تداوم، همبودی و طولی ارتباط بین افسردگی و رفتار ضد اجتماعی در دوران نوجوانی متوسط: یک مطالعه آینده نگر با پیگیری 2 سال گذشته

Abstract The study investigated continuity, comorbidity and longitudinal associations between depression Beck depression inventory (RBDI) and antisocial behaviour Youth self-report (YSR) in middle adolescence. Data were used from a community sample of 2070 adolescents who participated in a 2-year prospective follow-up study. The results indicate that both depression and antisocial behaviour had considerable continuity, and concurrent comorbidity between these disorders was strong. In contrast to several previous studies, antisocial behaviour did not predict subsequent depression, but conversely, depression predicted subsequent antisocial behaviour among girls. Among boys history of depression seemed to protect from subsequent antisocial behaviour. Gender differences in longitudinal associations are discussed.

Introduction Previous studies have established that both depression and antisocial behaviour in adolescence have considerable continuity (Fergusson, Lynskey, & Hoorwood, 1996; Kim et al., 2003; Wiesner, 2003). In a longitudinal study by Beyers and Loeber (2003) 6–35% of the variance of current depression was explained by prior depression, the corresponding variance predicting current delinquency was 25–34%. Similarly, Costello, Mustillo, Erkanli, Keeler, and Angold (2003) found that adolescents with a history of depression were four times more likely than those with no previous depression to have depression in the future. The corresponding odds ratio (OR) for antisocial behaviour among those with a history of antisocial behaviour was about 10. Girls showed more continuity of depression than boys, but no gender differences were found in the continuity of antisocial behaviour. In contrast, Ferdinand, Verhulst, and Wiznitzer (1995) found that continuity of both depression and antisocial behaviour did not differ by sex. According to Angold et al. (1999) and Costello et al. (2003) the stability described above represents homotypic continuity in which manifestations of the disorder in question change relatively little over time so that diagnosis remains the same at different assessments. High rates of concurrent comorbidity of adolescent depression and antisocial behaviour have been well established in both clinical (Biederman, Faraone, Mick, & Lelon, 1995; Goodyer, Herbert, Secher, & Pearson, 1997; Grilo et al., 1996) and population-based samples (Beyers & Loeber, 2003; Fergusson & Woodward, 2002; Loeber, Farrington, Stouthamer-Loeber, & Van Kammen, 1998; Ritakallio, Kaltiala-Heino, Kivivuori, & Rimpelä, 2005; Vermeiren, Deboutte, Ruchkin, & Schwab-Stone, 2002). In a meta-analysis of community studies Angold et al. (1999) found that after controlling for other comorbidities, conduct disorder was about seven times more common in depressed than in non-depressed adolescents. Previous research on gender differences in comorbidity, however, is inconsistent as some studies have found that depression is more prevalent among antisocial girls than boys (Costello et al., 2003; Flannery et al., 2001) while others have found the opposite (Maughan, Rowe, Messer, Goodman, & Meltzer, 2004). In studying developmental associations between depression and antisocial behaviour the concept of heterotypic continuity is pivotal. Heterotypic continuity refers to a continuous process in which one disorder generates manifestations of different forms over time or in which one disorder exposes adolescents to different disorders at different ages (Angold et al., 1999; Costello et al., 2003). There are at least three possible developmental models (acting out, failure and mutual influence model) to explain the heterotypic continuity between adolescent depression and antisocial behaviour. The models differ in whether they assume associations to be causal (one disorder creates an increased risk for the other) or non-causal (association based on non-specific risk factors) or disorders are reciprocally associated with each other (Caron & Rutter, 1991; Fergusson et al., 1996; Overbeek, Vollebergh, Meeus, Engels, & Luijpers, 2001; Wiesner, 2003). The acting out model proposes that depressed adolescents act out internalising problems and depression is masked out by antisocial behaviour. This model assumes that depression precedes antisocial behaviour and the cross-lagged associations are unidirectional and causal ( Capaldi, 1992; Overbeek et al., 2001). The failure model assumes that there is a causal, unidirectional association between antisocial behaviour and subsequent depression. Antisocial adolescents face many problems in social relations (e.g. rejection, lack of support, conflict with parents and teachers), have low social competence and coping skills and problems in academic attainment which in turn lead to failure experiences which increases vulnerability for depression ( Capaldi, 1992; Kiesner, 2002). According to the mutual influence model both depression and antisocial behaviour result from non-specific (shared or overlapping) risk factors, and disorders are also reciprocally reinforcing each other over time. This model assumes that the onset of one disorder increases vulnerability to the other and vice versa. Therefore, cross-lagged and bidirectional associations between disorders are possible ( Overbeek et al., 2001). As the current study investigated prospective associations between depression and