There is overwhelming evidence that psychopathology is the most significant risk factor for both suicide deaths and suicide attempts among adolescents (Brent et al., 1999; Groholt, Ekeberg, Wichstrom, & Haldorsen, 1997; Shaffer et al., 1996). Psychological autopsy studies have determined that the vast majority of adolescents who die by suicide have significant psychiatric problems, including previous suicidal behavior, depressive disorder, substance abuse, and conduct disorder. The initial onset of such problems often precedes the suicide by several years. Suicide risk among adolescents has also been established to increase with the number of psychiatric diagnoses, with comorbidity between affective disorders and substance abuse being of particular importance (Shaffer et al., 1996). One recent analysis (Gould, 2003) has suggested that eliminating psychopathology could prevent 78%–87% of youth suicides.

The psychiatric problems and gender-specific diagnostic profiles of youth who attempt suicide have been found to be similar to those who die by suicide, with relatively more females than males presenting with affective disorder (Brent et al., 1999; Shaffer et al., 1996) and more males than females having substance abuse, particularly among older male adolescents (Gould et al., 1998; Marttunnen, Avo, Henriksson, & Lönnqvist, 1991; Shaffer et al., 1996). Conduct disorder is also prevalent in young males with suicidal behavior, often comorbid with substance use disorders and anxiety and mood disorders (Brent et al., 1993a; Gould et al., 1998; Shaffer et al., 1996).

Panic disorder has been found to be related to suicidal behavior, particularly among girls (Gould et al., 1996, 1998). Among adults, however, no association has been found when controlling for comorbid depression (Warshaw, Dolan, & Keller, 2000). Other anxiety disorders such

as posttraumatic stress disorder have not been shown to be associated with suicidal behavior among young people when other comorbid psychiatric conditions are taken into consideration (Wunderlich, Bronisch, & Wittchen, 1998). Some studies have reported a relationship between bipolar disorder and both suicide deaths and suicide attempts among youth (Brent et al., 1988; Brent, Perper, & Moritz, 1993; Marttunnen et al., 1991; Shaffer & Hicks, 1994). Although suicide is relatively common among adults with bipolar disorder, given the relative rarity of the disorder, it accounts for only a small proportion of suicide deaths (Appleby, Cooper, Amos, & Faragher, 1999; Vijayakumar & Rajkumar, 1999). Suicidal ideation appears to be less directly related to psychopathology than either suicide attempts or suicide death (Andrews & Lewinsohn, 1992; Reinherz et al., 1995), perhaps because ideation, while occurring with higher frequency, is less persistent and may be fleeting. Gould et al. (1998) found suicidal ideation to be associated with depression in adolescent females and with depression and disruptive disorders in young males.

Aggressive-impulsive behavior (Apter, Plutchik, & van Praag, 1993; Gould et al., 1998; McKeown, et al., 1998; Sourander, Helstela, Haavisto, & Bergroth, 2001) has an increased association with suicidal behavior, particularly in the context of a mood disorder (Brent, Johnson, et al., 1994; Johnson, Brent, Bridge, & Connolly, 1998). Aggressive-impulsive behavior has been found to discriminate suicide attempters from psychiatric controls, and also appears to be related to familial transmission of suicidal behavior.

Hopelessness has also been implicated as an important factor associated with youth suicidal behavior, although its relationship is not independent of depression and depressed mood (Rotheram-Borus & Trautman, 1988; Rotheram-Borus, Trautman, Dopkins, & Shrout, 1990). Among a depressed subgroup that is at high risk for suicide, hopelessness may be an important marker. Pessimism, a negative cognitive style that may be related to hopelessness, has been found to characterize suicide attempters independent of depression (Lewinsohn, Rhode, & Seeley, 1996).

Studies among adults have consistently identified previous suicidal behavior to be the most important factor associated with risk of suicide (Hawton & Sinclair, 2003), with repetition of suicide attempts significantly increasing the risk of a fatal outcome (Sakinovsky, 2000). Studies of youth have likewise found previous suicide attempts to be one of the strongest predictors of both subsequent attempts and suicide deaths (McKeown et al., 1998; Shaffer et al., 1996; Wichstrom, 2000). This relationship is particularly strong among youth with mood disorders (Brent et al., 1999; Shaffer et al., 1996). The risk for future suicidal behavior has been estimated to increase 3–17 times when a previous attempt has occurred (Groholt et al., 1997). One study of a high school sample found that half of adolescent suicide attempters had made more than one attempt (Harkavy-Friedman, Asnis, Boeck, & DiFiore, 1987), and this finding is confirmed by data reported by the YRBS (CDC, n.d.b). Overall, it is estimated that one quarter to one third of adolescents who die by suicide had made a previous attempt (Brent et al., 1993a; Groholt et al., 1997; Shaffer at al., 1996).

