Burden attributable to child maltreatment in Australia

Child Abuse & Neglect 48 (2015) 208–220

Contents lists available at ScienceDirect

Child Abuse & Neglect

Research article

Burden attributable to child maltreatment in Australia Sophie E. Moore a , James G. Scott b,c , Alize J. Ferrari a,d,e , Ryan Mills f,g , Michael P. Dunne h , Holly E. Erskine a,d,e , Karen M. Devries i , Louisa Degenhardt e,j,k , Theo Vos e , Harvey A. Whiteford a,d,e , Molly McCarthy l , Rosana E. Norman m,n,∗ a

School of Public Health, University of Queensland, Herston, QLD, Australia Metro North Mental Health, Royal Brisbane and Women’s Hospital, Herston, QLD 4029 Australia c The University of Queensland Centre for Clinical Research, Herston, QLD 4029 Australia d Queensland Centre for Mental Health Research, Wacol, QLD, Australia e Institute for Health Metrics and Evaluation, University of Washington, Seattle, WA, USA f Department of Paediatrics, Logan Hospital, Metro South Hospital and Health Service, QLD, Australia g School of Medicine, University of Queensland, Herston, QLD, Australia h Children and Youth Research Centre, School of Public Health and Social Work, Queensland University of Technology, QLD, Australia i Gender Violence and Health Centre, London School of Hygiene and Tropical Medicine, London, United Kingdom j University of New South Wales, National Drug and Alcohol Research Centre, Sydney, NSW, Australia k Melbourne School of Population and Global Health, University of Melbourne, Melbourne VIC 3010, Australia l School of Criminology and Criminal Justice, Griffith University, Australia m Institute of Health and Biomedical Innovation, Queensland University of Technology, QLD, Australia n School of Public Health and Social Work, Queensland University of Technology, QLD, Australia b

a r t i c l e

i n f o

Article history: Received 5 February 2015 Received in revised form 1 May 2015 Accepted 6 May 2015 Available online 6 June 2015 Keywords: Child maltreatment Australia Child sexual abuse Child physical abuse Child emotional abuse Child neglect Depressive disorders Anxiety disorders Intentional self-harm

a b s t r a c t Child maltreatment is a complex phenomenon, with four main types (childhood sexual abuse, physical abuse, emotional abuse, and neglect) highly interrelated. All types of maltreatment have been linked to adverse health consequences and exposure to multiple forms of maltreatment increases risk. In Australia to date, only burden attributable to childhood sexual abuse has been estimated. This study synthesized the national evidence and quantified the burden attributable to the four main types of child maltreatment. Meta-analyses, based on quality-effects models, generated pooled prevalence estimates for each maltreatment type. Exposure to child maltreatment was examined as a risk factor for depressive disorders, anxiety disorders and intentional self-harm using counterfactual estimation and comparative risk assessment methods. Adjustments were made for co-occurrence of multiple forms of child maltreatment. Overall, an estimated 23.5% of self-harm, 20.9% of anxiety disorders and 15.7% of depressive disorders burden in males; and 33.0% of self-harm, 30.6% of anxiety disorders and 22.8% of depressive disorders burden in females was attributable to child maltreatment. Child maltreatment was estimated to cause 1.4% (95% uncertainty interval 0.4–2.3%) of all disability-adjusted life years (DALYs) in males, and 2.4% (0.7–4.1%) of all DALYs in females in Australia in 2010. Child maltreatment contributes to a substantial proportion of burden from depressive and anxiety disorders and intentional self-harm in Australia. This study demonstrates the importance of including all forms of child maltreatment as risk factors in future burden of disease studies. © 2015 Elsevier Ltd. All rights reserved.

∗ Corresponding author at: Institute of Health and Biomedical Innovation, School of Public Health and Social Work, Queensland University of Technology, 60 Musk Ave, Kelvin Grove, QLD 4059, Australia. http://dx.doi.org/10.1016/j.chiabu.2015.05.006 0145-2134/© 2015 Elsevier Ltd. All rights reserved.

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

209

Introduction Child maltreatment is a serious public health problem in Australia and worldwide, adversely affecting short- and longterm physical and mental health (Chen et al., 2010; Norman et al., 2012), social development and functioning (Alink, Cicchetti, Kim, & Rogosch, 2012; Kim & Cicchetti, 2003). Maltreatment of children is commonly divided into four main types: childhood sexual abuse (CSA), childhood physical abuse (CPA), childhood emotional abuse (CEA), and childhood neglect (CN) (Butchart, Phinney Harvey, Kahane, Mian, & Furniss, 2006). The definition of child maltreatment varies within Australian studies and globally, along with the definition of what constitutes a child (under 15, 16 or 18 years of age). These sources of variation, as well as other methodological factors such as the mode of data collection and type of sample assessed, lead to differences in the reported prevalence (Andrews, Corry, Slade, Issakidis, & Swanston, 2004). For instance, meta-analyses of CSA prevalence around the world have made use of estimates ranging from 8% to 31% in females and 3% to 17% in males; with pooled self-reported prevalence estimated higher in females (approximately 18% for females and 8% for males) (Barth, Bermetz, Heim, Trelle, & Tonia, 2013; Pereda, Guilera, Forns, & Gómez-Benito, 2009; Stoltenborgh, van Ijzendoorn, Euser, & Bakermans-Kranenburg, 2011). The prevalence of other forms of child maltreatment has not been as extensively investigated; however, available literature suggests that every year 4–16% of children are physically abused and 10% emotionally abused or neglected in high income countries (Gilbert et al., 2009). In recent meta-analyses of worldwide prevalence, the overall estimate was 23% for studies using self-report measures of CPA, with no apparent gender differences, (Stoltenborgh, Bakermans-Kranenburg, van IJzendoorn, & Alink, 2013) and 36% for studies using self-report measures of CEA (Stoltenborgh, Bakermans-Kranenburg, Alink, & van IJzendoorn, 2012). Regarding CN, estimated global prevalence based on a modest number of studies was 16% for physical neglect and 18% for emotional neglect with no apparent gender differences (Stoltenborgh, BakermansKranenburg, & van IJzendoorn, 2013). Currently there are no Australia-wide studies of the prevalence of child maltreatment. National Child Protection data exist, but these likely under-represent the true number of children experiencing maltreatment as they only capture cases where maltreatment has been reported to authorities, it can be substantiated, and the risk of harm to the child is deemed to be sufficiently high to justify intervention (Australian Institute of Health and Welfare, 2014). Empirical evidence has linked all forms of child maltreatment with adverse mental and physical health outcomes (Chen et al., 2010; Norman et al., 2012). Experiencing at least one form of child maltreatment may double the risk of developing mental health problems such as depressive and anxiety disorders, and increase the risk of self-harm and suicidal ideation and attempts (Afifi et al., 2008; Chen et al., 2010; Devries et al., 2014; Norman et al., 2012). The four types of child maltreatment are highly interrelated (Dong et al., 2004; Finkelhor, 2008; Finkelhor, Ormrod, & Turner, 2007; Jirapramukpitak, Prince, & Harpham, 2005; Mullen, Martin, Anderson, Romans, & Herbison, 1996; Teicher, Samson, Polcari, & McGreenery, 2006). In a cross-sectional study of the effect of multiple types of maltreatment in children and adolescents in Viet Nam, substantial proportions of participants reported exposure to multiple abuse (21% reported two types; 15% three types, and 6% four types of maltreatment) (Nguyen, Dunne, & Le, 2010). Similar levels of multi-type maltreatment have been found in Malaysia (Choo, Dunne, Marret, Fleming, & Wong, 2011). Exposure to multiple forms of maltreatment increases the child’s risk of developing later mental health problems (Edwards, Holden, Felitti, & Anda, 2003; Holt, Finkelhor, & Kantor, 2007; Rikhye et al., 2008; Schneider, Baumrind, & Kimerling, 2007; Teicher et al., 2006; Turner, Finkelhor, & Ormrod, 2006; Zoroglu et al., 2003). In Australia, as in most countries, there has not been a comprehensive assessment of the health consequences of child maltreatment at the national level. CSA was the only form of maltreatment included as a risk factor in the Australian Burden of Disease 2003 study (Begg et al., 2008) and in the Global Burden of Disease 2010 study (GBD 2010) (Lim et al., 2012). The omission of other types of child maltreatment limits our understanding of how these complex phenomena relate to each other, and how co-occurrence of multiple types may influence the overall burden. This is also an important gap in the international literature (Fang et al., 2015). In addition, to date, studies estimating the economic burden of child maltreatment (Fang, Brown, Florence, & Mercy, 2012; Fang et al., 2015) have not adjusted for co-occurrence of maltreatment types. The present analysis aims to derive estimates of the burden of depressive disorders, anxiety disorders and intentional self-harm attributable to CSA, CPA, CEA and CN in Australia adjusting for co-occurrence of multiple forms of child maltreatment. It is hoped that this will lead to quantification of the burden attributable to all forms of child maltreatment in future iterations of the Australian and global burden of disease studies and inform estimation of the economic impact of child maltreatment.

