Female offenders...

http://ijo.sagepub.com/Criminology

Therapy and Comparative International Journal of Offender

http://ijo.sagepub.com/content/early/2013/01/09/0306624X12472879The online version of this article can be found at:

DOI: 10.1177/0306624X12472879

published online 11 January 2013Int J Offender Ther Comp CriminolStephen J. Tripodi, Eyitayo Onifade and Carrie Pettus-Davis

Supportof Childhood Victimization, Childhood Neglect, and Childhood Positive

Nonfatal Suicidal Behavior Among Women Prisoners: The Predictive Roles

Published by:

http://www.sagepublications.com

can be found at:CriminologyInternational Journal of Offender Therapy and ComparativeAdditional services and information for

http://ijo.sagepub.com/cgi/alertsEmail Alerts:

http://ijo.sagepub.com/subscriptionsSubscriptions:

http://www.sagepub.com/journalsReprints.navReprints:

http://www.sagepub.com/journalsPermissions.navPermissions:

What is This?

- Jan 11, 2013OnlineFirst Version of Record >>

at FLORIDA STATE UNIV LIBRARY on March 1, 2013ijo.sagepub.comDownloaded from

http://ijo.sagepub.com/

http://ijo.sagepub.com/content/early/2013/01/09/0306624X12472879

http://www.sagepublications.com

http://ijo.sagepub.com/cgi/alerts

http://ijo.sagepub.com/subscriptions

http://www.sagepub.com/journalsReprints.nav

http://www.sagepub.com/journalsPermissions.nav

http://ijo.sagepub.com/content/early/2013/01/09/0306624X12472879.full.pdf

http://online.sagepub.com/site/sphelp/vorhelp.xhtml


International Journal ofOffender Therapy and

Comparative CriminologyXX(X) 1 –18

© The Author(s) 2013Reprints and permission:

sagepub.com/journalsPermissions.navDOI: 10.1177/0306624X12472879

http://ijo.sagepub.com

472879 IJOXXX10.1177/0306624X12472879International Journal of Offender Therapy and Comparative CriminologyTripodi et al.

1Florida State University, Tallahassee, USA2Washington University in St. Louis, MO, USA

Corresponding Author:Stephen J. Tripodi, College of Social Work, Florida State University, 296 Champions Way, University Center C, Tallahassee, FL 32306, USA Email: [email protected]

Nonfatal Suicidal Behavior Among Women Prisoners: The Predictive Roles of Childhood Victimization, Childhood Neglect, and Childhood Positive Support

Stephen J. Tripodi1, Eyitayo Onifade1, and Carrie Pettus-Davis2

Abstract

Women entering prison report high rates of childhood victimization. Women in prison also report higher rates of nonfatal suicidal behavior (self-reported suicide attempts) than women in the general population and similar rates to their male counterparts despite having significantly lower suicide rates than males in the general population. Yet, there is a dearth of research that addresses the relationship between childhood victimization and suicidality for women prisoners in the United States. The purpose of this study is (a) to assess the relationship between childhood victimization and nonfatal suicidal behavior for a random sample of women prisoners; (b) to investigate predictive differences between childhood physical victimization, childhood sexual victimization, childhood neglect, and childhood support; and (c) to determine whether women prisoners with higher frequencies of childhood victimization and neglect are more likely to have attempted suicide than women prisoners with lower frequencies. Results indicate that childhood victimization, neglect, and lack of support are all significantly associated with nonfatal suicidal behavior among women prisoners. Frequency of childhood neglect had a larger effect size than frequency of childhood physical victimization, childhood sexual victimization, and lack of support. The results of this study add to the growing body of literature on childhood victimization and suicidality in general, and nonfatal suicidal behavior for prisoner populations in



2 International Journal of Offender Therapy and Comparative Criminology XX(X)

particular. The article ends with a discussion on clinical implications; particularly the finding that frequency of childhood victimization, childhood neglect, and lack of childhood support matters when determining the risk of suicidality.

Keywords

women prisoners, childhood victimization, neglect, support, suicidality

Women are presently the fastest growing population in American prisons. There were 112,797 women incarcerated in state or federal prison in 2008 compared with just 12,300 women incarcerated in U.S. prisons in 1980 (Guerino, Harrison, & Sabol, 2011; Pew Center on the States, 2008; West & Sabol, 2009). While the U.S. population has increased during the same period, explaining part of the incarceration growth, the increase in women inmates far exceeds the overall population growth (Petersilia, 2005). The number of women in prison has increased dramatically since the “War on Drugs” was enacted. The enormous growth in imprisonment among women is particu-larly alarming because many of these women are the custodial parent of their children (Frye & Dawe, 2008). Also distressing is that women who end up in prison report high rates of lifetime victimization, particularly in childhood. Up to 78% of incarcerated women report having histories of physical and/or sexual victimization (McDaniels-Wilson & Belknap, 2008; Tripodi & Pettus-Davis, 2013). Given the high prevalence of victimization among these women, it is not surprising that disproportionately high rates of mental health and substance abuse problems are found among women in prison when compared with the general population (Hicks, Vaidyanathan, & Patrick, 2010; Houser, Belenko, & Brennan, 2011). Many women enter prison with significant untreated mental and physical health impairments but receive few treatment options during or after incarceration. Of particular concern, women in prison have higher rates of nonfatal suicidal behavior (self-reported suicide attempts) than women in the gen-eral population (Fazel & Danesh, 2002).

In multiple countries, including the United States, prison suicide rates are dispro-portionate compared with the general population for all prisoners in general and women prisoners in particular. In a recent study of 12 countries, suicide rates in prison were above 100 per 100,000 inmates compared with 21 suicides per 100,000 people in the general population (Fazel, Grann, Kling, & Hawton, 2011). Approximately 95% of the inmates in this study were males; however, important information was discovered with the women in their sample. In the United States, although women in the general population are at a lower risk to commit suicide than men, women inmates are just as likely to commit suicide as their male counterparts (Fazel et al., 2011).

