[Skip to Content]
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
October 5, 2009

A Longitudinal Study of Maternal Depression and Child Maltreatment in a National Sample of Families Investigated by Child Protective Services

Author Affiliations

Author Affiliations: Departments of Society, Human Development, and Health (Drs Conron, Koenen, Buka, and Gortmaker) and Epidemiology (Drs Koenen and Buka), Harvard School of Public Health, Boston, Massachusetts; Department of Psychiatry, Children's Hospital, Boston, Massachusetts (Dr Beardslee); and Department of Community Health, Brown University School of Medicine, Providence, Rhode Island (Dr Buka).

Arch Pediatr Adolesc Med. 2009;163(10):922-930. doi:10.1001/archpediatrics.2009.176

Objective  To assess whether a change in depression predicts a mother's change in maltreatment.

Design  Observational, repeated measures study.

Setting  National Survey of Child and Adolescent Well-being, 1999 to 2004.

Participants  Mothers who retained custody of a child aged 0 to 15 years following a maltreatment investigation and completed at least 2 of 3 surveys (n = 2386).

Main Exposure  Change in depression status between baseline and 18- and 36-month follow-ups, assessed with the Composite International Diagnostic Interview Short Form.

Main Outcome Measures  Change in psychological aggression, physical assault, and neglect between baseline and 18- and 36-month follow-ups, assessed with the Conflict Tactics Scale Parent-Child version.

Results  One-third (35.5%) of mothers experienced onset or remission of depression. Onset of depression was associated with an increase of 2.3 (95% confidence interval, 0.2-4.4) psychologically aggressive acts in an average 12-month period, but was not statistically significantly associated with change in physical assault or neglect.

Conclusion  Depression is positively associated with maternal perpetration of psychological aggression in high-risk families.

As evidence accumulates on the short- and long-term consequences of child maltreatment, pressure to identify effective prevention strategies mounts. Emotional, social, and behavioral problems are among the short-term manifestations of child maltreatment,16 while psychiatric disorder,711 irritable bowel syndrome, chronic pain disorders, obesity, diabetes, heart disease, and cancer are among the long-term correlates of neglect and abuse.1214 Child maltreatment affects many children. In 2006, 3.3 million referrals regarding approximately 6.0 million children were made to child protective service agencies in the United States.15 Unfortunately, few interventions have proven successful in reducing the incidence of maltreatment, especially among high-risk families.1619

Depression is common, particularly among women, and is associated with a 2- to 3-fold increased risk of perpetrating physical abuse, psychological aggression, and medical neglect.2022 At least 8% of US children live with a parent who is currently depressed.23 Depression is even more prevalent (23%) among custodial caregivers investigated for child maltreatment.24 Although an association between emotional distress and poor parenting has been documented in several observational studies,2022,2534 it remains unclear whether maternal depression is causally related to child maltreatment.

Most maltreatment research has relied on cross-sectional comparisons between depressed and nondepressed mothers of young children.21,2527,34 Even studies that collected temporally sequenced data relied on single measures of depression, maltreatment, and potential confounders.20,29 Such studies are vulnerable to bias due to confounding.35 Maternal depression is associated with many other risks (eg, substance abuse, limited social support, financial strain)3641 that are correlated with poor parenting—risks that could create a spurious association between depression and maltreatment. Women who were maltreated as children are more likely to be depressed in adulthood810 and to maltreat their own children.28,38 Intimate partner victimization, child behavior problems, and maternal depression frequently co-occur,4042 and each may increase the risk of maltreatment. Spurious associations will arise if depressed and nondepressed mothers differ on correlated risks that are not well controlled in statistical models.

In this study, we make use of longitudinal data collected from a nationally representative sample of families investigated for maltreatment to examine whether change in depression status predicts change in maltreatment. By examining whether a mother's behavior differs in the presence and absence of depression, we come closer to approximating the counterfactual condition43 and reduce bias due to confounding.35 We use the fixed effects regression approach articulated by Paul Allison.44 Bias due to temporally stable differences between depressed and nondepressed mothers (eg, childhood victimization) is effectively controlled using fixed effects regression.44 A correlate that does not change over time cannot cause change in maltreating behavior. Bias associated with correlates that do change within mothers over time, which could serve as alternative explanations for change in maltreatment, is reduced by modeling repeated measures of potential confounders.

Understanding the causal nature of the association between maternal depression and child maltreatment is important because scarce public resources must be wisely invested to reduce maltreatment. Effective interventions such as the Nurse Home Visiting Program45,46 and Parent-Child Interaction Therapy47 require significant public resources and/or high levels of parental motivation. Thus, cost and effort pose significant challenges to implementing such interventions on a broad scale.19 In contrast, depression is highly responsive to treatment48 and can be addressed with less parental effort in the existing health system.