Methods Exposure to child maltreatment was treated as a risk factor for disease and injury using counterfactual estimation and comparative risk assessment methods (Lim et al., 2012). This involved comparing the current local health status with the theoretical-minimum-risk exposure defined as a population not ever having been exposed to child maltreatment. Population attributable fractions (PAFs) were determined by the prevalence of ever having been exposed to these risk factors in the population and the relative risks (RR) of disease occurrence given exposure.

210

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

Risk Factor Definitions CSA was defined as the involvement in sexual activity that a child does not fully comprehend, is unable to give informed consent to, or for which the child is not developmentally prepared (Butchart et al., 2006). As per GBD 2010 (Lim et al., 2012), in this study CSA was defined as ever having experienced any form of CSA and hence would encompass all sub-types including unwanted non-contact, contact abuse, or intercourse (penetrative sexual abuse) in childhood. CPA was defined as the intentional use of physical force against a child that results in (or has a high likelihood of resulting in) harm to the child’s health, survival, development or dignity. This included hitting, beating, kicking, shaking, biting, strangling, scalding, burning, poisoning, and suffocating the child (Butchart et al., 2006). CEA involved both isolated incidents as well as a pattern of failure over time on the part of the parent or caregiver to provide a developmentally appropriate and supportive environment to the child. This included the restriction of movement, patterns of belittling, blaming, threatening, frightening, discriminating against or ridiculing and other non-physical forms of rejection or hostile treatment (Butchart et al., 2006; Hamarman & Bernet, 2000). CN included both isolated incidents as well as a pattern of failure over time on the part of a parent or caregiver to provide for the development and wellbeing of the child (where the parent is in a position to do so) in one or more of the following areas: health, education, emotional development, nutrition, shelter and safe living conditions (Butchart et al., 2006). Prevalence of Exposure We conducted a systematic review in line with the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines (Moher, Liberati, Tetzlaff, & Altman, 2009) to identify studies that provided data on the prevalence of exposure to CSA, CPA, CEA, or CN in Australia. Three electronic databases (PubMed, EMBASE, and PsycINFO) were searched up to 15 August 2014 using the terms, ‘child*’ along with ‘maltreatment’, ‘sexual abuse’, ‘emotional abuse’, ‘neglect’, ‘physical abuse’ as well as ‘Australia*’ and ‘prevalence’ (Supplementary Data Text S1). This systematic review incorporated studies meeting the following inclusion criteria: (1) the study reported original, empirical published research; (2) the study presented lifetime prevalence data (proportion of the population who have ever experienced one or more of the four sub-categories of child maltreatment in childhood); (3) maltreatment was reported by victims rather than informants (health professionals, teachers, family members); (4) the study was conducted in Australia and was largely representative of the Australian population. We included studies with different definitions of childhood (ranging from under 15 to under 18 years) as well as studies where childhood was not defined (e.g. maltreatment occurs “in childhood”). Studies of child maltreatment conducted in justice system, high risk or clinical samples [for example participants with border-line personality disorder (Goldman, D’Angelo, DeMaso, & Mezzacappa, 1992) or a history of depressive disorders (Gladstone et al., 2004)] were excluded. Where the same data were reported across different publications, the most informative article was selected: for example, studies reporting sex- or age-specific prevalence estimates were selected over those providing combined estimates. If multiple studies were carried out in the same population, the first published study was utilized ensuring sample independence with the inclusion of every participant only once in the subsequent meta-analyses. Reference lists of selected studies were screened for any other relevant studies. In addition to the systematic review of electronic databases, we surveyed Australian official statistics websites to identify country-level data for prevalence of exposure. Two relevant reports and data sources (Australian Bureau of Statistics, 2005; Australian Institute of Health and Welfare, 2014) were identified and reviewed. The full-text of studies that appeared to meet the inclusion criteria was retrieved for closer examination and assessed for eligibility. A standardized data extraction sheet was developed, and data retrieved included publication details, methodological characteristics such as sex, age of respondents, sample size, sample type, state in which the study was conducted, definition of childhood, exposure measures, type of abuse and prevalence reported. Quality of studies was assessed using a tool for assessing risk of bias in prevalence studies (Hoy et al., 2012) and converted to a proportional quality score (the total quality score divided by the maximum score possible) (Supplementary Data Text S1 and Supplementary Table S1). In the Child Protection Australia 2012–13 report, unit-record level data has been made available for analysis for the first time, providing unique counts of children receiving child protection services in each jurisdiction (Australian Institute of Health and Welfare, 2014). Substantiated cases are based on the type of abuse that was considered most severe or most likely to place the child at risk (the ‘primary’ type of maltreatment). Although official data could be used to estimate prevalence of substantiated child maltreatment in Australia for the four maltreatment categories (presented in Supplementary Table S1), they were not included in subsequent meta-analyses of prevalence as they are not based on retrospective self-reports and not fully representative of all cases of child maltreatment. The Personal Safety Survey presented CPA and CSA prevalence estimates that could be included in the meta-analysis (Australian Bureau of Statistics, 2005). Meta-analysis of Prevalence Some studies presented prevalence of maltreatment for both sexes combined. These estimates were converted to sex specific estimates in a 2-step process. First a female-to-male ratio was derived from studies identified in the systematic

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

211

Table 1 Results of meta-analysis of prevalencea of child maltreatment in Australia. Type of maltreatment Childhood sexual abuse primary analysis Any childhood sexual abuseb Maleb Femaleb Childhood sexual abuse subgroup analyses Age of respondents (children <18 yearsc ) Male Female Age of respondents (adults 18+ yearsc ) Male Female Definition of childhood <16 years* Maled Femaled Definition of childhood <18 years* Male Female Penetrative sexual abuse Male Female Non-penetrative sexual abuse Male Female Childhood physical abuse primary analysise Male Female Childhood emotional abuse primary analysise Male Female Childhood neglect primary analysise Male Female

Pooled prevalence % (95% CI)

Number of data points

I2 (%)

Cochran’s Q

Test for heterogeneity (p-value)

8.6 (5.3–12.7) 4.5 (2.6–6.9) 11.6 (6.2–18.3)

44 18 26

99.44 97.84 99.54

7,737.40 788.58 5,389.20

<0.01 <0.01 <0.01

6.6 (2.2–13.0) 6.4 (5.3–7.6) 7.7 (0.0–22.9) 8.8 (5.2–13.1) 4.4 (2.4–6.9) 11.8 (6.1–19.1) 10.8 (5.5–17.5) 5.5 (2.1–10.2) 14.1 (6.0–24.7) 5.7 (3.2–8.8) 3.7 (2.0–5.8) 7.8 (4.6–11.7) 6.4 (4.3–8.9) 5.2 (3.1–7.8) 6.9 (3.8–10.7) 21.8 (15.5–28.7) 10.4 (6.8–14.7) 26.8 (21.1–32.9) 8.9 (6.0–12.3) 6.7 (4.8–9.0) 9.9 (5.5–15.2) 8.7 (3.9–15.0) 7.0 (1.7–14.8) 9.1 (0.5–24.2) 2.4 (0.1–6.6) 2.0 (0.6–4.1) 3.5 (0.0–13.1)

4 2 2 40 16 24 26 9 17 6 3 3 9 3 6 12 4 8 16 6 10 9 4 5 10 4 6

97.93 0.00 99.31 99.48 98.05 99.55 99.54 98.21 99.59 97.40 89.40 96.03 94.04 88.01 95.26 98.43 93.71 96.62 98.29 92.45 98.65 97.14 94.67 97.99 97.36 81.04 98.38

145.28 0.06 145.16 7,571.87 769.77 5,158.26 5,402.01 447.26 3,911.91 192.44 18.87 50.41 134.15 16.68 105.44 702.49 47.68 206.83 876.61 66.19 664.26 279.98 56.28 199.03 341.07 15.82 308.07

