Childhood Abuse and Adulthood IPV, Depression, and High-Risk Drinking in Latinas

Abstract

Little is known about relationships between child maltreatment and adulthood IPV, depression, and risky drinking in Latinas. 548 Latinas in a sexual health RCT self-reported childhood physical, sexual, and emotional abuse, IPV, depression, and risky drinking. Childhood abuse was related to adulthood IPV, OR = 1.27, depression, OR = 2.02, and high-risk drinking, OR = 2.16]. Childhood emotional abuse was linked to depression, OR = 2.19; childhood physical abuse to risky drinking, OR = 2.62; and childhood sexual abuse to depression, OR = 2.78 and risky drinking, OR = 2.38. Results may inform prevention/intervention efforts for mental health nurses.

Child abuse and neglect is a pervasive and serious public health problem. In 2015, almost four million referrals were made to child protective services for child abuse or neglect (U.S. Department of Health and Human Services, 2017). Child abuse and neglect is associated with numerous negative long-term consequences, including lower educational attainment, under- or unemployment, poorer quality of life, and psychological distress symptoms (Banyard, Williams, & Siegel, 2001; Corso, Edwards, Fang, & Mercy, 2008; Currie & Widom, 2010; Fergusson, Boden, & Horwood, 2008). Child abuse is also associated with Intimate Partner Violence (IPV) in adulthood (Whitfield, Anda, Dube, & Felitti, 2003), depression (Lang, Stein, Kennedy, & Foy, 2004; Mandelli, Petrelli, & Serretti, 2015), and risky alcohol use, such as binge drinking (Alem, Soto, Baezconde-Garbanati, & Unger, 2015).

The current study sought to examine the relationship between child maltreatment and negative outcomes for U.S. Hispanic women. This is an important issue, as the literature has primarily focused on non-Hispanic white women (Newcomb, Munoz, & Carmona, 2009; Mandelli, Petrelli, & Serretti, 2015). Yet, there are about 56.5 million U.S. Hispanics, a heterogeneous group of immigrants and their descendants with a broad range of ethnic backgrounds, and national origins, who live across the U.S. Hispanics represent nearly 18% of the U.S. population, the largest racial/ethnic minority group in the U.S. (U.S. Census Bureau, 2016). The lack of empirical study in this area with Latina/Hispanic samples is concerning, as U.S. Hispanics are significantly less likely to access mental health treatment than non-Hispanic Whites (36% vs. 60%, respectively) (Glover, Pumariega, Holzer, Wise, & Rodriguez, 1999).

Effective and promising prevention strategies for child maltreatment have been described in the literature. In a systematic review on the prevention of child maltreatment, Mikton and Butchart (2009) identified four types of universal and selective interventions that have demonstrated success. These included home visiting, parent education, abusive head trauma prevention, and multi-component programs. However, it is unclear if these programs work equally well for Hispanic women. Few studies have specifically targeted Hispanic women or examined their results according to ethnicity. For example, in another systematic review of interventions designed to decrease child abuse in high-risk families findings were not reported according to Hispanic ethnicity (Levey et al., 2017).

It is possible that physiological changes occur in the brain of children that predispose them to these kinds of psychological consequences and risky behaviors in adulthood (Sinha, 2008). These biopsychosocial mechanisms need to be specifically explored in samples of Latinas, who may have unique cultural risk and protective factors, e.g., acculturation stress or familism, which may interact with biological factors (Sinha, 2008). Cultural-specific influences and experiences among Hispanic communities are likely to intersect with other vulnerabilities such as social and mental health conditions to provide a unique configuration needing to be addressed among Hispanic women. However, in order to address these, more research is needed to assess the nature of child abuse and neglect among Hispanic women, associated consequences later on in life, and barriers and facilitators to participation in prevention programs.