antisocial behaviour we focused on acting out and failure models. There are surprisingly few studies of the longitudinal associations between depression and antisocial behaviour in adolescence, and the results appear inconsistent ( Loeber et al., 2000). It should, however, be noted that due to differences in methodology (e.g. age range of the sample, sex ratio of the sample and period of follow-up), previous studies are difficult to compare. Most previous studies have focused on the transition from (late) childhood to early adolescence (e.g. Kiesner, 2002) or they have followed children through adolescence into adulthood (e.g. Roza, Hofstra, van der Ende, & Verhulst, 2003). Many studies conducted in community samples propose that antisocial behaviour precedes the onset of depression and that the effect is unidirectional (Capaldi, 1992; Feehan, McGee, & Williams, 1993; Fergusson et al., 2003; Kiesner, 2002; Rohde, Lewinsohn, & Seeley, 1991). Different results from these have been reported by Loeber, Russo, Stouthamer-Loeber, and Lahey (1994) who found that depression in early adolescent boys was associated with escalation to more serious delinquency and a variety of delinquent behaviour, but Vermeiren, Schwab-Stone, Ruchkin, De Clippele, and Deboutte (2002) found depression to be protective against recidivism. Harrington, Fudge, Rutter, Pickles, and Hill (1991) found that rates of adult depression were lower in those who as adolescents had comorbid antisocial behaviour and depression than in those who had had pure antisocial behaviour or pure depression in adolescence. Furthermore, some studies found no heterotypic continuity between these disorders (Costello et al., 2003; Lewinsohn et al., 1994; Overbeek et al., 2001; Roza et al., 2003) while others suggest that depression and antisocial behaviour affect each other in a bidirectional way (Beyers and Loeber, 2003). Studies on gender differences among the longitudinal associations between depression and antisocial behaviour in adolescence are relatively rare. The study by Wiesner (2003) found limited evidence that antisocial behaviour among boys preceded depression but not vice versa. Among girls a circular process was found as high level of antisocial behaviour preceded high level of depression, which in turn, preceded de-escalation of antisocial behaviour, which finally resulted in lower level of depression. In the study by Hofstra et al. (2002) antisocial behaviour in childhood and adolescence predicted only boys’ mood disorders in adulthood. Overbeek et al. (2001) found no gender differences, suggesting that there are no cross-lagged associations between depression and antisocial behaviour as homotypic continuity of disorders best explain the future course of depression and antisocial behaviour. In sum, the findings of previous studies on comorbidity and longitudinal associations between adolescent depression and antisocial behaviour have been inconsistent, especially with regard to gender differences. Although there are a few studies which includes both boys and girls, gender comparisons are still relatively uncommon. Given that the rates of both adolescent depression and antisocial behaviour and the course of these disorders differ among boys and girls (Maughan et al., 2004; Overbeek et al., 2001), it could be assumed that longitudinal associations between these disorders also differ between genders. Most previous studies have investigated longitudinal associations between adolescent depression and antisocial behaviour using only one developmental model as a theoretical framework. It would be interesting to simultaneously test different developmental models to explain the longitudinal association between depression and antisocial behaviour. Ascertaining the developmental associations between depression and antisocial behaviour might provide important information regarding the etiology and course of these disorders for use in prevention and treatment. The purpose of the present study was to investigate continuity, comorbidity and longitudinal associations between depression and antisocial behaviour using data from a 2-year longitudinal study with adolescent boys and girls. In detail, the study investigated (1) how common homotypic continuity of depression and antisocial behaviour is during a 2-year follow-up in middle adolescence, (2) whether there is concurrent comorbidity between depression and antisocial behaviour at age 15 and at age 17, (3) which developmental model—the failure model or the acting out model would best explain the longitudinal associations between depression and antisocial behaviour during a 2-year follow-up, and (4) whether there are any significant sex differences in any of these themes.

Conclusion Our findings indicated that antisocial behaviour and depression had considerable continuity in middle adolescence. Concurrent comorbidity between depression and antisocial behaviour was also remarkable. Among girls depression predicted subsequent antisocial behaviour in a 2-year span, but among boys depression seemed to protect them from subsequent antisocial behaviour. The present study offers implications for the prevention and treatment of adolescent depression and antisocial behaviour. Prevention should be initiated as soon as possible before the psychosocial adjustment of adulthood is impaired. In clinical settings, antisocial behaviour among adolescent girls should be considered as a possible marker of severe depression.