Study design and population

A longitudinal study using observational data collected for the National Survey of Child and Adolescent Well-being (NSCAW) was conceived to investigate whether change in maternal depression status predicts change in maltreating behavior. We hypothesized that depression would be positively associated with psychological aggression, physical assault, and child neglect. A representative sample of families (n = 5501) involved in maltreatment investigations completed by 92 child welfare agencies in 36 states was obtained.49 Participation in the NSCAW was 64.2%. Caregivers completed baseline surveys 2 to 6 months after the investigation closed in 1999 or 2000 and at 18 and 36 months. Interviews were conducted in the home by trained research staff.50 Details of the NSCAW are reported elsewhere.49 Approval to conduct secondary analysis of the NSCAW was granted by the Human Subjects Committee at the Harvard School of Public Health.

Permanent caregivers who participated in at least 2 of 3 waves of data collection were eligible for the current study. Custody for at least 12 months prior to NSCAW enrollment was an additional criterion for caregivers of children older than 1 year. These inclusion criteria were necessary to evaluate the change in 12-month maltreating behavior of the same caregiver over time. Despite a truncated reporting period for caregivers of children younger than 1 year at baseline, we opted to include these families in the study because rates of reported maltreatment are the highest for this age group.15 Men were excluded because they may experience depression differently from women51,52 and were too few in number to analyze separately. Thus, the final sample was restricted to women, most of whom were mothers.


The Composite International Diagnostic Interview Short-Form (CIDI-SF) scale53 was used to measure the 12-month incidence of a major depressive episode according to American Psychiatric Association diagnostic criteria.54 Mothers who endorsed 3 or more of 7 possible symptoms of a major depressive episode were classified as depressed.55 Psychometric research indicates good test-retest and interrater reliability and acceptable diagnostic validity of the CIDI long-form scale.56,57 The CIDI-SF diagnostic classifications are highly concordant with those obtained via the longer measure.53


Twelve-month psychological aggression, physical assault, and neglect were assessed on separate Conflict Tactics Scale Parent-Child version measures.58 Five threatening (eg, to hit or send the child away) and emotionally injurious (eg, calling the child dumb) behaviors were included on the psychological aggression subscale. The 13-item physical assault subscale included items such as shaking, hitting, and beating. Four forms of neglect, lack of supervision, and emotional, medical, and physical neglect were captured on the neglect index. Item responses for the 12-month incidence of each behavior (never, once, twice, 3 to 5, 6 to 10, 11 to 20, or more than 20 times) were assigned the values of 0, 1, 2, 4, 8, 15, and 25, respectively, and summed for each measure. Scores were treated continuously. Internal consistency reliability in this sample was adequate for the psychological aggression scale and poor59 for the physical assault and neglect measures (baseline Cronbach α coefficients were .65, .50, and .32, respectively). Evidence in support of the construct validity of each Conflict Tactics Scale Parent-Child version measure has been produced in several samples.58,60,61


Symptoms of substance dependence were measured with the CIDI-SF alcohol and drug dependence scales,53 summed, and treated continuously. The presence of a live-in intimate partner was treated as a binary variable (yes/no) and the number of children in the household (1 through 5 or more) was treated continuously. Employment status was assessed at baseline and at 36 months and treated as an ordinal variable (unemployed, part-time employment, full-time employment).

Economic status was operationalized as the income to needs ratio according to US census criteria62 and treated continuously. Annual family income was collected in $5000 intervals, beginning with less than $5000/y through more than $50 000/y. Income was recoded to the midpoint for each category, except for the highest. The highest income group was recoded to the size-adjusted median family income or, in a few instances, to the overall median family income so that every income was more than $50 000/y.63 Recoded family income was divided by size-adjusted poverty thresholds64 to obtain an income to needs ratio for each measurement occasion (2000, 2001-2002, and 2003).

Twelve-month exposure to physical intimate partner violence (IPV) was measured with the 9-item version of the Conflict Tactics Scale.65 The Conflict Tactics Scale was scored like the Conflict Tactics Scale Parent-Child version and then coded as an ordinal variable (none, minor, and severe violence). When children were younger than 2 years, child behavior problems were assessed with difficult temperament scales developed for the National Longitudinal Survey of Youth.66 Mothers completed the aggressive and delinquent or destructive behavior scales from the Child Behavior Checklist for children aged 2 to 3 and 4 to 18 years.67,68 Raw composite scores for child behavior problems were converted to z scores. This allowed for consistent interpretation of within-child change across instruments, despite differences in the composition and length of each instrument.

Statistical analysis

Analyses were conducted in SAS 9.269 (SAS Inc, Cary, North Carolina) using survey procedures. Sampling weights were used to address unequal probabilities of selection and participation. Taylor expansion method variance estimates accounted for the stratified and clustered survey design as well as the clustering of repeated observations within individuals.70,71 Data were analyzed according to the fixed-effects approach.44

Fixed-effects analyses estimate a fixed effect for every person in the data set and reduce bias in the estimate of an association between an exposure and an outcome by controlling for temporally stable (time-invariant) characteristics, both measured and unmeasured.44 Time-varying covariates were subject mean–centered so that associations between covariates and maltreatment outcomes were within-person estimates. Regression coefficients are appropriately interpreted as average within-mother associations between change in a covariate and change in an outcome. Mothers who were missing data for time-varying covariates or outcomes were excluded from the study sample. All tests of statistical significance were 2-tailed and relied on an α of .05.

