Validation of Generalized Anxiety Disorder 6 (GAD-6)—A Modified Structure of Screening for Anxiety in the Adolescent French Population

Abstract

Anxiety disorders remain underdiagnosed and undertreated, especially in child and adolescent populations. This study aimed to examine the construct validity of the Generalized Anxiety Disorder Scale 7 (GAD-7) in a sample of French adolescents by combining the Classical Test Theory (CTT) and the Item Response Theory (IRT) and to assess the invariance of items. A total of 284 adolescents enrolled in school in the Lorraine region were randomly selected to participate in a cross-sectional study. A psychometric evaluation was performed using a combination of CTT and IRT analyses. The study of psychometric properties of GAD-7 revealed poor adequation to the sample population, and engendered the deletion of one item (#7) and the merger of two response modalities (#2 and #3). These modifications generated the new GAD-6 scale, which had a good internal consistency reliability (Cronbach α = 0.85; PSI = 0.83), acceptable goodness-of-fit indices (χ2 = 28.89, df = 9, P = 0.001; RMSEA (90% CI) = 0.088 [0.054; 0.125]; SRMR = 0.063; CFI = 0.857), and an acceptable convergent validity (r = -0.62). Only one item (#5) had a consistent Differential Item Functioning (DIF) by gender. This study evaluated the structure of the GAD-7 scale, which was essentially intended at discriminating adolescent patients with high levels of anxiety, and adapted it to a population of adolescents from the general population. The GAD-6 scale presents better psychometric properties in this general population than the original GAD-7 version.