Lifetime suicide attempts for women prisoners range from 28% to 42% compared with approximately 9% in the general population (Chapman, Specht, & Cellucci, 2005; De Ravello, Abeita, & Brown, 2008). Despite high rates of suicidality among women prisoners, little is known about the predictors of suicidality for these women. Gaining a



Tripodi et al. 3

better understanding of predictors will allow for the development of targeted suicide prevention interventions in prison settings. This knowledge could also inform transi-tional reentry programming for women prisoners reentering community – recently released women prisoners are also at higher risk for suicidality (Pratt, Appleby, Piper, Webb, & Shaw, 2010).

Although some suicide prevention interventions exist for prisoners, these interven-tions are often based on the experiences of male prisoners (Dye, 2011; Hayward, Kravitz, Goldman, & Freeman, 2000) and do not take into account the unique needs of women prisoners. One particularly pertinent relationship for women prisoners is the association between childhood victimization and risk for suicidality in the general population. Women prisoners report high rates of childhood victimization—much higher than their male counterparts. Yet, there is only one published study that addresses the relationship between childhood victimization and suicidality for women prisoners in the United States (Clements-Nolle, Wolden, & Bargmann-Losche, 2009).

Clearly, additional studies are needed to develop a more thorough understanding of this relationship for women prisoners, and to gain deeper insight into how specific types of childhood victimization may partially engender suicidality. Thus, the purpose of this study is threefold: (a) to assess the relationship between childhood victimiza-tion and suicide attempts for a random sample of women prisoners; (b) to investigate predictive differences between childhood physical victimization, childhood sexual victimization, childhood neglect, and childhood support; and (c) to determine whether women prisoners with higher frequencies of childhood victimization and neglect are more likely to have experienced suicidality than women prisoners with lower frequencies.

BackgroundPredictors of Suicidal Thoughts, Suicide Attempts, and Successful Suicide

Research with samples from the general population indicates that experiencing childhood victimization is associated with suicidal thoughts, suicide attempts, and successful suicide (Joiner et al., 2007; Roy, 2004). Glowinski et al. (2001) inter-viewed approximately 3,000 female adolescent twins and found that childhood physical abuse was associated with attempting suicide. Ullman’s (2004) systematic review of suicidality among women in the general population similarly found that previous victimization is associated with suicidality. Joiner et al. (2007) found that childhood victimization was related to the lifetime number of suicide attempts for a sample drawn from the National Comorbidity Study, which is a probability sample of approximately 10,000 participants representing the population of the United States. Specifically, the effects on suicidality of childhood physical abuse (t = 5.98, p < .05) and sexual abuse (t = 5.40, p < .05) were similar to one another, and both were more strongly associated with lifetime suicide attempts than sexual molesta-




tion (t = 2.88, p < .05) and verbal abuse (t = 1.68, p = not significant). For their study, Joiner et al. (2007) measured sexual molestation as somebody touching the participants’ genitals when the participant did not want them to be touched and measured sexual abuse as being raped. Furthermore, Afifi et al. (2008) found that childhood victimization accounts for 16% of the variance for suicidal thoughts and 50% of the variance for suicide attempts among women in the general population.

Although there is a dearth of research assessing the relationship between child-hood victimization and suicidality among both men and women prisoner popula-tions, the existing studies have similar results as for the general population. Mandelli, Carli, Roy, Serretti, and Sarchiapone (2011), for example, assessed the influence of childhood trauma on the onset and repetition of nonfatal suicidal behavior with a sample of male prisoners in Italy. Using cluster analyses to clus-ter the men prisoners into two homogeneous groups for exposure to each type of early trauma measured (physical abuse, sexual abuse, and neglect), they found that inmates placed in the cluster with higher frequencies of childhood victimiza-tion and neglect were more likely to have exhibited behaviors considered suicidal.

There are also few empirical explanations for why women prisoners report higher rates of attempted suicide than women in the general population. Some researchers posit that the disproportionately high rates of childhood trauma may contribute to the greater prevalence of suicidality among criminal justice-involved adults (Harlow, 1999; Messina & Grella, 2006; Swogger, You, Cashman-Brown, & Conner, 2011). The only published study of victimization and suicidality among women prisoners in the United States assessed the relationship between childhood trauma and suicide attempts. Clements-Nolle et al. (2009) found women prisoners who had attempted suicide in the past and women prisoners who reported they may commit suicide in the future were more likely to have experienced physical abuse, sexual abuse, physical neglect, and emotional neglect in childhood.

Overall, while research indicates significant relationships between childhood victimization and suicidality, the literature is inconsistent regarding the relation-ships between different types of childhood victimization and suicidality for sam-ples from both the general population and from prisons. For women in the general population, Glowinski et al. (2001) and Brent et al. (1994) found that childhood physical victimization was associated with suicidality, but childhood sexual vic-timization was not. In later studies, Joiner et al. (2007) and Afifi et al. (2008) found that physical victimization and sexual victimization were associated with suicidality in the general population. The disparate findings regarding the differ-ent types of victimization and their relationships with suicidality are likely due to the researchers studying samples that differ from each other and measuring vic-timization differently.

The associations of both childhood neglect and childhood support with suicidality are understudied compared with the effects of childhood physical victimization and sexual victimization. In 2000, Twomey, Kaslow, and Croft (2000) stated there were no



Tripodi et al. 5

known empirical studies that assess the relationship between neglect and suicidality. Since 2000, Arata, Langhinrichsen-Rohling, Bowers, and O’Farrill-Swails (2005) studied more than 300 college students and found childhood neglect to be associated with suicidality. Brodsky and Stanley (2008) came to the same conclusion when study-ing the relationships between type of childhood victimization and suicidality for the general population. In a study with African American women, Meadows and Kaslow (2002) found that women who were neglected in childhood become more hopeless, and in turn are more likely to experience suicidality as adults.

