Effect of child abuse and neglect on risk behaviors in inner-city minority female adolescents and young adults

Abstract

Background:

Over six million children each year are referred to child protective services for child abuse (sexual, physical and emotional) and neglect (physical and emotional).

Objective:

While the relationship between child sexual abuse and sexually transmitted infections has been documented, there has been little research regarding the effects of other forms of maltreatment.

Participants and setting:

882 inner-city females aged 12–20 years of age seen at a large adolescent and young adult (AYA) health center in New York City between 2012–2017.

Methods:

History of maltreatment was assessed using the Childhood Trauma Questionnaire. Associations with depressive symptoms, antisocial behavior, peer deviancy, drug/alcohol use, and risky sexual behaviors were assessed.

Results:

History of maltreatment was common in our cohort of inner-city AYA females, with 59.6 % reporting any type of maltreatment, including sexual abuse (17.5 %), physical abuse (19.5 %) or neglect (26.2 %), and emotional abuse (30.7 %) or neglect (40.4 %). We observed significant associations between all forms of maltreatment and risk of depression, drug/alcohol use, antisocial behaviors, peer deviancy, and risky sexual risk behaviors (including having a higher number of sexual partners, having a sexual partner 5+ years older, and anal sex). Physical and emotional abuse were associated with having unprotected sex while under the influence of drugs/alcohol.

Conclusions and relevance:

Reporting a history of maltreatment was associated with an increased likelihood of engaging in risky sexual and antisocial behaviors, as well as depression in inner-city female youth. These data highlight the broad, lingering repercussions of all types of child maltreatment.

1. Introduction

Increasingly, research has demonstrated that adolescent and young adult (AYA) females who were the victims of sexual abuse are at higher risk of sexually transmitted infections (STIs) and unintended pregnancy (Noll, Shenk, & Putnam, 2009; Stoltenborgh, van Ijzendoorn, Euser, & Bakermans-Kranenburg, 2011; Wilson & Widom, 2008, 2009). In addition, victims of sexual abuse are more likely to engage in risky sexual behaviors during both adolescence and adulthood (Hillis, Anda, Felitti, & Marchbanks, 2001; Jones et al., 2010; Senn, Carey, & Coury-Doniger, 2011; Senn, Carey, & Vanable, 2008; van Roode, Dickson, Herbison, & Paul, 2009; Wekerle, Goldstein, Tanaka, & Tonmyr, 2017; Wilson, Emerson, Donenberg, & Pettineo, 2013). While past research concentrated on the effects of female sexual child abuse, other types of child maltreatment, including nonsexual forms of abuse and neglect that represent more than three-quarters of all reported cases of child maltreatment (Houck, Barker, Hadley, Menefee, & Brown, 2018; Lansford et al., 2002; Negriff, Schneiderman, & Trickett, 2015; Norman et al., 2012; Thibodeau, Lavoie, Hebert, & Blais, 2017; Walsh, Latzman, & Latzman, 2014) are increasingly recognized as having equivalent negative effects across indices of adolescent health (Rivera, Bray, Guastaferro, Kugler, & Noll, 2018; Vachon, Krueger, Rogosch, & Cicchetti, 2015; Yoon, Voith, & Kobulsky, 2018), including drug use, antisocial behavior, peer deviancy and psychopathology such as depressive symptoms (Oberlander et al., 2011).

As many as 20 % of U.S. children have experienced abuse or neglect over their lifetimes (Cummings & Berkowitz, 2014). Over six million children are reported each year to child protective services for maltreatment, with the majority of reported cases occurring in low-income, inner-city and minority populations (Diaz & Petersen, 2014). Cases of maltreatment are significantly higher for African-American children than those for non-Hispanic White children (Lee et al., 2012; Wilson et al., 2013). In addition, children from low socioeconomic status households experience some type of abuse and neglect at more than 5 times the rate of other children (Sedlak et al., 2010).

Urban and minority youth also report engaging in risky sexual behaviors at higher rates than the general population (Kann et al., 2016). Inner-city minority youth also experience negative sexual health outcomes at higher rates (Centers for Disease Control & Prevention, 2016) including acquiring STIs (Barton, Braxton, & Davis, 2016). The two most common STIs, chlamydia, and gonorrhea are more common among minority females than non-Hispanic White females of the same age (Centers for Disease Control & Prevention, 2016).

As part of an ongoing cohort study to assess behavioral and psychosocial risk factors for human papillomavirus (HPV) infection in sexually-active inner-city AYAs who received HPV vaccination at the Mount Sinai Adolescent Health Center (MSAHC) in New York, we introduced a standardized self-report screening survey for child maltreatment, including questions on history of physical and emotional abuse and neglect. We hypothesized that other forms of abuse and neglect will be associated with poorer psychosocial functioning and higher sexual risk behaviors independent of sexual abuse. The study’s high risk and large sample size afforded us to examine associations with risky psychosocial and sexual behaviors by type of maltreatment.

