Psychometric properties of the Japanese version of the Geriatric Anxiety Inventory for community-dwelling older adults

Background This study developed a Japanese version of the Geriatric Anxiety Inventory (GAI-J) and its short form (GAI-J-SF) to evaluate anxiety in older adults in Japan and assess these measures' psychometric properties with a cross-sectional design. Methods Participants (N = 400; mean age: 75 years) were community-dwelling older adults who answered a set of self-report questionnaires. They were recruited from a community centre for older persons in the Kanto region of Japan. Of the respondents, 100 participated in a follow-up survey to evaluate test-retest reliability. Item response theory was adopted to evaluate item parameters. Results Confirmatory factor analysis with categorical data suggested that, as with the original Geriatric Anxiety Inventory, the GAI-J/GAI-J-SF had a unifactor structure. Test-retest correlation and internal consistency analyses indicated that these scales had high reliability. Item response theory results showed that both measures' item parameters were acceptable. Correlations with the Penn State Worry Questionnaire, State Trait Anxiety Inventory-State Only, Generalized Anxiety Disorder-7 and Patient Health Questionnaire-9 were mostly consistent with our hypotheses. This supports the high convergent validity of the GAI-J/GAI-J-SF. Conclusions The findings indicate that the GAI-J and the GAI-J-SF have robust psychometric properties for assessing late-life anxiety in older Japanese adults. Future GAI-J studies in clinical groups are needed.