childhood adversity
have been applied (Nanni et al., 2012).
A recent meta-analysis has examined whether physical and sexual abuse in childhood were
associated with depression and anxiety in later life (Lindert et al., 2014). Although the results of
this analysis showed strong associations between sexual and physical abuse in childhood and
depression, the measures used to assess both childhood abuse and depression varied greatly.
Moreover, neither emotional abuse nor neglect were taken into account. Another recent meta-
analysis showed that maltreated individuals were twice as likely as those without a history of
childhood maltreatment to develop both recurrent and persistent depressive episodes (Nanni et
al., 2012). Again, the available data did not permit an examination of the specific contribution of
different maltreatment subtypes.
4
Indeed, although research has consistently documented a strong and significant relationship
between childhood maltreatment, in general, and depression, studies that have examined the
relationship between specific forms of adversity and depression development are sparse.
Currently, no review or meta-analysis has attempted to elucidate the association between a broad
variety of specific childhood experiences of abuse and neglect and depression among the
scientific publications in this field.
In contrast, current theories propose specific pathways from childhood adversities to
psychopathology, and propose that distinct types of events may cause specific symptoms in
individuals with specific vulnerabilities. For instance, the schema-based cognitive model of
depression and anxiety states that the type of emotion experienced depends on the thought and
belief content activated by life experiences (Beck, 2008). A recent study investigated three
dimensions of depression and anxiety symptomatology (general distress, anhedonic depression
and anxious arousal) and reported that only emotional neglect was associated with all the three
symptom dimensions, whilst sexual abuse was associated with general distress and anxious
arousal, and physical abuse with anxious arousal only (van Veen et al., 2013). Likewise,
attachment theory (Bowlby, 1982) postulates that caregiving figures are fundamental for the
development of the child´s representational or internal working models of the world. Thus,
experiences of early emotional abuse may be particularly maladaptive because negative
evaluation is supplied directly by the primary attachment figures, which may be more potent in
activating a more negative model of the self and the others (Shapero et al, 2014).
Ultimately, we propose that different types of childhood adversities, with a particular
emphasis on psychological types of maltreatment, may influence the development of distinct
etiological pathways in depression that would benefit from individually tailored interventions.
This is supported by recent findings suggesting that different types of caregiver-child
interactions were related to different emotion regulation pathways associated with depression
(O´Mahen et al., 2015). For example, in the study by O'Mahen and colleagues (2015), childhood
5
emotional neglect and childhood emotional abuse were associated with the emotion regulation
strategies that were most strongly related to depression (rumination and behavioural avoidance).
Further clarification of the specific role of psychological maltreatment may help inform parents,
teachers, health care workers and the general public in recognising such maltreatment, but also
the likely detrimental impact on the developing child. This may help to overcome the
underestimation of the impact of emotional abuse/neglect, leading to better recognition and more
adequate interventions to prevent long-term disorders (Schimmenti and Bifulco, 2015).
Thus, the present meta-analysis aimed to estimate the specific association between depression
(recurrent or persistent) and different types of childhood maltreatment (antipathy, neglect,
physical abuse, sexual abuse, and psychological abuse) assessed with the same measure.
Specifically, we chose the Childhood Experience of Care and Abuse interview (CECA) (Bifulco
et al., 1994), a measure with a 20-year standing and considered the ‘gold standard’ in this area of
international research (Thabrew et al., 2012).
According to findings from the literature, forms of maltreatment centering on themes of
parents’ rejection, and their perception of failure, or unworthiness in the child seem to be
associated with a high vulnerability for developing depression (Gibb et al., 2003). Thus, we
hypothesized that antipathy (involving parental criticism and hostility), neglect (withholding
material care), and psychological abuse (coercive sadistic control) as defined by Bifulco and
colleagues (Bifulco et al., 1994; Moran et al., 2002) would present a stronger association with
depression than other forms of maltreatment (physical and sexual abuse).
Additionally, evidence has consistently demonstrated that adolescents with a history of
childhood maltreatment have a greater vulnerability for becoming depressed or suicidal than
individuals without such a history (Bifulco et al., 2014; Dunn et al., 2013; Lumley and Harkness,
2007; Moretti and Craig, 2013; Ystgaard et al., 2004). Furthermore, the risk of depression and
suicide attempts in maltreated individuals seem to be higher in adolescence than in adulthood
(Brown et al., 1999; Dunn et al., 2013). Thus, this meta-analysis further aimed to find specific
6
associations between childhood maltreatment types and depression, distinguishing adult from
adolescent samples.
Method
Search strategies
A systematic search strategy was used to identify relevant studies. A three-step literature
search was conducted. First, a search of PubMed, PsycINFO, ISI Web of Knowledge, and
Scopus (Elsevier) was performed to identify studies putatively reporting childhood maltreatment
assessed with the Childhood Experience of Care and Abuse (CECA) interview. The search was
conducted between 1st and 28th November 2013 with an update on 5th August 2015, specifying
as the start date of publication the year 1994 (the publication year of the CECA interview). The
following search terms were used: “CECA”, “childhood experience of care and abuse”, and
“childhood abuse”, along with “depression”. Secondly, a database search was performed using
the authors´ names of all articles that were identified within the first step. Thirdly, reference lists
of articles included within the review were manually checked for any studies not retrieved by the
computerized literature search.
Selection criteria
To achieve a high standard of reporting, the PRISMA guidelines were adopted (Moher et al.,
2009). The PRISMA checklist is presented as Appendix 1. Studies were selected for inclusion in
the meta-analysis only when reported as an original research paper in a peer-reviewed journal.
Studies were included independent of the language in which they were published; however, all
proved to have been published in English.
As regard to depression assessment, only studies that assessed depression as a diagnostic
category (rather than depressive symptoms) were included. In addition, the use of interview-
based diagnoses was utilized as a selection criterion. We also included studies independent of the
type of depression assessed (persistent or recurrent) or time of onset. Most studies had assessed
7
major depressive disorder (MDD), recurrent or persistent; only one study had related childhood
maltreatment with the first onset of depression.