Although time-varying covariates were correlated, correlation coefficients were smaller than 0.50, indicating that multicollinearity did not substantially affect our ability to examine relationships between depression and maltreatment.72 The linear regression assumption of normality of the residuals was reasonably met. Sensitivity analyses (not shown) comparing estimates from models with and without children younger than 1 year indicate that including children younger than 1 year in our sample, despite a truncated baseline maltreatment reporting period, did not appreciably alter our results.


A total of 2683 mothers who retained custody of a child aged 0 to 15 years following a maltreatment investigation, were custodians for at least 1 year prior to NSCAW participation, or were parents of a child younger than 1 year and completed at least 2 surveys were eligible for the current study. Most (90.4%; n = 2386) provided complete data for the variables used in this study and thus were included in our sample. Follow-up was 92.4% at 18 months and 89.3% at 36 months.

Our sample was diverse by maternal age, race/ethnicity, and educational attainment, as well as child age and sex (Table 1). Neglect was the most frequent form of alleged maltreatment at the event that determined study eligibility. Allegations of physical neglect and/or inadequate supervision were made on behalf of 54.2% of children. Physical abuse was alleged regarding one-third (32.7%) of the sample, while emotional abuse was alleged on behalf of 10.7% of children.

Table 1. 
Baseline Characteristics of Custodial Mothers, Their Children Aged 0 to 15 Years, and the Index Maltreatment Event
Baseline Characteristics of Custodial Mothers, Their Children Aged 0 to 15 Years, and the Index Maltreatment Event

Approximately one-fourth of mothers met symptom criteria for a depressive episode per wave of data collection (Table 2). Episodes lasted an average of 12 to 14 weeks. Approximately one-third (35.5%) of mothers experienced onset or remission of depression during the 36-month study; 15.2% experienced depression onset, 12.9% experienced the remission of an episode, and 7.4% experienced both onset and remission. At baseline, mothers reported an average of 18.8 psychologically aggressive acts, 8.0 acts of physical assault, and 3.4 neglectful acts during the prior 12 months. The average number of reported acts of psychological aggression and physical assault declined over time.

Table 2. 
Maternal Depression, Maltreating Behavior, and Other Time-Varying Characteristics by Wave in the NSCAW
Maternal Depression, Maltreating Behavior, and Other Time-Varying Characteristics by Wave in the NSCAW

Fixed-effects regression results (Table 3) indicate that change in depression status was associated with an average change of 3.10 (95% CI, 0.86-5.35) psychologically aggressive acts within a 12-month period (model 1). This association was slightly attenuated (models 2-5) on introduction of controls for household composition, family socioeconomic status, and intimate partner violence. The introduction of child behavior problems in model 6 resulted in a noticeable reduction in the magnitude of the depression regression coefficient, although the association between maternal depression and psychological aggression (β = 2.28; 95% CI, 0.17-4.38) remained statistically significant. Results from bivariate analyses (not shown) indicate that relations between maternal depression and psychological aggression held in both directions. Mothers who reported onset of depression reported an increase in psychological aggression, while those who experienced remission of a depressive episode reported a reduction in psychological aggression.

Table 3. 
Fixed-Effects Linear Regression Models of the Relationship Between Changes in Maternal Depression and Maltreating Behavior
Fixed-Effects Linear Regression Models of the Relationship Between Changes in Maternal Depression and Maltreating Behavior

Change in depression was not statistically significantly associated with change in physical assault in the unadjusted model (model 1), nor after subsequent adjustment for time-varying covariates (models 2-6). Change in depression was statistically significantly associated with change in neglect in an unadjusted model (model 1); however, this association was attenuated after adjusting for symptoms of alcohol and drug dependence (model 2). Other significant findings are described below.

Change in employment status (none to part-time, part-time to full-time), assessed at baseline and the 36-month follow-up, predicted a decrease of 1.19 (95% CI, −2.29 to −0.08) acts of psychological aggression, while the addition of an intimate partner to the household was associated with 2.40 (95% CI, −4.50 to −0.30) fewer acts of psychological aggression, adjusting for depression and all other time-varying covariates (model 6). Increased exposure in level (none to minor, minor to severe) of physical intimate partner victimization predicted an average increase of 1.49 (95% CI, 0.66-2.31) acts of physical assault and 0.98 (95% CI, 0.02-1.95) neglectful acts per 12-month period. An increase in child behavior problems was associated with increases in both psychological aggression (4.42 acts; 95% CI, 3.24-5.59) and physical assault (1.79 acts; 95% CI, 0.96-2.61).


This study assessed the contribution of maternal depression to the perpetration of 3 types of child maltreatment in a nationally representative sample of families investigated for maltreatment. Our findings suggest that depression is a causal determinant of psychological aggression in high-risk mothers of children aged 0 through 15 years, but not of physical assault or neglect. These findings are partially consistent with published research.

