Different cultures often express identical symptoms in different ways. The original four-factor structure of the Center for Epidemiologic Studies Depression Scale (CES-D) might not be appropriate for all cultural groups. This study aimed to investigate the factor structure of the CES-D among Japanese Brazilians and to examine relationships between the CES-D and Japanese language proficiency. Participants were Japanese Brazilians (n = 136). Exploratory and confirmatory factor analyses were conducted to determine the structure of the Portuguese CES-D. The relationships were examined using the generalized linear model. We tested the five CES-D models on Japanese Brazilians and found these models did not satisfactory meet goodness-of-fit criteria. Therefore, we performed post-hoc factor analyses on the CES-D and determined a best-fit structure composed of four factors, namely: ‘Depressed Affect/Somatic Symptoms’, ‘Loneliness/Sadness’, ‘(Absence of) Positive Affect’ and ‘Interpersonal Difficulties’. Two items functioned differently from the original model and the structure differed from models developed among Brazilian or Japanese samples. Internal reliability and construct validity were acceptable. Japanese language proficiency, a measure of culture retention, was significantly related to subscores for the absence of positive affect. This suggests a complex relationship between language and cultural background that can affect reporting on instruments such as the CES-D.
|Number of pages
|International Journal of Culture and Mental Health
|Published - 2015 Oct 2
- confirmatory factor analysis
- factor structure
- Japanese Brazilians
- Japanese language proficiency