Assessment of anxiety in older adults: psychometric properties and relationships with self-reported functional impairment.

OBJECTIVE

The current study aimed to examine the psychometric properties of two geriatric anxiety measures: the Geriatric Anxiety Inventory (GAI) and the Geriatric Anxiety Scale (GAS). This study also aimed to determine the relationships of these measures with two measures of functional ability and impairment: the Barkley Functional Impairment Scale (BFIS) and the Everyday Cognition Scale (E-Cog).

DESIGN

Confirmatory factor analyses (CFA) were used to analyze the factor structures of the GAI and GAS in older adults. Tests for dependent correlations were used to examine the relationship between anxiety scales and functioning.

SETTING

Amazon's Mechanical Turk.

PARTICIPANTS

348 participants (aged 55-85, M= 62.75 (4.8), 66.5% female) with no history of psychosis or traumatic brain injury.

RESULTS

CFAs supported the previously demonstrated bifactor solution for the GAI. For the GAS, the previously demonstrated three-factor model demonstrated a good-to-excellent fit. Given the high correlation between the cognitive and affective factors (r =.89), a bifactor solution was also tested. The bifactor model of the GAS was found to be primarily unidimensional. Tests for dependent correlations revealed that the GAS demonstrated stronger relationships with measures of self-reported functional impairment than the GAI.

CONCLUSIONS

The current study provides further psychometric validation of the factor structure of two geriatric anxiety measures in an older adult sample. The results support previous work completed on the GAI and the GAS. The GAS was more strongly correlated with self-reported functional impairment than the GAI, which may reflect differences in content in the two measures.