Studies have documented an association between parental depression and psychological aggression in both a general household sample73 and in higher-risk samples.22,28 Several studies describe a relationship between maternal depression and physical abuse20 including spanking,73 harsh parenting,28 and the perpetration of any severe physical assault in a period of 12 months.22,27 Less research has been conducted on determinants of neglect. Depressed parents identified through a general population survey were no more likely to report neglect than nondepressed parents.20 In contrast, depressed mothers who participated in a cohort study of women with extensive childhood trauma histories were more likely to report any 12-month neglect than mothers with fewer depressive symptoms.27

There are several possible explanations for differences between our null physical assault and neglect findings and the significant findings of others. Notably, our repeated measures study was better able to control for bias due to confounding than prior research. Yet, other limitations may have resulted in type II errors (failing to reject the null hypothesis).74

Other aspects of mothers' lives were associated with change in child maltreatment in our study, adjusting for depression. A decline in psychological aggression was associated with an increase in the amount of time mothers worked. Some research indicates that employment may provide respite from the stress of parenting and/or may increase the amount of time children spent in structured and safe child care environments that foster social behavior.75

The addition of an intimate partner to the household was associated with a decrease in the perpetration of psychological aggression in our study, independent of family economic status and IPV. Lower rates of physical and psychological aggression toward very young children were reported by low-income mothers in nonviolent relationships compared with single mothers enrolled in a home visitation program.22 The presence of a nonabusive partner may reduce the likelihood of maltreatment by providing positive relationship models and by reducing strain on mothers.76

Several studies have documented the co-occurrence of IPV and child maltreatment,77 but few have examined the role of IPV in predicting maltreatment. One study of families who participated in a nurse home visiting program found that the presence of IPV during infancy was positively associated with agency-reported maltreatment in the first 5 years of life.78 Our study may be the first to document an association between change in exposure to IPV and change in physical assault and neglect.

Perceived child behavior problems were positively associated with increases in psychological aggression and physical assault in our study. Relations between child behavior and parenting have been noted by others. In fact, child behavior, parenting, and maternal depression appear to be reciprocally related.40 Interventions that assist parents in effectively disciplining and positively engaging with children describe improvements in parenting practices, child behavior, and parental mental health.47,79,80

Specific strengths of our study merit highlighting. Our study adds to prior study by using a fixed-effects approach44 to analyze longitudinal data and by evaluating prospective relationships between depression and 3 separate forms of maltreatment in a nationally representative, high-risk sample. Our focus on maltreatment rather than parenting style or degree of response directs attention to parenting behavior that is socially unacceptable rather than less than optimal. Information about maltreatment and substance use was collected via automated, computerized methods. Stigmatized behavior is more frequently reported when recorded through self-administered automated methods than when collected by an interviewer.81

Several limitations warrant consideration. Underreporting of physical assault and neglect in our sample may have reduced our ability to detect an association between change in depression and both types of maltreatment by creating a floor effect.74 In addition, low reliability, particularly of the assault and neglect measures, reduced the precision of our estimates and our ability to detect statistically significant relations.82 Low reliability of the neglect measure may indicate that neglect is a multidimensional construct59,83; thus, analyses of neglect subtypes may yield different results than those described here. Controlling for conditions that are caused by depression (eg, mediators) would inappropriately reduce our estimate of the strength of associations between maltreatment types and depression; however, controlling for such conditions (eg, substance abuse, child behavior problems) may also (appropriately) attenuate associations because they are confounders. Thus, we opted to include them in our models and to present estimates that are somewhat conservative. Estimates presented in final models may also be too conservative because maternal depression is associated with the overreporting of child behavior problems.84,85

Although changes in depression status and psychological aggression occurred during the same 12-month period, we cannot be certain that a change in depression preceded a change in aggression. Unfortunately, NSCAW data (collected at 18-month intervals) would not allow us to use a lagged term to address temporality in a manner that is consistent with our understanding of the lag between mood and behavior. Finally, our findings are generalizable to mothers of children who were the subject of a maltreatment report.

Our findings strongly suggest that reducing maternal depression may reduce children's exposure to psychological aggression, yet approaches to affect such change were not estimated in our study. Moreover, further investigation of the association between maternal depression and physical assault and neglect is needed, given our study limitations. A randomized controlled trial could address many of these limitations and provide stronger evidence regarding a causal link between maternal depression and child maltreatment. Depressed, high-risk mothers who are randomly assigned to receive high-quality depression treatment could be compared with those assigned to standard care on their maltreating behavior.

Our study found that maternal depression and IPV victimization are associated with the perpetration of child maltreatment in families investigated by child protective service agencies. An accurate tool with which to identify maltreating parents is lacking17; however, research indicates that adults can be screened quickly and effectively for depression and IPV in primary care settings.8691 Routine screening is recommended by the American Academy of Pediatrics,92 the American Association of Family Physicians,93 and the US Preventive Services Task Force.94 This study contributes additional evidence in support of these recommendations.

Correspondence: Kerith J. Conron, ScD, MPH, Department of Society, Human Development, and Health, Harvard School of Public Health, 677 Huntington Ave, Boston, MA 02115 (kconron@post.harvard.edu).

Accepted for Publication: March 6, 2009.