<0.01 0.81 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01 <0.01

a Prevalence estimates for CSA, CEA, CPA and CN presented in this table may also include exposure to other types of child maltreatment and hence cannot be added to estimate overall prevalence of child maltreatment. b in this study CSA was defined as ever having experienced any form of sexual abuse in childhood and hence these estimates were used in subsequent base analyses. c Reported age ranges that could fit into more than one category were allocated to the most representative one. d These CSA prevalence estimates were used in sensitivity analysis. e Data not available to carry out subgroup analyses for non-sexual maltreatment. * Studies where definition of childhood was <15 years or <17 years or “not stated” were excluded from this subgroup analysis.

review presenting sex specific estimates for the same maltreatment types. This ratio was then applied to estimates for both sexes combined to derive sex specific estimates for all studies included in subsequent meta-analyses. Meta-XL version 2.0 (Barendregt & Doi, 2014), a meta-analysis add-in tool for Microsoft Excel, was used to pool prevalence estimates using a quality effects model, which is a modified version of the fixed-effects inverse variance method that additionally allows giving greater weight to studies of high quality versus studies of lesser quality by using the quality scores assigned to each study (Barendregt, Doi, Lee, Norman, & Vos, 2013; Norman et al., 2012). To address the effects of important study characteristics and explore heterogeneity, we additionally conducted several pre-specified subgroup analyses (depending on data availability) by the following: age of respondents reporting abuse (adults versus children), definition of childhood (under 16 years versus under 18 years of age) and type of sexual abuse (penetrative versus non-penetrative). Pooled prevalence estimates for CSA, CPA, CEA and CN presented in Table 1 are non-additive and may also include exposure to other forms of maltreatment. Splitting Pooled Prevalence Using Estimates of Co-occurrence In order to account for the co-occurrence of maltreatment types, pooled prevalence estimates were split into the prevalence of single exposure and multiple combinations of exposure to abuse sub-types following the method previously used for combined exposure to CSA and intimate partner violence (Begg et al., 2008; Norman et al., 2010). The proportion of co-occurrence was calculated using a re-analysis of data from the Mater Hospital University of Queensland Study of Pregnancy (MUSP). MUSP is a birth cohort which prospectively collected a wide range of psychosocial variables from the first prenatal clinic visit on 7,223 children born between 1981 and 1984 at The Mater Mother’s Hospital in Brisbane,

212

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220 Table 2 Proportion of co-occurrence (%) of child maltreatment types from MUSP data. Type of maltreatment Sexual abuse Sexual only Sexual and physical Sexual and emotional Sexual and neglect Sexual, physical and emotional Sexual, physical and neglect Sexual, emotional and neglect Sexual, emotional, neglect and physical Total (sexual abuse) Physical abuse Physical only Physical and emotional Physical and neglect Physical and sexual Physical, emotional and neglect Physical, sexual and neglect Physical, sexual and emotional Physical, sexual, emotional and neglect Total (physical abuse) Emotional abuse Emotional only Emotional and neglect Emotional and physical Emotional and sexual Emotional, physical and neglect Emotional, neglect and sexual Emotional, physical and sexual Emotional, neglect, physical and sexual Total (emotional abuse) Neglect Neglect only Neglect and physical Neglect and sexual Neglect and emotional Neglect, sexual and physical Neglect, sexual and emotional Neglect, emotional and physical Neglect, physical, sexual and emotional Total (neglect)

Proportion of co-occurrence (%) 42.9 6.8 4.0 8.2 3.4 3.4 7.5 23.8 100.0

20.9 25.8 8.1 3.5 26.1 1.7 1.7 12.2 100.0

9.0 13.9 27.7 2.2 28.1 4.1 1.9 13.1 100.0

26.4 8.5 4.5 13.7 1.9 4.1 27.9 13.0 100.0

MUSP = Mater Hospital University of Queensland Study of Pregnancy.

Australia. Data on substantiated child maltreatment for all four sub-types of child maltreatment were obtained from the relevant state child protection agency, and linked to the original dataset up to and including the 14 year follow-up. The rate of follow-up at 14 years was 71.6% (Mills et al., 2013; Najman, Bor, O’Callaghan, Williams, Aird, & Shuttlewood, 2005). The MUSP proportions of co-occurrence of the four forms of maltreatment (Table 2) were applied to pooled prevalence estimates (Table 1) to derive mutually exclusive exposure categories. Due to insufficient age specific data, we applied the age pattern for CSA for Australia from GBD 2010 derived using the DisMod-MR tool (Vos et al., 2012) to prevalence estimates for all ages combined to estimate prevalence in five year age groups (Supplementary Table S2). Health Outcomes We included outcomes for which there was sufficient evidence for a causal effect: depressive disorders, anxiety disorders, and intentional self-harm (Andrews et al., 2004; Chen et al., 2010; Devries et al., 2014; Lim et al., 2012; Norman et al., 2012). Depressive disorders included cases of major depressive disorder (MDD) and dysthymia meeting Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR) (American Psychiatric Association, 2000) diagnostic criteria (Ferrari et al., 2013). According to DSM-IV-TR (American Psychiatric Association, 2000), anxiety disorders include generalized anxiety disorder; panic disorder and agoraphobia; social phobia; specific phobia; obsessive–compulsive disorder; post-traumatic stress

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

213

Table 3a Relative risk (RR) estimates from original meta-analyses.a Health outcome and type of maltreatment

RR (95% CI)

Depressive disorders Childhood sexual abuse Childhood physical abuse Childhood emotional abuse Childhood neglect

1.68 (1.49–1.88) 1.55 (1.34–1.79) 2.32 (1.92–2.81) 1.59 (1.20–2.11)

Anxiety disorders Childhood sexual abuse Childhood physical abuse Childhood emotional abuse Childhood neglect

2.53 (2.04–3.13) 1.47 (1.20–1.79) 3.17 (1.65–6.10) 1.38 (1.20–1.59)

Intentional self-harm Childhood sexual abuse Childhood physical abuse Childhood emotional abuse Childhood neglect

2.07 (0.77–5.52) 1.85 (1.49–2.28) 3.22 (1.89–5.50) 1.63 (1.16–2.28)

a Sources: Odds ratios (ORs) for CSA and depressive disorders: Karen M. Devries personal communication, 2015, CSA and self-harm: (Devries et al., 2014), CSA and anxiety disorders: (Chen et al., 2010); and for CPA, CEA and CN and depressive disorders, anxiety disorders and intentional self-harm: (Norman et al., 2012). ORs from original meta-analyses converted to RR estimates (Di Pietrantonj, 2006).

disorder; separation anxiety disorder; and anxiety disorders not otherwise specified. Sub-types of anxiety disorders tend to co-occur (Baxter, Vos, Scott, Ferrari, & Whiteford, 2014). Following GBD 2010 conventions (Baxter et al., 2014), anxiety disorders were assessed as one comprehensive disorder group, consisting of ‘any’ anxiety disorder in order to more accurately capture burden attributable to child maltreatment from all anxiety disorders. Intentional self-harm was defined as cases meeting International Classification of Diseases (ICD)-10 codes (X60-X84) for intentional self-inflicted poisoning or injury (World Health Organization, 1992) and included attempted (non-fatal self-harm) and completed suicides. Relative Risk Estimates Odds ratios (ORs) for CSA and depressive disorders (Karen M. Devries, personal communication, 2015), CSA and intentional self-harm (Devries et al., 2014), CSA and anxiety disorders (Chen et al., 2010) and CPA, CEA and CN and depressive disorders, anxiety disorders and intentional self-harm (Norman et al., 2012) from previous meta-analyses were first converted to RR estimates for use in PAF calculations in this study to avoid overestimating attributable burden. This was done using an imputation method which reconstructs four-fold tables and event frequency values from published OR’s and their 95% confidence intervals (CI), given the sample sizes (Di Pietrantonj, 2006). For CSA, CPA, CEA and CN we then repeated the meta-analyses using reconstructed RR estimates. Weighted summary measures were computed using MetaXL version 2.0 with RRs chosen as the principal summary measure (Table 3a). Relative Risk Estimates for Multiple Forms of Maltreatment These pooled RRs (Table 3a) were adjusted to derive RR estimates for single as well as combined exposure to multiple forms of maltreatment categories once again adapting the method used to adjust RR estimates for combined exposure to CSA and IPV (Begg et al., 2008; Norman et al., 2010). In MUSP, anxiety and depression was assessed by using the Youth Self-Report (YSR) questionnaire at the 14-year follow-up. We used a 10% cutoff to create a dichotomous variable by treating the highest 10% of scores as anxious and depressed “cases” consistent with criteria listed by Achenbach (Achenbach & Edelbrock, 1983) for using the YSR for research purposes. Intentional self-harm was assessed using two questions from the YSR questionnaire at the 14 year follow up, “I deliberately try to hurt or kill myself” and “I think about killing myself”. If a participant answered ‘sometimes’ or ‘often’ to either of these two questions, they were categorized as having engaged in intentional self-harm. All analyses of MUSP data were conducted using Stata 12.0 (StataCorp, 2011). Logistic regression models were fitted, adjusting for a priori identified potential confounders including age, sex, maternal age, maternal marital status and family income. For consistency of measurement, the ORs for combined exposure to multiple forms of abuse from MUSP data were first converted to RRs using the previously outlined method (Di Pietrantonj, 2006). MUSP RRs for depression and anxiety disorders and intentional self-harm with exposure to 1, 2, 3 or 4 forms of substantiated maltreatment compared with no maltreatment are presented in Table 3b. These MUSP RRs for multiple forms of maltreatment along with RRs from original meta-analyses for CSA, CPA, CEA and CN (unadjusted for combined exposure) (Table 3a) and the proportion of co-occurrence of child maltreatment types from MUSP data (Table 2) were used to derive RRs for CSA only, CPA only, CEA only, CN only

