Never too late to be anxious: Validation of the geriatric anxiety inventory, Italian version

Aim. The aim of this work was to validate the Italian version of GAI (GAI-It) and its short form (GAI-It SF) in an over 65-population. Methods. In 3 recruitment areas across Italy, two raters reciprocally blind to results assessed eligible subjects; a semi-structured diagnostic clinical interview was performed by a psychiatrist. Results. Among the 76 enrolled subjects (mean age 72.7 +/- 6.8 years), anxiety symptoms were very common: 69.7% (moderate/severe HADS-Anxiety), 76.3% (moderate/severe STAI-state), 71.0% (moderate/severe STAI-trait), 61.8% (GAI), 55.3% (GAI-SF). Sensitivity, specificity and positive predictive value of GAI confirmed a good reliability of the Italian version, with Cronbach's Alpha equal to 0.93 for GAI-It and to 0.77 for GAI-It SF, indicating a very good and good construct validity, respectively, of the scales. The Pearson correlation index demonstrated a moderately positive correlation among GAI, GAI-SF and STAI. Conclusions. Our data confirm the validity of GAI-It as a valuable instrument to assess anxiety in an elderly population, for clinical and research purposes. Keywords