بدرفتاری نوجوانان و تاثیر آن بر رفتار ضداجتماعی بزرگسالان جوان

Abstract Statement of problem Childhood maltreatment is known to be a risk factor for a range of later problems, but much less is known about adolescent maltreatment. The present study aims to investigate the impact of adolescent maltreatment on antisocial behavior, while controlling for prior levels of problem behavior as well as sociodemographic characteristics.

Introduction Many disciplines, including social work, developmental and clinical psychology, criminology, sociology, medicine, and psychiatry recognize adolescence as a critical developmental stage within the life course. A recent review of research on adolescent development acknowledges the expansion of research on adolescence in the last decade (Steinberg & Morris, 2001). Several adolescent experiences, for example, family conflict and disruption (Dornbusch et al., 1985; Harold, Fincham, Osborne, & Conger, 1997; Thornberry, Smith, Rivera, Huizinga, & Stouthamer-Loeber, 1999), criminal victimization (Lauritsen, Sampson, & Laub, 1991), or gang involvement (Thornberry, Krohn, Lizotte, Smith, & Tobin, 2003) have been linked to negative consequences, including increased risk of criminal involvement, substance use, and problems in life-course transitions. However, research on an important and related topic—adolescent abuse and neglect and their consequences—has not garnered much attention in the research literature. A search of major journals from several different disciplines over the past decade reveals very little research that deals specifically with this topic. Recent studies have begun to illuminate the relationship between adolescent maltreatment and negative behavioral and psychological development in adolescence (e.g., see Ireland, Smith, & Thornberry, 2002; Thornberry, Ireland, & Smith, 2001). Taken together, these studies suggest that any substantiated maltreatment experience during adolescence increases the risk of general delinquency, violence, drug use, alcohol-related problems, internalizing problems, externalizing problems, and depressive symptoms, as well as arrest during adolescence compared to those never maltreated. The present analysis moves beyond previous research on the consequences of adolescent maltreatment in two respects. First, prospective, longitudinal studies have not investigated the consequences of adolescent maltreatment on adult outcomes. Second, prior studies have not examined the relationship between adolescent maltreatment and either contemporaneous or longitudinal outcomes while controlling for previous involvement in offending. In this study, we respond to these gaps in the literature by focusing specifically on adolescent maltreatment and its consequences in early adulthood while controlling for earlier involvement in offending.

Results The first key analytic question is the distribution of adolescent maltreatment in an urban community sample. In the RYDS sample, 198 subjects (20% of the entire sample) experienced substantiated maltreatment at some point between birth and age 18. Furthermore, the rate of substantiated adolescent maltreatment for this sample of urban adolescents is just below 10%. In all, 82 adolescents had a substantiated case of maltreatment that occurred between the ages of 12 and 17, which accounts for about 40% of all substantiated cases of maltreatment. About one third (35%) of the adolescent maltreatment cases also had a substantiated maltreatment incident prior to 12, while about 65% had an official record of maltreatment for the first time in adolescence. In these analyses, groups that experienced maltreatment only in adolescence are combined with those who experienced maltreatment in childhood as well as adolescence in order to maximize sample numbers. Prior analysis suggests little difference in outcomes between these groups and we wanted to maximize sample size (Ireland et al., 2002). Among the maltreated adolescents, 14 (1.6%) experienced sexual abuse, 36 (4.1%) experienced physical abuse, and 32 (3.6%) were neglected or emotionally abused (see Table 1). In all but two