Poster #173 - Examining Racial-Ethnic Differences in the Factor Structure of Mental Disorders in the DSM-5

• In Event: 3-003 - Poster Session 09
  In Poster Session: PS 09 Section - Race, Ethnicity, Culture, Context

Integrative Statement

INTRODUCTION: There is accumulating evidence that mental disorders in the Diagnostic and Statistical Manual for Mental Disorders (DSM) are organized hierarchically (i.e., general factors representing internalizing and externalizing disorders), in part due to their heterogenous presentations and frequent comorbidity within individuals (Caspi et al., 2014; Krueger & Markson, 2006). However, few studies have investigated whether this hierarchy varies by racial-ethnic status, despite known differences in prevalence and comorbidity patterns of mental disorders across African Americans (AA) and Caucasians in particular (e.g. Rosenthal & Wilson, 2012). One study found structural invariance of 11 DSM-IV disorders between AA and Caucasian subgroups (i.e., presence of internalizing and externalizing factors), suggesting no “ethnicity bias” in the DSM (Eaton et al, 2013). However, this study was limited to adult populations using outdated DSM-IV criteria. Further research is needed to understand the structural hierarchy of mental disorders between AA and European American subgroups using DSM-5 criteria, and in younger populations.

METHOD. We used data from the Philadelphia Neurodevelopmental Cohort (PNC), which consists of subjects 8-21 years of age (N=9370). A multi-group confirmatory factor analysis (CFA) was conducted to test the structural invariance across 13 mental disorders (using DSM-5 criteria) between AA (n=3088) and Caucasians (n=5147). In the CFA, ten disorders were a priori classified as “Internalizing” disorders: Major Depressive Disorder, Generalized Anxiety Disorder, Anorexia, Bipolar Disorder, Specific Phobia, Social Anxiety Disorder, Agoraphobia, Separation Anxiety Disorder, PTSD, Obsessive Compulsive Disorder. The remaining 3 were “Externalizing” disorders: Attention-Deficit/Hyperactivity Disorder, Oppositional Defiant Disorder, and Conduct Disorder. Subject sex and age were entered as covariates.

RESULTS. First, we compared unidimensional, multidimensional, and bifactor models using the combined sample (N=9370). A general factor with Internalizing and Externalizing subfactors emerged as the optimal model (χ2 = 292.42, p<.001, CFI=.98, TLI=.97, RMSEA=.01). Consistent with Eaton et al. (2013), we found evidence of structural invariance between the constrained and unconstrained model in Caucasian versus the AA subgroups (ΔCFI=0, ΔTLI= .003, ΔRMSEA=-.001, Δχ2=31.68, p=.06), suggesting that this latent structure was statistically equivalent across race-ethnicity. All factor loadings of the mental disorders were significant, with the exception of Bipolar Disorder and Anorexia, which did not load on to the Internalizing factor for either subgroup.

DISCUSSION. Crucially, this study extends the work of others (i.e., Eaton et al., 2013) in a younger population using updated DSM-5 criteria regarding the structural invariance of mental disorders across two major racial-ethnic subgroups in the United States. These results suggest that differences in prevalence rates and comorbidities between racial-ethnic subgroups can best be attributed to variations in broader groupings of the mental disorders (i.e., general factor, Internalizing and Externalizing factors) rather than in discrete DSM-5 categories. Future work in the area should focus on identifying the risk factors for latent factors within racial-ethnic subgroups that have lead to higher rates of the discrete mental disorders in AA relative to Caucasians (Brown, 2003)

Authors

• Quanfa He, University of Wisconsin-Madison
  Presenting Author

• James J. Li, University of Wisconsin-Madison
  Non-Presenting Author