The final study sample consisted of 94 women (63 African Americans and 31 Latinas) who were relatively young (age M=35.1, SD=9.3 years) and had limited economic resources: 68 percent reported little or no income, 60 percent were unemployed, 95 percent were living below the monthly eligibility standards for poverty (USDA, 2009
), and more than half were unemployed or unable to work. In addition to these substantial vulnerabilities, they also reported long-standing stressors, including childhood adversities (M=4.88, SD=2.9) and moderately high chronic burden (M=30.16, SD=10.8). Slightly over half of the women (55%) described their relationship status as “dating”, with only 20% reported living with a partner. Though most of the women in our sample were mothers of at least one child, African American women had significantly more children (M=2.10, SD=1.97) than Latinas (M=1.29, SD=1.66) [F(1,92) = 3.83, p
Statistical analyses comparing the two ethnic groups on all demographic, CSA characteristics, covariates (i.e., relationship to the perpetrator, age of occurrence, duration of the abuse) and outcome variables yielded no significant differences. Thus, ethnic groups were combined in the regression model analyses. Exploratory analyses run separately for each ethnic group were conducted to assess evidence for distinct patterns to be tested in future studies.
Characteristics of First-Incident Child Sexual Abuse
Participants reported between 1 to 5 distinct incidents of CSA (M=1.5, SD=0.9). A small percentage (11 percent) of the sample reported that the first incident lasted less than 1 week, and 33 percent estimated that it lasted about one week. Of the remaining participants (57 percent), the duration of the first CSA incident ranged from 2 weeks to many years (M= 4.2, SD=3.9 years). In most cases, there was one perpetrator (typically a family member). The mean age at the first incident was 9.58 years (SD=3.93), and was more likely to have occurred prior to or during elementary school years (57 percent) than during adolescence (42 percent). Chi-square analysis for peri-trauma variables revealed that only age at the time of the first CSA incident was linked to disclosure, with younger victims more likely to be non-disclosers than those victimized as teens (38 percent vs. 25 percent) (X2(2) =12.47, p=.002).
Frequency distributions and violations of normality resulted in the need to categorize the type of first CSA experience as either severe (digital penetration, attempted rape, rape, oral copulation, and anal sex) or moderately severe (fondling and frottage). A severe first-incident child sexual abuse was reported by 53 percent of the sample.
Depression Symptoms and Child Sexual Abuse Characteristics
Nearly two-thirds of the women (63.8%) had scores indicating the presence of depression (CES-D ≥16), despite being recruited in non-clinical community venues. Mean depressive symptoms and prevalence of clinically significant symptoms (CES-D score >21) as a function of the severity of CSA, disclosure (yes or no), response to disclosure (low unsupportive response or high unsupportive) and self-blame at the time (low or high) are presented in . None of these variables were moderated by the victim’s age at the first incident, the duration of the episode, or by the relationship to the perpetrator. Severe incidents of CSA were associated with significantly greater depression scores than moderately severe CSA [F(1,90)=4.77, p=.03]. The prevalence of clinically meaningful symptom levels was greater among those with severe (60 percent) compared to those with moderately CSA (40 percent) [X2(1)=3.48, p=.05]. More than half of the women (52 percent, N=49) were categorized as non-disclosers. Although ANOVA indicated no differences between the discloser and non-discloser groups on mean depression scores, there was a trend for greater prevalence of clinically significant symptoms among disclosers (60.5 percent) compared to non-disclosers (42.9 percent) [X2(1)=2.44, p=.08.] There were also no differences in depression scores as a function of the level of unsupportive responses among those who disclosed their CSA experiences. Women who reported high self-blame at the time of the first CSA incident had greater rates of clinically significant symptoms (64.7 percent) than those reporting low-self-blame (43.9 percent) [X2(1)=3.71, p=.05]. Also, there was a trend for self-blame to be associated with higher overall levels of depressive symptoms [F(1,89)=3.49, p=.07].
Peri- and Post-Trauma Models
Linear regression analyses were conducted to predict levels of depression, with chronic burden and childhood adversities treated as covariates. CSA severity, disclosure status and self-blame at the time of the incident were entered in invariant order. Primary models included the total sample, while models for each ethnic group were run separately for exploratory purposes. To examine the effects of CSA severity and of disclosure, analyses were conducted in two steps, as shown in , beginning with the most parsimonious model examining CSA severity and disclosure as predictors.
Controlling for the covariates of chronic burden (p<.01) and childhood adversity (p<.01), there was a significant effect for CSA severity (p<.05), but not for disclosure. Adding ethnicity to the model yielded no significant main effect or interactions. However, exploratory examination of the model run separately for each ethnic group revealed two sets of findings. First, chronic burden was significantly associated with depressive symptoms for Latinas (p<.001), but not for African American women. Second, severe CSA was strongly related to depression among Latinas (p<.001), but not among African American women. These findings should be treated with caution because of the relatively small sample sizes.