214

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

Table 3b Relative Risk (RR) estimates with exposure to 1, 2, 3 or 4 forms of substantiated maltreatment compared with no exposure to maltreatment from MUSP data. Types of maltreatment

N in total MUSP cohort

No maltreatment 1 type 2 types 3 types 4 types

6,703 218 162 96 35

N completing 14-year YSR 4,912 125 86 39 10

Adjusted RR (95% CI) a Depressive/anxiety disorders

Adjusted RR (95% CI) a Intentional self-harm

1.00 1.66 (0.96–2.87) 3.10 (1.75–5.51) 3.15 (1.32–7.51) 0.98 (0.11–8.85)b

1.00 1.28 (0.75–2.19) 2.56 (1.46–4.47) 2.63 (1.13–6.11) 1.52 (0.26–8.78)b

YSR = Youth Self-Report questionnaire; MUSP = Mater Hospital University of Queensland Study of Pregnancy. a Source: Re-analysis of MUSP data. All risks were adjusted for subject age, sex, maternal age, maternal marital status and family income. ORs converted to RR estimates (Di Pietrantonj, 2006). b Due to small numbers in 4 forms of abuse category, we assumed the same RR as for 3 forms of abuse for this category in our analyses.

and the different combinations of exposure categories (Supplementary Tables S3–S5) for the related health outcomes. The same adjusted RRs were used for males and females and across all age groups. Calculation of Population Attributable Fractions and Attributable burden The estimated RRs given CSA only, CEA only, CPA only and CN only as well as the combined exposure states, were paired with their corresponding prevalence estimates to calculate PAFs using the following formula (Levin, 1953). PAF =

P(RR − 1) P(RR − 1) + 1

where ‘P’ is the prevalence of child maltreatment and ‘RR’ is the relative-risk of health outcomes from original published meta-analyses adjusted for co-occurrence of different child maltreatment sub-types using MUSP data as outlined above. PAFs were applied to estimates of the burden of disease in Australia from GBD 2010 (Institute for Health Metrics and Evaluation, 2013) for the selected health outcomes (depressive and anxiety disorders and intentional self-harm), measured in disabilityadjusted life years (DALYs) [DALY = years of life lost due to premature mortality (YLL) + years lived with disability (YLD)] by age and sex. Sensitivity Analysis A sensitivity analysis was also carried out using prevalence of exposure to CSA based on the subgroup analyses of studies where childhood was defined as <16 years (see Table 1) and then following the same method to calculate PAFs as outlined above. Uncertainty Analysis In order to present uncertainty ranges around point estimates reflecting the main sources of sampling uncertainty in the calculations, we made use of Monte Carlo simulation–modelling techniques and the Ersatz software version 1.3 (Barendregt, 2014). Beta distributions were specified for prevalence estimates. For RR estimates, we used the Ersatz random function ErRelativeRisk (Barendregt, 2010). For each corresponding output variable, 95% uncertainty intervals were calculated bounded by the 2.5th and 97.5th percentiles of 10,000 iteration values generated. Results Systematic Review of Child Maltreatment Prevalence in Australia A total of 687 articles were identified by the search of electronic databases with two additional reports identified from official statistics websites (Australian Bureau of Statistics, 2005; Australian Institute of Health and Welfare, 2014) and one article identified from scanning the reference lists of other articles (Goldman & Goldman, 1988). After duplicates were removed, the remaining 557 records were screened for prevalence of at least one form of child maltreatment that occurred within childhood in Australia. Out of 55 articles assessed for eligibility, 23 studies satisfied the pre-determined inclusion criteria and provided prevalence estimates for CSA, CPA, CN or CEA for use in subsequent meta-analyses of prevalence (see Supplementary Table S1 and Fig. S1) (Australian Bureau of Statistics, 2005; Bergen, Martin, Richardson, Allison, & Roeger, 2004; Chu, Williams, Harris, Bryant, & Gatt, 2013; de Visser, Smith, Rissel, Richters, & Grulich, 2003; Dinwiddie et al., 2000; Dunne, Purdie, Cook, Boyle, & Najman, 2003; Eslick, Koloski, & Talley, 2011; Fleming, 1997; Goldman & Goldman, 1988; Goldman & Padayachi, 1997; Hollingsworth, Callaway, Duhig, Matheson, & Scott, 2012; Mamun et al., 2007; Mazza, 1996,

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

215

Table 4 Estimated burden attributable to all forms of child maltreatment combined, Australia 2010. Health outcome

Male PAF

Depressive disorders Anxiety disorders Intentional self-harm

15.7% 20.9% 23.5%

Female

Attributable DALYs 10,863 8,265 16,748

PAF 22.8% 30.6% 33.0%

Attributable DALYs 26,030 23,370 6,480

Total 95% uncertainty interval

35,876 11,000–62,000

55,881 16,000–95,000

% of total burden 95% uncertainty interval

1.4% 0.4–2.3%

2.4% 0.7–4.1%

PAF = Population Attributable Fraction; DALY = disability-adjusted life years.

2001; Moore, Romaniuk, Olsson, Jayasinghe, Carlin, & Patton, 2010; Mouzos & Makkai, 2004; Najman, Dunne, Purdie, Boyle, & Coxeter, 2005; Nelson et al., 2006; Price-Robertson, Smart, & Bromfield, 2010; Reeve & van Gool, 2013; Rosenman & Rogers, 2004; Straus & Savage, 2005; Watson & Halford, 2010).

Prevalence of Child Maltreatment in Australia Pooled prevalence estimates based on retrospective self-reports by type of child maltreatment are presented in Table 1. The test for heterogeneity was significant, with p < 0.01 for all primary analyses of CSA, CPA, CN and CEA. Pooled self-reported prevalence of CSA was estimated to be higher in females and the female-to-male ratio of sexual abuse prevalence derived from studies identified in the systematic review was 2.2 (95% CI: 1.7–2.8). There were only slight to no gender differences for the other types of child maltreatment [female-to-male ratio 1.0 (0.9–1.1) for CPA; 1.2 (0.9–1.6) for CEA; and 1.0 (1.0–1.0) for CN]. Although prevalence of CSA across all age groups was higher in subgroup analyses of studies where childhood was defined as <16 years [5.5% (2.1–10.2%) for males and 14.1% (6.0–24.7%) for females used in the sensitivity analysis] compared with all studies included in the primary analysis which included different definitions of childhood [4.5% (2.6–6.9%) for males and 11.6% (6.2–18.3%) for females], these differences were not statistically significant. According to MUSP data, CEA seldom occurs in isolation and co-occurs frequently with CPA and CN. The proportion of co-occurrence of CPA, CEA and CN was between 26% and 28% (Table 2). We split pooled prevalence from Table 1 into the different combinations of exposure to maltreatment using MUSP proportions of co-occurrence of child maltreatment types. The highest prevalence across all ages was for single exposure to CSA (2.6% for males and 7.8% for females) followed by the CPA and CEA combination (1.8% for males and 2.5% for females), single exposure to CPA (1.9% for males and 2.3% for females) and the CPA, CEA and CN combination (1.4% for males and 2.0% for females) of maltreatment. Prevalence of all child maltreatment types peaked in the 25–34 year age group in both males and females and then decreased with increasing age (Supplementary Table S2).