Author Contributions: Dr Conron had full access to the National Survey of Child and Adolescent Well-being data analyzed in this study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Conron, Gortmaker, and Buka. Acquisition of data: Conron and Gortmaker. Analysis and interpretation of data: Conron, Gortmaker, Koenen, and Beardslee. Drafting of the manuscript: Conron and Beardslee. Critical revision of the manuscript for important intellectual content: Conron, Beardslee, Koenen, Buka, and Gortmaker. Statistical analysis: Conron and Gortmaker. Obtained funding: Conron. Administrative, technical, and material support: Beardslee. Study supervision: Gortmaker, Koenen, and Buka.

Financial Disclosure: None reported.

Funding/support: This study was supported by training grant MH15161 from the National Institutes of Health (Dr Conron).

Disclaimer: This manuscript includes data from the National Survey of Child and Adolescent Well-Being, which was developed under contract with the Administration on Children, Youth, and Families, US Department of Health and Human Services (ACYF/DHHS). The data were obtained from the National Data Archive on Child Abuse and Neglect. The information and options expressed herein reflect solely the position of the authors and do not indicate the support or endorsement of ACYF/DHHS. Dr Conron conducted this work as a graduate student at the Harvard School of Public Health and revised the manuscript at the Family Research Lab, University of New Hampshire.