Two alternative models suggest different reasons for how maltreatment may relate to later negative outcomes. Additive risk models suggest that the number of types of abuse experienced during childhood, regardless of the specific type of traumas, is significantly associated with negative outcomes (Felitti et al., 1998). Specificity models, however, suggest that particular types of abuse are related to specific negative outcomes. For example, in a meta-analysis of the long-term health consequences of childhood abuse, Norman and colleagues (2012) noted that childhood physical abuse, emotional abuse, and neglect where all associated with increased odds of a depressive disorder in adulthood, but the specific type of abuse was not universally related to negative health outcomes. Further, childhood physical abuse and neglect was associated with problem drinking, but emotional abuse was not. This is an interesting finding given that emotional abuse had a larger relationship with depressive disorder (OR = 3.85), than physical abuse (OR = 2.04) or neglect (OR = 2.11). Another study showed that number of adverse childhood experiences, including abuse, were related to greater substance use in adulthood in a sample of young adult Hispanics (Allem, Soto, Beazconde-Garbanati, & Unger, 2015). Mandelli, Petrelli and Serretti (2015) suggested in a recent meta-analysis that emotional abuse and neglect had the strongest links to depression in adulthood, compared to other types of trauma in the early years. Understanding whether experiencing multiple types of abuse in childhood increases risk for multiple negative outcomes (additive risk models), or whether abuse has potentially causal relationships with specific negative outcomes (specificity models), is critical to inform intervention efforts and inform tailored person-centered treatments.

In this study, we focused on the relationship between child maltreatment and IPV, depression, and drinking in adulthood. All of these outcomes have empirical support as consequences of child maltreatment for Non-Hispanic white women, but less so for U.S. Hispanic women. More specifically, we tested three hypotheses: 1) Exposure to any type of childhood abuse is positively related to IPV, depression, and high-risk drinking; 2) The number of child abuse experiences is positively related (i.e., has an additive effect) to IPV, depression, and high-risk drinking; and 3) Childhood physical abuse, sexual abuse, and emotional abuse are each positively related to IPV, depression, and high-risk drinking (i.e., specificity model).

Methods

Participants

Data for this study was collected from the initial assessment of 548 U.S. Hispanic women in the second (of three) randomized trial of SEPA (Salud, Educación, Prevención y Autocuidado/Health, Education, Prevention and Self-care), a sexual health group intervention designed to reduce sexual risks and increase healthy sexual behaviors, e.g., condom use and communication about HIV/STI with sexual partners, for Hispanic women (Author 2012). To be eligible, women had to be between the ages of 18 and 50 and report sexual activity within the three months prior to intake, and self-identify as Hispanic/Latina. Both the University of Miami and Florida Department of Health and Human Services Institutional Review Boards approved the study. Participants were recruited from 1) community-based social services (e.g., English classes, childcare, job development and placement, and health education) organizations for Hispanics, 2) an urban Florida Department of Health site, 3) flyers posted in the community, and 4) public service messages delivered via mass media. Women received $50 for participation in the assessment. Characteristics of women in this sample are summarized in Table 1 and reported elsewhere (Author 2012). Women in this study were born in a number of nations and territories, including Colombia (186, 34%), Cuba (70, 13%), Peru (45, 8%), the Dominican Republic (33, 6%), Nicaragua (29, 5%), Puerto Rico (28, 5%), Honduras (27, 5%), Venezuela (21, 4%), Argentina (14, 3%), Mexico (13, 2%), El Salvador (12, 2%), Ecuador (9, 2%), and seven others that were each 1% or less of the sample.

Table 1.

Characteristics of Women in the Sample (N = 548).

Variables	M	SD
Age, years	38.48	8.53
Years in U.S.	10.12	9.11
Years of Education	13.37	3.45
Number of Children	1.61	1.36
	n	%
High School Education	404	74
Employed	180	33
Born in U.S.	40	7
Living with a Spouse/Partner	380	69
Family Income < $2000/month	376	69

Note. Years in U.S. only for women born outside U.S.

Measures

All measures were available in Spanish and English and had been used with Hispanic samples in past research. Assessments were conducted between January 2008 and April 2009. Questions were asked via face to face interviews in English or Spanish and assessors documented participant responses on a computer using an internet-based software system (eVelos). Anchor sheets with each item and response choices were provided to participants so they could read along with assessors. All assessors were bilingual in English and Spanish.

Childhood abuse was assessed using a Violence Assessment developed for a previous randomized trial with Latinas (Author 2005, Author 2008). Three types of abuse experiences during childhood were assessed with this measure by asking participants to self-report exposure to emotional/verbal (including yelling, name calling, threats, stalking, possessiveness), physical (physically abused or beaten), or sexual (sexually abused, raped/forced or sexually assaulted) abuse as a child (i.e., prior to 18 years of age). The measure allowed participants to describe as many incidents as they had experienced with different perpetrators for each type of abuse. For analysis, this data was coded in three ways: 1) the presence of any reported child abuse, 2) the number of types of child abuse reported (range 0–3), and 3) presence of any of the three types of child abuse (emotional, yes or no; physical, yes or no, or sexual, yes or no).

