Parental Sexual Abuse and Suicidal Behaviour Among Women With Major Depressive Disorder

Abstract

Objective

Women with major depressive disorder (MDD) and childhood sexual abuse histories have an increased risk for suicidal behaviours, but it is unclear whether specific abuse characteristics contribute to risk. We aimed to examine the contributions of abuse characteristics to lifetime history of suicide attempts and multiple suicide attempts, independent of posttraumatic stress disorder and borderline personality disorder.

Method

Women with MDD and sexual abuse histories (n = 106) were assessed regarding sexual abuse characteristics, psychiatric diagnoses, and suicide attempts.

Results

In multivariate logistic regressions, the odds of having multiple suicide attempts increased 12.27-fold when childhood sexual abuse was perpetrated by a parent figure or a parent, compared with a nonparent.

Conclusions

Parental perpetration of sexual abuse increases the likelihood of multiple suicide attempts among women outpatients. The relationship of the perpetrator to the abused woman is important in suicide risk evaluation and treatment planning.

Women with childhood sexual abuse histories have an elevated risk for suicide attempts^1–6 and comprise a large segment of the outpatient psychiatric population.^7–9 Recognizing that risk for suicidal behaviour is multi-determined, the extent to which it can be ascribed to specific characteristics of childhood sexual abuse, independent of established diagnostic correlates such as PTSD and BPD,^10–16 is unclear. A few studies in clinical samples have not found that sexual abuse was associated with suicide attempt histories after controlling for PTSD and personality disorders,^11,17,18 but none examined specific characteristics of the abuse in relation to suicide attempts.

In our study, we examined 4 abuse characteristics known to be associated with suicide attempt histories: abuse involving intercourse,^10,19–21 a parent figure or parent perpetrator,^19,22 abuse involving force,^19,20 and age of onset.¹⁹ People with histories of single compared with multiple attempts were distinguished, as the latter, have a higher risk for future suicidal behaviour.^23–25 We expected that abuse involving intercourse, a parent figure or parent perpetrator, or force would be associated with suicide attempts. To our knowledge, this is the first study among women with MDD and sexual abuse histories to examine the relative contributions of abuse characteristics to suicide attempt histories, independent of current PTSD, BPD, and other covariates.

Method

Data were obtained from prior²⁶ and ongoing trials of a psychosocial treatment for women with reported childhood sexual abuse histories and MDD. The study protocol was approved by the Institutional Review Board of the University of Rochester Medical Center prior to implementing the study, and was reviewed and re-approved annually. Only pretreatment baseline data are reported.

Subjects

The sample included 106 women with childhood sexual abuse histories and current MDD in a community mental health centre. The mean age of participants was 35 years (SD 9.81, range 19 to 57); 53 (50.0%) were African American, 53 (50.0%) were White, and 12 (11.3%) identified an Hispanic ethnic origin. Fifty-five women (51.9%) had children aged 18 years and younger; 41 of these mothers were living without a spouse or partner. Sixty-five women (61.3%) were unemployed and 31 (29.2%) were employed full-time. Thirty-eight (35.8%) women had some college education, 50 (47.2%) had a high school diploma or equivalent, and 18 (17.0%) had less formal education. Sixty women (56.6%) had public assistance support as their major source of income, which included social services, disability benefits, or unemployment benefits; the other 46 women (43.4%) had private sources of income.

Data Collection

Over a 42-month period from 2003 to 2006, 1080 women seeking treatment in a community mental health centre were screened for potential study eligibility during intake interviews. Eligibility criteria for referral to the study were depressive symptoms and a self-reported history of sexual abuse before the age of 18 years. A clinical research coordinator, certified in human subjects protection, conducted the informed consent process with every woman at the beginning of the pretreatment assessment. The participant read the consent form, or had it read to her by the research coordinator, and was asked questions to determine her understanding of the consent. Women provided consent to participate by signing the consent form and were provided a signed copy to retain.