Bolger  KEPatterson  CJ Developmental pathways from child maltreatment to peer rejection.  Child Dev 2001;72 (2) 549- 568PubMed
Boney-McCoy  SFinkelhor  D Is youth victimization related to trauma symptoms and depression after controlling for prior symptoms and family relationships? a longitudinal, prospective study.  J Consult Clin Psychol 1996;64 (6) 1406- 1416PubMed
Dubowitz  HPapas  MABlack  MMStarr  RH  Jr Child neglect: outcomes in high-risk urban preschoolers.  Pediatrics 2002;109 (6) 1100- 1107PubMed
Jaffee  SRCaspi  AMoffitt  TETaylor  A Physical maltreatment victim to antisocial child: evidence of an environmentally mediated process.  J Abnorm Psychol 2004;113 (1) 44- 55PubMed
Kim  JCicchetti  D A longitudinal study of child maltreatment, mother-child relationship quality and maladjustment: the role of self-esteem and social competence.  J Abnorm Child Psychol 2004;32 (4) 341- 354PubMed
Lansford  JEDodge  KAPettit  GSBates  JECrozier  JKaplow  J A 12-year prospective study of the long-term effects of early child physical maltreatment on psychological, behavioral, and academic problems in adolescence.  Arch Pediatr Adolesc Med 2002;156 (8) 824- 830PubMed
Brown  JCohen  PJohnson  JGSmailes  EM Childhood abuse and neglect: specificity of effects on adolescent and young adult depression and suicidality.  J Am Acad Child Adolesc Psychiatry 1999;38 (12) 1490- 1496PubMed
McCauley  JKern  DEKolodner  K  et al.  Clinical characteristics of women with a history of childhood abuse.  JAMA 1997;277 (17) 1362- 1368PubMed
Molnar  BEBuka  SLKessler  RC Child sexual abuse and subsequent psychopathology: results from the National Comorbidity Survey.  Am J Public Health 2001;91 (5) 753- 760PubMed
Widom  CSDuMont  KCzaja  SJ A prospective investigation of major depressive disorder and comorbidity in abused and neglected children grown up.  Arch Gen Psychiatry 2007;64 (1) 49- 56PubMed
Widom  CSMarmorstein  NRWhite  HR Childhood victimization and illicit drug use in middle adulthood.  Psychol Addict Behav 2006;20 (4) 394- 403PubMed
Felitti  VJAnda  RFNordenberg  D  et al.  Relationship of childhood abuse and household dysfunction to many of the leading causes of death in adults: the Adverse Childhood Experiences (ACE) Study.  Am J Prev Med 1998;14 (4) 245- 258PubMed
Springer  KWSheridan  JKuo  DCarnes  M The long-term health outcomes of childhood abuse: an overview and a call to action.  J Gen Intern Med 2003;18 (10) 864- 870PubMed
Williamson  DFThompson  TJAnda  RFDietz  WHFelitti  V Body weight and obesity in adults and self-reported abuse in childhood.  Int J Obes Relat Metab Disord 2002;26 (8) 1075- 1082
 Child Maltreatment 2006  US Department of Health and Human Services, Administration on Children, Youth and Families Web site. http://www.acf.hhs.gov/programs/cb/pubs/cm06/cm06.pdf. Accessed September 1, 2008
Barlow  JJohnston  IKendrick  DPolnay  LStewart-Brown  S Individual and group-based parenting programmes for the treatment of physical child abuse and neglect.  Cochrane Database Syst Rev 2006;3CD005463
MacMillan  HL Child maltreatment: what we know in the year 2000.  Can J Psychiatry 2000;45 (8) 702- 709PubMed
Chaffin  MBonner  BLHill  RF Family preservation and family support programs: child maltreatment outcomes across client risk levels and program types.  Child Abuse Negl 2001;25 (10) 1269- 1289PubMed
Klevens  JWhitaker  DJ Primary prevention of child physical abuse and neglect: gaps and promising directions.  Child Maltreat 2007;12 (4) 364- 377PubMed
Chaffin  MKelleher  KHollenberg  J Onset of physical abuse and neglect: psychiatric, substance abuse, and social risk factors from prospective community data.  Child Abuse Negl 1996;20 (3) 191- 203PubMed
Weissman  MMFeder  APilowsky  DJ  et al.  Depressed mothers coming to primary care: maternal reports of problems with their children.  J Affect Disord 2004;78 (2) 93- 100PubMed
Windham  AMRosenberg  LFuddy  LMcFarlane  ESia  CDuggan  AK Risk of mother-reported child abuse in the first 3 years of life.  Child Abuse Negl 2004;28 (6) 645- 667PubMed
Weissman  MMGammon  GDJohn  K  et al.  Children of depressed parents: increased psychopathology and early onset of major depression.  Arch Gen Psychiatry 1987;44 (10) 847- 853PubMed
 National survey of child and adolescent well-being (NSCAW): CPS sample component wave 1 data analysis report.  Children's Bureau of the Administration on Children and Families, US Department of Health and Human Services Web site. http://www.acf.hhs.gov/programs/opre/abuse_neglect/nscaw/reports/cps_sample/cps_report_revised_090105.pdf. Published April 2005. Accessed September 1, 2008
Zuravin  SJ Severity of maternal depression and three types of mother-to-child aggression.  Am J Orthopsychiatry 1989;59 (3) 377- 389PubMed
McLearn  KTMinkovitz  CSStrobino  DMMarks  EHou  W The timing of maternal depressive symptoms and mothers' parenting practices with young children: implications for pediatric practice.  Pediatrics 2006;118 (1) e174- e18210.1542/10.1542/peds.2005-1551
Banyard  VLWilliams  LMSiegel  JA The impact of complex trauma and depression on parenting: an exploration of mediating risk and protective factors.  Child Maltreat 2003;8 (4) 334- 349PubMed
Dubowitz  HBlack  MMKerr  MA  et al.  Type and timing of mothers' victimization: effects on mothers and children.  Pediatrics 2001;107 (4) 728- 735PubMed
Brown  JCohen  PJohnson  JGSalzinger  S A longitudinal analysis of risk factors for child maltreatment: findings of a 17-year prospective study of officially recorded and self-reported child abuse and neglect.  Child Abuse Negl 1998;22 (11) 1065- 1078PubMed
Jackson  APBrooks-Gunn  JHuang  CCGlassman  M Single mothers in low-wage jobs: financial strain, parenting, and preschoolers' outcomes.  Child Dev 2000;71 (5) 1409- 1423PubMed