cases, the victim of sex abuse was female. More balanced gender distributions are observed for other types. The majority of subjects experienced multiple subtypes of maltreatment, however. For example, of 14 subjects sexually abused during adolescence, 4 experienced only sexual abuse. Of those not sexually abused, 42 subjects maltreated in adolescence were physically abused, but only 12 of these adolescents experienced only physical abuse. Only 19 of the 43 neglected in adolescence experienced neglect only. Chi-square tests were used to compare the prevalence of control variables in maltreated and non-maltreated groups. The control variables are associated as expected with adolescent maltreatment. For example, chronic family poverty results in a higher prevalence of adolescent maltreatment compared to families not living in chronic poverty (14.6% vs. 7.8%, p < .05), and similar patterns are observed for parental educational attainment and caregiver changes. Additionally, about 18% of the females experience adolescent maltreatment, while about 6% of the males experience adolescent maltreatment (p < .05). Finally, race/ethnicity is significantly related to maltreatment: African American and Hispanic subjects are more likely to be maltreated in adolescence (13.5% and 11.7%, respectively) compared to White subjects (6.0%). Living in high poverty neighborhoods, however, is unrelated to adolescent maltreatment in this sample. Table 2 examines the short-term behavioral consequences of experiencing any adolescent maltreatment. The first row indicates that, after taking control variables and previous involvement in the behavior into consideration, experiencing any adolescent maltreatment increases the risk of arrest, violent offending, and drug use in late adolescence. The odds of arrest among those maltreated in adolescence are over two times the odds of arrest for those never maltreated (OR = 2.29, CI = 1.32, 3.97), and the same is true for violent offending (OR = 2.11, CI = 1.26, 3.54). Finally, the odds of drug use in late adolescence are more than one and a half times greater (OR = 1.76, CI = 1.05, 2.95). However, there is no relationship between adolescent maltreatment and general offending in late adolescence after taking into account control variables and prior general offending. Table 2. Adolescent maltreatment and late adolescent outcomes Arrest General offending Violent offending Drug use Odds ratio CI Odds ratio CI Odds ratio CI Odds ratio CI Any adolescent maltreatment 2.29** (1.32, 3.97) 1.63 (.99, 2.68) 2.11** (1.26, 3.54) 1.76* (1.05, 2.95) Early adolescent offendinga 3.87** (2.61, 5.74) 4.21** (3.09, 5.76) 6.04** (4.29, 8.50) 8.29** (5.79,11.87) Parent did not complete HS 1.64* (1.13, 2.40) .95 (.69, 1.29) .80 (.57, 1.13) .76 (.54, 1.09) Community poverty 1.00 (.99, 1.02) 1.03** (1.02, 1.04) 1.03** (1.02, 1.04) .99 (.98, 1.01) Chronic family poverty .99 (.66, 1.49) 1.08 (.76, 1.52) 1.14 (.78, 1.66) 1.37 (.93, 2.01) Caregiver changes 1.98** (1.35, 2.91) 1.50* (1.07, 2.11) 1.08 (.75, 1.56) 1.37 (.95, 1.98) Female .25** (.17, .37) .45** (.33, .62) .42* (.30, .61) .67* (.47, .95) African American 3.74** (1.99, 7.03) 1.19 (.78, 1.82) 1.18 (.72, 1.94) 1.30 (.80, 2.12) Hispanic 3.96** (1.93, 8.13) 1.14 (.67, 1.95) 1.41 (.76, 2.58) 1.04 (.57, 1.91) −2 log likelihood 571.14 802.41 817.50 670.52 Model chi-square 115.95 120.99 148.13 123.75 N 666 666 666 666 a The lagged behavioral measure for each equation varies depending upon the dependent variable. For example, when predicting late adolescent arrest, the lagged early adolescent measure is “early adolescent arrest.” The model chi-square for each equation is statistically significant. * p < .05 (one-tailed). ** p < .01(one-tailed). Table options Most control variables are related, as expected, to at least one of the dependent variables. Children of parents who did not complete high school are more likely to be arrested. Community poverty is related to general and violent offending, although family poverty is unrelated to any of the outcomes. Caregiver changes impact arrest and general offending. Young women are