Population Attributable Fractions (PAFs) PAFs for depressive and anxiety disorders and intentional self-harm by mutually exclusive child maltreatment categories, age and sex, are presented in Supplementary Tables S3–S5. In females, single exposure to CSA, followed by dual exposure to CPA and CEA and the CPA, CEA and CN combination yielded the highest PAFs and this was a function of the high prevalence and RRs for these combinations of maltreatment. For males, highest PAFs were usually for dual exposure to CPA and CEA and the combination of CPA, CEA, and CN for most health outcomes (Supplementary Tables S3–S5). For exposure to all forms of child maltreatment combined (Table 4), the highest PAF was for intentional self-harm in females (33.0%), followed by anxiety disorders in females (30.6%), intentional self-harm in males (23.5%), depressive disorders in females (22.8%), anxiety disorders in males (20.9%) and depressive disorders in males (15.7%).

Attributable Burden Overall, child maltreatment was an important risk factor in Australia and accounted for an estimated 1.4% (95% uncertainty interval 0.4–2.3%) of all DALYs in males, and approximately 2.4% (0.7–4.1%) of all DALYs in females, in 2010 (Table 4). Slightly higher estimates of attributable burden were obtained in the sensitivity analysis where only studies which defined childhood as <16 years were used to estimate prevalence of CSA with all forms of child maltreatment accounting for 37,206 DALYs (11,000–64,000) in males and 58,990 DALYs (18,000–100,000) in females compared with the base analysis [35,876 DALYs (11,000–62,000) in males and 55,881 DALYs (16,000–95,000) in females (Table 4)]. Age-specific estimates of attributable burden are presented in Supplementary Table S6.

216

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

Discussion Although comparative risk assessment methodology has previously been applied to CSA in Australia, this study is novel in measuring the impact of the four main types of child maltreatment on mental health. In the absence of nationally representative data for Australia, our systematic review identified 23 studies reporting on the prevalence of child maltreatment that could be included in the meta-analysis. The majority of these studies reported on the prevalence of CSA, highlighting the dearth of studies reporting the prevalence of CEA, CPA and CN. Our pooled CSA prevalence estimate across all ages for females based on studies where childhood was defined as <16 years [14.1% (6.0–24.7%)] was similar to age standardized prevalence estimates for Australia from GBD 2010 where childhood was defined as <15 years [females: 14.1% (12.4–16.1%)] but lower for males [5.5% (2.1–10.2%) in this study versus 9.1% (7.9–10.3%) in GBD 2010] (Institute for Health Metrics and Evaluation, 2013). A large disparity existed between our pooled prevalence estimates based on retrospective self-reports and those derived from Child Protection 2012–13 official data (Australian Institute of Health and Welfare, 2014) where only 0.85% of children were substantiated as cases of maltreatment. Child Protection data usually captures the most severe cases of child maltreatment and substantiated cases under-represent the true number of children experiencing abuse and neglect due to challenges in identifying and substantiating maltreatment (Brown, Frederico, Hewitt, & Sheehan, 1998; Wildeman et al., 2014). There is also unreliability in self-reporting of child maltreatment in studies, and biases are more likely towards underreporting rather than over-reporting of abuse (Fergusson, Horwood, & Woodward, 2000) and hence our estimates of child maltreatment prevalence for Australia may underestimate the true prevalence. In Child Protection Australia 2012–13 data, co-occurrence of primary types of child maltreatment with a secondary type was included for the first time. Since proportions of co-occurrence in our study were derived from the MUSP study where child protection data had been linked to the original dataset up to and including the 14 year follow-up, we expected these to be similar: low for CSA and other types of child maltreatment but high for CPA, CEA and CN. In Australian Institute of Health and Welfare child protection reports, where CEA was the primary type of substantiated maltreatment, CN co-occurred in about a third (32%) of cases and for substantiated CPA, there was also a high proportion of co-occurrence with CEA and CN (37% and 26% respectively) (Australian Institute of Health and Welfare, 2014) but data on co-occurrence of three or four forms of maltreatment were not available. The misclassification of emotional neglect as emotional abuse in child protection records may explain the high overlap between neglect and emotional abuse. This study shows that a significant proportion of depressive and anxiety disorders and intentional self-harm in Australia is attributable to maltreatment in childhood. As these findings are novel for Australia, there are no studies available for a direct comparison of attributable burden. A study from the USA reported that childhood adversity (including all forms of child maltreatment) explained 26% and 32% of lifetime mood and anxiety disorders, respectively (Green et al., 2010); however, this study did not provide separate estimates for males and females, and included other childhood adversity such as parental mental illness. GBD 2010 (Lim et al., 2012) estimated that in Australia, CSA explained 7% and 10% of the burden of depressive disorders and 9% and 14% of self-harm burden in males and females, respectively. This is comparable to our study when we compute results for CSA by adding all mutually exclusive categories containing sexual abuse (5% and 11% of the burden of depressive disorders and 8% and 16% of self-harm burden in males and females, respectively). Our more inclusive analysis found all forms of child maltreatment combined accounted for 1.4% and 2.4% of total male and female DALYs in Australia while just the CSA component accounted for 0.3% and 0.7%. This latter estimate is comparable to GBD 2010 (0.6% and 0.7% of total male and female DALYs) (Lim et al., 2012) which only included CSA as a risk factor and no other forms of child maltreatment. Limitations Our study had some limitations. There are inconsistencies in how each form of maltreatment is defined and measured, as well as inconsistencies in how childhood is defined across studies. Studies included in this meta-analysis of prevalence used various definitions of childhood ranging from <15 to <18 years and included studies where the definition of childhood was not stated which is in alignment with RR estimates for non-sexual maltreatment (Norman et al., 2012), CSA and depressive disorders (Karen M. Devries, personal communication, 2015) and CSA and self-harm (Devries et al., 2014). RR estimates for CSA and anxiety disorders were derived from a meta-analysis where CSA was defined as abuse occurring at or before age 18 years (Chen et al., 2010). While the meta-analysis of prevalence included only studies where child maltreatment was self-reported, which is subject to recall bias, some studies included in the original meta-analyses used to derive RR estimates, used official records of confirmed abuse whereas other studies relied on self-report. The definition and measurement of health outcomes also varied across studies used to derive RR estimates (Norman et al., 2012). Depressive and anxiety disorders in the MUSP study were assessed using symptom scales rather than diagnostic instruments and self-harm was based on self-reported suicidal behaviour at 14 years of age. In this analysis, RR estimates for multiple forms of maltreatment derived from previous meta-analyses and MUSP data were then applied to burden of disease estimates for Australia from GBD 2010 for selected health outcomes meeting DSM and ICD-10 diagnostic criteria. In addition, published studies and meta-analyses concerned with single forms of victimization such as CSA may have overestimated the unique association between single forms of child maltreatment and adverse health outcomes because they did not adequately control for other types of victimization (Finkelhor et al., 2007). Unfortunately, very few studies explore

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

217

the interrelationships among victimizations and the contribution of these interrelationships to mental health outcomes. The few studies identified investigating exposure to multiple types of child maltreatment and health outcomes typically either excluded one of the four forms of child maltreatment (CN in particular), had small sample sizes, had inadequate control for confounding, or did not examine depressive and anxiety disorders or intentional self-harm as the outcomes. We dealt with this issue in this analysis by adjusting the original RRs from published meta-analyses, reporting on single forms of maltreatment, to derive RR estimates for combined exposure to multiple forms of maltreatment. Furthermore, exposure to child maltreatment often co-occurs within the context of other family dysfunction, social deprivation, and other environmental stressors that are also associated with mental disorders. It is possible that, since we often relied on studies of self-reported child maltreatment in adults, participants have been further exposed to other interpersonal violence including community violence, bullying and intimate partner violence. For example, it is common for intimate partner violence to occur in women who have been previously exposed to CSA, and women who experience both adult and childhood abuse are at a higher risk of depressive disorders and other mental disorders (McGuigan & Middlemiss, 2005; Messman-Moore, Long, & Siegfried, 2000). Hence, some of the effect of child maltreatment on mental disorders may still be explained by confounding. A further limitation was our inability to examine the effect of severity or frequency of abuse on health outcomes due to the paucity of data available. Also, in most studies included in the meta-analysis of prevalence, child maltreatment was reported by adults across wide age groups (such as 18–74 years) making it difficult to derive an age pattern and hence we applied the GBD 2010 age pattern for CSA for Australia to all maltreatment types to derive age-specific estimates. Nevertheless, a similar decrease in prevalence after the age of 25–34 years was observed for self-reported exposure to adult physical and sexual violence in the Personal Safety Survey (Australian Bureau of Statistics, 2005). An additional limitation was the application of exposure data from different time periods (1988–2013) to burden of disease estimates for 2010 for Australia from GBD 2010. Finally, only selected mental health outcomes for which there was convincing evidence for a causal link were included in this analysis. Other mental, substance use and physical disorders associated with child maltreatment but for which the current evidence is not as strong are not included. In addition, this estimate of attributable burden includes the long-term health consequences but does not include the direct physical injury burden of child maltreatment. Data on perpetratorvictim relationships were not available, and hence, we were unable to distinguish fatal (homicides) and nonfatal injuries due to child maltreatment from other interpersonal violence in children 0–18 years of age in Australia. The focus of this study is on health outcomes but there are also other non-health consequences of child maltreatment including poor educational outcomes; poor employment outcomes; higher levels of healthcare utilization; and criminal behaviour (Gilbert et al., 2009; Widom, 1989) which were not included.