As regards the measurement of childhood adversities, the gold standard criterion in this area
of international research has previously been considered the Childhood Experience of Care and
Abuse (CECA; Bifulco et al., 1994) interview. The CECA is a retrospective semi-structured
interview that uses an investigator-based approach to rating. Behavioral indicators of
parent/perpetrator actions instead of the subject's own feelings or reports of incident severity are
taken into account. All scales have 4 points (marked, moderate, some, and little/none), with
repeated incidents of abuse recorded and repeated neglect or antipathy in different household
arrangements, as determined by new parent figures. For certain analyses the scales are
dichotomized, with ‘marked or moderate’ denoting severe abuse or neglect. In most analyses the
‘peak’ severe rating of the neglect or abuse is utilised where there are repeat occurrences.
The core domains are defined as follows:
Antipathy: parental hostility, coldness or rejection shown toward the child, including
scapegoating the child in contrast to treatment of siblings.
Neglect: the failure of parents to provide for the child's basic material needs (food, clothing,
shelter, and protection) and developmental needs (interest in school, friends, child's happiness,
health, and well-being).
Physical Abuse: violence directed towards the child by a household member (including
parents, surrogate parents, siblings, or relatives in the household); inclusion criteria involve
hitting about the head or being hit hard around the body with the hands/fists, being hit with an
implement, kicked, bitten, or burned, or threats or use of a gun or knife, with severity determined
by the intensity of the attack and its frequency.
Sexual Abuse: age-inappropriate sexual contact by any adult or older peer is included,
whether related or not, acquainted or not. Sexual abuse includes a range of sexual contact
including intercourse, violation or penetration with an object, oral sex, touching of
8
breasts/genitals, as well as requiring the child to watch sexual activity or pornography, and
verbal solicitations for sex or age-inappropriate verbal content. Severity is determined by the
extent of sexual contact as well as relationship to perpetrator, with higher ratings given for
family members and trusted authority figures or family friends.
Psychological abuse: episodes of humiliation, terrorization, cognitive disorientation,
exploitation, or corruption of the child or intentional deprivation of needs or valued objects, from
parents usually in the context of a parental, highly controlling and domineering relationship with
the child. Severity is determined by the range of such experiences and their frequency.
The CECA is a reliable measure of childhood experience in both adults and adolescents.
CECA interviewers receive extensive training (see www.cecainterview.com) and ratings have
been demonstrated to have excellent psychometric properties, with satisfactory interrater
reliability and convergent validity between siblings (Bifulco et al., 2002; Bifulco et al., 1997).
Coded variables
Variables for each article included in the meta-analysis were: rates of antipathy, neglect,
physical abuse, sexual abuse, psychological abuse, and a composite index of childhood
maltreatment (defined in most of the included studies as the presence of at least one type of
maltreatment rated as ‘marked’ or ‘moderate’); year of publication; mean age of participants;
gender (% of females); characteristics of sample (whether clinical or community based,
adolescents or adults), and depression assessment.
Database
The literature search yielded a total of 44 potential studies. Of these, 15 were omitted because
of complete or partial sample overlap; in these cases we included the study that offered the larger
subsample or, alternatively, the full report of relevant information. Another 13 studies conducted
the research only with a depressed sample; thus, they were excluded due to a lack of a
comparison group. Finally, three studies were omitted because they did not use a depression
assessment consistent with our selection criteria and one because it was only theoretical. In the
9
end, our search resulted in 12 primary studies with 34 coefficients and a total of 4372
participants (2918 women and 1454 men) (see the PRISMA flow chart in Fig. 1).
Unfortunately, some of the proposed variables were not assessed or reported in various
studies. All studies did provide information about the origin of the sample, prevalence of female
participants, mean age, procedure for depression assessment, and type of sample. However,
concerning childhood experiences of abuse and neglect, most of the studies assessed different
scales of the CECA interview; thus, different meta-analyses were conducted including different
studies (see Table 1).
Insert Figure 1 and Table 1about here
Coding of variables
All studies were coded by the first author. Moreover, a student assistant independently
extracted data from the same eligible articles. Inconsistencies were resolved in consensus
meetings and confirmed with the authors of the primary studies when necessary. The coding
process was standardized by employing detailed coding rules. We coded (a) characteristics of
publication (year of publication), (b) sample information (sample size; country of origin;
percentage of females; mean age, adult or adolescent samples, and clinical or population-based
samples), (c) effect sizes, and (d) the type of abuse. The interrater agreement ranged from 0.89
(for type and characteristics of the samples) to 1.00 (for most of the other variables).
Within the 12 studies included, nine of them used samples from the United Kingdom, one
from Canada, one from the United States of America, and one a culturally mixed sample. Within
the 12 studies included in this meta-analysis, eight studies were conducted with adult samples
and four with adolescent samples. Additionally, eight studies used high-risk or clinical samples
and four studies population-based samples.
Effect sizes were coded for different types of abuse and negect assessed by the CECA or a
composite index of childhood maltreatment. Specifically, nine studies reported information
about the composite index, whereas eight studies reported specific information about the
10
different CECA scales: five studies reported the effects size for antipathy, six for neglect, six for
physical abuse, six for sexual abuse, and only two for psychological abuse. A detailed list of
coded characteristics, percentage of missing data, and interrater agreement is provided in the
Supplemental Material.
Analytic strategy
The standardized mean difference (Cohen’s d) effect size statistic was used in this meta-
analysis. Cohen’s d is particularly useful when comparing the effects of two nonexperimentally
defined groups on an outcome that is not uniformly operationalized across studies (Cohen, 1988;
Lipsey and Wilson, 2000). When a study reported a dichotomous outcome in terms of an odds
ratio, the odds ratio was transformed using the appropriate calculations to make it directly
comparable to the d statistic (Borenstein et al., 2009). Mean effect sizes were estimated by the
meta-analytic procedure, using the techniques outlined by Hunter and Schmidt (Hunter and
Schmidt, 2004) and the associated software package (Schmidt and Lee, 2004).