IPV was assessed with the Partner-to-You scale from the Revised Conflict Tactics Scale short form (Straus & Douglas, 2004), adapted for Latinas in this study by reducing the number of items, simplifying wording, and shortening the number of possible response options to from the original seven to four (never, one time, two times, or three or more times). The measure assessed the frequency of 12 behaviors in the past three months (e.g., insulted you, beat you up, forced you to have sex). In this sample, Cronbach’s alpha on this scale was.86. To account for positive skew in the analysis, IPV was coded as 1 = any reported partner violence; 0 = no reported partner violence. That is, if the participant reported any of the twelve events happened one or more times, then IPV was coded as 1.

Depression was assessed via the Center for Epidemiologic Studies Scale (CES-D) (Radloff, 1977). This scale has 20 questions asking participants to report the frequency (i.e., number of days in the past week) of experiencing symptoms of depression (e.g., I had crying spells, I felt sad). This scale is widely used in population-based and community studies and has been translated and validated in Spanish (Roberts 1980). Responses to these questions are added for a total score ranging from 0 to 40 points. Scores of 16 and above indicate a likelihood of clinical depression (1 = depression, 0 = no depression), which was used to account for positive skew. Other transformations were attempted, but did not result in approximately normal distributions for depression. The CES-D had very good reliability in this study (Cronbach’s α = .94).

High-risk drinking.

The presence of high-risk drinking (of alcohol) was assessed with the CAGE (Cut down, Annoyed, Guilty, Eye opener) screening measure (Ewing, 1984). Women could respond with yes or no to four questions based on their lifetime experiences. Low risk drinking was defined as responding negatively to three or four questions or not drinking alcohol, high risk drinking was defined as responding affirmatively to any two of the four questions (Ewing, 1984). For analysis, a dummy-coded variable was used, with low-risk drinking as the comparison group.

Analysis Plan

We examined the links between childhood abuse experience and adult outcomes using path analysis in Mplus 7.11 (Muthén & Muthén, 2013). Path analysis is an extension of regression models that allows multiple independent and dependent variables to be included in the model at the same time. Examining adult outcomes (IPV, depression, high risk drinking), a series of path analyses examined the influence of 1) any abuse experiences, 2) additive abuse (number of child abuse experiences), and 3) the influence of specific types of child abuse (emotional, physical, and sexual). Step 1 tested the relationship between a single predictor (Any Childhood Abuse) and each of the adult outcome variables (IPV, depression, high risk drinking). Step 2 tested the Additive Model of Childhood Abuse by examining the relationship between a predictor variable defined as the number of types of reported childhood abuse (no abuse, emotional, physical, and/or sexual) and each of the adult outcome variables (IPV, depression, high risk drinking). Step 2 tested the Specificity Model of Childhood Abuse by examining the separate relationships of three predictor variables (emotional abuse vs. no emotional abuse; physical vs no physical abuse; sexual abuse vs. no sexual abuse) to each of the three adult outcome variables (IPV, depression, high risk drinking).

Results

Abuse in Childhood

About one-quarter (136, 25%) of women in this sample experienced at least one type of abuse during childhood, with 91 (17%) reporting verbal abuse during childhood, 84 (15%) reporting physical abuse in childhood, and 81 (15%) reporting sexual abuse during childhood. Overall, 412 (75%) of women reported experiencing no childhood abuse, 48 (9%) reported one type of abuse, 56 (10%) reported two types of abuse, and 32 (6%) all three types of child abuse. Figure 1 shows the proportion of women who reported each type of abuse during childhood, as well as IPV, depression, and high-risk drinking during adulthood.

Figure 1.

The proportions of women who reported each type of abuse during childhood, as well as IPV, depression, and high-risk drinking during adulthood.

IPV, Depression, and High-risk Drinking in Adulthood

Nearly two-thirds (350, 64%) of women reported experiencing IPV with their current partner. Almost half (256, 47%) scored over the clinical cutoff point for depression. About one-fifth (105, 19%) reported high-risk drinking. Childhood abuse and adulthood IPV, depression, and high-risk drinking are shown in Table 2.

Table 2.

Women Reporting Child Abuse and Adulthood IPV, Depression, and High-risk Drinking.