The perceived support experienced in childhood may help to explain differences in suicidality; however, there is limited extant research on the relationship between child-hood support and suicidality. Kerr, Preuss, and King (2006) found that low support received in childhood is associated with increased adolescent mental health problems, including suicidality. Recent research with prisoners indicates a relationship between lack of support in childhood and suicidality as an adult. Marzano, Hawton, Rivlin, and Fazel (2011) conducted a study with women prisoners in England and found that women who attempted suicide were more likely to feel they were not supported in childhood by their family than women prisoners who did not harm themselves. Other recent studies indicate that all categories of childhood victimization as categorized in previous studies were associated with suicidality (cf. Mandelli et al. 2011: male pris-oners; Clements-Nolle et al. 2009: women prisoners).

This study represents the second (published) investigation of victimization and sui-cidality with women prisoners in the United States. We build on the important prelimi-nary research of Clements-Nolle and colleagues (2009) by assessing the relationship between specific types of childhood victimization and suicidality by using type of victimization as a predictor variable in regression models. We also assess the predic-tive role of childhood neglect and childhood support on suicidality. We used a random sample of incarcerated women in North Carolina to answer the following research questions:

Research Question 1: Are histories of childhood physical victimization, childhood sexual victimization, childhood neglect, and childhood support associated with nonfatal suicidal behavior for this sample of incarcerated women?

Research Question 2: Are women prisoners who report higher frequencies of childhood victimization (physical victimization and sexual victimization) more likely to experience nonfatal suicidal behavior than women prisoners with lower frequencies of childhood victimization?

Research Question 3: Are women prisoners who report higher frequencies of childhood neglect more likely to experience nonfatal suicidal behavior than women prisoners with lower frequencies of childhood neglect?

Research Question 4: Are women prisoners who report lower frequencies of support in childhood more likely to experience nonfatal suicidal behavior than women prisoners with higher frequencies of support in childhood?




MethodParticipantsA random sample of 125 women from two prisons in North Carolina voluntarily par-ticipated in the study. One of the prisons is a maximum-, medium-, and minimum-level facility that holds approximately 1,300 prisoners, and the other prison is a minimum security prison that holds approximately 200 prisoners. There were 630 potential par-ticipants (women scheduled to be released from prison 30-120 days after the data col-lection period), of which 229 were randomly sampled. The purpose of this random sampling was simply to shorten the list of eligible participants to then randomly select women from the prisons. Of those 229 potential participants, 150 women were ran-domly selected and asked to participate, and 125 women agreed to do so yielding an 83% response rate. To be eligible, the prisoner had to be between 30 and 120 days from release at time of data collection, at least 18 years old, and cognitively functioning to the extent that they were able to indicate they understood the nature of the study and what being a study participant entailed. Participants had to be between 30 and 120 days from release because this particular study is part of a larger study assessing the relation-ships between childhood victimization and recidivism with mental health and sub-stance use problems as mediating variables. Thus, to assess recidivism as the outcome variable in the larger study, women had to be scheduled to release from prison soon after data collection. All procedures were approved by the Florida State University and North Carolina Department of Correction Human Subject Review Boards.

Five members of the research team conducted the interviews by reading out loud the measurements and writing down the participants’ responses. Two of the interview-ers have a PhD in social work, one has a PhD in community psychology, and two have a masters of social work degree. All members of the research team had at least a gradu-ate degree in social work or related field so they could recognize whether the partici-pant was having a traumatic reaction to the questions being asked. It was protocol for the researchers to refer the participant to mental health counseling if the researcher considered it appropriate or if the participant desired counseling.

The demographics and mean victimization scores for the sample are included in Table 1. The average age for the sample was 34 (SD = 9.94) with a range of 19 to 62. In all, 66 participants were Caucasian (52.8%), 54 participants were African American (43.2%), and 5 participants were Hispanic (4%). African Americans were dispropor-tionately represented in North Carolina prisons compared with the state of North Carolina. A total of 70% of North Carolina citizens are White, 21.3% are African American, 6.5% are Hispanic, and 1.2% are American Indian.

MeasuresPredictor Variables

Self-reported frequency of childhood trauma. The researchers used a 20-item version of the Childhood Trauma Questionnaire (CTQ) to measure childhood victimization



Tripodi et al. 7

(Bernstein, Fink, Handelsman, & Foote, 1994). When completing the CTQ, the par-ticipant was asked how frequently certain events happened to them as a child on a 5-point Likert-type scale, ranging from never to very frequently. For example, one item on the CTQ states, “When I was growing up, people in my family hit me so hard that it left me with bruises or marks,” and another item states, “When I was growing up, someone threatened to hurt me or tell lies about me unless I did something sexual to them.” The total CTQ score is based on the cumulative frequency of childhood victimization. In addition, self-reported frequency of childhood physical victimiza-tion, childhood sexual victimization, childhood neglect, and childhood support are four of the subscales within the CTQ.

The reliability and validity of the CTQ—including criterion validity, convergent validity, and discriminant validity—has been consistently evaluated with research on groups such as psychiatric patients and college students (Bernstein et al., 2003; Bernstein & Fink, 1998). The CTQ has high internal consistency for this study (α= .943). The internal consistency for the four subscales of interest for this study are as follows: Sexual Abuse subscale, α = .977; Support subscale, α = .937; Physical Abuse subscale, α = .920; and Neglect, α = .686.

The CTQ subscales measure the frequency of each construct—0 represents no pres-ence of the construct, and the upper bound represents high frequency of the construct. The sample’s mean score for the Physical Abuse subscale was 7.54 (range = 0-20, SD = 7.54), the mean score for the Sexual Abuse subscale was 6.26 (range = 0-20, SD = 6.26), the mean score for the Neglect subscale was 1.82 (range = 0-15, SD = 3.11), and the mean score for the Support subscale was 16.28 (range = 0-20, SD = 5.47).

Table 1. Characteristics of the Sample

M SD

Age at incarceration (N = 125) 34.3 9.94CTQ total score (N = 121) 19.21 19.03CTQ Physical Abuse subscale (N = 122) 7.54 6.53CTQ Sexual Abuse subscale (N = 122) 6.26 8.22CTQ Neglect subscale (N = 123) 1.82 3.11CTQ Support subscale (N = 123) 16.28 5.47

Frequency %

Race (N = 125) African American 54 43.2 Caucasian 66 52.8 Hispanic 5 4.0Substance use disorder (N = 124) 77 62.1Attempted suicide (N = 124) 32 25.8

Note: CTQ = Childhood Trauma Questionnaire.