2. Methods

2.1. Study population and setting

The Mount Sinai Adolescent Health Center (MSAHC) uses a unique model that integrates medical, sexual and reproductive health, dental, optical, and behavioral and mental health. The services provided cover a continuum of wellness, including health education and risk reduction, as well as prevention and treatment services. All services are confidential and free to patients. The center is located on the border of East Harlem and supports an underserved population of children, adolescents, and young adults (ages 10–24).

A full description of the study design has been published previously (Braun-Courville et al., 2014). Briefly, females were eligible to participate if they: 1) were between 13 and 19 years of age at time of consent, 2) had ever engaged in vaginal or anal intercourse, and 3) intended to get or had already received an FDA approved HPV vaccine (GARDASIL®). Females pregnant at time of recruitment, or who had terminated a pregnancy within the 4 weeks prior to enrollment, were asked to return at a later date and not included in the current analysis. Participants returned every six months for gynecological examination and completed a self-administered questionnaire on sexual behaviors and psychosocial risk measures (described in more detail below). To assess the burden of child abuse and neglect in the population, we used the Childhood Trauma Questionnaire (Bernstein & Fink, 1998) for subjects entering or returning for a follow-up visit in the ongoing cohort surveillance study (Braun-Courville et al., 2014). Similarly, some scales (CES-D, peer deviancy module, and sex under influence of drugs or alcohol) were introduced later in the study; these outcomes were analyzed at enrollment for new subjects and at first assessment for returning subjects.

2.2. Ethics

Written informed consent was collected from all participants prior to enrollment. This study was approved by the Institutional Review Board at The Icahn School of Medicine at Mount Sinai.

2.3. Assessment of child maltreatment type

The Childhood Trauma Questionnaire (CTQ)(Bernstein & Fink, 1998) is a 28-item self-report inventory that was used to assess retrospectively if our research participants have been exposed to abuse (Sexual, Physical and Emotional) and neglect (Physical and Emotional) during their childhood. Study participants were asked to report based on their experiences ‘When I was growing up…”. The CTQ has been validated extensively for use in retrospective studies as a screening tool for child abuse and neglect, and has been used as the gold standard for testing of other child maltreatment survey tools and interview techniques (Bernstein et al., 2003; DiLillo et al., 2006, 2010; Lobbestael, Arntz, Harkema-Schouten, & Bernstein, 2009). The scale includes 5 items representing each type of abuse/neglect and a 3-item minimization/denial scale for detecting false-negative child maltreatment reports. Subjects responded to statements about their childhood experiences, using a 5-point scale. The severity of each type of abuse and neglect was then quantified and grouped to characterize none to minimal, low to moderate, and severe to extreme maltreatment using for following cut-point (Bernstein & Fink, 1998): physical abuse (≤7, 8–9, ≥10); sexual abuse (≤5, 6–7, ≥8); emotional abuse (≤8, 9–12, ≥13); physical neglect (≤7, 8–9, ≥10); and emotional neglect (≤9, 10–14, ≥15) respectively.

2.4. Assessment of sexual behaviors and psychosocial factors

Self-administered questionnaires consisting of items to assess risk behaviors are completed at enrollment and every six months as part of the base study protocol. These included questions regarding demographics, recent and lifetime history of sexual behaviors, contraceptive use, history of pregnancies, STIs and anogenital warts, and use of cigarettes, alcohol, and illicit substances. The risk of alcohol and substance abuse was also assessed using the 6-item American Academy of Pediatrics’ Committee on Substance Abuse behavioral health screening scale (CRAFFT) (Knight, Sherritt, Shrier, Harris, & Chang, 2002). A CRAFFT score of 2 or more was considered positive, indicating a need for additional screening for alcohol and drug/substance abuse and dependence (Knight et al., 2002). History of unprotected sex while under the influence of drugs or alcohol, and having sex in exchange for money was also asked.

In addition, the questionnaires included a number of validated scales to assess depressive symptoms and antisocial behavior. Depression was assessed using the 20-item Center for Epidemiologic Studies-Depression (CES-D)(Gotlib & Cane, 1989) scale that has been previously validated for female AYA populations including minority groups (Garrison, Addy, Jackson, McKeown, & Waller, 1991; Roberts, Lewinsohn, & Seeley, 1991; Rushton, Forcier, & Schectman, 2002). A CES-D score of 16 or more was considered as positive for depression based on evidence from population studies (Lewinsohn, Seeley, Roberts, & Allen, 1997).

