archives@tulane.edu / Exposure to violence among children and adolescents has been linked to negative outcomes including posttraumatic stress disorder, externalizing problems, low academic achievement, and impaired social-emotional adjustment. Aggressive behaviors specifically are shown to be strongly predicted by both direct and indirect violence exposure. Aggressive behaviors have consequences on later adjustment, making the identification of protective factors imperative to bolster support for children whose environments are affected by violence. The current study examined the effect of neighborhood violence on overt aggression as well as the potential of school climate variables to buffer the impact of violence on students’ aggressive behaviors. It was hypothesized that higher rates of neighborhood violence in the year preceding the study would predict higher levels of aggressive behaviors reported by teachers and students. It was further hypothesized that student perceptions of school safety, teacher-student relationships, and academic engagement would moderate this relationship, such that the relationship between neighborhood violence and aggressive behaviors would be lower in the context of positive school climate indicators than in the context of negative school climate indictors. Results did not support either hypothesis. However, results suggested that teacher-student relationships may reduce aggressive behaviors over time, which provides promising evidence that can inform teacher efforts and schoolwide opportunities to mitigate negative outcomes for their students. / 1 / Lea Petrovic

Background: Preterm birth is a major public health concern in the US. Previous studies have suggested that quality of the mother-father relationship, social support, and neighborhood violence may be associated with preterm birth; however, findings are equivocal. Objectives: The main objectives of this dissertation were: 1) to determine the modifying effect of perceived residential environment on the association between quality of mother–father relationship and preterm birth in a sample of African-American women, 2) to examine whether the receipt of social support modifies the association between neighborhood violence exposure and preterm birth in a nationally representative sample of US women, and 3) to determine the extent to which neighborhood violence mediates the association between neighborhood deprivation and preterm birth in a geographic cohort of women in Richmond city, Virginia. Methods: Data were obtained from three sources – 1) Life-course Influences on Fetal Environments (LIFE) study, 2) National Longitudinal Study of Adolescent to Adult Health, and 3) live birth records, police crime reports and census data for Richmond city, Virginia. Multivariable log-binomial regression models were used to examine the modifying effect of perceived residential environment on the association between quality of mother–father relationship and preterm birth, as well as the modifying effect of social support on the association between neighborhood violence exposure and preterm birth. Multilevel structural equation modeling was used to examine the mediational influence of neighborhood violence on the association between neighborhood deprivation and preterm birth. Results: For the association between neighborhood violence exposure and preterm birth, maternal receipt of social support modified the association [(Tertile 1: adjusted prevalence ratio (APR)=1.12; 95% CI=1.11-1.13, p<.0001); (Tertile 2: APR=1.07; 95% CI=1.06-1.08, p<.0001); and (Tertile 3: APR=0.88; 95% CI=0.86-0.89, p<.0001)] in a nationally representative sample of US women. No significant interaction was observed between any domain of the mother–father relationship and perceived maternal residential environment (all p > 0.05) in a sample of African American women. Additionally, no significant association was found between the quality of mother–father relationship and preterm birth (Trust domain: APR=1.03, 95% CI=0.99-1.07; dependability domain: APR=1.01, 95% CI=0.98-1.06; criticism domain: APR=1.03, 95% CI=0.99-1.07). The association between neighborhood deprivation and preterm birth in a geographic cohort of women in Richmond city, Virginia, was not mediated by neighborhood violence (β=0.063, 95% CI= –0.025, 0.151). Conclusions: Rates of preterm birth in women exposed to neighborhood violence may be improved by providing adequate social support during the pregnancy period. Insufficient evidence was found to support the modifying effect of perceived residential environment on the association between the mother-father relationship and preterm birth, as well as the mediational effect of neighborhood violence on the association between neighborhood deprivation and preterm birth. Future studies are needed to confirm these findings.