Subgroups for different types of abuse and neglect and the composite index of maltreatment
were created, therefore enabling us to examine the relationship between different types of
maltreatment and depression. Separate analyses were conducted to determine the specific effects.
In addition to the number of studies included (k), the total sample size (n), and the mean effect
sizes (Means d), the effect sizes’ standard deviations (SDd) of the mean effect size, the 95%
confidence intervals (95%CI) and 80% credibility intervals (80%CV) were calculated to
determine whether the mean effect size was significantly different from zero. We decided to
report also credibility intervals because they express the distribution of effect sizes in a random-
effects model, indicating whether results can be generalized or should be further differentiated
by investigating potential moderators (Hunter and Schmidt, 2004, p. 205). We corrected for
variance due to artifacts, specifically by sampling error, predictor, and criterion unreliability, and
we reported the percentage of variance accounted for by artifacts (%Var). Finally, we reported
the fail-safe N statistic that addressed the problem of publication and availability bias. It
11
estimates the number of nonidentified studies with null effects that would be necessary to reduce
the effect size to a nonsignificant value, defined in this study as an effect size of .05 (Hunter and
Schmidt, 2004, p. 500). According to Rosenthal (Rosenthal, 1991), a file-drawer effect exists
when the fail-safe N is less than five times the number of published studies plus ten. Thus, a low
fail-safe N indicates publication bias.
Additionally, since Hunter and Schmidt have argued that if the percentage of variance is less
than 75%, the search for moderators is recommended (2004; p. 401), we followed this guideline
to determine whether moderator variables existed.
Results
A summary of estimated mean effect sizes and the related statistics is given in Table 2 and
Figure 2. For the relationship between the composite index of childhood maltreatment and
depression (k=9; n=3591), the mean effect size across studies was d=.431 (95%CI=-.316–1.178).
This effect size indicates that there was a small-to-medium association between the composite
index of childhood abuse and depression. However, the confidence intervals for this estimated
associations included zero, so this result could be interpreted as non-significant.
When looking at individual experiences, the mean effect size for the association between
antipathy and depression across studies (k=5; n=842) was d=.513 (95%CI=.201–.829), indicating
that there was a medium association between antipathy and depression. The mean effect sizes for
a relationship between neglect (k=6; n=1040) and physical abuse (k=6; n=1045) and depression
were large, respectively d=.813 (95%CI=.609–1.017) and d=.810 (95%CI=.690–.930). The mean
effect sizes for the association of psychological abuse with depression were significantly
stronger (d=.932; 95%CI=.930–.934) than for the other kind of maltreatment and the composite
index. For the relation between sexual abuse and depression (k=6, n=3120) the mean effect size
across studies was d=.500 (95%CI=.224–.776), indicating that there was a medium association
between sexual abuse and depression. None of the confidence intervals for the estimated
12
associations presented in Figure 2 included zero, so all results could be interpreted as reaching
significance.
Furthermore, file-drawer analyses yielded values for the fail-safe N that largely exceeded the
number of studies included in the meta-analysis (ranging from 35 to 92), indicating that many
unpublished studies would need to exist in order to reduce the mean effect size below the level of
significance (p>.05). Using Rosenthal’s standards, these N indicate that the relationships found
were likely not attributable to publication bias.
Insert Table 2 and Figure 2 about here
Potential moderators
Table 3 presents the results of the analysis addressing whether the association between
specific childhood adversities and depression outcome varied as a function of adult or adolescent
samples. A total of 26 effect sizes were included in this analysis.
By utilizing the composite childhood abuse/neglect index, a stronger association with
depression was found in adolescent samples (d=.766) than in adult samples (d=.399). The fail-
safe N were 43 and 42 respectively for adolescent and adult samples, indicating that the
relationships between adolescents/adults and depression were not attributable to publication bias.
In terms of specific types of maltreatment, antipathy was more strongly related to depression
within adult samples (d=.688) than within adolescent samples (d=.345). However, file-drawer
analyses for adolescent samples yielded a value for the fail-safe N of 12. According to
Rosenthal’s standards, this N indicates that the association found may be attributable to a bias in
the literature toward publishing significant findings. No differences were found for the
association between neglect and depression in adolescent or adult samples, in which d was large
within both kinds of samples (d=.868 and d=.732 respectively). The file-drawer analyses confirm
the validity of these results. With the available data we could not conduct an analysis for sexual
abuse (only one study analyzed sexual abuse in an adolescent sample) or psychological abuse.
13
Additionally, as shown in Table 4, we tested whether the type of sample (clinical/high risk
community vs nonclinical) might be a potential moderator of the relationship between childhood
maltreatment and depression. The results showed that the mean effect size for the composite
index of maltreatment in childhood was larger in clinical/high-risk samples (d=.712) than in
population-based samples (d=.322). However, as showed by the file-drawer analyses, the result
on population-based samples may be affected by a bias in the literature toward publishing
significant findings. Unfortunately, with the available data we could not test whether the type of
sample (clinical/high risk vs nonclinical) moderates the relationships between the specific forms
of maltreatment and depression. Similarly, gender could not be tested as potential moderator
because data were lacking.
Insert Table 3 and Table 4 about here
Discussion
Experiences of abuse and neglect in childhood have been consistently associated with higher
rates of adult depression. However, a full understanding of the relationship between childhood
maltreatment and depression cannot be achieved until different types of abuse and neglect are
differentiated or considered in combination. The present meta-analysis addressed the
heterogeneity in the results of previous studies concerning a relationship between childhood
maltreatment and depression and thereby allowed a differential investigation of different types of
maltreatment experiences.