less likely than their male counterparts to be arrested, to engage in general and violent offending, and to use drugs. Finally, risk of arrest is higher for African American and Hispanic youth, although their risk of self-reported offending is not significantly greater. All lagged dependent variables are significantly related to the outcomes, as expected. Repeating all analyses using a general control variable that includes prior delinquency or drug use does not change the substantive consistency of the results, with the exception that adolescent maltreatment moves to insignificance when predicting general offending in early adulthood, and physical abuse moves to significance when predicting arrest in early adulthood. Table 3 extends the analysis into early adulthood to determine if the short-term effects identified in late adolescence persist into a respondent's early 20s. The first row of Table 3 shows that experiencing any adolescent maltreatment increases the odds of arrest, general offending, violent crime, and drug use in early adulthood when compared to those who were never maltreated, after taking into consideration the control variables including prior involvement in similar behaviors. These multivariate results support the conclusion that experiencing any adolescent maltreatment increases the risk of problematic behavioral adaptations into early adulthood. For example, experiencing any adolescent maltreatment increases the odds of drug use in early adulthood by 2.14 (CI = 1.22, 3.73) times compared to those who were never maltreated. However, the strongest impact of adolescent maltreatment in early adulthood is seen for arrest: the odds of arrest are almost two and a quarter times higher (OR = 2.24, CI = 1.31, 3.85) for those who have been maltreated as adolescents compared to those who have never been maltreated. Overall, these results indicate that the consequences of adolescent maltreatment persist into early adulthood (controlling for similar behavior at earlier ages as well as sociodemographic factors)—presumably after the maltreatment has ceased. The magnitude of the risk remains relatively unchanged from late adolescence to early adulthood, so there is no clear indication that the negative effects of adolescent maltreatment substantially decay over time. For example, the odds ratio for risk of arrest is 2.29 (CI = 1.32, 3.97) in late adolescence (Table 2); it is 2.24 (CI = 1.31, 3.85) in early adulthood (Table 3). There is an increase in risk for drug use and general offending in early adulthood. Finally, the controls included in the estimated equations presented in Table 3 in general behave in a manner similar to the results presented in Table 2. Table 3. Adolescent maltreatment and early adult outcomes Arrest General offending Violent offending Drug use Odds ratio CI Odds ratio CI Odds ratio CI Odds ratio CI Any adolescent maltreatment 2.24** (1.31, 3.85) 1.75* (1.04, 2.93) 2.03* (1.23, 3.35) 2.14* (1.22, 3.73) Late adolescent offendinga 5.70** (3.92, 8.29) 3.22** (2.38, 4.35) 3.35** (2.39, 4.70) 13.15** (8.91, 19.42) Parent did not complete HS 2.01** (1.38, 2.93) .55** (.40, .75) .79 (.56, 1.12) 1.19 (.84, 1.68) Community poverty 1.01 (.99, 1.02) 1.00 (.99, 1.01) .99* (.97, 1.00) 1.02* (1.00, 1.03) Chronic family poverty 1.15 (.77, 1.73) 1.13 (.79, 1.62) 1.41 (.96, 2.06) .85 (.58, 1.25) Caregiver changes 1.15 (.78, 1.71) 1.26 (.88, 1.80) .83 (.57, 1.22) 1.66* (1.14, 2.43) Female .28** (.19, .42) .54** (.39, .74) .92 (.65, 1.32) .42** (.30, .60) African American 1.16 (.69, 1.95) .39** (.24, .63) 1.13 (.69, 1.85) .75 (.46, 1.21) Hispanic .81 (.43, 1.54) .47* (.26, .83) 1.26 (.70, 2.28) .81 (.45, 1.45) −2 log likelihood 593.701 753.67 671.37 672.97 Model chi-square 142.54 106.04 49.70 201.28 N 659 637 637 637 a The lagged behavioral measure for each equation varies depending upon the dependent variable. For example, when predicting early adult arrest, the lagged late adolescent measure is “late adolescent arrest.” The model chi-square for each equation is statistically significant. * p < .05 (one-tailed). ** p < .01 (one-tailed). Table options