Demographics. Age is a continuous variable based on the participants’ age at the time of data collection. Race is a dichotomous variable portraying whether the partici-pant is a minority (coded as 1) or otherwise (coded as 0). Race and age were consid-ered important to include in the models because prior researchers have found an association between these variables and nonfatal suicidal behavior (Clements-Nolle et al., 2009; Moscicki, 1995; Shah, 2007). Clements-Nolle et al. (2009) found that White prisoners were more likely to report having attempted suicide than minority prisoners, and Shah (2007) found a positive correlation between age and risk of sui-cide. Race and age, however, were not significant predictor variables in any of the four logistic regression models in this particular study.

Substance use problems. Substance use problems have also been associated with suicidality in prison and nonprison populations (Clements-Nolle et al. 2009; Spokas, Wenzel, Brown, & Beck, 2012; Spokas, Wenzel, Stirman, Brown, & Beck, 2009). Whether the participant has a substance use problem is defined as whether they meet the criteria for a substance abuse disorder as assessed by the Substance Abuse Module (SAM; Cottler, Robins, & Helzer, 1989). The SAM is a 38-item measure that assesses Diagnostic and Statistical Manual of Mental Disorders (4th ed.; American Psychiatric Association, 1994) diagnostic criteria for present and lifetime substance use disorders. The reported reliability kappa values range from .82 to .84. For this study, whether the participant met the criteria for a substance abuse disorder is a dichotomous variable based on responses to the SAM.

Outcome VariableNonfatal suicidal behavior. Nonfatal suicidal behavior was measured using the psychi-

atric information section of the Addiction Severity Index (ASI; McLellan et al., 1992). The ASI is designed to address many potential problem areas that may co-occur with substance use disorders, including psychiatric problems. Although the primary purpose of the ASI is to assess substance use, the psychiatric subscale provided applicable sui-cidality data. When completing this scale, the researchers asked participants whether they have attempted suicide in the past 30 days and whether they have attempted sui-cide ever in their lifetime. No participants reported that they attempted suicide in the past 30 days; thus, the outcome variable for this study is whether the participant has ever attempted suicide, which is our definition of nonfatal suicidal behavior.

Analyses Binary logistic regression allows for the assessment of the relationship between mul-tiple predictor variables (age, race, substance abuse disorder, childhood victimization, childhood neglect, and childhood support) and a dichotomous outcome variable (whether the participant has attempted suicide; Hosmer & Lemeshow, 2000; Kutner, Nachtsheim, Neter, & Li, 2005). The logistic regression produces an odds ratio (OR). ORs are interpretable in terms of the probabilities of the dependent variable (whether the participant had attempted suicide) being equal to 1 when all other variables in the



Tripodi et al. 9

model are held equal. The data in this study fit the assumptions of the logistic regres-sion that the outcome variables were binary and the sample was independently and randomly selected.

We used logistic regression analyses to assess the influence of physical victimiza-tion, sexual victimization, neglect, and childhood support on whether the participants have ever attempted suicide. Prior to analyzing the logistic regression analyses, we calculated the Hosmer and Lemeshow statistic (Hosmer & Lemeshow, 2000) to deter-mine the models’ goodness of fit. All Hosmer and Lemeshow statistics were not sig-nificant, indicating goodness of fit for all of the regression models.

Whether the participant ever attempted suicide is the dependent variable in all of the logistic regression models. The predictor variables for Model 1 are age, race, sub-stance abuse disorder, and self-reported frequency of childhood physical victimiza-tion. The predictor variables for Model 2 are age, race, substance abuse disorder, and self-reported frequency of childhood sexual victimization. The predictor variables for Model 3 are age, race, substance abuse disorder, and self-reported frequency of child-hood neglect. The predictor variables for Model 4 are age, race, substance abuse dis-order, and self-reported frequency of positive support received in childhood.

Predictor variables were entered using a forced entry approach. For each model, we examined the following: case to predictor variable ratio, multicollinearity by examin-ing the variance inflation factors (VIF), model fit, and relationships between predictor variables and the dependent variable.

ResultsFour binary regression models were conducted to answer the following four research questions: (a) Are histories of childhood physical victimization, childhood sexual vic-timization, childhood neglect, and childhood support associated with nonfatal suicidal behavior for this sample of incarcerated women? (b) Are women prisoners who report higher frequencies of childhood victimization (physical victimization and sexual victim-ization) more likely to experience nonfatal suicidal behavior than women with lower frequencies of childhood victimization? (c) Are women prisoners who report higher frequencies of childhood neglect more likely to experience nonfatal suicidal behavior than women with lower frequencies of childhood neglect? (d) Are women prisoners who report lower frequencies of support in childhood more likely to experience nonfatal suicidal behavior than women with higher frequencies of support in childhood?

The first regression model had four predictor variables: age, race, substance abuse disorder, and self-reported history of childhood physical victimization. The 122 cases for the analysis—three cases had missing data and were excluded from the analysis— satisfied the minimum case to predictor variable ratio for logistic regression (Hosmer & Lemeshow, 2000). There were 122 cases and four predictor variables, for a ratio of 30.5:1.

Multicollinearity was assessed by examining the VIF (Hosmer & Lemeshow, 2000). A VIF around 10.0 indicates problems of multicollinearity. The VIFs for the variables in this analysis ranged from 1.034 to 1.119; multicollinearity was not a problem.




The presence of a relationship between the dependent variable and the combination of predictor variables is based on the statistical significance of the model chi-square for the –2 log-likelihood differences between the model with the predictor variables and the model without the predictor variables (Hosmer & Lemeshow, 2000). The probability of the model χ2(20.594) was p < .001. The null hypothesis that there is no difference between the model with only a constant and the model with the predictor variables was rejected, supporting the existence of a relationship between the predictor variables and the dependent variable.