The association between the composite index of abuse/neglect and depression was moderate,
(d=.43) and, although it failed to reach statistical significance, nearly equal to the overall effect
sizes reported in previous meta-analyses of child abuse and depression (d=.45) (Nanni et al.,
2012). The results are also comparable with a previous meta-analysis on childhood abuse/neglect
and physical health problems (d=.42) (Wegman and Stetler, 2009). Compared to Nanni and
colleagues´ (2012) study, where the association between childhood maltreatment and depressive
disorder was tested in 16 studies with 23,544 participants, our strict selection criteria including
14
the choice of a clinical interview (CECA) resulted in nine studies with a reduced sample size of
3591 participants for the composite index of abuse/neglect. This may have contributed to non-
significant findings despite similar effect sizes. Further, the effects of the nine studies included in
the present analysis showed great heterogeneity, which might also have contributed to non-
significance.
Our findings on specific types of childhood maltreatment revealed that some types were
particularly strongly associated with depression outcome. In particular, and consistent with our
hypothesis, psychological abuse presented a stronger association with depression than other
forms of maltreatment. Psychological abuse as defined by Bifulco and colleagues (2002) is
concerned with cruelty demonstrated by verbal and nonverbal acts from a close other in a
position of power or responsibility for the child and it was associated with feelings of shame in
childhood and depression in adulthood (Bifulco et al., 2002; Moran et al., 2002). This result
confirms an emerging body of literature suggesting a significant relationship between emotional
maltreatment and depression (e.g., Gibb et al., 2001; Kim and Cicchetti, 2006; Liu et al., 2009).
Furthermore, findings from this meta-analysis extend the evidence of prior studies demonstrating
the influence not only of psychological abuse but also of neglect on depressive disorders, here
defined as the degree to which the caretaker does not provide for the child's basic emotional and
material needs. We should highlight here that neglect and psychological abuse likely represent
the two extreme polarities of maltreatment in a child. On the one hand, neglect is the most
relevant form of maltreatment “by omission”, in which the child is deprived of basic responses to
his or her needs of protection, care, and love from caregivers; on the other, psychological abuse
is a perfect representative of maltreatment “by commission”, in which caretakers voluntarily
degrade, humiliate, and terrorize their young in order to have power and control over them. In
both instances this may result in a child’s feeling of powerlessness and reduced self-esteem,
which may easily foster depression in later life. According to attachment theory (Bowlby, 1983),
attachment figures help develop representational models of the relational world. Thus, those who
15
have lived with neglect and/or psychological abuse may be at risk of developing a more negative
self-model, becoming prone to internalizing symptoms (Shapero et al., 2014).
As regards physical and sexual abuse, findings from this meta-analysis may help to clarify the
earlier debate about the specific influence of these forms of abuse on depression. In particular,
the stronger association found between physical abuse and depression confirms the results of a
recent meta-analysis, in which all studies of childhood physical abuse and depression found
increased odd ratios of depression among those reporting physical abuse (Lindert et al., 2014). In
contrast, the association between sexual abuse and depression was not as strong as that with
others forms of maltreatment. This result, although not in contradiction to previous findings
(Cutajar et al., 2010; Dube et al., 2005; Kendler et al., 2000; Molnar et al., 2001), highlights the
importance of not focusing on one form of abuse alone, e.g. sexual abuse, but rather on a broad
variety of adverse childhood experiences, which might better explain the potential early
pathways that lead to depression later in life.
The effects of specific childhood maltreatments were not equally large across adult and
adolescent samples. The types of samples (adult or adolescent) moderated the effect sizes
between the composite index of maltreatment and depression (larger in adolescent samples) and
between antipathy and depression (larger in adult samples). The effect of neglect and physical
abuse was similarly large in both kinds of samples.
These findings of the composite index suggest that adolescence may be a sensitive
developmental period during which several experiences of maltreatment need to be assessed,
given the profound influences on the risk for depression (Lumley and Harkness, 2007). It also
points to the greater recency of the experience of maltreatment for adolescents, in whom
subsequent protective factors have not yet emerged as in adult samples. One possible explanation
for the stronger association between antipathy and depression observed in adult samples than in
adolescents might be related to the nature of this maltreatment. Indeed, according to Bifulco and
colleagues’ definition (1994), antipathy reflects criticism, dislike, coldness, rejection, or hostility
16
shown by a parent in the household towards the child; thus, it regards the daily parent-child
interaction and not single or specific episodes of abuse. One could argue that during adolescence,
in which the parent-child relationship is still ongoing, individuals struggle to adequately reflect
on the idea of parental antipathy and have difficulties realizing the pernicious nature of this kind
of experience. Another hypothesis could be that antipathy may be more prone to recall-bias than
other types of maltreatment and therefore shows larger effects among adults. However, these
data resulted from a small number of studies and the findings of the moderator analyses
concerning sample characteristic need to be interpreted with caution. Indeed, regarding the
specific association between antipathy and depression in adolescent samples, there was a low
fail-safe N, that indicates that this finding may be affected by publication bias. Thus, it may limit
the conclusion that can be drawn from this moderator.
Using a type of sample (clinical/high risk vs nonclinical) variable as moderator resulted in a
higher association between the composite index of childhood maltreatment and depression for
clinical/high risk than for nonclinical samples. This result is in contrast to the results of a
previous meta-analysis (Nanni et al., 2012), in which studies in population samples showed a
stronger association between childhood maltreatment and depression than studies in clinical
samples (odds ratio=2.75 versus odds ratio=1.78). This may occur because a different distinction
was made in this analysis – both clinical samples and high risk community samples selected for
vulnerability factors were included. The latter may vary from clinical samples in terms of aspects
of disorder (comorbidity, personality disorder), but also factors associated with service-
utilisation. Thus, further studies are needed to allow further interpretations about this result in
terms of risk and clinical selection criteria.
