A psychometric evaluation of the Hospital Anxiety and Depression Scale for the medically hospitalized elderly

作者：Helvik Anne Sofie^*; Engedal Knut; Skancke Randi H; Selbaek Geir

来源：Nordic Journal of Psychiatry, 2011, 65(5): 338-344.

DOI：10.3109/08039488.2011.560684

• Factor analysis
• Principal component analysis
• Psychometric properties
• Quality of life
• Validation

Background: Few psychometric studies of the Hospital Anxiety and Depression Scale (HADS) scale have been performed with clinical samples of elderly individuals. Methods: The participants were 484 elderly (65-101 years, 241 men) patients in an acute medical unit. The HADS, the Montgomery-Aasberg Depression Rating Scale (MADRS) and questionnaires assessing quality of life, functional impairment, and cognitive function were used. The psychometric evaluation of the HADS included the following analyses: 1) the internal construct validity by means of principal component analysis followed by an oblique rotation and corrected item-total correlation; 2) the internal consistency reliability by means of the alpha coefficient (Cronbach's) and 3) concurrent validity by means of Spearman's rho. Results: We found a two-factor solution explaining 45% of the variance. Six of seven items loaded adequately (>= 0.40) on the HADS-A subscale (item 7 did not) and five of seven items loaded adequately on the HADS-D subscale (items 8 and 10 did not). Cronbach's alpha for the HADS-A and HADS-D subscale was 0.78 and 0.71, respectively. The correlation between HADS-D and the MADRS, a measure of the concurrent validity, was 0.51. Conclusion: The HADS appears to differentiate well between depression and anxiety. The internal consistency of the HADS in a sample of elderly persons was as satisfactory as it is in samples with younger persons. In contrast to younger samples, item 8 ("I feel as if I have slowed down") did not load adequately on the HADS-D subscale. This may be attributed to the way elderly people experience and describe their symptoms.

• 出版日期2011-10