Psychometric properties of the state part of the Spielberger State-Trait Anxiety Inventory (STAI) in geriatric patients

Background A conspicuously high score on the state part of the State-Trait Anxiety Inventory (STAI) has been observed among geriatric inpatients who a re neither demented nor critically ill; 43% of them had a sumscore that, ac cording to Spielberger's criteria, would reflect clinically relevant anxiet y symptoms. Objectives To explore the reasons for this high score. Methods 101 geriatric inpatients and 68 healthy controls of similar age, li ving at home and recruited through senior citizen centres participated in a controlled cross-sectional study. Results High item-scores were more frequent on the symptom-negative items t han on the symptom-positive items. Multigroup factor analysis produced two factors termed 'well-being' and 'nervousness', which had a moderate correla tion (0.61). The intercept was much higher on 'well-being' than on 'nervous ness', showing that a lack of well-being contributes significantly to the h igh score on the STAL This confounds the sumscore. However, the geriatric i npatients nevertheless had a high score on the factor 'nervousness'. Female controls scored higher than males on both factors, whereas among the geria tric patients neither age nor gender related to them. Conclusions The STAI state sumscore is a biased indicator of anxiety in ger iatric inpatients owing to confounding by wellbeing. The most important cau se for the observed high score on the STAI state instrument in geriatric pa tients relates to a reduced well-being. Copyright (C) 2001 John Wiley & Son s, Ltd.