The fail-safe N was high for most of the other meta-analytic results, indicating that the
relationships between abuse/maltreatment and depression detected are not attributable to a bias
in the literature toward publishing significant findings.
17
Our meta-analysis revealed that the association between the composite index of maltreatment
and depression was lower than all associations of specific experiences of maltreatment and
depression. Additionally, this association was the only one in which the CI included zero. This
finding is a likely subject to bias and might be caused by the greater heterogeneity (adolescent or
adult, clinical or nonclinical) of the nine studies included in the analysis of the composite index.
Indeed, as discussed above, the moderation analyses showed that associations with depression
were higher in adolescent samples than in adults. Additionally, using a type of sample
(clinical/high risk vs nonclinical) variable as moderator resulted in a higher association for
clinical/high risk than for nonclinical samples, and indicated potential publication-bias for non-
clinical samples. In addition, studies with weak effect sizes tended to only report the composite
index, which may lead to a selective underestimation of this effect.
Limitations and Strengths
The results of this meta-analysis should be evaluated in the context of several potential
limitations. First, assessing the type of abuse and neglect presents numerous problems. Several
authors have suggested that multiple forms of abuse are likely to occur together (Finkelhor et al.,
2007; Wolfe and McGee, 1994). Since it was not known whether the studies included individuals
who exclusively experienced one form of abuse or individuals who experienced multiple forms
of abuse, conclusions about the unique contributions of each type of abuse from these results
would be hasty.
Furthermore, some studies included in our meta-analysis did not report detailed data on
important aspects of abuse, such as frequency, duration, or severity. However, the CECA´s
ranking rules take into account several characteristic of abuse (e.g., severity, frequency, duration,
and people involved) to determine the rates, and the CECA is only used after prior training for
reliability. Thus, we can suppose that the experiences of maltreatment reported in the studies
included in the present meta-analysis should be comparable in regard to the severity of childhood
adversity experienced by the participants. Indeed, although it limited the studies eligible for the
18
present meta-analysis, our strict criteria for selecting the primary studies should guarantee a
reliable comparison of childhood abuse and neglect among different studies, thus increasing the
validity of our results.
Finally, the problem of small numbers of available studies restricted the analyses for various
potential moderators. It is likely that the search for other potential moderators, such as clinical
versus population-based samples or gender differences, might produce more detailed findings.
Unfortunately, with the available data we could not examine the effect of other potential
confounders. These results should therefore be interpreted as preliminary insights into what
could be promising directions for future research.
Research and clinical implications
Overall, the results of this meta-analysis point to the importance of considering several types
of maltreatment experiences as risk factors for an outcome of depression with a particular focus
on the more “silent” forms of maltreatment such as psychological abuse and neglect. This meta-
analysis adds information to the relationship between child maltreatment and depression, but it
also highlights potential gaps in the literature that need to be addressed. Future research should
focus on experiences of emotional or psychological abuse and neglect since the number of
studies were still small for detailed analyses. Further research should also include a more
representative sample of the population, including larger numbers of males and older adults.
Additionally, more information on other potential moderators, such as age at time of
maltreatment, duration of maltreatment, and severity of abuse/neglect, should be included. This
would help better characterize the circumstances under which and the individuals for whom a
greater vulnerability for developing depression as an effect of childhood maltreatment exists.
Finally, a number of clinical implications of these findings should be highlighted. Information
about the specific history of childhood maltreatment may help to identify individuals who are at
high risk of developing depression. Clinicians may consider that a routine inquiry concerning
childhood maltreatment could add important prognostic information to their assessment, and this
19
enquiry needs to go beyond the assessment of physical and sexual abuse. These results suggest
that a history of psychological maltreatment may be an important marker in targeting depression
prevention efforts in populations.
Reference
Studies preceded by an * were included in the meta-analysis.
Abela, J., Skitch, S., 2007. Dysfunctional attitudes, self-esteem, and hassles: Cognitive
vulnerability to depression in children of affectively ill parents. Behav. Res. Ther. 45, 1127–
1140.
Alloy, A., Abramson, L., Smith, J., et al, 2006. Role of parenting and maltreatment histories in
unipolar and bipolar mood disorders: Mediation by cognitive vulnerability to depression. Clin
Child Fam Psych. 9, 23–64.
Beck, A.T., 2008. The evolution of the cognitive model of depression and its neurobiological
correlates. American Journal of Psychiatry 165, 969–977.
*Bifulco, A., Bernazzani, O., Moran, P.M., Ball, C., 2000. Lifetime stressors and recurrent
depression: preliminary findings of the Adult Life Phase Interview (ALPHI). Soc. Psychiatry
Psychiatr. Epidemiol. 35, 264–275.
Bifulco, A., Brown, G.W., Harris, T.O., 1994. Childhood Experience of Care and Abuse
(CECA): a retrospective interview measure. J. Child Psychol. Psychiatry 35, 1419–1435.
Bifulco, A., Brown, G.W., Lillie, A., Jarvis, J., 1997. Memories of childhood neglect and abuse:
corroboration in a series of sisters. J. Child Psychol. Psychiatry 38, 365–374.
Bifulco, A., Brown, G.W., Moran, P., Ball, C., Campbell, C., 1998. Predicting depression in
women: the role of past and present vulnerability. Psychol. Med. 28, 39–50.
*Bifulco, A., Moran, P., Jacobs, C., Bunn, A., 2009. Problem partners and parenting: exploring
linkages with maternal insecure attachment style and adolescent offspring internalizing
disorder. Attach. Hum. Dev. 11, 69–85.
20
*Bifulco, A., Moran, P.M., Baines, R., Bunn, A., Stanford, K., 2002. Exploring psychological
abuse in childhood: II. Association with other abuse and adult clinical depression. Bull.
Menninger Clin. 66, 241–258.
*Bifulco, A., Moran, P.M., Ball, C., Jacobs, C., Baines, R., Bunn, A., Cavagin, J., 2002.