Antisocial behavior was measured using the 13-item Antisocial/Delinquency/Violence module of the National Longitudinal Study of Adolescent Health (ADDHealth), including lying, delinquent acts (stealing, trespassing, graffiti), and use of violence (physical fights, using a weapon) in the past 12 months (Kathleen, Udry, Bearman, & Harris, 2009). An antisocial index score of 3 or more was defined as positive, while a score less than or equal to 2 was considered as none or mild (Pechorro, Moreira, Basto-Pereira, Oliveira, & Ray, 2019). This cutoff was found to be associated with risk of HPV in a previous study in this population (Linares et al., 2015).

Peer deviance and delinquency behaviors were assessed using the 17-item Chicago Youth Development Study Self-Report Delinquency Scale adapted from the ADDHealth study survey. This scale evaluated involvement with deviant peers by asking youth to report the extent to which their friends engaged in 17 delinquent behaviors in the prior 3 months, including violent behaviors, drug, and alcohol use, and getting arrested (Miller, Gorman-Smith, Sullivan, Orpinas, & Simon, 2009). Peer deviancy was defined on a categorical scale with any report of being involved in gang fights, rape, cocaine use, or getting arrested considered as severe peer deviancy, while a report of any of the other 14 delinquent behaviors was considered as mild or moderate peer deviancy.

Sexual behavior outcomes were abstracted as collected in the study questionnaire and included: lifetime number of sexual partners (1, 2, 3–4, 5–9, 10+), anal sex (never vs. ever), and ever having unprotected sex while under influence of drugs or alcohol (no vs. yes).

2.5. Clinical history and physical exam

All study participants received a comprehensive gynecological examination that included: sexual, reproductive, behavioral and psychosocial history, immunization update, blood and urine testing/screening (as indicated), and age-specific health education. All subjects were screened for evidence of sexual and physical abuse annually by trained adolescent medicine physicians, and suspected cases were referred immediately for counseling with a mental health expert at MSAHC.

2.6. Statistical analyses

Distribution of demographic, social and behavioral characteristics were numerically presented as frequencies and percentages. We examined the distribution of the CTQ by severity (none-minimal vs. low-moderate vs. severe) for each of the five maltreatment types (sexual abuse, physical abuse, emotional abuse, physical neglect and emotion neglect)(Litrownik et al., 2005). The distribution of five types of maltreatment by response variables are provided in the supplemental material.

Cross-sectional associations between reported maltreatment type, depression and risky behaviors were examined using multivariable logistic regression for binary outcomes adjusting for age, race/ethnicity, lifetime number of sexual partners and psychosocial indicators including socio-economic-status (using previously described index (Linares et al., 2015), family support (operationalized as an adult in the household), and educational attainment (operationalized as attending school or college) for all outcomes except lifetime sexual partners. In addition, CTQ minimization/denial scale variable was added in all models as a covariate to address underreporting. Exposure to maltreatment type was categorized as follows: (i) those reporting exposure to any single maltreatment type (primary exposed group), (ii) those reporting exposure to other types of maltreatment (adjusting variable), and (iii) those reporting no abuse or neglect (reference group).

Odds ratios (OR) and 95 % confidence intervals (CI) were estimated for the above psychosocial and sexual outcomes, by comparing subjects reporting none-to-minimal vs. low-to-extreme maltreatment type, assuming independent relationships between each maltreatment type. Cumulative logit models were fit to assess the associations between maltreatment exposure and ordinal response variables: peer deviancy (severe, mild-moderate or none), and lifetime number of sexual partners (10+, 5–9, 3–4, 2 or ≤1). Proportional odds assumptions were examined using the score test. For models that did not pass the proportional odds test, an unconstrained partial proportional odds model was fit and assessed using either Hosmer-Lemeshow or Pearson goodness of fit tests (Ananth & Kleinbaum, 1997; Hosmer, Lemeshow, & Sturdivant, 2013). P-values of < 0.05 were considered to have poor fit.

To further examine the effects of physical and emotional abuse and neglect remained in the absence of sexual abuse we repeated the analyses above excluding AYAs who reported a history of sexual abuse. In addition, we ran sensitivity analyses with multiple imputation using a fully-conditional specification approach, under a missing at random assumption, with 10 imputed data sets to accommodate for missing data.

All participants completed the measures once, including 38 % who completed the survey at enrollment (visit 1), 33 % within the first year (between visits 2 and 5), and the remaining at a later visit. Given the varied assessment over visits, we evaluated if there were any cohort effects (i.e., due to the staggered administration of the surveys) but did not find any significant differences.