Conclusions This is the first study to quantify the burden of mental disorders attributable to all forms of child maltreatment in Australia and one of few internationally attempting to deal with the complexity of exposure to multiple types of maltreatment. It is important for future burden of disease studies to include all forms of child maltreatment as risk factors for mental disorders and self-harm. This study provides a method by which disease burden from all forms of child maltreatment can be calculated enabling governments and policy makers to more accurately measure the disease burden and economic costs of maltreatment and enable evaluation of interventions specific to certain types of maltreatment in childhood. Experiencing multiple forms of child maltreatment is common (Choo et al., 2011; Dong et al., 2004; Green et al., 2010; Schneider et al., 2007) hence children identified as exposed to one form of child maltreatment should also be screened for other types. The deleterious health effects of child maltreatment are now well known and given the costs to society in terms of health care and disability burden, preventive strategies proven to be effective in reducing the risk of child maltreatment warrant substantial investment.

Conflict of Interest All authors declare no conflicts of interest.

References Achenbach, T. M., & Edelbrock, C. (1983). Manual for the Child Behavior Checklist and Revised Child Behavior Profile. Burlington: University of Vermont Department of Psychiatry. Afifi, T. O., Enns, M. W., Cox, B. J., Asmundson, G. J. G., Stein, M. B., & Sareen, J. (2008). Population attributable fractions of psychiatric disorders and suicide ideation and attempts associated with adverse childhood experiences. American Journal of Public Health, 98(5), 946–952. http://dx.doi.org/10.2105/AJPH.2007.120253 Alink, L. R., Cicchetti, D., Kim, J., & Rogosch, F. A. (2012). Longitudinal associations among child maltreatment, social functioning, and cortisol regulation. Developmental Psychology, 48(1), 224–236. http://dx.doi.org/10.1037/a0024892 American Psychiatric Association. (2000). Diagnostic and Statistical Manual of Mental Disorders (DSM-IV-TR). Washington, DC: American Psychiatric Association.

218

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

Andrews, G., Corry, J., Slade, T., Issakidis, C., & Swanston, H. (2004). Child sexual abuse. In M. Ezzati, A. D. Lopez, A. Rodgers, & C. J. L. Murray (Eds.), Comparative quantification of health risks: Global and regional burden of disease attributable to selected major risk factors (Vol. 2) (pp. 1851–1940). Geneva World Health Organization. Australian Bureau of Statistics. (2005). Personal Safety Survey. Canberra: Australian Bureau of Statistics. Australian Institute of Health and Welfare. (2014). Child protection Australia 2012–13. Child Welfare series no.58. Cat. no. CWS 49. Canberra: AIHW. Barendregt, J. J. (2010). The effect size in uncertainty analysis. International Society for Pharmacoeconomics and Outcomes Research, 13, 388–391. http://dx.doi.org/10.1111/j.1524-4733.2009.00686.x Barendregt, J. J. (2014). Ersatz version 1.3. Retrieved from http://www.epigear.com Barendregt, J. J., & Doi, S. A. (2014). MetaXL version 2.0. Retrieved from http://www.epigear.com Barendregt, J. J., Doi, S. A., Lee, Y. Y., Norman, R. E., & Vos, T. (2013). Meta-analysis of prevalence. Journal of Epidemiology & Community Health, 67(11), 974–978. http://dx.doi.org/10.1136/jech-2013-203104 Barth, J., Bermetz, L., Heim, E., Trelle, S., & Tonia, T. (2013). The current prevalence of child sexual abuse worldwide: A systematic review and meta-analysis. International Journal of Public Health, 58(3), 469–483. http://dx.doi.org/10.1007/s00038-012-0426-1 Baxter, A. J., Vos, T., Scott, K. M., Ferrari, A. J., & Whiteford, H. A. (2014). The global burden of anxiety disorders in 2010. Psychological Medicine, 1–12. http://dx.doi.org/10.1002/mpr.1444 Begg, S., Vos, T., Barker, B., Stevenson, C., Stanley, L., & Lopez, A. D. (2008). Burden of disease and injury in Australia in the new millennium: Measuring health loss from diseases, injuries and risk factors. The Medical Journal of Australia, 188(1), 36–40. Bergen, H. A., Martin, G., Richardson, A. S., Allison, S., & Roeger, L. (2004). Sexual abuse, antisocial behaviour and substance use: Gender differences in young community adolescents. Australian and New Zealand Journal of Psychiatry, 38(1–2), 34–41. http://dx.doi.org/10.1111/j.1440-1614.2004.01295.x Brown, T., Frederico, M., Hewitt, L., & Sheehan, R. (1998). Child abuse and the family court. Australian Institute of Criminology. Butchart, A., Phinney Harvey, A., Kahane, T., Mian, M., & Furniss, T. (2006). Preventing child maltreatment: A guide to action and generating evidence. Geneva: World Health Organization and International Society for Prevention of Child Abuse and Neglect. Chen, L. P., Murad, M. H., Paras, M. L., Colbenson, K. M., Sattler, A. L., Goranson, E. N., Elamin, M. B., Seime, R. J., Shinozaki, G., Prokop, L. J., & Zirakzadeh, A. (2010). Sexual abuse and lifetime diagnosis of psychiatric disorders: Systematic review and meta-analysis. Mayo Clinic Proceedings, 85(7), 618–629. http://dx.doi.org/10.4065/mcp.2009.0583 Choo, W. Y., Dunne, M. P., Marret, M., Fleming, M., & Wong, Y. -L. (2011). Victimization experiences of adolescents in Malaysia. Journal of Adolescent Health, 49(6), 627–634. http://dx.doi.org/10.1016/j.jadohealth.2011.04.020 Chu, D. A., Williams, L. M., Harris, A. W. F., Bryant, R. A., & Gatt, J. M. (2013). Early life trauma predicts self-reported levels of depressive and anxiety symptoms in nonclinical community adults: Relative contributions of early life stressor types and adult trauma exposure. Journal of Psychiatric Research, 47(1), 23–32. http://dx.doi.org/10.1016/j.jpsychires.2012.08.006 de Visser, R. O., Smith, A. M., Rissel, C. E., Richters, J., & Grulich, A. E. (2003). Sex in Australia: experiences of sexual coercion among a