Raver  CC Does work pay psychologically as well as economically? the role of employment in predicting depressive symptoms and parenting among low-income families.  Child Dev 2003;74 (6) 1720- 1736PubMed
Linver  MRBrooks-Gunn  JKohen  DE Family processes as pathways from income to young children's development.  Dev Psychol 2002;38 (5) 719- 734PubMed
Pachter  LMAuinger  PPalmer  RWeitzman  M Do parenting and the home environment, maternal depression, neighborhood, and chronic poverty affect child behavioral problems differently in different racial-ethnic groups?  Pediatrics 2006;117 (4) 1329- 1338PubMed
Solantaus  TLeinonen  JPunamaki  RL Children's mental health in times of economic recession: replication and extension of the family economic stress model in Finland.  Dev Psychol 2004;40 (3) 412- 429PubMed
Fitzmaurice  GMLaird  NMWare  JH Applied Longitudinal Analysis.  Hoboken, NJ John Wiley & Sons, Inc2004;
Grant  BFStinson  FSDawson  DA  et al.  Prevalence and co-occurrence of substance use disorders and independent mood and anxiety disorders: results from the National Epidemiologic Survey on Alcohol and Related Conditions.  Arch Gen Psychiatry 2004;61 (8) 807- 816PubMed
Dube  SRAnda  RFFelitti  VJCroft  JBEdwards  VJGiles  WH Growing up with parental alcohol abuse: exposure to childhood abuse, neglect, and household dysfunction.  Child Abuse Negl 2001;25 (12) 1627- 1640PubMed
Locke  TFNewcomb  M Child maltreatment, parent alcohol and drug-related problems, polydrug problems, and parenting practices: a test of gender differences and four theoretical perspectives.  J Fam Psychol 2004;18 (1) 120- 134PubMed
Dearing  ETaylor  BAMcCartney  K Implications of family income dynamics for women's depressive symptoms during the first 3 years after childbirth.  Am J Public Health 2004;94 (8) 1372- 1377PubMed
Elgar  FJMcGrath  PJWaschbusch  DAStewart  SHCurtis  LJ Mutual influences on maternal depression and child adjustment problems.  Clin Psychol Rev 2004;24 (4) 441- 459PubMed
Koverola  CPapas  MAPitts  SMurtaugh  CBlack  MMDubowitz  H Longitudinal investigation of the relationship among maternal victimization, depressive symptoms, social support, and children's behavior and development.  J Interpers Violence 2005;20 (12) 1523- 1546PubMed
Kitzmann  KMGaylord  NKHolt  ARKenny  ED Child witnesses to domestic violence: a meta-analytic review.  J Consult Clin Psychol 2003;71 (2) 339- 352PubMed
Rothman  KJGreenland  S Modern Epidemiology. 2nd ed. Philadelphia, PA Lippincott-Raven Publishers1998;
Allison  PD Fixed Effects Regression Methods for Longitudinal Data using SAS.  Washington, DC SAS Press2005;
Kitzman  HOlds  DLHenderson  CR  Jr  et al.  Effect of prenatal and infancy home visitation by nurses on pregnancy outcomes, childhood injuries, and repeated childbearing: a randomized controlled trial.  JAMA 1997;278 (8) 644- 652PubMed
Olds  DHenderson  CR  JrCole  R  et al.  Long-term effects of nurse home visitation on children's criminal and antisocial behavior: 15-year follow-up of a randomized controlled trial.  JAMA 1998;280 (14) 1238- 1244PubMed
Chaffin  MSilovsky  JFFunderburk  B  et al.  Parent-child interaction therapy with physically abusive parents: efficacy for reducing future abuse reports.  J Consult Clin Psychol 2004;72 (3) 500- 510PubMed
Miranda  JSchoenbaum  MSherbourne  CDuan  NWells  K Effects of primary care depression treatment on minority patients' clinical status and employment.  Arch Gen Psychiatry 2004;61 (8) 827- 834PubMed
NSCAW Research Group, Methodological lessons from the National Survey of Child and Adolescent Well-Being: the first three years of the USA's first national probability study of children and families investigated for abuse and neglect.  Child Youth Serv Rev 2002;24 (6/7) 513- 541
National Survey of Child and Adolescent Well-Being, Combined Waves 1-4: Data File User's Manual, Restricted Release Version.  Ithaca, NY Cornell University2004;
Kuehner  C Gender differences in unipolar depression: an update of epidemiological findings and possible explanations.  Acta Psychiatr Scand 2003;108 (3) 163- 174PubMed
Piccinelli  MWilkinson  G Gender differences in depression [critical review].  Br J Psychiatry 2000;177486- 492PubMed
Kessler  RCAndrews  GMroczek  DUstun  TBWittchen  HU The world health organization composite international diagnostic interview short-form (CIDI-SF).  Int J Methods Psychiatr Res 1998;7 (4) 171- 185
American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders. 4th ed. Washington, DC American Psychiatric Association1994;
Haarasilta  LMarttunen  MKaprio  JAro  H The 12-month prevalence and characteristics of major depressive episode in a representative nationwide sample of adolescents and young adults.  Psychol Med 2001;31 (7) 1169- 1179
Wittchen  HU Reliability and validity studies of the WHO–Composite International Diagnostic Interview (CIDI): a critical review.  J Psychiatr Res 1994;28 (1) 57- 84PubMed
Andrews  GPeters  L The psychometric properties of the Composite International Diagnostic Interview.  Soc Psychiatry Psychiatr Epidemiol 1998;33 (2) 80- 88PubMed
Straus  MAHamby  SLFinkelhor  DMoore  DWRunyan  D Identification of child maltreatment with the Parent-Child Conflict Tactics Scales: development and psychometric data for a national sample of American parents.  Child Abuse Negl 1998;22 (4) 249- 270PubMed
DeVellis  RF Scale Development: Theory and Applications: Applied Social Research Methods Series Volume 26. 2nd ed. Thousand Oaks, CA Sage Publications2003;
Bennett  DSSullivan  WMLewis  M Relations of parental report and observation of parenting to maltreatment history.  Child Maltreat 2006;11 (1) 63- 75PubMed