Variables	n	%
Childhood Emotional Abuse	91	17
Childhood Physical Abuse	84	15
Childhood Sexual Abuse	81	15
Adulthood IPV	350	64
Adulthood Depression	256	47
Adulthood High-risk Drinking	64	12

Relationships between Childhood and Adulthood

Any Childhood Abuse

Women that reported experiencing any type of child abuse were more likely to report IPV in adulthood, b = 0.64, SE = 0.22, p = .004, OR = 1.89, 95%CI[1.23, 2.91], depression in adulthood, b = 1.33, SE = 0.21, p < .001, OR = 3.78, 95%CI[2.49, 5.75], and high-risk drinking, b = 1.76, SE = 0.28, p < .001, OR = 5.81, 95%CI[3.33, 10.14].

Additive Childhood Abuse.

The number of the three types of childhood abuse experiences was positively related to IPV in adulthood, b = 0.24, SE = 0.11, p = .027, OR = 1.27, 95%CI[1.03, 1.56], depression in adulthood, b = 0.70, SE = 0.11, p < .001, OR = 2.02, 95%CI[1.62, 2.52], and high-risk drinking, b = 0.77, SE = 0.13, p < .001, OR = 2.16, 95%CI[1.68, 2.77]. That is, women who reported a greater number of childhood abuse experiences also reported greater IPV, depression, and high-risk drinking during adulthood.

Specificity:

Types of Childhood Abuse. Figure 2 shows the relationships between three types of childhood abuse and IPV, depression, and high-risk drinking in adulthood. Childhood emotional abuse was significantly related to adulthood depression, b = 0.79, SE = 0.32, β = .15, p = .015, OR = 2.19, 95%CI[1.17, 4.12], but not to IPV, b = 0.35, SE = 0.33, β = .07, p = .285, OR = 1.42, 95%CI[0.75, 2.68], or high-risk drinking, b = 0.50, SE = 0.39, β = .10, p = .204, OR = 1.64, 95%CI[0.77, 3.52]. Childhood physical abuse was significantly related to high-risk drinking in adulthood, b = 0.96, SE = 0.39, β = .18, p = .013, OR = 2.62, 95%CI[1.22, 5.63], but not to IPV, b = 0.32, SE = 0.34, β = .06, p = .340, OR = 1.38, 95%CI[0.71, 2.68], or depression, b = 0.34, SE = 0.34, β = .06, p = .315, OR = 1.40, 95%CI[0.73, 2.70]. Childhood sexual abuse was significantly related to adulthood depression, b = 1.02, SE = 0.32, β = .19, p = .001, OR = 2.78, 95%CI[1.50, 5.16]; and high-risk drinking, b = 0.87, SE = 0.36, β = .16, p = .015, OR = 2.38, 95%CI[1.18, 4.81], but not IPV, b = 0.02, SE = 0.31, β = .00, p = .944, OR = 1.02, 95%CI[0.56, 1.87]. Table 3 shows these relationships, with significant results in bold.

Figure 2.

Relationships between abuse experiences in childhood and intimate partner violence, depression, and high-risk drinking in adulthood. IPV = Intimate Partner Violence. Significant paths are solid lines; non-significant paths are dashed lines. Standardized coefficients are shown. *p < .05, **p < .01

Table 3.

Results of Path Analysis to Test the Specificity Model with Three Types of Child Abuse Predicting Adulthood IPV, Depression, and High-risk Drinking.

	b	SE	β	p	OR
Childhood Emotional Abuse
Depression	0.79	0.32	.15	.015	2.19
IPV	0.35	0.33	.07	.285	1.42
High-risk Drinking	0.50	0.39	.10	.204	1.64
Childhood Physical Abuse
Depression	0.34	0.34	.06	.315	1.40
IPV	0.32	0.34	.06	.340	1.38
High-risk Drinking	0.96	0.39	.18	.013	2.62
Childhood Sexual Abuse
Depression	1.02	0.32	.19	.001	2.78
IPV	0.02	0.31	.00	.944	1.02
High-risk Drinking	0.87	0.36	.16	.015	2.38

Note. Bold shows significant relationships.

