Factor structure, measurement invariance, and clinical change benchmarks of the Generalized Anxiety Disorder-7 (GAD-7) in pregnancy and postpartum

This is the authors' abstract. We don't add key points for this paper.

The Generalized Anxiety Disorder-7 (GAD-7) scale is widely used to measure anxiety symptom severity during the perinatal period. However, measurement invariance between pregnant and postpartum women is under-studied, differential item functioning has not been assessed, and score precision across the anxiety continuum is unknown in perinatal samples. The objective was to examine the factor structure and measurement properties of the GAD-7 in perinatal women, a Data were drawn from an ongoing investigation in Italy that began in 2021. Dimensionality was assessed using exploratory graph analysis and confirmatory factor analysis (CFA); measurement invariance and differential item functioning were evaluated via item-level effect size ( d MACS ) and iterative multigroup CFA across perinatal phases and trimesters; and item response theory was applied to examine scale information and reliability. The GAD-7 demonstrated strict one-dimensionality, scalar measurement invariance across the two perinatal phases and across five finer-grained perinatal phases (trimesters 1–3, early and late postpartum), and excellent classical and IRT reliability. Score precision and clinical change benchmarks have been provided. The results are restricted to Italian (92%) or Italian-speaking (8%) and may not generalize to other cultures. Participants have predominantly high levels of education. The GAD-7 possesses a unidimensional latent structure that is invariant across the entire perinatal continuum and can be used in perinatal research, mental-health surveillance, intervention trials, and routine obstetric care. • The GAD-7 demonstrated scalar measurement invariance across the antepartum and postpartum phases and across five finer-grained perinatal phases (trimesters 1–3, early and late postpartum). • The GAD-7 provides the most information in the moderate to high anxiety range and remains acceptably precise across nearly the entire observed score distribution. • Score precision and change benchmarks are provided for antepartum, postpartum, and cross-phase comparisons.