Straus  MAHamby  SL Measuring physical and psychological maltreatment of children with the Conflict Tactics Scales. Kaufman  Kantor GJasinksi  JL Out of the Darkness Contemporary Perspectives on Family Violence. Thousand Oaks, CA Sage Publications1997;119- 135
 How the Census Bureau measures poverty.  US Census Bureau Web site. http://www.census.gov/hhes/www/poverty/povdef.html. Accessed April 13, 2007
 Table F-8: size of family, all races by median and mean income, 1947 to 2005.  US Census Bureau Web site. http://www.census.gov/hhes/www/income/histinc/f08ar.html. Accessed November 9, 2007
 Poverty thresholds.  US Census Bureau Web site. http://www.census.gov/hhes/www/poverty/threshld.html. Accessed April 13, 2007
Straus  MADouglas  EM A short form of the Revised Conflict Tactics Scales, and typologies for severity and mutuality.  Violence Vict 2004;19 (5) 507- 520PubMed
Mott  FLBaker  PCBall  DEKeck  CCLenhart  SM The NLSY Children 1992: Description and Evaluation.  Columbus, OH Ohio State University Center for Human Resource Research1995;
Achenbach  TM Manual for the Child Behavior Checklist 2-3 and 1991 Profile.  Burlington, VT University of Vermont Department of Psychiatry1991;
Achenbach  TM Manual for the Child Behavior Checklist 4-18 and 1991 Profile.  Burlington, VT University of Vermont Department of Psychiatry1991;
SAS Institute Inc, SAS Version 9.2.  Cary, NC SAS Institute Inc2008;
Biemer  BChrist  SWheeless  SWiesen  C National Survey of Child and Adolescent Well-being (NSCAW).  Ithaca, NY Cornell University2005;
Williams  RL A note on robust variance estimation for cluster-correlated data.  Biometrics 2000;56 (2) 645- 646PubMed
Afifi  AAClark  V Computer-Aided Multivariate Analysis. 3rd ed. NY Chapman & Hall/CRC1997;
Regalado  MSareen  HInkelas  MWissow  LSHalfon  N Parents' discipline of young children: results from the National Survey of Early Childhood Health.  Pediatrics 2004;113 (6) ((suppl)) 1952- 1958PubMed
Shadish  WRCook  TDCampbell  DT Experimental and Quasi-experimental Designs for Generalized Causal Inference.  Boston, MA Houghton Mifflin Co2002;
Epps  SRHuston  AC Effects of a poverty intervention policy demonstration on parenting and child behavior: a test of the direction of effects.  Soc Sci Q 2007;88 (2) 344- 365
Belsky  J Etiology of child maltreatment: a developmental-ecological analysis.  Psychol Bull 1993;114 (3) 413- 434PubMed
Osofsky  JD Prevalence of children's exposure to domestic violence and child maltreatment: implications for prevention and intervention.  Clin Child Fam Psychol Rev 2003;6 (3) 161- 170PubMed
McGuigan  WMPratt  CC The predictive impact of domestic violence on three types of child maltreatment.  Child Abuse Negl 2001;25 (7) 869- 883PubMed
Carr  A Evidence-based practice in family therapy and systemic consultation I: child-focused problems.  J Fam Ther 2000;22 (1) 29- 60
Bearslee  WRWright  EJGladstone  TRForbes  P Long-term effects from a randomized trial of two public health preventive interventions for parental depression.  J Fam Psychol 2007;21 (4) 703- 713PubMed
Perlis  TEDes Jarlais  DCFriedman  SRArasteh  KTurner  CF Audio-computerized self-interviewing versus face-to-face interviewing for research data collection at drug abuse treatment programs.  Addiction 2004;99 (7) 885- 896PubMed
Fox  J Applied Regression Analysis, Linear Models and Related Methods.  Thousand Oaks, CA Sage Publications1997;
Dubowitz  HPitts  SCBlack  MM Measurement of three major subtypes of child neglect.  Child Maltreat 2004;9 (4) 344- 356PubMed
Chilcoat  HDBreslau  N Does psychiatric history bias mothers' reports? an application of a new analytic approach.  J Am Acad Child Adolesc Psychiatry 1997;36 (7) 971- 979PubMed
Briggs-Gowan  MJCarter  ASSchwab-Stone  M Discrepancies among mother, child, and teacher reports: examining the contributions of maternal depression and anxiety.  J Abnorm Child Psychol 1996;24 (6) 749- 765PubMed
Olson  ALDietrich  AJPrazar  G  et al.  Two approaches to maternal depression screening during well child visits.  J Dev Behav Pediatr 2005;26 (3) 169- 176PubMed
Dubowitz  HPrescott  LFeigelman  SLane  WKim  J Screening for intimate partner violence in a pediatric primary care clinic.  Pediatrics 2008;121 (1) e85- e91PubMed
Sohal  HEldridge  SFeder  G The sensitivity and specificity of four questions (HARK) to identify intimate partner violence: a diagnostic accuracy study in general practice.  BMC Fam Pract 2007;849PubMed
MacMillan  HLWathen  CNJamieson  E  et al. McMaster Violence Against Women Research Group, Approaches to screening for intimate partner violence in health care settings: a randomized trial.  JAMA 2006;296 (5) 530- 536PubMed
Shefet  DDascal-Weichhendler  HRubin  O  et al.  Domestic violence: a national simulation-based educational program to improve physicians' knowledge, skills and detection rates.  Med Teach 2007;29 (5) e133- e138PubMed
Dubowitz  HFeigelman  SLane  W  et al.  Screening for depression in an urban pediatric primary care clinic.  Pediatrics 2007;119 (3) 435- 443PubMed
American Academy of Pediatrics Committee on Child Abuse and Neglect, The role of the pediatrician in recognizing and intervening on behalf of abused women.  Pediatrics 1998;101 (6) 1091- 1092PubMed
American Academy of Family Physicians, Violence [position paper].  American Academy of Family Physicians We site. http://www.aafp.org/online/en/home/policy/policies/v/violencepositionpaper.html. Accessed October 24, 2008
US Preventive Services Task Force, Screening for depression: recommendations and rationale.  Agency for Healthcare Research and Quality Web site. http://www.ahrq.gov/clinic/3rduspstf/depression/depressrr.htm. Accessed September 1, 2008