The final analytic question is whether the short-term and longer-term effects of adolescent maltreatment vary by the type of maltreatment experienced. Table 4 reports the results obtained from considering specific types of maltreatment instead of a global measure of any adolescent maltreatment. The top panel of Table 4 considers the relationship between adolescent neglect, adolescent physical abuse, and adolescent sex abuse and late adolescent outcomes, while the bottom panel examines the relationship between these types of maltreatment and early adult outcomes. Recall that the specific types of maltreatment considered do not represent pure types of maltreatment but may include multiple maltreatment types. The most consistent pattern is that adolescent neglect significantly increases the odds of three outcomes in late adolescence—arrest, general offending, and violent crime, although not the odds of drug use. Physical abuse increases the odds of violent crime (OR = 2.54, CI = 1.17, 5.48) and drug use (OR = 3.66, CI = 1.78, 7.52) compared to those never maltreated, but physical abuse does not significantly increase the odds of either arrest or general offending. Results indicate that those who experience adolescent sexual abuse do not have increased odds for any outcome in late adolescence compared to those never maltreated. Table 4. Types of adolescent maltreatment and outcomes in late adolescence and early adulthooda Arrest General offending Violent offending Drug use Odds ratio CI Odds ratio CI Odds ratio CI Odds ratio CI A. Late adolescent outcomes Adolescent neglect 4.37** (1.96, 9.75) 3.23* (1.39, 7.52) 3.59** (1.61, 8.01) .96 (.41, 2.25) Adolescent physical abuse 1.41 (.61, 3.30) 1.12 (.55, 2.31) 2.54* (1.17, 5.48) 3.66** (1.78, 7.52) Adolescent sex abuse 1.67 (.56, 4.97) 1.13 (.43, 2.99) .59 (.19, 1.79) .97 (.35, 2.66) −2 log likelihood 568.00 799.27 720.42 665.05 Model chi-square 119.09 124.13 97.08 129.21 N 666 666 666 666 B. Early adult outcomes Adolescent neglect 2.36* (1.03, 5.42) 1.32 (.60, 2.88) 1.76 (.82, 3.78) 2.55*** (1.10, 5.90) Adolescent physical abuse 2.15 (1.00, 4.64) 1.36 (.63, 2.93) 2.17* (1.05, 4.50) .99 (.41, 2.41) Adolescent sex abuse 2.23 (.76, 6.61) 5.04* (1.50, 16.91) 2.30 (.85, 6.24) 5.74** (1.68, 19.63) −2 log likelihood 593.68 750.46 671.20 668.65 Model chi-square 142.57 109.25 49.87 205.59 N 659 637 637 637 a The control variables presented in Table 2 and Table 3 are included in each of the estimated equations presented above, but the coefficients are not reported in the interest of space. Similarly, the lagged measures of earlier behaviors are also included in each equation as before, but are not reported. The model chi-square for each equation is statistically significant. * p < .05 (one-tailed). ** p < .01 (one-tailed). Table options The second panel of Table 4 examines outcomes in early adulthood. Although neglect had fairly robust effects in late adolescence, its negative effects have somewhat dissipated in early adulthood. Neglect increases the risk of arrest and drug use in early adulthood but not general or violent offending. The impact of physical abuse among maltreated adolescents has also dissipated somewhat—physical abuse is only related to violent offending in early adulthood. Finally, although experiencing sexual abuse among maltreated adolescents was unrelated to our measures of antisocial behavior in late adolescence, by early adulthood sexual abuse is linked to general offending (OR = 5.04, CI = 1.50, 16.91) and drug use (OR = 5.74, CI = 1.68, 19.63). Overall, five of the coefficients for specific types of maltreatment are significant in the top half of Table 4, and five coefficients are significant in the equations in the bottom half of Table 4 although the pattern of effects for different subtypes changes somewhat over time. We present these findings with caution, however, because we used a one-tailed test of significance and the number of cases of each type of maltreatment is relatively small. Coefficients for drug use appear especially unstable as indicated by the rather large estimated confidence intervals.