3. Results

3.1. Study population characteristics

The study sample consisted of 882 AYA females from a predominantly minority population with a median age of 18 years at entry (with a mean of 17.6 years and standard deviation of 1.31), most of whom (65.5 %) were enrolled in middle or high school (Table 1). Reflecting on the enrollment criteria, all participants were sexually active at the time of enrollment; nearly 40 % reported having vaginal sex before the age of 14, and 41.2 % reported having had a total of two or more male sexual partners. Other high-risk sexual behaviors included having unprotected sex while under the influence of drugs or alcohol (reported by 39.1 % of subjects) and having sex in exchange for money, drugs, shelter, or food (1.6 %). Twenty-four percent of subjects reported having had sex with a partner at least five years older.

Table 1.

Study subject characteristics.

	N	%
Age mean (SD)	17.6 (1.31)
13–15	62	7.0
16–17	307	34.8
18–20	513	58.2
Race/Ethnicity *
Any Hispanic	533	60.5
Non-Hispanic African-American	315	35.8
Other non-Hispanic	33	3.8
Education
≤10th grade	209	23.7
11–12th grade	369	41.8
High-School Graduate	138	15.6
Some College	165	18.7
Vaginal Sex Lifetime Partners
1	517	58.8
2	210	23.9
3 +	153	17.4
Anal Sex Lifetime Partners
0	642	73.1
1	166	18.9
2+	70	8.0
Age at first sexual intercourse
> 14 years	539	61.1
< 14 years	343	39.1
HPV vaccine at baseline
Vaccine naïve	107	12.1
1–3 doses	775	87.9
History of Pregnancy
No	584	66.2
Yes	219	24.8
Any contraception†
Most times to Always	636	72.1
Never/Rarely/sometimes	246	27.9
Condom use during vaginal intercourse
Most times to Always	397	46.4
Never/Rarely/Sometimes	458	53.6
Ever had sex with partner 5 years older or more
Never	566	64.2
Ever	212	24.0
Unprotected sex while under influence of drugs or alcohol
Never	545	61.8
Ever	337	39.1
Received money, drugs, shelter or food in exchange for sex
Never	868	98.4
Ever	14	1.6
Antisocial behavior
No	624	70.8
Yes	257	29.1
Alcohol/Substance use
No	566	64.2
Yes	310	35.2
Peer Deviancy
No	122	13.8
Mild/Moderate	559	63.4
Severe	201	22.8
Depression
No	646	73.2
Yes	236	26.8
Lifetime # of sexual partners
1	154	17.5
2	161	18.3
3–4	259	29.4
5–9	212	24.0
10+	96	10.9

^†Including hormonal and barrier contraceptive methods;

^‡Including non-full term. Total N (%) may not add to 882 for all data.

3.2. Reported history of child maltreatment

History of maltreatment was reported by 59.6 % of subjects in our sample. Using established CTQ scale cut-points for abuse and neglect described above, the prevalence of each type of maltreatment was the following: emotional abuse (30.7 %) / neglect (40.4 %), physical abuse (19.5 %) / neglect (26.2 %), and sexual abuse (17.5 %; Fig. 1). Over a third (37.5 %) of AYA females reported being victims of at least two maltreatment types, with almost 5 % reporting enduring all 5 types of maltreatment (Fig. 2). Given the multiplicity of experiences that may be captured by self-reported emotional neglect, the proportion of the sample reporting only this form of maltreatment was examined. The proportion reporting a history of emotional neglect only were 8.9 % (n = 79). Using the CTQ denial scale, we assessed potential for underreporting of abuse/neglect; overall, 32 % of subjects exhibited some denial (score 1–3) with approximately 10 % exhibiting a high level of denial for any maltreatment.

Fig. 1.

Bar chart shows proportions of subjects exposed to each of the five types of abuse and neglect by severity of victimization (low, moderate and severe).

Fig. 2.

Sunburst plot showing combined history of child sexual abuse (red) by history of child physical abuse (light green) and neglect (dark green) and child emotional abuse (light orange) and neglect (dark orange) reported by inner-city adolescent and young adult women.

3.3. Association between child maltreatment and psychosocial risk factors

Results from multivariable logistic regression analyses comparing single maltreatment type are shown for psychosocial factors (Table 2) measured in the study cohort, adjusted for age, race/ethnicity, socio-economic status, family support, attending college or school, exposure to other maltreatment types, and denial. We observed statistically significant increased odds of depression and drug/alcohol use in AYA females reporting a history of all forms of maltreatment type. History of abuse and neglect was also significantly associated with increased peer deviancy.

Table 2.

Adjusted associations between history of child maltreatment type and psycho-social risk factors in sexually-active inner-city adolescent and young adult females.