Inclusion criteria were current MDD, established through the SCID-I,²⁷ and a sexual abuse history, established through a structured clinical interview.²⁸ Consensus diagnostic conference meetings were conducted by a clinical psychologist (study principal investigator), a psychiatrist (study co-investigator), and the research staff who conducted the research interviews. Consensus multiaxial DSM-IV diagnoses were made based on SCID interview data and record review. Sexual abuse was defined as: any unwanted sexual contact; or any sexual contact with a family member aged 5 or more years older than the patient that occurred prior to the patient being aged 18 years. Sexual contact was defined as physical contact of a sexual nature, ranging from fondling to sexual intercourse. All participants reported non-consensual sexual contact before they were aged 18 years by at least 1 perpetrator. Exclusion criteria were psychosis, schizophrenia, bipolar disorder, mental retardation, and active substance abuse or dependence.

Among the 163 women screened as potentially eligible, 133 agreed to participate in a baseline evaluation designed to establish eligibility for the trial. Twenty-three women did not meet inclusion criteria and 4 did not complete the baseline evaluation, leaving 106 participants who met criteria for MDD and a history of childhood sexual abuse.

Materials

Childhood sexual abuse severity was characterized using the CTQ.²⁹ The CTQ is a self-report instrument that instructs participants to rate the frequency with which events occurred when they were growing up, using a 5-point scale ranging from never true (1) to very often true (5). The abuse severity cut-point guidelines recommended by Bernstein et al²⁹ were used: none (score of 5 or less), low (score of 6 to 7), moderate (score of 8 to 12), and severe (score greater than 12). Most of the women (91.5%: 97/106) had experienced severe to extreme sexual abuse.

Childhood sexual abuse characteristics were assessed using a structured clinical interview.²⁸ Directions to this interview state that questions concern

non-consensual sexual contact you may have experienced in childhood or adolescence. By non-consensual contact, we mean: You did not willingly agree to the contact; you were forced.

The relationship of the perpetrator(s) is queried with:

Before you were 18 years old, did any of the following people behave in a sexual ways towards you?

which is followed by a checklist of possible relationships ranging from familial, to friend or acquaintance, to person in a position of authority, to stranger. Questions about the age of onset of the sexual abuse, whether the abuse involved intercourse, and whether the abuse involved physical force followed. Variables of interest were coded dichotomously (1 compared with 0): abuse onset before the age of 6 years, compared with abuse onset after the age of 6 years; abuse involving force, compared with no force; a parent figure or parent perpetrator, compared with another perpetrator; and abuse involving intercourse, compared with no intercourse. The parent figure or parent perpetrator category included parents, step-parents, and a mother’s intimate partner(s). Physical force includes a range of acts, from holding the person down to physical violence.

A suicide attempt history was measured using the Lifetime Suicide Attempt Self-Injury Interview.³⁰ This measure provides a count and categorization of suicide attempts. Suicide attempts were defined as self-harming behaviours with intent to die or ambivalence about intent to die. Suicide attempt status was coded dichotomously: presence of at least one suicide attempt in the past (ever-attempters) compared with the absence of lifetime suicide attempts (non-attempters). Multiple suicide attempter status was coded in a binary format: a history of 2 or more lifetime suicide attempts (multiple-attempters) compared with a history of 1 suicide attempt (single-attempters). The woman’s age at the time of the first lifetime suicide attempt was recorded.

Current PTSD was evaluated with the SCID-I Axis I disorders.²⁷ PTSD diagnosis was coded in a dichotomous format: as present (1) when the criteria for current PTSD or PTSD in partial remission were met, and as absent (0) when the criteria for these diagnoses were not met. The SCID-II personality disorders, BPD module,³¹ was used to assess BPD. The SCID-II is a semi-structured interview developed for the assessment of DSM-IV personality disorders; in our study, only the BPD module was used. As with PTSD, a dichotomous coding system (present, compared with absent) was used for BPD.

Several covariates were entered into the multivariate analyses. A history of alcohol or substance dependence^10,32,33 and chronic depression³⁴ were controlled because of their documented associations with suicidal behaviour.³⁵ Both diagnoses were established by consensus based on the SCID-I²⁷ and were coded dichotomously. Chronic depression was defined as: current major depressive episode of at least 2 years’ duration, current MDD superimposed on a pre-existing dysthymic disorder, or recurrent MDD with incomplete remission between episodes over a minimum of a 2-year duration prior to treatment.³⁶