Discussion

The results document that among U.S. Hispanic women, abuse in childhood is related to IPV, depression, and high-risk drinking. This study adds to the mounting evidence linking child abuse to these conditions in adulthood. Results were consistent with both the additive and specificity theoretical models. In this study, both the presence of any reported abuse and a greater number of types of abuse was related to IPV, depression, and high risk drinking among Hispanic women, lending support to the additive model. It should be noted the only the additive effects of childhood abuse experiences were linked to adulthood IPV, and none of the specific types of child abuse were related to adulthood IPV—unlike adulthood depression or high-risk drinking. In support of the specificity model, this study identified four significant relationships between specific types of childhood abuse and adult adverse health variables among Latinas, also lending support to the specificity model. Adult depression was related to childhood sexual and emotional abuse. Adult high-risk drinking was related to childhood physical and sexual abuse. This is consistent with previous research conducted with other populations ) and models that link exposure to stress (e.g., childhood trauma) to long-term psychological consequence (e.g., depression) in adulthood (e.g., Lang et al., 2004; Mandelli, Petrelli, & Serretti, 2015).

Unlike studies supporting the link between childhood emotional abuse and IPV in adulthood (Parks, Kim, Day, Garza & Larky, 2011; Whitefield et al., 2003), emotional abuse was not linked to IPV in this study. This may have been because IPV was only assessed during the participant’s current relationship, and not for all relationships. Previous studies supporting links between abuse during childhood and adulthood have used measures to capture lifetime exposure to victimizations during adulthood (Parks et al., 2011). Like reported by others, we did not find a relationship between emotional childhood abuse and problem drinking (Norman et al., 2012). Further research should aim to explain the differential effects that childhood emotional abuse has on health outcomes.

There are several limitations of this research study. First, this study used a cross-sectional design which collected data on child abuse from recall, and IPV, depression, or high-risk drinking in adulthood at one time point. Consequently, exposure-response and dose-response relationships could not be fully tested over time. Small subgroups, 15 – 17% reported abuse during childhood, and a majority (64%) reported abuse in adulthood, which could have affected the analysis and results. Second, data was collected through self-reports which are subject to reporting and recall bias. We had relatively high numbers of women with depression using the self-report CES-D than would be expected in a community sample. This study did not have a corroborating depression measure, such as a clinical interview which limits our ability to examine the meaning of scores. There is evidence for validity of the cutoff used in this study. For example, Thomas and colleagues (2001) showed that the cutoff of 16 had good sensitivity (.95) and specificity (.70) for major depressive disorder in a sample of low-income women in primary care. Another study with U.S. Latinos/Latinas in primary care showed similar levels of sensitivity (.90) and specificity (.84) for major depression (Robison et al., 2002). Future research into self-report measurement of depression for Latinas is warranted. Third, study staff engaged a non-random sample of Hispanic women from the community for the randomized control trial. Because a randomized trial does not necessarily require a random sample, results from this study may not be representative of the larger community. Hispanics are a large, heterogeneous group of people with a variety of backgrounds, language preferences, values, countries of origin, and lived experience. The women from this study were mostly immigrant women from Colombia, Cuba. and elsewhere in Latin America. Their experiences are likely different from other Hispanic/Latina women in the U.S., although the sample did include a small number of women born in Puerto Rico and Mexico. This study does provide preliminary evidence for the need for larger, multisite studies in this area. We recommend this type of study be replicated in multiple settings with a probability sample from the community, and with subgroups of Latinas.

Intervention programs that reduce and/or prevent childhood abuse, or improve protective factors that can buffer against childhood abuse, are likely to help Latinas well into adulthood. There is empirical evidence that home visitation programs, parent education, abusive head trauma prevention (i.e., programs preventing shaken baby syndrome), and multicomponent programs (i.e., programs including include parenting skills, family support, child care and preschool education) are effective in preventing child abuse (Mikton & Butchart, 2009). More research is needed to identify if these approaches are also effective in preventing child abuse, and how to best work within the diverse cultural milieus of U.S. Hispanic families to adapt these interventions and programs to the unique needs and strengths of multiple communities.

There a various issues to consider when developing tailored child-maltreatment prevention programs for Hispanic women communities. First, there appears to be differences in the risk and protective factors present between immigrant and U.S. born populations. For example, researchers have noted that U.S.-born Hispanic women are almost twice as likely to report adverse child experiences as immigrants (Caballero, Johnson, Munoz Buchanan & Ross DeCamp et al., 2017). More research is needed to better explain this decaying protective pattern in child maltreatment across generations. Additionally, attention must also be paid to addressing barriers and facilitators to participation in these interventions. Previous research has indicated that Hispanic families with high levels of family stress, and therefore potentially at greater risk for child maltreatment, are less likely to participate in prevention programs (Perrino et al., 2016). More research is needed to identify effective strategies to engage high risk Hispanic sub-groups in interventions aimed at preventing child maltreatment.