Maltreatment exposure*		Depressive symptoms‡	Antisocial behavior	Drug or Alcohol use‖	Peer Deviancyβ
Sexual abuse	Complete Case	5.1 (3.2, 8.5)	1.8 (1.1, 2.8)	3.0 (1.9, 4.6)	4.2 (2.6, 6.8)
		CH = 3.1, p = 0.93	CH = 14.9, p = 0.06	CH = 5.9, p = 0.64	Cp = 1666.7, 0.19
	MI estimates	5.2 (3.2, 8.5)	1.8 (1.1, 2.8)	3.0 (1.9, 4.6)	3.3 (2.1, 5.1)
Physical abuse	Complete Case	4.9 (3.0, 7.9)	1.9 (1.2, 2.9)	3.4 (2.3, 5.3)	4.9 (3.1, 7.9)
		CH = 6.9, p = 0.54	CH = 18.6, p = 0.02	CH = 5.8, p = 0.67	Cp = 1682.2, p = 0.16
	MI estimates	4.9 (3.0, 7.9)	1.9 (1.2, 3.0)	3.5 (2.3, 5.4)	4.9 (3.1, 7.9)
Emotional abuse	Complete Case	6.4 (4.0, 10.0)	1.7 (1.1, 2.5)	3.1 (2.1, 4.6)	3.0 (2.0, 4.4)
		CH = 10.4, p = 0.23	CH = 12.2, p = 0.14	CH = 3.2, p = 0.92	Cp = 1681.1, p = .20
	MI estimates	6.4 (4.1, 10.0)	1.7 (1.1, 2.5)	3.2 (2.2, 4.7)	3.0 (2.0, 4.4)
Physical neglect	Complete Case	3.7 (2.3, 5.7)	1.6 (1.0, 2.4)	2.5 (1.7, 3.6)	2.3 (1.4, 3.6)
		CH = 2.5 p = 0.96	CH = 9.3, p = 0.32	CH = 6.7, 0.56	Cp = 1674.9, p = 0.22
	MI estimates	3.6 (2.3, 5.7)	1.6 (1.1, 2.4)	2.5 (1.7, 3.6)	2.2 (1.4, 3.5)
Emotional neglect	Complete Case	4.4 (2.8, 6.7)	1.5 (1.0, 2.2)	2.0 (1.4, 2.9)	2.3 (1.5, 3.6)†
		CH = 4.8, p = 0.78	CH = 9.3, p = 0.28	CH = 6.8, 0.55	Cp = 1652.5, 0.24
	MI estimates	4.4 (2.8, 6.8)	1.5 (1.0,2.2)	2.0 (1.4, 3.0)	2.7 (1.7, 4.2)
Excluding subjects reporting no history of sexual abuse
Physical abuse	Complete Case	2.7 (1.5, 4.9)	1.9 (1.1, 3.3)	3.1 (1.9, 5.2)	3.2 (1.9, 5.4)
		CH = 3.9, p = 0.87	CH = 15.2, 0.06	CH = 3.1, p = 0.92	Cp = 1376.9, p = 0.29
	MI estimates	2.7 (1.5, 4.8)	1.9 (1.1, 3.3)	3.1 (1.9, 5.2)	3.2 (1.9, 5.4)
Emotional abuse	Complete Case	5.1 (3.1, 8.4)	1.5 (0.9, 2.4)	3.1 (2.0, 4.9)	2.4 (1.6, 3.8)
		CH = 8.5, P = 0.38	CH = 13.6, 0.09	CH = 2.2, p = 0.97	Cp = 1355.9, p = 0.45
	MI estimates	5.1 (3.1, 8.4)	1.5 (102, 2.4)	3.2 (2.0, 4.9)	2.8 (1.7, 4.9)&
Physical neglect	Complete Case	2.7 (1.6, 4.5)	1.6 (1.0, 2.5)	2.0 (1.3, 3.2)	1.3 (0.8, 1.9)
		CH = 2.6, p = 0.95	CH = 19.8, 0.27	CH = 4.5, p = 0.80	Cp = 1389.6, p = 0.25
	MI estimates	2.7 (1.6, 4.4)	1.6 (1.0, 2.5)	2.0 (1.3, 3.1)	1.3 (0.8, 1.9)
Emotional neglect	Complete Case	3.5 (2.2, 5.5)	1.4 (0.9, 2.1)	1.8 (1.2, 2.8)	1.5 (1.0, 2.2)
		CH = 3.9, p = 0.86	CH = 11.4, p = 0.17	CH = 6.8, p = 0.55	Cp = 1376.8, p = 0.22
	MI estimates	3.5 (2.2, 5.5)	1.4 (0.9, 2.1)	1.8 (1.2, 2.8)	1.5 (1.0, 2.2)

C_p- Pearson Goodness of fit Test Chi-square and corresponding p-value. MI- Multiple imputation.