Because we wanted to determine whether the remote event of sexual abuse was predictive over and above abuse experiences in adulthood,^19,37,38 sexual assault and IPV in adults aged 18 years and older were assessed via single questions on the TLEQ³⁹ and coded dichotomously (present, compared with absent). Adult exposure to IPV and sexual assault were assessed by 2 items on the TLEQ. Adulthood sexual trauma was assessed by the question, “After your 18th birthday: Did anyone touch sexual parts of your body or make you touch sexual parts of their body against your will or without consent?” Women were classified as having experienced adult IPV according to their response to the question, “Have you ever been slapped, punched, kicked, beaten up, or otherwise physically hurt by your spouse (or former spouse), a boyfriend/girlfriend, or some other intimate partner?” The TLEQ has been shown to have good convergent validity with other traumatic life events interviews³⁹; for women, the percentage of occurrence and nonoccurrence agreement of traumatic life events was 91%. Test–retest reliability with various populations was also shown, with overall agreement rates for adult sexual assault and IPV ranging from 79% to 83%.

Statistical Methods

Independent variables were 4 childhood sexual abuse characteristics: intercourse, a parent figure perpetrator, physical force, and onset before the age of 6 years. The 2 dichotomous dependent variables were lifetime history of: suicide attempt (no attempt, compared with 1 or more attempts) and multiple suicide attempts (1 attempt, compared with more than 1 attempt). Chi-square analyses were conducted to examine associations between independent and dependent variables, as well as independent variables and covariates. In 2 multivariate analyses, logistic regression was used to examine independent associations between abuse characteristics and the 2 dependent variables, controlling for demographics (age and race), current PTSD, BPD, adult sexual assault, adult IPV, history of alcohol and (or) substance dependence, and chronic depression. Models were evaluated using the Hosmer-Lemeshow goodness-of-fit test.⁴⁰ The SPSS Version 15 (SSPS Inc, Chicago, IL) for Windows was used for statistical analyses.

Results

Sample Characteristics

Sixty-eight women (64.2%) had a diagnosis of PTSD and 36 (34.0%) had a diagnosis of BPD. Fifty-nine women (55.7%) reported at least 1 lifetime suicide attempt (ever-attempters), and 29 (27.4%) reported multiple suicide attempts. The overwhelming majority of first lifetime suicide attempts were preceded by childhood sexual abuse (57 out of 59 women; 96.6%). The most common suicide attempt method used was intentional overdose.

Regarding sexual abuse characteristics, 97 women (91.5%) reported severe sexual abuse on the CTQ. Abuse involved sexual intercourse for 78 women (73.6%) and physical force for 73 (68.9%) women. Thirty-two women (30.2%) reported sexual abuse by a parent or a parent figure. Forty-nine women (46.2%) reported sexual abuse before the age of 6 years.

Univariate Analyses

Women with a history of suicide attempts (ever-attempters) were more likely to have a current PTSD diagnosis (44 of 59 ever-attempters; 74.6%) than those without a history of suicide attempts (24 of 47 non-attempters; 51.1%), (χ² = 6.29, df = 1, P = 0.01) (Table 1). Women who had attempted suicide multiple times (17 of 29; 58.6%) were more likely to be diagnosed with BPD, compared with those who had made 1 suicide attempt (7 of 30; 23.3%), χ² = 7.61, df = 1, P = 0.006). A greater proportion of the former had a history of childhood sexual abuse with a parent figure or parent perpetrator (12 of 29; 41.4%, compared with 5 of 30; 16.7%), (χ² = 4.39, df = 1, P = 0.04).

Table 1.

Univariate analyses: differences between groups of ever-attempters (ever-A), compared with non-attempters (non-A), and single-attempters (single-A), compared with multiple-attempters (multiple-A) in terms of childhood sexual abuse, severity parameters, and diagnostic covariates

	Model 1, n (%)				Model 2, n (%)
Variable	Ever-A (n = 59)	Non-A (n = 47)	χ2a	P	Single-A (n = 30)	Multiple-A (n = 29)	χ2a	P
Childhood sexual abuse severity parameters
Intercourse	43 (72.9)	35 (74.5)	0.03	0.86	22 (73.3)	21 (72.4)	0.01	0.94
Parent figure	15 (25.4)	10 (21.3)	0.25	0.62	4 (13.3)	11 (37.9)	4.71	0.03
Force used	32 (54.2)	22 (46.8)	0.58	0.45	12 (40)	15 (51.7)	0.82	0.37
Onset before age 6 years	22 (46.8)	24 (40.7)	0.40	0.53	9 (30)	15 (51.7)	2.88	0.09
Diagnostic covariates
BPD	24 (40.7)	12 (25.5)	2.68	0.10	7 (23.3)	17 (58.6)	7.61	0.006
PTSD	44 (74.6)	24 (51.1)	6.29	0.01	20 (66.7)	24 (82.8)	2.01	0.16