While prior attempts figure prominently in the histories of many young persons who die by suicide, the nature of the linkage between earlier and subsequent suicidal behavior is less clear. Existing evidence suggests, however, that previous suicide attempts convey an increased risk for suicide death even after controlling for psychiatric risk factors (Brent et al., 1999; Shaffer et al., 1996).

An additional personal factor that has been suggested to be associated with youth suicide is homosexuality. A number of individual studies (Faulkner & Cranston, 1998; Fergusson, Horwood, & Beautrais, 1999; Garofalo et al., 1998; Remafedi, French, Story, Resnick, & Bloom, 1998; Russell & Joyner, 2001; Wichstrom & Hegna, 2003) as well as a comprehensive review article (McDaniel, Purcell, & D'Augelli, 2001) report increased rates of nonlethal suicidal behavior among youth with same-sex sexual orienta

tion. Wichstrom and Hegna (2003) found that the suicide attempt rate was higher for those with same-sex orientation regardless of whether they had actually had same-sex sexual contact. They found that those who had had same-sex sexual contact had the highest rate. Stigmatization, victimization, isolation, and parental rejection have been identified as factors in suicidal behavior among gay, lesbian, and bisexual youth (McDaniel et al., 2001).

There is no empirical evidence that links suicide deaths among youth to sexual orientation (Shaffer, Fisher, Hicks, Parides, & Gould, 1995). Research in this area is likely to be limited by methodological challenges, notably inaccuracies in the reporting of sexual orientation or sexual behavior (McDaniel et al., 2001; Russell, 2003).

Some studies have linked youth suicidal behavior to parental psychopathology (Brent et al., 1988; Brent, Perper, et al., 1994; Gould et al., 1996), although others have found rates of family psychopathology among young suicide attempters and completers to be similar to those of other clinical samples (see Wagner, 1997; Wagner et al., 2003). Both suicide and suicide attempts have been found to be increased in families in which a parent has died by suicide or attempted suicide, even when controlling for the impact of parental psychopathology (Brent, Bridge, Johnson, & Connolly, 1996; Brent et al., 2002; Glowinski et al., 2001). This relationship may be mediated by familial transmission of impulsive aggression (Brent et al., 2003).

The mechanisms through which psychopathology and suicidality among parents may influence youth suicidal behavior are not yet clear. Although little is currently known about the genetics of youth suicide, adult studies suggest that biological factors play a significant role in suicide. Neurobiological abnormalities, in particular lower levels of central nervous system serotonin (5-HT), have been implicated in aggressive impulsivity and suicidal behavior in adults (Oquendo & Mann, 2000). Postmortem studies of the brains of adult suicide victims have also shown higher levels of 5-HT receptors than in normal controls (Arango et al., 1990; Mann, Stanley, McBride, & McEwen, 1986).

Postmortem studies of youth who have died by suicide are rare, and therefore the implications of adult findings for understanding youth suicide are not clear. One postmortem study involving 15 teenage suicide victims and 15 normal matched control subjects found significantly higher levels of 5-HT receptor expression in the prefrontal cortex and hippocampus of those who had died by suicide, suggesting that this abnormality may be a marker of adolescent as well as adult suicide (Pandey et al., 2002). These authors noted that higher levels of serotonin receptor expression have been implicated in alterations in emotion, stress, and cognition, which suggests promising avenues of exploration for understanding the neurobiology of youth suicide. Further studies are needed to confirm these findings and to clarify genetic relationships.

The familial expression of suicidality is likely a function of example as well as biology. Certain factors related to the family environment such as lack of family support and parent–child conflict have been found to contribute to risk for youth suicidal behavior (see Wagner, 1997; Wagner et al., 2003 for reviews). Particularly among younger suicide victims, parent–child conflict has been identified as a common precipitant to suicidal behavior (Brent et al., 1999; (Groholt, Ekeberg, Wichstrom, & Haldorsen, 1998).

There is considerable evidence that various negative life events contribute independently to youth suicide (see Wagner, 1997; Wagner et al., 2003), over and above other risk factors (Gould et al., 1996; Johnson, Cohen, et al., 2002). Physical abuse, for example, has been demonstrated to increase risk in case–control (Brent, Johnson, et al., 1994; Brent et al., 1999) and longitudinal studies (Brown, Cohen, Johnson, & Smailes, 1999; Johnson et al., 2002; Silverman, Reinherz, & Giaconia, 1996), and has been associated with