We examined the relationships between the predictor variables and the dependent variable by analyzing the significance of the Wald test of the beta coefficient and the interpretations of the ORs for significant relationships. In Model 1, as indicated in Table 2, having a substance abuse disorder (OR = 3.018, p = .046) and history of physical victimization (OR = 1.123, p = .001) were both significant predictors of sui-cide attempts. This indicates that women with substance use disorders were approxi-mately 3 times more likely to have attempted suicide. A 1-unit increase in frequency of self-reported childhood physical victimization was associated with a 12.3% increased likelihood of having attempted suicide, all else being equal.

The second regression model assessed age, race, substance abuse disorder, and fre-quency of childhood sexual victimization as the predictor variables. The 122 cases for the analysis satisfied the minimum case to predictor variable ratio for logistic regres-sion. Three cases were missing data and excluded from the analysis. There were 122 cases and four predictor variables, for a ratio of 30.5:1. There were no problems of multicollinearity. The VIFs for the predictor variables ranged from 1.025 to 1.084.

The logistic regression model indicated a relationship between the dependent vari-able and the combination of predictor variables. The probability of the model χ2(26.526) was p < .001. The null hypothesis that there is no difference between the model with

Table 2. Childhood Victimization (CTQ Physical Abuse Subscale and Sexual Abuse Subscale) and Suicide Attempts

Models 1 and 2

Model 1 Model 2

Predictors B (SE) OR (95% CI) p Wald B (SE) OR (95% CI) p Wald

Age 0.026 (0.02) 1.027 [0.979, 1.076] .272 1.207 0.016 (0.03) 1.016 [0.968, 1.067] .516 0.423Race 0.002 (0.47) 1.002 [0.396, 2.535] .996 0.000 −0.205 (0.49) 0.815 [0.314, 2.116] .675 0.177Substance abuse

disorder1.104 (0.55) 3.018 [1.019, 8.937] .046 3.975 1.581 (0.58) 4.861 [1.554, 15.201] .007 7.228

CTQ Physical Abuse subscale

0.116 (1.13) 1.123 [1.050, 1.202] .001 11.324 NA NA NA NA

CTQ Sexual Abuse subscale

NA NA NA NA 0.015 (0.03) 1.122 [1.060, 1.188] <.001 18.556

Note: CTQ = Childhood Trauma Questionnaire; OR = odds ratio; CI = confidence interval.



Tripodi et al. 11

only a constant and the model with the predictor variables is rejected. Women with substance use disorders were approximately 4.8 times more likely to have attempted suicide (OR = 4.9, p = .007). A 1-unit increase in frequency of childhood sexual vic-timization was associated with being 12.2% more likely to have attempted suicide (OR = 1.122, p < .001), all else being equal.

The third regression model assessed age, race, substance abuse disorder, and child-hood neglect as the predictor variables. The 123 cases for the analysis satisfied the minimum case to predictor ratio for logistic regression. There were 123 cases and four predictor variables, for a ratio of 30.75:1. There were no problems of multicollinearity. The VIFs for the predictor variables ranged from 1.063 to 1.096.

The logistic regression model indicated a relationship between the dependent vari-able and the combination of predictor variables. The probability of the model χ2(24.525) was p < .001. The null hypothesis that there is no difference between the model with only a constant and the model with the predictor variables was rejected. As indicated in Table 3, women with a substance abuse disorder in this model were approximately 3.7 times more likely to have attempted suicide (OR = 3.655, p = .022). A 1-unit increase in frequency of childhood neglect was associated with being 32.1% more likely to have attempted suicide (OR = 1.321, p < .001), all else being equal.

The fourth regression model assessed age, race, substance abuse disorder, and sup-port in childhood as the predictor variables. The 123 cases satisfied the minimum case to predictor variable ratio, which is 30.75:1. There were no problems of multicol-linearity. The VIFs for the predictor variables ranged from 1.097 to 1.125.

The logistic regression model indicated a relationship between the dependent vari-able and the combination of predictor variables. The probability of the model χ2(23.217) was p < .001. The null hypothesis that there is no difference between the model with only a constant and the model with the predictor variables is rejected. As indicated in

Table 3. Childhood Neglect and Support (CTQ Neglect Subscale and Support Subscale) and Suicide Attempts

Models 3 and 4

Model 3 Model 4

Predictors B (SE) OR (95% CI) p Wald B (SE) OR (95% CI) p Wald

Age 0.015 (0.03) 1.015 [0.967, 1.065] .554 0.351 0.008 (0.03) 1.008 [0.960, 1.059] .745 0.106Race 0.153 (0.48) 1.166 [0.454, 2.993] .750 0.102 0.432 (0.49) 1.540 [0.584, 4.058] .382 0.736Substance abuse

disorder1.296 (0.57) 3.655 [1.208, 11.059] .022 5.265 1.320 (0.56) 3.743 [1.244, 11.258] .019 5.517

CTQ Neglect subscale

0.278 (0.077) 1.321 [1.136, 1.536] <.001 13.017 NA NA NA NA

CTQ Support subscale

NA NA NA NA . −0.156 (0.04) 0.856 [0.786, 0.931] <.001 13.001

Note: CTQ = Childhood Trauma Questionnaire; OR = odds ratio; CI = confidence interval.




Table 3, women with a substance abuse disorder in this model were approximately 3.7 times more likely to have attempted suicide (OR = 3.743, p = .019). A 1-unit increase in frequency of support in childhood was associated with being 14.4% less likely to have ever attempted suicide (OR = .856, p < .001), all else being equal.