^adf = 1

Multivariate Analyses

Among the hypothesized childhood sexual abuse predictors, abuse perpetrated by a parent figure or parent increased the odds of having made multiple suicide attempts 12.27-fold (P = 0.01) (Table 2). Abuse involving force or intercourse did not increase the odds of either multiple or lifetime suicide attempts.

Table 2.

Multivariate analyses: separate logistic regression analyses of variables predicting suicide attempter status: model 1 ever-attempter (ever-A), compared with non-attempter (non-A) status, and model 2 multiple-attempter (multiple-A), compared with single-attempter status (single-A), controlling for covariates^a

	Model 1 Ever-A compared with non-A				Model 2 Multiple-A compared with single-A
Variable	B	Wald (df =1)	OR (CI 95%)	P	B	Wald (df =1)	OR (CI 95%)	P
Childhood sexual abuse
Intercourse	−0.45	0.65	0.64 (0.21–1.92)	0.42	−0.55	0.45	0.58 (0.12–2.85)	0.50
Parent figure	−0.66	1.41	0.52 (0.18–1.53)	0.24	2.51	6.20	12.27 (1.71–88.33)	0.01
Physical force	0.26	0.23	1.29 (0.45–3.68)	0.63	0.02	<0.001	1.02 (0.19–5.36)	0.98
Onset before age 6 years	−0.75	2.49	0.47 (0.19–1.20)	0.11	0.86	1.40	2.36 (0.57–9.74)	0.24

^aThe covariates included age, race or ethnic origin (White as the reference group), history of drug or alcohol dependence, chronic depression, adulthood sexual assault, and IPV.

Model 1. According to the Hosmer-Lemeshow²⁹ goodness-of-fit test, the overall model fit was acceptable (n = 106; χ² = 6.21, df = 8, P = 0.62). Overall classification was 70.8%; correction classification rates were 63.8% for no history of suicide attempts and 76.3% for ever-attempter status.

Model 2. Overall model fit was acceptable (8, n = 106; χ² = 18.65, df = 8, P = 0.02), and overall classification was 83.1%; 86.7% for single-attempter status and 79.3% for multiple-attempter status

One covariate was associated with increased odds of multiple attempts: younger women were more likely to have made multiple attempts (OR 0.92; 95 % CI 0.85 to 1.00; P = 0.05). Three covariates were associated with ever-attempter status: PTSD (OR 3.61; 95% CI 1.22 to 10.70; P = 0.02); chronic depression (OR 2.64; 95% CI 0.996 to 6.99, P = 0.05); and White racial status (OR 2.97; 95% CI 1.17 to 7.55; P = 0.02).

Discussion

In this sample of women with MDD and sexual abuse histories, having a parent or parent figure as the perpetrator was associated with dramatically increased odds of multiple suicide attempts. This finding is consistent with a prior study,²² which demonstrated a relation between the kinship to the perpetrator and the number of suicide attempts in a community sample. Our study, in contrast, focuses on a clinical sample of women at high risk for suicidal behaviour and seeks to inform the clinical assessment of suicide risk, apart from psychiatric diagnoses of BPD and PTSD. We have demonstrated that perpetrator status has a strong and independent relation to suicide attempt histories. PTSD was associated with having ever attempted suicide, but not with multiple attempts. Although we observed an association between BPD and multiple-attempter status in univariate analyses, that relation is confounded by one or more variables in the multivariate analysis. It may also be that certain features of BPD, such as impulsivity,⁴¹ rather than the diagnosis per se, are related to propensity to multiple suicidal behaviour attempts.