representative sample of adults. Australian and New Zealand Journal of Public Health, 27(2), 198–203. http://dx.doi.org/10.1111/j.1467-842X.2003.tb00808.x Devries, K. M., Mak, J., Child, J. C., Falder, G., Bacchus, L. J., Astbury, J., & Watts, C. (2014). Childhood sexual abuse and suicidal behavior: A meta-analysis. Pediatrics, 133(5), 1–14. http://dx.doi.org/10.1542/peds.2013-2166 Di Pietrantonj, C. (2006). Four-fold table cell frequencies imputation in meta analysis. Statistics in Medicine, 25(13), 2299–2322. http://dx.doi.org/ 10.1002/sim.2287 Dinwiddie, S., Heath, A. C., Dunne, M. P., Bucholz, K. K., Madden, P. A., Slutske, W. S., Bierut, L. J., Statham, D. B., & Martin, N. G. (2000). Early sexual abuse and lifetime psychopathology: A co-twin-control study. Psychological Medicines, 30(1), 41–52. http://dx.doi.org/10.1017/S0033291799001373 Dong, M., Anda, R. F., Felitti, V. J., Dube, S. R., Williamson, D. F., Thompson, T. J., Loo, C. M., & Giles, W. H. (2004). The interrelatedness of multiple forms of childhood abuse, neglect, and household dysfunction. Child Abuse & Neglect, 28(7), 771–784. http://dx.doi.org/10.1016/j.chiabu.2004.01.008 Dunne, M. P., Purdie, D. M., Cook, M. D., Boyle, F. M., & Najman, J. M. (2003). Is child sexual abuse declining? Evidence from a population-based survey of men and women in Australia. Child Abuse & Neglect, 27(2), 141–152. http://dx.doi.org/10.1016/S0145-2134(02)00539-2 Edwards, V., Holden, G., Felitti, V., & Anda, R. (2003). Relationship between multiple forms of childhood maltreatment and adult mental health in community respondents: Results from the Adverse Childhood Experiences study. American Journal of Psychiatry, 160(8), 1453–1460. http://dx.doi.org/10.1176/appi.ajp.160.8.1453 Eslick, G. D., Koloski, N. A., & Talley, N. J. (2011). Sexual, physical, verbal/emotional abuse and unexplained chest pain. Child Abuse & Neglect, 35(8), 601–605. http://dx.doi.org/10.1016/j.chiabu.2011.04.007 Fang, X., Brown, D. S., Florence, C. S., & Mercy, J. A. (2012). The economic burden of child maltreatment in the United States and implications for prevention. Child Abuse & Neglect, 36(2), 156–165. http://dx.doi.org/10.1016/j.chiabu.2011.10.006 Fang, X., Fry, D., Brown, D., Mercy, J., Dunne, M., Butchart, A., Corso, P. S., Maynzyuk, K., Dzhygyr, Y., Chen, Y., McCoy, A., & Swales, D. (2015). The burden of child maltreatment in the East Asia and Pacific Region. Child Abuse and Neglect, 42, 146–162. http://dx.doi.org/10.1016/j.chiabu.2015.02.012 Fergusson, D. M., Horwood, L. J., & Woodward, L. J. (2000). The stability of child abuse reports: A longitudinal study of the reporting behaviour of young adults. Psychological Medicine, 30(3), 529–544. http://dx.doi.org/10.1017/S0033291799002111 Ferrari, A. J., Charlson, F. J., Norman, R. E., Patten, S. B., Freedman, G., Murray, C. J., Vos, T., & Whiteford, H. A. (2013). Burden of depressive disorders by country, sex, age, and year: findings from the global burden of disease study 2010. PLoS Medicine, 10(11), e1001–e1547. http://dx.doi.org/10.1371/ journal.pmed.1001547 Finkelhor, D. (2008). Childhood victimization: Violence, crime, and abuse in the lives of young people. New York: Oxford University Press, Inc. Finkelhor, D., Ormrod, R. K., & Turner, H. A. (2007). Poly-victimization: A neglected component in child victimization trauma. Child Abuse & Neglect, 31, 7–26. http://dx.doi.org/10.1016/j.chiabu.2006.06.008 Fleming, J. (1997). The prevalence of childhood sexual abuse in a community sample of Australian women. The Medical Journal of Australia, (166), 65–68. Gilbert, R., Widom, C. S., Browne, K., Fergusson, D., Webb, E., & Janson, S. (2009). Burden and consequences of child maltreatment in high-income countries. The Lancet, 373, 68–81. http://dx.doi.org/10.1016/S0140-6736(08)61706-7 Gladstone, G. L., Parker, G. B., Mitchell, P. B., Malhi, G. S., Wilhelm, K., & Austin, M. P. (2004). Implications of childhood trauma for depressed women: An analysis of pathways from childhood sexual abuse to deliberate self-harm and revictimization. American Journal of Psychiatry, 161(8), 1417–1425. http://dx.doi.org/10.1176/appi.ajp.161.8.1417 Goldman, R. J., & Goldman, J. D. (1988). The prevalence and nature of sexual abuse. Australian Journal of Sex, Marriage & Family, 9, 94–106. Goldman, J. D., & Padayachi, U. K. (1997). The prevalence and nature of child sexual abuse in Queensland, Australia. Child Abuse & Neglect, 21(5), 489–498. http://dx.doi.org/10.1016/S0145-2134(97)00008-2 Goldman, S. J., D’Angelo, E. J., DeMaso, D. R., & Mezzacappa, E. (1992). Physical and sexual abuse histories among children with borderline personality disorder. American Journal of Psychiatry, 149(12), 1723–1726. http://dx.doi.org/10.1176/ajp.149.12.1723 Green, J. G., McLaughlin, K. A., Berglund, P. A., Gruber, M. J., Sampson, N. A., Zaslavsky, A. M., & Kessler, R. C. (2010). Childhood adversities and adult psychiatric disorders in the national comorbidity survey replication. I: Associations with first onset of DSM-IV disorders. Archives of General Psychiatry, 67(2), 113–123. http://dx.doi.org/10.1001/archgenpsychiatry.2009.186 Hamarman, S., & Bernet, W. (2000). Evaluating and reporting emotional abuse in children: Parent-based, action-based focus aids in clinical decision-making. Journal of the American Academy of Child and Adolescent Psychiatry, 39(7), 928–930. http://dx.doi.org/10.1097/00004583-200007000-00023 Hollingsworth, K., Callaway, L., Duhig, M., Matheson, S., & Scott, J. (2012). The association between maltreatment in childhood and pre-pregnancy obesity in women attending an antenatal clinic in Australia. PLOS ONE, 7(12), e51868. http://dx.doi.org/10.1371/journal.pone.0051868 Holt, M. K., Finkelhor, D., & Kantor, G. K. (2007). Multiple victimization experiences of urban elementary school students: Associations with psychosocial functioning and academic performance. Child Abuse & Neglect, 31(5), 503–515. http://dx.doi.org/10.1016/j.chiabu.2006.12.006