Childhood adversity, parental vulnerability and disorder: examining inter-generational
transmission of risk. J. Child Psychol. Psychiatry 43, 1075–1086.
Bifulco, A., Schimmenti, A., Moran, P., Jacobs, C., Bunn, A., Rusu, A.C., 2014. Problem
parental care and teenage deliberate self-harm in young community adults. Bull. Menninger
Clin. 78, 95–114.
Borenstein, M., Hedges, L.V., Higgins, J., Rothstein, H.R., 2009. Introduction to Meta-Analysis,
1 ed. John Wiley & Sons, Chichester, U.K.
Bowlby, J., 1983. Attachment: Attachment and Loss Volume One, 2nd edition. ed. Basic Books,
New York.
*Brown, G.W., Craig, T.K.J., Harris, T.O., Handley, R.V., Harvey, A.L., Serido, J., 2007. Child-
specific and family-wide risk factors using the retrospective Childhood Experience of Care &
Abuse (CECA) instrument: a life-course study of adult chronic depression - 3. J. Affect.
Disord. 103, 225–236.
*Brown, G.W., Moran, P., 1994. Clinical and psychosocial origins of chronic depressive
episodes. I: A community survey. Br. J. Psychiatry J. Ment. Sci. 165, 447–456.
Brown, J., Cohen, P., Johnson, J.G., Smailes, E.M., 1999. Childhood abuse and neglect:
specificity of effects on adolescent and young adult depression and suicidality. J. Am. Acad.
Child Adolesc. Psychiatry 38, 1490–1496.
Chen, J., Cai, Y., Cong, et al. 2014. Childhood sexual abuse and the development of recurrent
major depression in Chinese women. PloS One 9, e87569.
Cohen, J., 1988. Statistical Power Analysis for the Behavioral Sciences, 2 edition. ed. Routledge,
Hillsdale, N.J.
21
Cutajar, M.C., Mullen, P.E., Ogloff, J.R.P., Thomas, S.D., Wells, D.L., Spataro, J., 2010.
Psychopathology in a large cohort of sexually abused children followed up to 43 years. Child
Abuse Negl. 34, 813–822.
Dube, S.R., Anda, R.F., Whitfield, C.L., Brown, D.W., Felitti, V.J., Dong, M., Giles, W.H.,
2005. Long-term consequences of childhood sexual abuse by gender of victim. Am. J. Prev.
Med. 28, 430–438.
Dunn, E.C., McLaughlin, K.A., Slopen, N., Rosand, J., Smoller, J.W., 2013. Developmental
timing of child maltreatment and symptoms of depression and suicidal ideation in young
adulthood: results from the National Longitudinal Study of Adolescent Health. Depress.
Anxiety 30, 955–964.
Fergusson, D.M., Boden, J.M., Horwood, L.J., 2008. Exposure to childhood sexual and physical
abuse and adjustment in early adulthood. Child Abuse Negl. 32, 607–619.
Finkelhor, D., Ormrod, R.K., Turner, H.A., 2007. Polyvictimization and trauma in a national
longitudinal cohort. Dev. Psychopathol. 19, 149–166.
Gibb, B.E., Alloy, A., Abramson, L., 2001. History of childhood maltreatment, depressogenic
cognitive style, and episodes of depression in adulthood. Cogn. Ther Res 25, 425–446.
Gibb, B.E., Butler, A.C., Beck, J.S., 2003. Childhood abuse, depression, and anxiety in adult
psychiatric outpatients. Depress. Anxiety 17, 226–228.
*Harkness, K.L., Bruce, A.E., Lumley, M.N., 2006. The role of childhood abuse and neglect in
the sensitization to stressful life events in adolescent depression. J. Abnorm. Psychol. 115,
730–741.
*Hill, J., Pickles, A., Burnside, E., Byatt, M., Rollinson, L., Davis, R., Harvey, K., 2001. Child
sexual abuse, poor parental care and adult depression: evidence for different mechanisms. Br.
J. Psychiatry J. Ment. Sci. 179, 104–109.
Hunter, J.E., Schmidt, F.L., 2004. Methods of Meta-Analysis: Correcting Error and Bias in
Research Findings, 2nd edition. ed. SAGE Publications, Inc, Thousand Oaks, Calif.
22
Kendler, K.S., Bulik, C.M., Silberg, J., Hettema, J.M., Myers, J., Prescott, C.A., 2000.
Childhood sexual abuse and adult psychiatric and substance use disorders in women: an
epidemiological and cotwin control analysis. Arch. Gen. Psychiatry 57, 953–959.
Kim, J., Cicchetti, D., 2006. Longitudinal trajectories of self-system processes and depressive
symptoms among maltreated and nonmaltreated children. Child Dev. 77, 624–639.
*Lenze, S.N., Xiong, C., Sheline, Y.I., 2008. Childhood adversity predicts earlier onset of major
depression but not reduced hippocampal volume. Psychiatry Res. 162, 39–49.
Lindert, J., von Ehrenstein, O.S., Grashow, R., Gal, G., Braehler, E., Weisskopf, M.G., 2014.
Sexual and physical abuse in childhood is associated with depression and anxiety over the life
course: systematic review and meta-analysis. Int. J. Public Health 59, 359–372.
Lipsey, M.W., Wilson, D., 2000. Practical Meta-Analysis, 1 edition. ed. SAGE Publications, Inc,
Thousand Oaks, Calif.
Liu, R.T., Alloy, L.B., Abramson, L.Y., Iacoviello, B.M., Whitehouse, W.G., 2009. Emotional
maltreatment and depression: prospective prediction of depressive episodes. Depress. Anxiety
26, 174–181.
Lumley, M.N., Harkness, K.L., 2007. Specificity in the Relations among Childhood Adversity,
Early Maladaptive Schemas, and Symptom Profiles in Adolescent Depression. Cogn Ther Res
31.