^*Odds ratios (95 % confidence intervals) are shown comparing none-to-minimal vs. low-to-extreme abuse or neglect for those with the maltreatment type shown compared to those reporting no abuse or neglect (reference group), adjusting for age, race/ethnicity, lifetime number of sexual partners (for selected outcomes), psychosocial indicators, denial, and exposure to other forms of abuse or neglect.

^‡CES-D score (> 16).

^‖CRAFFT scale (> 1).

^βADDHealth Peer Deviancy score (severe vs. mild + moderate vs. none).

^†Estimates by partial proportional odds model shown for Severe vs Mild/Moderate/None, those for Severe/Mild/Moderate vs None were 1.2 (0.7–1.9) [MI estimates: 1.3 (0.8, 2.1)] for Emotional neglect.

^&Estimates by partial proportional odds model shown for Severe vs Mild/Moderate/None, those for Severe/Mild/Moderate vs None MI estimates: 2.2 (1.1, 4.7)] C_H Hosmer and Lemeshow Goodness-of-Fit Test Chi-square with 8 degrees of freedom, and corresponding p-value.

3.4. Association between child maltreatment and risky sexual behaviors

With respect to risky sexual behaviors, AYA females who reported a history of physical or emotional abuse during childhood were more likely to: (i) report a higher number of lifetime sexual partners, (ii) have engaged in anal sex, (iii) report having had sex with an older partner (≥5 years), and (iv) admit to ever having unprotected sex while under the influence of drugs or alcohol (after adjusting for the same covariates as above), compared to those reporting no history of abuse or neglect (Table 3). No significant associations were seen for age at first vaginal intercourse, use of contraception, and history of pregnancy (not shown).

Table 3.

Adjusted associations between history of child maltreatment type and risky sexual behaviors in inner-city adolescent and young adult females.

Maltreatment exposure*		Lifetime # of sexual partnersβ	Anal sex ever	Sexual partner ≥ 5 years older‡	Unprotected sex under influence‖
Sexual abuse	Complete Case	2.3 (1.3, 4.2)†	1.7 (1.0, 2.7)	1.9 (1.1, 3.3)	1.4 (0.9, 2.1)
		Cp = 1326.3, p = 0.04	CH = 3.9, p = 0.87	CH = 3.4, p = 0.90	CH = 7.1, p = 0.52
	MI estimates	2.3 (1.3, 4.2)†	1.7 (1.0, 2.7)	2.1 (1.3, 3.6)	1.3 ((0.9, 2.1)
Physical abuse	Complete Case	1.6 (1.1, 2.3)	2.1 (1.3, 3.3)	2.0 (1.2, 3.4)	1.6 (1.1, 2.5)
		Cp = 1318.9, p = 0.02	CH = 9.5, p = 0.30	CH = 4.1, p = 0.85	CH = 4.2, p = 0.83
	MI estimates	1.6 (1.0, 2.4)	2.1 (1.3, 3.3)	2.2 (1.3, 3.7)	1.6 (1.1, 2.4)
Emotional abuse	Complete Case	1.5 (1.1, 2.1)	1.8 (1.2, 2.7)	1.8 (1.1, 2.8)	1.5 (1.0, 2.1)
		Cp = 1344.9, p = 0.01	CH = 10.1, p = 0.26	CH = 3.2, p = 0.92	CH = 4.2, p = 0.84
	MI estimates	1.5 (1.1, 2.2)	1.8 (1.1, 2.7)	1.9 (1.2, 3.1)	1.4 (1.0, 2.1)
Physical neglect	Complete Case	1.4 (1.0, 1.9)	1.5 (1.0, 2.3)	1.5 (0.9, 2.4)	1.3 (0.9, 1.8)
		Cp = 1385.7, p = 0.01	CH10.1, p = 0.26	CH = 5.1, p = 0.74	CH = 8.7, p = 0.36
	MI estimates	1.4 (1.0, 1.9)	1.5 (1.0, 2.3)	1.6 (1.0, 2.6)	1.2 (0.8, 1.8)
Emotional neglect	Complete Case	1.5 (1.1, 2.1)	1.6 (1.1, 2.4)	1.6 (1.0, 2.6)	1.1 (0.8, 1.6)
		Cp = 1336.7, p = 0.02	CH = 8.3, p = 0.41	CH = 3.2, p = 0.92	CH = 11.6, p = 0.17
	MI estimates	1.6 (1.1, 2.1)	1.6 (1.1, 2.4)	1.8 (1.1, 2.8)	1.1 (0.8, 1.6)
Excluding subjects reporting no history of sexual abuse
Physical abuse	Complete Case	1.6 (1.0, 2.4)	2.4 (1.4, 4.1)	2.4 (1.3, 4.5)	2.0 (1.2, 3.4)
		Cp = 1194.0, p = 0.01	CH = 9.7, p = 0.28	CH = 1.6, p = 0.99	CH = 14.2, p = 0.08
	MI estimates	1.6 (1.0, 2.4)	2.4 (1.4, 4.1)	2.5 (1.3, 4.9)	2.0 (1.2, 3.4)
Emotional abuse	Complete Case	1.6 (1.1, 2.2)	2.0 (1.2, 3.2)	1.8 (1.0, 3.0)	1.5 (1.0, 2.3)
		Cp = 1219.50, p = 0.01	CH = 10.7, p = 0.22	CH = 3.1, p = 0.93	CH = 8.6, p = 0.37
	MI estimates	1.5 (1.1, 2.2)	2.0 (1.2, 3.2)	1.8 (1.0, 3.1)	1.5 (1.0, 2.4)
Physical neglect	Complete Case	1.4 (0.9, 1.9)	1.3 (0.8, 2.1)	1.4 (0.8,2.3)	1.3 (0.9, 2.0)
		Cp = 1231.2, p = 0.01	CH = 1.9, p = 0.98	CH = 3.3, p = 0.91	CH = 10.1, p = 0.25
	MI estimates	1.4 (1.0, 1.9)	1.2 (0.8, 2.0)	1.4 (0.8, 2.4)	1.3 (0.8, 2.0)
Emotional neglect	Complete Case	1.6 (1.1, 2.2)	1.5 (0.9, 2.3)	1.6 (0.9, 2.6)	1.2 (0.8, 1.7)
		Cp = 1212.2, p = 0.01	CH = 12.3, p = 0.14	CH3.9, p = 0.87	CH = 5.8, p = 0.56
	MI estimates	1.5 (1.1, 2.1)	1.5 (1.0, 2.3)	1.7 (1.0, 2.8)	1.2 (0.8, 1.7)

