The relation of child sexual abuse and depression in young women: comparisons across four ethnic groups

There is a growing controversy over whether child (under age 18) sexual abuse (CSA) is related to adult adjustment. Numerous studies have reported that CSA is related to adult mental health problems including depression (Burnam et al., 1988; Ernst, Angst, & Foldenyi, 1993; Fromuth, 1986; Gidycz & Koss, 1989; Mullen, Martin, Anderson, Romans, & Herbison, 1993; Roesler & McKenzie, 1994; Saunders, Villeponteaux, Lipovsky, Kilpatrick, & Veronen, 1992; Stein, Golding, Siegel, Burnam, & Sorenson, 1988; Yama, Tovey, & Fogas, 1993). A review of the long-term effects of CSA found that six out of eight studies reported a relationship with depression (Beitchman et al., 1992).

Despite many reports of a relation between CSA and adult adjustment, there is accumulating evidence to contradict these claims. In a review of seven studies using national probability samples, Rind and Tromovitch (1997) argued that CSA either is not related to adjustment or that the relation is confounded by background factors associated with both CSA and adjustment and, furthermore, that any relation between CSA and adjustment has been exaggerated by studies of clinical samples. Thus, some studies have found that CSA accounts for little or no variance in adjustment when background characteristics are controlled (e.g., Finkelhor, Hotaling, Lewis, & Smith, 1990; Higgins & McCabe, 1994). On the other hand, other studies have found that CSA accounts for a significant amount of variance in adjustment despite controlling for background characteristics (e.g., Conte & Schuerman, 1987; Mullen et al., 1993; Roesler & McKenzie, 1994; Stein et al., 1988; Yama et al., 1993).

Differences in how CSA is defined may account for many of the conflicting reports of the relation of CSA and adjustment. Studies that define CSA based solely on the "partners'" relative ages, without regard for whether a sexual experience was wanted, account for most of the cases in which no relation was found between CSA and adjustment (e.g., Finkelhor et al., 1990; Higgins & McCabe, 1994; Rind & Tromovitch, 1997). In addition, sampling may account for some of the conflicting reports of the relation between CSA and adjustment. With few exceptions (e.g., Russell, Shurman, & Trocki, 1988; Stein et al., 1988; Wyatt, 1990, 1992), research on CSA has been limited to samples of non-Hispanic white women, with most samples coming from colleges or clinics. Cultural differences among ethnic groups are likely to play significant roles in the meaning of CSA to victims and their families and in how women and their families adapt after CSA (e.g., Fontes, 1993, 1995; Wyatt, 1990). Few studies of the relation of CSA to adjustment have considered whether this relation might differ by ethnicity. The purpose of this study was to examine the relation of CSA and depression in young women and to compare this relation for African American, Mexican American, Native American, and non-Hispanic white women.

Few studies of CSA have controlled for family or background factors associated with adult adjustment despite evidence that many risk factors for depression also may be risk factors for CSA (e.g., Alexander & Lupfer, 1987; Ernst et al., 1993; Finkelhor et al., 1990; Mullen et al., 1993; Nash, Hulsey, Sexton, Harralson, & Lambert, 1993). That is, factors such as family disruption, dysfunction, or violence may increase the likelihood both of CSA and depression. Thus, some or all of the often reported relation of CSA and depression could be spurious due to uncontrolled common background factors. Although some studies that controlled for background factors reported no relation between CSA and adjustment, many others have found a relationship even after controlling for confounding factors (e.g., Bifulco, Brown, & Adler, 1991; Boney- McCoy & Finkelhor, 1995; Burnam et al., 1988; Ernst et al., 1993; Fromuth, 1986; Gidycz & Koss, 1989; Mullen et al., 1993; Nash et al., 1993). Controlling for potentially confounding background factors does not provide a test of a causal relation between CSA and adjustment; only well-designed longitudinal studies can make this determination. Instead, controlling for background factors provides a conservative test (i.e., probable underestimate) of whether CSA is related to adjustment (Briere, 1988) and, therefore, deserving of continued attention by researchers and clinicians. However, because of the well-known interrelations among background factors, CSA, and adjustment, failing to control for background factors will lead to overestimates of the relation of CSA and adjustment (e.g., Beitchman, Zucher, Hood, Da Costa, & Akman, 1991; Rind & Tromovitch, 1997).

Finally, most studies of CSA have used convenience samples, usually of white, middle- class college students (e.g., Alexander & Lupfer, 1987; Gidycz & Koss, 1989; Jackson, Calhoun, Amick, Maddever, & Habif, 1990; Yama et al., 1993) or women in treatment for mental health or sexual problems (e.g., Briere & Runtz, 1986; van Gijseghem & Gauthier, 1994). Studies of high-functioning populations such as college students should contribute to underestimates of the relation of CSA and adjustment while studies of clinical populations should lead to overestimates of this relation (e.g., Briere & Runtz, 1986; Rind & Tromovitch, 1997). Few studies have examined the role that ethnicity might play in either a woman's risk for experiencing CSA or in how CSA is related to adjustment. In one of these, Wyatt (1985) reported a higher incidence of CSA for non-Hispanic white women than African American women. Another study (Siegel, Sorenson, Golding, Burnam, & Stein, 1987) reported a higher incidence for non-Hispanic whites than for Hispanics. Russell et al. (1988) reported that African American women had greater negative effects from CSA than non-Hispanic whites. Stein et al. (1988) reported a stronger relation of CSA to adjustment for Hispanic women than for non-Hispanic whites. Of the studies that included minority women in sufficient numbers for making comparisons, only Stein et al. controlled for background factors when making these comparisons.