Overall, the binary logistic regression models indicated that substance abuse, child-hood victimization, childhood neglect, and childhood support are all associated with nonfatal suicidal behavior. Women in the sample who met the criteria for a substance abuse disorder were more likely to have attempted suicide at least once in their lives than women who did not. Women in the sample who experienced childhood physical victimization or childhood sexual victimization were more likely to have attempted suicide than women who had not experienced childhood victimization. In addition, frequency was an important factor when assessing this relationship between childhood victimization and nonfatal suicidal behavior—women who reported more frequent childhood victimization were more likely to have attempted suicide than women who reported less frequent childhood victimization. Finally, both childhood support and childhood neglect were significantly associated with nonfatal suicidal behavior. Women in the sample who reported more frequent childhood support were less likely to have attempted suicide, and women in the sample who reported more frequent childhood neglect were more likely to have attempted suicide. Frequency of childhood neglect had the largest effect size of all the predictor variables.

DiscussionThe primary purposes of this study were to add to the knowledge base on the influence of childhood victimization, childhood neglect, and childhood support on suicidality for women prisoners by (a) assessing the associations between different types of childhood experiences and suicidality, (b) evaluating whether childhood neglect and childhood support are associated with suicidality, (c) determining whether women prisoners with higher frequencies of childhood victimization are more likely to have experienced suicidality than women with lower frequencies, and (d) determining whether women prisoners with higher frequencies of childhood neglect, and lower frequencies of support, are more likely to have experienced suicidality. Consistent with other research on suicidality (Clements-Nolle et al., 2009), we found women who experienced childhood physical victimization, childhood sexual victimization, or childhood neglect were more likely to have attempted suicide at least once in their lives. We also found that women who perceived themselves as being supported in childhood were significantly less likely to have experienced suicidality.

We found that higher frequencies of both childhood physical victimization and childhood sexual victimization were associated with a higher likelihood of suicidality. Frequency was also important when assessing the influence of neglect and support on suicidality. Women prisoners in this sample who reported higher frequencies of neglect were more likely to experience suicidality, and higher frequencies of childhood sup-port were associated with a reduced chance of having experienced suicidality.



Tripodi et al. 13

The results of this study add to the growing body of literature on childhood victim-ization and suicidality in general (Dube et al., 2001; Enns et al., 2006) and suicidality for prisoner populations in particular (Clements-Nolle et al., 2009; Mandelli et al., 2011). Clements-Nolle et al. (2009) conducted an important study in 2009 that assessed the relationship between childhood victimization and past suicide attempts and found a significant relationship between the two constructs. This particular study builds on their work by including four different regression models with the two different types of childhood victimization, childhood neglect, and childhood support as the predictor variables. Thus, we were able to look at the influence of the two primary types of childhood victimization (physical victimization and sexual victimization), along with neglect and support, on suicidality in a systematic and robust manner. We were also able to build on previous work assessing predictors of suicidality for women prisoners by analyzing the influence of frequencies of childhood victimization, neglect, and childhood support. This allows a more nuanced evaluation of the relationship between childhood experiences and suicidality and potentially a more targeted and specific approach when developing programs to clinically work with women prisoners who have suffered childhood victimization and neglect.

Other important clinical considerations arose from the results of this study involv-ing the finding that frequency of childhood victimization and frequency of childhood neglect is associated with a higher likelihood of suicidality for this particular sample of incarcerated women. Trauma researchers have long recognized that many factors contribute to the pervasive and lasting negative consequences of childhood victimiza-tion and neglect—and frequency does not necessarily predict pervasive and intrusive consequences. However, our data suggest frequency may be clinically relevant for women prisoners, and thus, intervention development should occur with this aware-ness in mind. For example, as suicide prevention interventions are developed and refined for women prisoners, interventions should be tested that consider different levels of treatment or different dosage of treatment. The study results suggest that treatment planning should include assessment of not only a history of childhood vic-timization and neglect but also the frequency of these experiences and the extent to which the related trauma is still intrusive to the women’s functioning.

Study results also suggest that perceived childhood support from childhood caregiv-ers may perform a protective function against suicidality for these women. There are two potential clinical implications that arise from this result. First, if positive support in childhood is in fact protective, positive support in adulthood may also be protective. This suggests that intervention development and testing should incorporate social sup-port as a possible mechanism in preventing suicidality. Naturally occurring support has been found to be effective in terms of reducing recidivism and improving postrelease outcomes for male prisoners (cf. for a review, Pettus-Davis, Howard, Roberts-Lewis, & Scheyett, 2011). Although these study results do not speak to the potential for preven-tive interventions for children of women prisoners, the results suggest further research into how women prisoners with children may provide positive support to their children as a targeted suicidality prevention effort. There are existing parenting programs in




prisons throughout the United States, many of which are effective in teaching women prisoners the skills and thoughts necessary to be successful and positive parents on release (Tripodi, Bledsoe, Kim, & Bender, 2011). Suicidality programs during incar-ceration have the potential to offer women prisoners an opportunity not only to recover from their own emotional problems but also to learn and use positive parenting prac-tices that can support positive future outcomes of their children.

Also of importance in this study, although not one of our primary research ques-tions, is the relationship between substance abuse and suicidality. Women prisoners in this sample meeting the criteria for a substance use disorder were significantly more likely to report suicidality. Although we are not able to establish temporal order between substance abuse and suicidality, the two are highly correlated, so we recom-mend that future researchers conduct a study determining whether the onset of sub-stance abuse predicts suicidality for women prisoners.

The strengths and limitations of this study should be considered when interpreting the findings. The primary strength of this study is that the sample was randomly selected from a list of all women eligible to participate. This allows generalization from our sample to the population of women close to release from the two prisons in North Carolina where the sample was drawn. In addition, those with severe mental illness were not excluded from our sample like many studies with prisoners, again, allowing our sample to adequately represent the population. A third strength is that the CTQ allowed us to look at the frequency of childhood victimization as a continu-ous variable, rather than simply asking the participants whether they were victimized as children.

