Factorial Validity and Invariance of the 7-Item Generalized Anxiety Disorder Scale (GAD-7) Among Populations With and Without Self-Reported Psychiatric Diagnostic Status

Frontiers in Psychology

Satomi Doi¹  Yoshitake Takebayashi²  Kumiko Muramatsu³  Masaru Horikoshi⁴  Masaya Ito⁴ 

[1] Department of Global Health Promotion, Tokyo Medical and Dental University, Tokyo, Japan;Department of Health Risk Communication, School of Medicine, Fukushima Medical University, Fukushima, Japan;Graduate School of Clinical Psychology, Niigata Seiryo University, Niigata, Japan;National Center for Cognitive Behavior Therapy and Research, National Center of Neurology and Psychiatry, Tokyo, Japan;

关键词: anxiety;    factor structure;    measurement invariance;    Japanese;    factorial validity;   

DOI  :  10.3389/fpsyg.2018.01741

来源: DOAJ

【 摘 要 】

The 7-item Generalized Anxiety Disorder Scale (GAD-7) is commonly used to monitor anxiety symptoms. However, its factor structure has been inconsistent among competing models: unidimensional, two-dimensional, or higher order models. Additionally, it is unknown whether the scale has measurement invariance between populations with and without self-reported psychiatric diagnostic status. Participants were Japanese adults with self-reported anxiety disorder (AD; n = 479), self-reported AD and major depressive disorder (MDD; n = 314), or without self-reported psychiatric diagnostic status (self-reported non-MDD/AD; n = 654), who completed this questionnaire on the Internet. Confirmatory factor analyses showed the higher order model had similar fit indices to the unidimensional and two-dimensional factor models. For the higher order model of GAD-7, metric invariance was supported between the self-reported non-MDD/AD and self-reported AD status groups, and scalar invariance was supported between the self-reported AD status and self-reported AD with MDD status groups. Moreover, convergent and discriminant validity were consistent with previous findings in Western cultures. These results suggest that factor loadings are equivalent and the construct has the same meaning between the self-reported non-MDD/AD and self-reported AD status groups, and the total or sub-scale scores were comparable between self-reported AD status and self-reported AD with MDD status groups. The major limitation of this study is that the participants’ diagnoses were self-reported, not confirmed by clinical structured interview. Further studies that incorporate clinical structured interviews are needed.

【 授权许可】

Unknown