MacMillan, H.L., Fleming, J.E., Streiner, D.L., et al., 2001. Childhood abuse and lifetime
psychopathology in a community sample. Am. J. Psychiatry 158, 1878–1883.
Moher, D., Liberati, A., Tetzlaff, J., Altman, D.G., PRISMA Group, 2009. Preferred reporting
items for systematic reviews and meta-analyses: the PRISMA statement. J. Clin. Epidemiol.
62, 1006–1012.
Molnar, B.E., Buka, S.L., Kessler, R.C., 2001. Child sexual abuse and subsequent
psychopathology: results from the National Comorbidity Survey. Am. J. Public Health 91,
753–760.
23
Moran, P.M., Bifulco, A., Ball, C., Jacobs, C., Benaim, K., 2002. Exploring psychological abuse
in childhood: I. Developing a new interview scale. Bull. Menninger Clin. 66, 213–240.
Moretti, M.M., Craig, S.G., 2013. Maternal versus paternal physical and emotional abuse, affect
regulation and risk for depression from adolescence to early adulthood. Child Abuse Negl. 37,
4–13.
Musliner, K.L., Singer, J.B., 2014. Emotional support and adult depression in survivors of
childhood sexual abuse. Child Abuse Negl. 38, 1331–1340.
Nanni, V., Uher, R., Danese, A., 2012. Childhood maltreatment predicts unfavorable course of
illness and treatment outcome in depression: a meta-analysis. Am. J. Psychiatry 169, 141–
151.
O’Mahen, H, Karl, A, Moberly, N, Fedock, G, 2015. The association between childhood
maltreatment and emotion regulation: Two different mechanisms contributing to depression?
J Affect Disord. 174, 287–295.
*Pickles, A., Aglan, A., Collishaw, S., Messer, J., Rutter, M., Maughan, B., 2010. Predictors of
suicidality across the life span: the Isle of Wight study. Psychol. Med. 40, 1453–1466.
Rosenthal, R., 1991. Meta-analysis: a review. Psychosom. Med. 53, 247–271.
Schimmenti, A., Bifulco, A., 2015. Linking lack of care in childhood to anxiety disorders in
emerging adulthood: the role of attachment styles. Child Adolesc Ment Health, 20, 41–48.
Schmidt FL, Lee HA, 2004. Software for the Hunter-Schmidt meta-analysis methods.,
Department of Management & Organization, Iowa City: ed. University of Iowa.
Shapero, B.G., Black, S.K., Liu, R.T., Klugman, J., Bender, R.E., Abramson, L.Y., Alloy, L.B.,
2014. Stressful life events and depression symptoms: the effect of childhood emotional abuse
on stress reactivity. J. Clin. Psychol. 70, 209–223.
Thabrew, H., de Sylva, S., Romans, S.E., 2012. Evaluating childhood adversity. Adv.
Psychosom. Med. 32, 35–57.
24
*Tousignant, M., Habimana, E., Biron, C., Malo, C., Sidoli-LeBlanc, E., Bendris, N., 1999. The
Quebec Adolescent Refugee Project: psychopathology and family variables in a sample from
35 nations. J. Am. Acad. Child Adolesc. Psychiatry 38, 1426–1432.
van Veen, T., Wardenaar, K.J., Carlier, I.V., Spinhoven, P., Penninx, B.W., Zitman, F.G., 2013.
Are childhood and adult life adversities differentially associated with specific symptom
dimensions of depression and anxiety? Testing the tripartite model. J Affect Disord 146, 238–
45.
*Webster, J.J., Palmer, R.L., 2000. The childhood and family background of women with
clinical eating disorders: a comparison with women with major depression and women
without psychiatric disorder. Psychol. Med. 30, 53–60.
Wegman, H.L., Stetler, C., 2009. A meta-analytic review of the effects of childhood abuse on
medical outcomes in adulthood. Psychosom. Med. 71, 805–812.
Widom, C.S., DuMont, K., Czaja, S.J., 2007. A prospective investigation of major depressive
disorder and comorbidity in abused and neglected children grown up. Arch. Gen. Psychiatry
64, 49–56.
Wolfe D, McGee R., 1994. Dimensions of child maltreatment and their relationship to adolescent
adjustment. Dev. Psychopathol. 6, 165–81.
Ystgaard, M., Hestetun, I., Loeb, M., Mehlum, L., 2004. Is there a specific relationship between
childhood sexual and physical abuse and repeated suicidal behavior? Child Abuse Negl. 28,
863–875.
Table 1. Summary of studies and their characteristics
Author(s)
and
Country
Sample
Type of
sample
N
Mean
Age or
%
Female
Depression
Measure
Type of childhood
adversities
25
publication
year
Range
(Years)
assessed
Brown &
Moran
(1994)
England
Adult
High-risk or
clinical
404
18-50
100%
PSE
Composite index of
childhood
maltreatment*
Tousignant
et al. (1999)
Mixed
Adolescent
High-risk or
clinical
203
15.7
51.7%
DISC
Physical abuse
Bifulco et al.
(2000)
England
Adult
High-risk or
clinical
198
33.5
100%
PSE &
SCAN
Composite index of
childhood
maltreatment
Webster &
Palmer
(2000)
England
Adult
High-risk or
clinical
160
34
100%
Clinical
assessment
Antipathy, Neglect,
Physical and Sexual
abuse, Composite
index of childhood
maltreatment
Hill et al.
(2001)
England
Adult
Population-
based
198
30.8
100%
SADS
Neglect and Sexual
abuse
Bifulco et al.
(2002)
England
Adolescent
High-risk or
clinical
277
20
49.5%
PSE & SCID
Antipathy, Neglect,
Physical and Sexual
Abuse, Composite
index of childhood
maltreatment
Bifulco et al.