The primary limitation to this study is that we were not able to assure temporal order between childhood victimization and suicide attempts. It is possible that the suicide attempt preceded the childhood victimization for some of the participants. We also suspect that mental health problems act as a mediating variable and helps explain the relationships between childhood victimization, neglect, and support on suicidality. However, we were not able to assess psychiatric diagnoses as a mediating variable because of temporal issues—We had no way of knowing whether the participant was diagnosed before or after their suicide attempt. In addition, while the 83% response rate can be considered a strength to the study, we do not have information on the 17% who declined to participate. It is unknown how this affects our sample. The third limi-tation is that the information comes from self-reports. As Mandelli et al. (2011) point out, although the CTQ has been shown to be reliable and valid, it does contain the following limitations: (a) It does not provide a reliable estimate of the presence of victimization and (b) it does not provide information on the severity and duration of each type of victimization. Finally, the single-item self-reported measure for nonfatal suicidal behavior may not be psychometrically valid.

Future researchers should take careful steps to assure temporal order between childhood victimization and suicide attempts, perhaps by asking the participants if they have attempted suicide since the age of 18 and using a childhood victimization measure for victimization such as the CTQ. In addition, obtaining information from



Tripodi et al. 15

women in prison from different states in different geographic areas of the United States would make the sample more generalizable to all women in U.S. prisons. This study looked at two prisons in one state, so generalization to all women prisoners in the United States cannot be confirmed. Finally, although substance use disorders are included in these regression models as predictor variables, future researchers should look at both substance use problems and psychiatric diagnoses as mediating variables, to assess whether substance use and mental health problems is a link between child-hood victimization and suicidality for women prisoners.

Acknowledgments

Special thanks to David Edwards of the North Carolina Department of Correction, and both the administration and staff at the two prisons where data collection took place.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study was supported by a grant from the National Institute of Mental Health (Grant P30 MH079920).

References

Afifi, T. O., Enns, M. W., Cox, B. J., Asmundson, G. J., Stein, M. B., & Sareen, J. (2008). Pop-ulation attributable fractions of psychiatric disorders and suicide ideation and attempts associated with adverse childhood experiences. American Journal of Public Health, 98, 946-952.

American Psychiatric Association. (1994). Diagnostic and statistical manual of mental disor-ders (4th ed.). Washington, DC: Author.

Arata, C. M., Langhinrichsen-Rohling, J., Bowers, D., & O’Farrill-Swails, L. (2005). Single versus multi-type maltreatment: An examination of the long-term effects of child abuse. Journal of Aggression, Maltreatment, & Trauma, 11, 29-52.

Bernstein, D. P., & Fink, L. (1998). Childhood Trauma Questionnaire: A retrospective self-report manual. San Antonio, TX: The Psychological Corporation.

Bernstein, D. P., Fink, L., Handelsman, L., & Foote, J. (1994). Initial reliability and validity of a new retrospective measure of child abuse and neglect. American Journal of Psychiatry, 151, 1132-1136.

Bernstein, D. P., Stein, J. A., Newcomb, M. D., Walker, E., Pogge, D., Ahluvalia, T., & Zule, W. (2003). Development and validation of a brief screening version of the Childhood Trauma Questionnaire. Child Abuse and Neglect, 27, 169-190.

Brent, D. A., Perper, J. A., Moritz, G., Liotus, L., Schweers, J., Balach, L., & Roth, C. (1994). Familiar risk factors for adolescent suicide: A case-control study. Acta Psychiatrica Scan-dinavica, 89, 52-58.




Brodsky, B. S., & Stanley, B. (2008). Adverse childhood experiences and suicidal behavior. Psychiat-ric Clinics of North America, 31, 223-235.

Chapman, A. L., Specht, M. W., & Cellucci, A. J. (2005). Borderline personality disorder and self-harm: Does experiential avoidance play a role? Suicide and Life-Threatening Behavior, 35, 388-399.

Clements-Nolle, K., Wolden, M., & Bargmann-Losche, J. (2009). Childhood trauma and risk for past and future suicide attempts among women in prison. Women’s Health Issues, 19, 185-192.

Cottler, L. B., Robins, L. N., & Helzer, J. E. (1989). The reliability of the CIDI-SAM: A comprehen-sive substance abuse interview. British Journal of Addiction, 84, 801-814.

De Ravello, L., Abeita, J., & Brown, P. (2008). Breaking the cycle/mending the hoops: Adverse child-hood experiences among incarcerated American Indian/Alaska Native women in New Mexico. Health Care Women International Journal, 29, 300-315.

Dube, S. R., Anda, R. F., Felitti, V. J., Chapman, D. P., Williamson, D. F., & Giles, W. H. (2001). Child-hood abuse, household dysfunction, and the risk of attempted suicide throughout the lifespan. Journal of the American Medical Association, 286, 3089-3096.

Dye, M. H. (2011). The gender paradox in prison suicide rates. Women & Criminal Justice, 21, 290-307.

Enns, M. W., Cox, B. J., Afifi, T. O., De Graaf, R., Have, M. T., & Sareen, J. (2006). Childhood adver-sities and risk for suicidal ideation and attempts: A longitudinal population-based study. Psycho-logical Medicine, 36, 1769-1778.

Fazel, S., & Danesh, J. (2002). Serious mental disorder in 23000 prisoners: A systematic review of 62 surveys. Lancet, 359, 545-550.

Fazel, S., Grann, M., Kling, B., & Hawton, K. (2011). Prison suicide in 12 countries: An ecological study of 861 suicides during 2003-2007. Social Psychiatry and Psychiatric Epidemiology, 46, 191-195. doi:10.1007/s00127-00010-00184-00124

Frye, S., & Dawe, S. (2008). Interventions for women prisoners and their children in the post-release period. Clinical Psychologist, 12, 99-108.

Glowinski, A. L., Bucholz, K. K., Nelson, E. C., Fu, Q., Madden, P. A., Reich, W., & Heath, A. C. (2001). Suicide attempts in an adolescent female twin sample. Journal of the Ameri-can Academy of Child & Adolescent Psychiatry, 40, 1300-1307.

Guerino, P., Harrison, P. M., & Sabol, W. J. (2011). Prisoners in 2010 (Report No. NCJ 236096). Retrieved from http://bjs.ojp.usdoj.gov/content/pub/pdf/p10.pdf

Harlow, C. W. (1999). Prior abuse reported by inmates and probationers (NCJ Publication No. 1728797). Rockville, MD: U.S. Department of Justice.

