Introduction: Anxiety may occur as an emotion, symptom, an anxiety disorder of its own, or as a part of other physical and psychiatric disorders. This may cause diagnostic problems when using the ICD-10 and DSM-IV-R criteria. There is also reason to believe that anxiety scales do not measure solely anxiety symptoms, because factors may vary between populations.
Objective: To present the factor structure of the state part of State-Trait Anxiety Inventory (STAI) and the Hopkins Symptom Checklist (HSCL-25), and a literature review of how anxiety rating scales measured anxiety in elderly suffering from physical disorders.
Design: Cross-sectional and literature review.
Materials and methods: The state part of the STAI was developed to measure anxiety symptoms. The HSCL-25 is a subset of the SCL-90, and is meant mainly to measure psychiatric symptomatology. These were factors analyzed based upon polychoric correlations (all items scored 1-4 in both instruments).
Four datasets were used: One coming from of 101 geriatric in-patients followed up during 12 months after discharge from hospital (examined with STAI), one from 77 geriatric in-patients interviewed three months after discharge from hospital (examined with STAI and HSCL-25), one from 68 elderly healthy controls (examined with STAI and HSCL-25), and one randomized selected group of elderly Norwegians living outside institutions and homes for elderly (examined with HSCL-25). The literature databases searched for anxiety symptom scales were MEDLINE, PSYCHINFO and CINAHL.
Results: The STAI assessments in geriatric in-patients suffering from physical disorders showed two clear concepts named after their items as Nervousness and well-being. In healthy elderly controls STAI measured one concept – nervousness. Using a mean item score of 2.0 as the threshold value for caseness, the prevalence of the STAI factor “nervousness” – was 37% and “well-being” 78%. Factor analyses of the HSCL-25 resulted in a cluster of concepts named anxiety, depression, somatic, and suicidal ideation. There were significant correlations between these factors (anxiety, depression, somatic and suicidal ideation) and physical disorders and mobility in the elderly.
The literature review revealed that factor analysis of anxiety rating scales tend to miss out in published paper.
Conclusions: Anxiety rating scales tend to measure symptoms of anxiety, depression, suicidal ideation, lack of well-being and chronic disorders. There is a confounding link between these factors and physical disorders. To obtain proper measurements, the scales must be subjected to factor analysis, and preferably in all populations where the scales are used. A literature-review verified that this methodology has been little used in previously published papers. This could contribute to why anxiety is difficult to diagnose in elderly suffering from physical disorders.
Back to S020 Anxiety in the Elderly; Current Status and an Agenda for the Future
Back to The Eleventh International Congress