(2002)
England
Adult
High-risk or
clinical
204
35
100%
PSE &
SCAN
Antipathy, Neglect,
Physical, Sexual,
and Psychological
abuse
Harkness et
al. (2006)
Canada
Adolescent
Population-
based
103
15.5
65%
K-SADS
Composite index of
childhood
maltreatment
Author(s)
and
publication
Country
Sample
Type of
sample
N
Mean
Age or
Range
%
Female
Depression
Measure
Type of childhood
adversities assessed
26
year
(Years)
Brown et al.
(2007)
England
Adult
Population-
based
198
34
100%
SCAN
Composite index of
childhood
maltreatment
Lenze et al.
(2008)
USA
Adult
High-risk or
clinical
55
48.5
100%
DIGS
Antipathy, Neglect,
Physical, Sexual,
and Psychological
abuse, Composite
index of childhood
maltreatment
Bifulco et al.
(2009)
England
Adolescent
High-risk or
clinical
146
20.7
53.4
SCID
Antipathy, Neglect,
Physical abuse,
Composite index of
childhood
maltreatment
Pickles et
al. (2010)
England
Adult
Population-
based
2226
44.6
50%
SADS
Sexual abuse and
Composite index of
childhood
maltreatment
Note: PSE - Present State Examination (Wing et al, 1974); DISC – Diagnostic Interview Schedule for
Children Version 2.25 (Breton et al., 1995); SCAN – Schedule for Clinical Assessment in
Neuropsychiatry (Wing et al., 1990); SADS – Schedule for Affective Disorders and Schizophrenia
(Spitzer & Endicott, 1975); SCID - Structured Clinical Interview for DSM-IV (First, Gibbon, Spitzer, &
Williams, 1996); K-SADS - Schedule for Affective Disorders and Schizophrenia child and adolescent
version (Kaufman, Birmaher, Brent, Rao, & Ryan, 1996); DIGS - Diagnostic Interview for Genetic
Studies (Nurnberger et al., 1994).
*Severe neglect, or physical or sexual abuse.
27
Table 2. Meta-analytic results of the relations between childhood adversities and depression
Variable
k
N
Mean
d
SD
d
95% CI
80% CV
% Var
Nfs
File-
Drawer
Effect?
Composite index of childhood
maltreatment *
9
3591
.431
.381
-.316 –
1.178
-.057 –
.919
6.62
69
No
Antipathy
5
842
.515
.160
.201 –
.829
.311 –
.720
49.07
47
No
Neglect
6
1040
.813
.104
.609 –
1.017
.679 –
.947
69.70
92
No
Physical abuse
6
1045
.810
.061
.690 –
.930
.732 –
.887
87.08
91
No
Psychological abuse
2
259
.932
.000
.930 –
.934
.932 –
.932
100.00
35
No
Sexual abuse
6
3120
.500
.141
.224 –
.776
.320 –
.681
28.58
54
No
Note. k = number of samples; N = total sample size; Mean d = estimate of the true score effect sizes
corrected for artifacts; SD d = Standard deviation of true score effect size; 95% CI = 95% confidence
interval; 80% CV = 80% credibility interval; % Var = percentage of variance due to artifacts; Nfs = fail-
safe n (for the reduction of d to a trivial effect of .05).
*Severe neglect, or physical or sexual abuse.
Table 3. Meta-analytic results of specific adult or adolescent samples among studies of
childhood adversities and depression
Variable
k
N
Mean
d
SD
d
95% CI
80% CV
% Var
Nfs
File-
Drawer
Effect?
Composite index of childhood
28
maltreatment*
Adolescent
3
526
.766
.076
.617 –
.915
.670 –
.863
81.11
43
No
Adult
6
3241
.399
.388
-.361 –
1.159
-.097 –
.896
4.79
42
No
Antipathy
Adolescent
2
423
.345
.000
.343 –
.347
.345 –
.345
100.00
12
Yes
Adult
3
419
.688
.104
.484 –
.892
.554 –
.821
73.71
38
No
Neglect
Adolescent
2
423
.732
.000
.730 –
.734
.732 –
.732
100.00
27
No
Adult
4
617
.868
.156
.562 –
1.174
.668 –
.1.068
53.80
65
No
Physical abuse
Adolescent
3
626
.871
.311
.261 –
1.481
.832 –
.911
95.62
49
No
Adult
3
419
.717
.000
.715 –
.719
.717 –
.717
100.00
40
No
Note. k = number of samples; N = total sample size; Mean d = estimate of the true score effect sizes
corrected for artifacts; SD d = Standard deviation of true score effect size; 95% CI = 95% confidence
interval; 80% CV = 80% credibility interval; % Var = percentage of variance due to artifacts; Nfs = fail-
safe n (for the reduction of d to a trivial effect of .05).
*Severe neglect, or physical or sexual abuse
Table 4. Meta-Analytic results of clinical/high-risk or population-based samples among studies
of childhood adversities and depression
29
Variable
k
N
Mean
d
SD
d
95% CI
80% CV
%
Var
Nfs
File-
Drawer
Effect?
Composite index of childhood
maltreatment*
Clinical or high-risk samples
6
1240
.712
.249
.224 –
1.200
.394 –
1.030
25.06
79
No
Population-based samples
3
2527
.322
.371
-.405 –
1.049
-.153 –
.797
3.38
16
Yes
Note. k = number of samples; N = total sample size; Mean d = estimate of the true score effect sizes
corrected for artifacts; SD d = Standard deviation of true score effect size; 80% CV = 80% credibility
interval; % Var = percentage of variance due to artifacts; Nfs = fail-safe n (for the reduction of d to a
trivial effect of .05).
*Severe neglect or physical or sexual abuse
highlights
Associations between depression and specific types of maltreatment remain ambiguous
Various definitions and measures of maltreatment reduce comparability of studies
This meta-analysis using the CECA interview can overcome these limitations
Psychological abuse followed by neglect were the strongest correlates of depression
Those more “silent” types of childhood maltreatment require more clinical attention
30
31