Significant associations are shown in bold.

C_H Hosmer and Lemeshow Goodness-of-Fit Test Chi-square with 8 degrees of freedom, and corresponding p-value.

C_p Pearson Goodness of fit Test Chi-square and corresponding p-value. MI- Multiple imputation.

^*Odds ratios (95 % confidence intervals) are shown comparing none-to-minimal vs. low-to-extreme abuse or neglect for those with the maltreatment type shown compared to those reporting no abuse or neglect (reference group), adjusting for age, race/ethnicity, lifetime number of sexual partners (for selected outcomes), psychosocial indicators, denial, and exposure to other forms of abuse or neglect.

^βEstimates from proportional odds models (10+ vs < 10, ≥5 vs < 5, ≥3 vs < 3, ≥2 vs 1).

^‡Had ever had a sexual partner 5 years or older at enrollment (yes vs. no).

^‖Unprotected sex while under influence of drugs or alcohol.

^†Estimate by partial proportional odds model for 10+ vs < 10; estimates for lower outcome categories were 1.2 (0.8–1.9) [MI: 1.2 (0.8, 1.9)] for > = 5 vs < 5, 1.5 (0.9–1.8) [MI: 1.2 (0.8, 1.8)] for > = 3 vs < 3, and 1.5 (0.9–2.6) [MI: 1.5 (0.9, 2.5)] for > = 2 vs < 2.

Given the potentially pervasive impact of sexual abuse, we also assessed the effects of other forms of abuse and neglect among (N = 726) AYAs who reported no history of sexual abuse (Tables 2 and 3). Physical and emotional abuse histories were significantly associated with psychosocial factors and risky sexual behaviors among AYAs who reported these experiences, but not sexual abuse. Similarly AYAs reporting physical and emotional neglect without sexual abuse showed significant associations with psychosocial factors depression and drug or alcohol use. Estimates for all model covariates are provided in the supplemental material.

4. Discussion

There is increasing support for the broad detrimental effects of all forms of child maltreatment on AYA health (Lansford et al., 2002; Vachon et al., 2015). Emerging research indicates that exposure to child nonsexual abuse and neglect has long lasting consequences on AYA psychosocial functioning and sexual behaviors (Negriff et al., 2015; Norman et al., 2012). Accordingly, identifying childhood factors that predict high-risk behaviors during adolescence and early adulthood is a priority (Oshri, Tubman, & Burnette, 2012; Sedlak et al., 2010; Sena et al., 2015) particularly among inner-city minority youth that tend to engage in risky sexual behaviors at higher rates than the general population (Kann et al., 2016).

In a large cohort of 882 inner-city minority AYA females, the current study examined the associations between type of child maltreatment (encompassing three types of abuse and two types of neglect), psychosocial functioning (depression, antisocial behavior, peer deviancy, substance/alcohol use), and risky sex behaviors. The current data show not only that exposure to child nonsexual (physical and emotional) abuse is associated with high-risk sexual, depression, and antisocial behaviors, but that these associations are close in magnitude to that of exposure to child sexual abuse. Furthermore, our study demonstrates that exposure to child neglect itself (physical and emotional) after adjusting for exposure to other forms of maltreatment, is associated with similar detrimental outcomes, with associations often just as strong as those seen for child abuse victims.

