attributable risk (PAR). The PAR for child sexual abuse was 9–20 percent of suicide attempts (Brown et al., 1999; Fergusson et al., 1996; Molnar et al., 2001a). This means that, independent of psychopathology and other known risk factors, child sexual abuse accounts for 9–20 percent of suicide attempts in adults. The study by Molnar and colleagues (2001a), from the National Comorbidity Survey, is especially significant because it is nationally representative of the U.S. population. Their analysis of serious suicide attempts revealed a PAR from child sexual abuse of 9–12 percent, and a PAR from mental disorders of 70–80 percent. The latter figure means that 70–80 percent of suicide attempts are associated with mental disorders.

On the basis of their findings, Molnar and colleagues suggested that a substantial proportion of suicide risk is missed by sole reliance on the presence of psychopathology. This point is discussed again later in the chapter.



Until recently, gender effects were not found to modify child sexual abuse as a risk factor for suicide. A meta-analysis, covering studies from 1981–1995, did not find a gender effect in the impact of child sexual abuse on suicidality (Paolucci et al., 2001). Nevertheless, two more recent and large population-based studies (Edgardh and Ormstad, 2000; Molnar et al., 2001a) found that child sexual abuse placed males at greater risk for suicide attempts. These two studies were consistent in finding odds that were 4–11 times higher among males and 2–4 times higher among females. While another large, population-based study did not find a gender effect (Dinwiddie et al., 2000), the odds ratios for both males and females hovered around 7, a figure in the mid-range of the other population based studies. It also found the population prevalence of child sexual abuse to be somewhat lower than other studies.8 Thus, the newer body of evidence suggests that once sexual abuse occurs, males appear to be at higher risk of suicide attempts, but findings are not uniform.

For other types of childhood trauma, gender effects have not been reported in large, population-based studies. In a clinical sample of alcoholic inpatients (n=802), physical abuse displayed a gender effect: suicide attempts were significantly associated with physical abuse in men, but not in women (Windle et al., 1995).


In general, the prevalence of child sexual abuse is higher for females (12-17 percent) than for males (5-8 percent) (Gorey and Leslie, 1997; Molnar et al., 2001b).

The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement