A SYSTEMATIC EVALUATION OF PTSD SYMPTOMS AND COMORBID PSYCHOPATHOLOGY: RESULTS FROM A REPRESENTATIVE SAMPLE OF CHILDREN AND ADOLESCENTS ASSESSED 6 MONTHS AFTER THE 9/11 WORLD TRADE CENTER ATTACK

BACKGROUND. In contrast to the abundant literature on Posttraumatic Stress Disorder (PTSD) in adults, few studies focused on children and adolescents. Ongoing issues in PTSD research in youths are 1) the insufficient specificity of DSM Criterion A, 2) the lack of support for a unique relationship between trauma and PTSD, 3) PTSD phenotypic heterogeneity, 4) the non-specificity of PTSD symptom, 5) the high rates of comorbid disorders, 6) the inclusion of dimensional scaling of PTSD in DSM-V, and 6) differential item functioning of PTSD critieria. OBJECTIVES. The overall aim of this work is to address these ongoing issues in PTSD research. In particular: 1) to identify specific variables associated with PTSD, 2) to examine if specific variables are differently associated with PTSD and MDD, 3) to explore the role of grief reactions in the development of PTSD and Depression, 4) to examine the heterogeneity of the PTSD syndrome, 5) to examine patterns of comorbidity between PTSD indicators and other psychopathological symptoms, 6) to inform the dimensional scaling of PTSD, and 7) to determine if PTSD indicators display differential item functioning across groups defined by gender and age. METHODS. Participants and study design: a unique citywide, large, random, representative sample of 8,236 New York City public school students in grades 4 through 12, assessed 6 months following the 9/11 World Trade Ceenter (WYC) attack. Measures: a self-report questionnaire assessing: 1) demographic variables, 2) PTSD, 3) other psychiatric disorders, 4) exposure to the WTC attack (direct, indirect and media exposure), 5) functional impairment, 6) grief reactions, and 7) other variables. Statistical Analysis. Several methods were used, including: 1) multiple correspondence analysis (MCA), 2) multivariate logistic regressions (MLR), 3) latent class analysis (LCA), 4) confirmatory factor analysis (CFA), 5) item response theory (IRT), and 6) differential item functioning (DIF). RESULTS. MCA: Exploratory analyses with MCA identified axes of variation in psychopathology that guided following analyses. MLR: Different risk factors were associated with PTSD and Depression; however, grief mediated the relationship between loss of a loved one and both PTSD and Depression. LCA: A four-class PTSD model best described PTSD symptom structure; classes differed quantitatively and qualitatively. The latent structure of PTSD varied across empirically defined subgroups characterized by different patterns of exposure, and across subgroups defined by gender and age. LCA on the whole set of psychiatric symptoms yielded a 6-class model as the best fitting one. Three severe disturbance classes were identified, defined by different combinations of internalizing and externalizing symptoms. Classes varied across groups defined by gender and age. CFA: A one factor model confirmed the unidimensionality of PTSD symptoms. IRT: PTSD criteria show different severity and discrimination parameters that might be usefull for the dimensional scaling of PTSD. DIF: PTSD items function differently in subgroups defined by age and gender. DISCUSSION. The results are discussed in light of epidemiological and biological findings, and in relation to the ongoing issues in PTSD research. Implications for treatment/prevention of PTSD and for genetic studies of PSTD are provided. Implications for the future edition of the DSM (DSM-V) are discussed as well.