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

219

Hoy, D., Brooks, P., Woolf, A., Blyth, F., March, L., Bain, C., Baker, P., Smith, E., & Buchbinder, R. (2012). Assessing risk of bias in prevalence studies: Modification of an existing tool and evidence of interrater agreement. Journal of Clinical Epidemiology, 65(9), 934–939. http://dx.doi.org/10.1016/j.jclinepi.2011.11.014 Institute for Health Metrics and Evaluation. (2013). GBD data visualisation tools. Retrieved from http://www.healthmetricsandevaluation.org/gbd/ visualizations/country Jirapramukpitak, T., Prince, M., & Harpham, T. (2005). The experience of abuse and mental health in the young Thai population: A preliminary survey. Social Psychiatry and Psychiatric Epidemiology, 40(12), 955–963. http://dx.doi.org/10.1007/s00127-005-0983-1 Kim, J., & Cicchetti, D. (2003). Social self-efficacy and behavior problems in maltreated and nonmaltreated children. Journal of Clinical Child and Adolescent Psychology, 32(1), 106–117. http://dx.doi.org/10.1207/S15374424JCCP3201 10 Levin, M. L. (1953). The occurrence of lung cancer in man. Acta Unio Internationalis Contra Cancrum, 9(3), 531–541. Lim, S. S., Vos, T., Flaxman, A. D., Danaei, G., Shibuya, K., Adair-Rohani, H., & Ezzati, M. (2012). A comparative risk assessment of burden of disease and injury attributable to 67 risk factors and risk factor clusters in 21 regions, 1990–2010: A systematic analysis for the Global Burden of Disease Study 2010. The Lancet, 380, 2224–2260. http://dx.doi.org/10.1016/S0140-6736(12)61766-8 Mamun, A. A., Alati, R., O’Callaghan, M., Hayatbakhsh, M. R., O’Callaghan, F. V., Najman, J. M., Williams, J. M., & Bor, W. (2007). Does childhood sexual abuse have an effect on young adults’ nicotine disorder (dependence or withdrawal)? Evidence from a birth cohort study. Addiction, 102(4), 647–654. http://dx.doi.org/10.1111/j.1360-0443.2006.01732.x Mazza, D. (1996). Physical, sexual and emotional violence against women: A general practice-based prevalence study. Medical Journal of Australia, 164(1), 14. Mazza, D. (2001). The physical, sexual and emotional violence history of middle-aged women: A community-based prevalence study. Medical Journal of Australia, 175(4), 199. McGuigan, W. M., & Middlemiss, W. (2005). Sexual abuse in childhood and interpersonal violence in adulthood: A cumulative impact on depressive symptoms in women. Journal of Interpersonal Violence, 20(10), 1271–1287. http://dx.doi.org/10.1177/0886260505278107 Messman-Moore, T. L., Long, P. J., & Siegfried, N. J. (2000). The revictimization of child sexual abuse survivors: An examination of the adjustment of college women with child sexual abuse, adult sexual assault, and adult physical abuse. Child Maltreatment, 5(1), 18–27. http://dx.doi.org/ 10.1177/1077559500005001003 Mills, R., Scott, J., Alati, R., O’Callaghan, M., Najman, J. M., & Strathearn, L. (2013). Child maltreatment and adolescent mental health problems in a large birth cohort. Child Abuse & Neglect, 37(5), 292–302. http://dx.doi.org/10.1016/j.chiabu.2012.11.008 Moher, D., Liberati, A., Tetzlaff, J., & Altman, D. G. (2009). Preferred reporting items for systematic reviews and meta-analyses: The PRISMA Statement. PLoS Medicine, 6(6), e1000097. http://dx.doi.org/10.1371/journal.pmed1000097 Moore, E. E., Romaniuk, H., Olsson, C. A., Jayasinghe, Y., Carlin, J. B., & Patton, G. C. (2010). The prevalence of childhood sexual abuse and adolescent unwanted sexual contact among boys and girls living in Victoria, Australia. Child Abuse & Neglect, 34(5), 379–385. http://dx.doi.org/10.1016/j.chiabu.2010.01.004 Mouzos, J., & Makkai, T. (2004). Women’s experiences of male violence: Findings from the Australian component of the International Violence Against Women Survey (IVAWS). Canberra, ACT: Australian Institute of Criminology. Mullen, P. E., Martin, J. L., Anderson, J. C., Romans, S. E., & Herbison, G. P. (1996). The long-term impact of the physical, emotional, and sexual abuse of children: A community study. Child Abuse & Neglect, 20(1), 7–21. http://dx.doi.org/10.1016/0145-2134(95)00112-3 Najman, J. M., Bor, W., O’Callaghan, M., Williams, G. M., Aird, R., & Shuttlewood, G. (2005). Cohort profile: The Mater-University of Queensland Study of Pregnancy (MUSP). International Journal of Epidemiology, 34(5), 992–997. http://dx.doi.org/10.1093/ije/dyi119 Najman, J. M., Dunne, M. P., Purdie, D. M., Boyle, F. M., & Coxeter, P. D. (2005). Sexual abuse in childhood and sexual dysfunction in adulthood: An Australian population-based study. Archives of Sexual Behavior, 34(5), 517–526. http://dx.doi.org/10.1007/s10508-005-6277-6 Nelson, E. C., Heath, A. C., Lynskey, M. T., Bucholz, K. K., Madden, P. A., Statham, D. J., & Martin, N. G. (2006). Childhood sexual abuse and risks for licit and illicit drug-related outcomes: A twin study. Psychological Medicine, 36(10), 1473–1483. http://dx.doi.org/10.1017/S0033291706008397 Nguyen, H. T., Dunne, M. P., & Le, A. V. (2010). Multiple types of child maltreatment and adolescent mental health in Viet Nam. Bulletin of the World Health Organization, 88(1), 22–30. http://dx.doi.org/10.2471/BLT.08.060061 Norman, R. E., Schneider, M., Bradshaw, D., Jewkes, R., Abrahams, N., Matzopoulos, R., & Vos, T. (2010). Interpersonal violence: An important risk factor for disease and injury in South Africa. Population Health Metrics, 8(32) http://dx.doi.org/10.1371/journal.pmed.1001349 Norman, R. E., Byambaa, M., De, R., Butchart, A., Scott, J., & Vos, T. (2012). The long-term health consequences of child physical abuse, emotional abuse, and neglect: A systematic review and meta-analysis. PLoS Medicine, 9(11), e1001349. Pereda, N., Guilera, G., Forns, M., & Gómez-Benito, J. (2009). The prevalence of child sexual abuse in community and student samples: A meta-analysis. Clinical Psychology Review, 29(4), 328–338. http://dx.doi.org/10.1016/j.cpr.2009.02.007 Price-Robertson, R., Smart, D., & Bromfield, L. (2010). Family is for life: How childhood experiences within the family help or hinder the lives of young adults. Family Matters, 85, 7–17. Reeve, R., & van Gool, K. (2013). Modelling the relationship between child abuse and long-term health care costs and wellbeing: Results from an Australian community-based survey. Economic Record, 89(286), 300–318. http://dx.doi.org/10.1111/1475-4932.12044 Rikhye, K., Tyrka, A. R., Kelly, M. M., Gagne, G. G., Jr., Mello, A. F., Mello, M. F., Price, L. H., & Carpenter, L. L. (2008). Interplay between childhood maltreatment, parental bonding, and gender effects: Impact on quality of life. Child Abuse & Neglect, 32(1), 19–34. http://dx.doi.org/10.1016/j.chiabu.2007.04.012 Rosenman, S., & Rodgers, B. (2004). Childhood adversity in an Australian population. Social Psychiatry and Psychiatric Epidemiology, 39(9), 695–702. http://dx.doi.org/10.1007/s00127-004-0802-0 Schneider, R., Baumrind, N., & Kimerling, R. (2007). Exposure to child abuse and risk for mental health problems in women. Violence and Victims, 22(5), 620–631. http://dx.doi.org/10.1891/088667007782312140 StataCorp. (2011). Stata Statistical Software: Release 12. College Station, TX: StataCorp LP. Stoltenborgh, M., van Ijzendoorn, M. H., Euser, E. M., & Bakermans-Kranenburg, M. J. (2011). A global perspective on child sexual abuse: Meta-analysis of prevalence around the world. Child Maltreatment, 16(2), 79–101. http://dx.doi.org/10.1177/1077559511403920 Stoltenborgh, M., Bakermans-Kranenburg, M. J., Alink, L. R. A., & van IJzendoorn, M. H. (2012). The universality of childhood emotional abuse: A meta-analysis of worldwide prevalence. Journal of Aggression, Maltreatment and Trauma, 21(8), 870–890. http://dx.doi.org/10.1080/10926771.2012.708014 Stoltenborgh, M., Bakermans-Kranenburg, M. J., & van IJzendoorn, M. H. (2013). The neglect of child neglect: A meta-analytic review of the prevalence of neglect. Social Psychiatry and Psychiatric Epidemiology, 48, 345–355. http://dx.doi.org/10.1007/s00127-012-0549-y Stoltenborgh, M., Bakermans-Kranenburg, M. J., van IJzendoorn, M. H., & Alink, L. R. A. (2013). Cultural–geographical differences in the occurrence of child physical abuse? A meta-analysis of global prevalence. International Journal of Psychology, 48(2), 81–94. http://dx.doi.org/10.1080/ 00207594.2012.697165 Straus, M. A., & Savage, S. A. (2005). Neglectful behavior by parents in the life history of university students in 17 countries and its relation to violence against dating partners. Child Maltreatment, 10(2), 124–135. http://dx.doi.org/10.1177/1077559505275507 Teicher, M. H., Samson, J. A., Polcari, A., & McGreenery, C. E. (2006). Sticks, stones, and hurtful words: Relative effects of various forms of childhood maltreatment. American Journal of Psychiatry, 163(6), 993–1000. http://dx.doi.org/10.1176/ajp.2006.163.6.993 Turner, H. A., Finkelhor, D., & Ormrod, R. (2006). The effect of lifetime victimization on the mental health of children and adolescents. Social Science & Medicine, 62(1), 13–27. http://dx.doi.org/10.1016/j.socscimed.2005.05.030 Vos, T., Flaxman, A. D., Naghavi, M., Lozano, R., Michaud, C., Ezzati, M.,.., & Moradi-Lakeh, M. (2012). Years lived with disability (YLDs) for 1160 sequelae of 289 diseases and injuries 1990–2010: A systematic analysis for the Global Burden of Disease Study 2010. The Lancet, 380(9859), 2163–2196. http://dx.doi.org/10.1016/S0140-6736(12)61729-2 Watson, B., & Halford, W. K. (2010). Classes of childhood sexual abuse and women’s adult couple relationships. Violence and Victims, 25(4), 518–535. http://dx.doi.org/10.1891/0886-6708.25.4.518

220

S.E. Moore et al. / Child Abuse & Neglect 48 (2015) 208–220

Widom, C. S. (1989). The cycle of violence. Science, 244(4901), 160–166. http://dx.doi.org/10.1126/science.2704995 Wildeman, C., Emanuel, N., Leventhal, J. M., Putnam-Hornstein, E., Waldfogel, J., & Lee, H. (2014). The prevalence of confirmed maltreatment among us children, 2004 to 2011. JAMA Pediatrics, 168(8), 706–713. http://dx.doi.org/10.1001/jamapediatrics.2014.410 World Health Organization. (1992). International Classification of Diseases and Related Health Conditions (10th ed.). Geneva: World Health Organization. Zoroglu, S. S., Tuzun, U., Sar, V., Tutkun, H., Savac¸s, H. A., Ozturk, M., Alyanak, B., & Kora, M. E. (2003). Suicide attempt and self-mutilation among Turkish high school students in relation with abuse, neglect and dissociation. Psychiatry and Clinical Neurosciences, 57(1), 119–126. http://dx.doi.org/10.1046/j.1440-1819.2003.01088.x

Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/ j.chiabu.2015.05.006.