According to our findings, women with MDD who have been sexually abused by a parent or parent figure may be at a higher risk for multiple suicide attempts, and close monitoring for suicide risk may be warranted. Prior studies have indicated a higher risk for future suicidal behaviour with increasing numbers of suicide attempts.^23–25 Sexual assault by a parent or parent figure is typically accompanied by other family adversities, including parental conflict²¹ and parental psychopathology.¹⁰ Joiner et al²⁰ has described a person’s suicide attempts as a response to perceived burdensomeness, thwarted belonging, and acquired capacity for self-harm. The pain, shame, and alienation of the abused child growing up in such profoundly troubled families may be the bellwether for interpersonal risk factors in adulthood.

Several limitations in these data should be acknowledged. First, the study’s cross-sectional design precludes causal conclusions. Longitudinal studies are warranted to address hypotheses regarding causal pathways. Second, findings may not be generalizable to men and to women with sexual abuse histories not seeking mental health treatment,⁴² and to those diagnosed with psychosis, schizophrenia, bipolar disorder, mental retardation, and active substance abuse or dependence. Although our sample is limited to women with MDD and childhood sexual abuse histories, our research⁴³ has found that this group represents about 15% to 20% of treatment-seeking women in community mental health care. Third, biases in retrospective reporting of abuse and suicide attempt histories are possible. However, empirical evidence supports the reliability and validity of retrospective reports of suicide attempts.³⁰ A recent large-scale study⁴⁴ found that a self-report measure of childhood sexual abuse had long-term stability, construct validity, and evidence of familial corroboration. Reviews of the validity of retrospective reporting of childhood abuse have concluded that false positives are rare, whereas false negative are not rare.^45,46 In our study, false negatives (that is, denying sexual abuse by a parent when it did occur) would mitigate against finding between-group (parent abuse, compared with no parent abuse) differences. Further, studies have shown that the presence of psychiatric diagnoses, including depression, do not affect the reliability of childhood abuse reports.^5,47 Fourth, we acknowledge that the age of abuse onset dichotomization (that is, onset before the age of 6 years, compared with abuse onset after the age of 6 years) can be construed as arbitrary, yet some type of categorization is necessary for scientific communication, and this one had been used effectively in prior research.¹⁹ Fifth, associations between sexual abuse characteristics other than parent figure or parent perpetrator and suicide attempt history may be attenuated by the sample size. Finally, adult exposure to IPV or sexual assault was assessed through single-item questions from the TLEQ.³⁹ Other forms of childhood abuse or trauma, which are frequently co-occurring with sexual abuse, could have an effect on suicidal behaviours, as has been shown in several studies.⁴⁸ Data on PTSD chronicity, which could be associated with multiple-attempter status, were not available.

Conclusions

Our findings suggest that the clinical assessment of suicide risk among women with MDD and histories of sexual abuse could be informed by the identification of particular sexual abuse characteristics. People who attempt suicide more than once are more likely to die by suicide eventually,^25,49 and typically have poor problem solving skills, difficulty regulating emotions, and more conflictual relationships.²³ More specifically, current findings suggest that women abused by a parent or parent figure could require ongoing monitoring for suicide risk, intensive safety planning, and, possibly, developmentally informed treatment. We recognize that legitimate concerns can be raised in response to advising clinicians and researchers to inquire about specifics of childhood trauma histories,⁵⁰ including the potential destabilizing effects of traumatic memories and the time constraints of brief psychotherapy. Nevertheless, our findings support the view that identifying specific characteristics of abuse could be critical in determining the level of risk for repetitive attempts and suicide.²⁰

Clinical Implications.

• Parental perpetration of sexual abuse is strongly associated with multiple suicide attempts among women with MDD.

• The effect of a parent figure or parent perpetrator on suicide attempt status was independent of borderline personality disorder and posttraumatic stress disorder.

• Findings suggest the importance of attending to the features of sexual abuse in suicide risk evaluation and treatment planning.

Limitations

• Our study used a cross-sectional design and a retrospective reporting of life events.

• Definitive, causal conclusions regarding the effects of sexual abuse on suicidal behaviour would require longitudinal data.

• The implications of the findings may be limited to clinical populations of women with MDD and childhood sexual abuse histories.

Abbreviations

BPD

borderline personality disorder

CTQ

Childhood Trauma Questionnaire

DSM

Diagnostic and Statistical Manual of Mental Disorders

IPV

intimate partner violence

MDD

major depressive disorder

PTSD

posttraumatic stress disorder

SCID

Structured Clinical Interview for DSM-IV

TLEQ

Traumatic Life Events Questionnaire