From an intervention perspective, a broad vulnerability skill-based approach (Houck et al., 2018) such as increasing emotion self-regulation and coping with negative affect, may help mitigate the impact of child abuse and neglect on co-occurring sexual risk-taking, depression, peer deviance, and substance/alcohol use (Cary & McMillen, 2012; Jones et al., 2013; Proctor et al., 2017; Wilson & Widom, 2011). Adding to recent studies about the harmful effects of sexual and nonsexual abuse on AYA risky sexual practices (Abajobir, Kisely, Williams, Strathearn, & Najman, 2018; Thibodeau et al., 2017; Thompson et al., 2017; Werner et al., 2018) our study expands emerging literature to an urban inner-city female AYA clinic population (Wekerle et al., 2017). This is important as the majority of reported cases of abuse and neglect occur in low-income, inner-city and minority populations (Diaz & Petersen, 2014). Rates of STIs are also highest among inner-city minority youth (Braxton et al., 2018).

The results from this study require validation; we acknowledge that external validity may be confined to individuals who are more receptive to receiving medical and psychological care as the study sample was comprised of AYA patients who were seeking services from MSAHC, and had consented to participate in a prospective HPV research study (Carey, Senn, Vanable, Coury-Doniger, & Urban, 2008). All subjects were sexually active at the time of enrollment, and therefore more likely to present with sexual risk behaviors than other AYA female populations. Child abuse victims also have an increased risk of re-victimization (Barnes, Noll, Putnam, & Trickett, 2009; Finkelhor, Ormrod, & Turner, 2007). While we did not assess for incidents of new sexual or physical abuse in our study, those with a history of abuse were more likely to report having had older sexual partners (≥5 years). While such reports may be indicative of problematic relationships, it is unknown whether they constituted abuse as timing was not ascertained. Mediational analyses will be needed to identify direct and indirect influences underlying linkages between child maltreatment history and later risky sexual practices (Yoon et al., 2018).

There are also limitations surrounding measurement of child abuse and neglect. First, it was impossible to verify the temporal order with reported history of both risk behaviors and abuse/neglect, making it difficult to establish causation. Second, the gathering of information around maltreatment experiences, such as involvement with child protection services, timing, and sequence of maltreatment and concurrent poly-victimization, which were not assessed here, may have revealed additional components affecting sexual and other health outcomes. Third, there is some controversy in the field regarding accuracy in recalling traumatic events that occurred in childhood. However, there is strong evidence supporting the validity of retrospective measures (Kendall-Tackett & Becker-Blease, 2004). Results from a large study of adults show recall of child abuse and neglect to be highly reliable when assessed on two separate occasions, approximately a year apart (Dube, Williamson, Thompson, Felitti, & Anda, 2004). In addition, it’s important to note that the majority of evidence to date on the long-term effects of child abuse comes from retrospective studies in adults.

5. Conclusion

While research has demonstrated that AYA females who were the victims of sexual abuse are at higher risk of STIs (Stoltenborgh et al., 2011; Wilson & Widom, 2008) growing evidence supports the detrimental effects of nonsexual types of abuse (physical and emotional) and child neglect; which represent more than three-quarters of all reported cases of child maltreatment (Lansford et al., 2002; Negriff et al., 2015; Norman et al., 2012; Vachon et al., 2015). Beyond the well-known effects of child sexual abuse, the current data show the extensive burden across child nonsexual abuse and neglect types (adjusting for other forms of maltreatment) ranging from a two- to four-fold increased risk on multiple indices of psychosocial dysfunction and sexual behavior. Inner-city AYA females reporting a history nonsexual abuse (physical or emotional) types and neglect were significantly more likely to report depression, using drugs or alcohol, engaging with peers who tend to have deviant or delinquent behaviors, having higher number and older sexual partners, and having unprotected sex while under the influence of drugs or alcohol, compared to those reporting no history of abuse or neglect. The findings suggest that intervening broadly with maltreated youth is important to reduce sexual risky behavior among inner-city minority females (Barton et al., 2016).

Abbreviations:

CDC

Centers for Disease Control and Prevention

CES-D

Center for Epidemiologic Studies-Depression

CTQ

Childhood Trauma Questionnaire

CI

confidence interval

HPV

human papillomavirus

MSAHC

Mount Sinai Adolescent Health Center

AddHealth

National Longitudinal Study of Adolescent to Adult Health

OR

odds ratio

STI

sexually transmitted infection