Hayward, T. W., Kravitz, H. W., Goldman, L. B., & Freeman, A. (2000). Characteristics of women in jail and treatment orientations: A review. Behavior Modification, 24, 307-324.

Hicks, B. M., Vaidyanathan, U., & Patrick, C. J. (2010). Validating female psychopathy subtypes: Differences in personality, antisocial and violent behavior, substance abuse, trauma, and mental health. Personality Disorders: Theory, Research, and Treatment, 1, 38-57.

Hosmer, D. W., & Lemeshow, S. (2000). Applied logistic regression (2nd ed.). New York, NY: Wiley.Houser, K. A., Belenko, S., & Brennan, P. K. (2011). The effects of mental health and substance

abuse disorders on institutional misconduct among female inmates. Justice Quarterly. Advance online publication. doi:10.1080/07418825.2011.641026



Tripodi et al. 17

Joiner, T., Sachs-Ericsson, N., Wingate, L., Brown, J., Anestis, M., & Selby, E. (2007). Child-hood physical and sexual abuse and lifetime number of suicide attempts: A persistent and theoretically important relationship. Behaviour Research and Therapy, 45, 539-547.

Kerr, D. C., Preuss, L. J., & King, C. A. (2006). Suicidal adolescents’ social support from family and peers: Gender-specific associations with psychopathology. Journal of Abnor-mal Child Psychology, 34, 103-114.

Kutner, M. H., Nachtsheim, C. J., Neter, J., & Li, W. (2005). Applied linear statistical models (5th ed.). New York, NY: McGraw-Hill Irwin.

Mandelli, L., Carli, V., Roy, A., Serretti, A., & Sarchiapone, M. (2011). The influence of childhood trauma on the onset and repetition of suicidal behavior: An investigation in a high risk sample of male prisoners. Journal of Psychiatric Research, 45, 742-747.

Marzano, L., Hawton, K., Rivlin, A., & Fazel, S. (2011). Psychosocial influences on prisoners suicide: A case-control study of near lethal self-harm in women prisoners. Social Science & Medicine, 72, 874-883.

McDaniels-Wilson, C., & Belknap, J. (2008). Extensive sexual violation and sexual abuse histories of incarcerated women. Violence Against Women, 14, 1090-1127.

McLellan, A. T., Kushner, H., Metzger, D., Peters, F., Smith, I., Grissom, G., & Argeriou, M. (1992). The fifth edition of the Addiction Severity Index. Journal of Substance Abuse Treatment, 9, 199-213.

Meadows, L. A., & Kaslow, N. J. (2002). Hopelessness as mediator of the link between reports of a history of child maltreatment and suicidality in African American women. Cognitive Therapy and Research, 26, 657-674.

Messina, N., & Grella, C. (2006). Childhood trauma and women’s health outcomes in a Cali-fornia prison. American Journal of Public Health, 96, 1842-1848.

Moscicki, E. K. (1995). Epidemiology of suicidal behavior. Suicide and Life Threatening Behavior, 25, 22-35.

Petersilia, J. (2005). From cell to society: Who is returning home? In J. Travis & C. Visher (Eds.), Prisoner reentry and crime in America (pp. 15-49). New York, NY: Cambridge University Press.

Pettus-Davis, C., Howard, M. O., Roberts-Lewis, A., & Scheyett, A. (2011). Naturally occur-ring social support in interventions for former prisoners with substance use disorders: Conceptual framework and program model. Journal of Criminal Justice, 36, 461-530.

Pew Center on the States. (2008). One in 100: Behind bars in America 2008. Retrieved from http://www.pewcenteronthestates.org/report_detail.aspx?id=35904

Pratt, D., Appleby, L., Piper, M., Webb, R., & Shaw, J. (2010). Suicide in recently released prisoners: A case-control study. Psychological Medicine, 40, 827-835.

Roy, A. (2004). Relationship of childhood trauma to age of first suicide attempt and number of attempts in substance dependent patients. Acta Psychiatrica Scandinavica, 109, 121-122.

Shah, A. (2007). The relationship between suicide rates and age: An analysis of multinational data from the World Health Organization. International Psychogeriatrics, 19, 1141-1152.

Spokas, M., Wenzel, A., Brown, G. K., & Beck, A. T. (2012). Characteristics of individuals who make impulsive suicide attempts. Journal of Affective Disorders, 136, 1121-1125.




Spokas, M., Wenzel, A., Stirman, S. W., Brown, G. K., & Beck, A. T. (2009). Suicide risk factors and mediators between childhood sexual abuse and suicide ideation among male and female suicide attempters. Journal of Traumatic Stress, 22, 467-470.

Swogger, M. T., You, S., Cashman-Brown, S., & Conner, K. R. (2011). Childhood physical abuse, aggression, and suicide attempts among criminal offenders. Psychiatric Research, 185, 363-367.

Tripodi, S. J., Bledsoe, S. E., Kim, J. S., & Bender, K. (2011). Effects of correctional-based programs for female inmates: A systematic review. Research on Social Work Practice, 21, 15-31.

Tripodi, S. J., & Pettus-Davis, C. (2013). Histories of childhood victimization and subsequent mental health problems, substance use, and sexual victimization for a sample of incarcer-ated women in the US. International Journal of Law and Psychiatry, 36, 30-40.

Twomey, H. B., Kaslow, N. J., & Croft, S. (2000). Childhood maltreatment, object relations, and suicidal behavior in women. Psychoanalytic Psychology, 17, 313-335.

Ullman, S. E. (2004). Sexual assault victimization and suicidal behavior in women: A review of the literature. Aggression and Violent Behavior, 9, 331-351.

West, H. S., & Sabol, W. J. (2009). Prison inmates at Midyear 2008: Statistical tables (Report No. NCJ 225619). Retrieved from http://bjs.ojp.usdoj.gov/content/pub/pdf/pim08st.pdf



Female offenders...

Documents