Official URL: https://doi.org/10.1002/gps.1289

Background: There has been little attention to the underlying dimensional structure of the Hospital Anxiety and Depression Scale (HADS) in analyses involving individuals older than 65 years of age despite its routine application in this age group. Methods: The factor structure of the HADS was investigated using a sample of 680 veterans with limb amputations who were aged at least 66 years (mean 79.0, SD 5.02; range 66-92) Four models were specified and estimated using Lisrel 8.54. Model 1 specified a uni-dimensional structure. Model 2 specified a correlated two-factor model. Model 3 specified a correlated two-factor model but with cross factor loadings for item 7. Model 4 specified a three-factor model after Clark and Watson's (1991) tripartite theory of anxiety and depression. Results: Models 2, 3 and 4 were found to adequately fit the data. However, on balance, model 4 was found to be a better explanation of the data than the alternative specifications. Conclusions: The results indicate that factor models for the HADS developed in younger samples are replicable with older adults, thus supporting the suggestion that the HADS structure is invariant for age. However, considering the composition of the current sample, i.e. veterans with limb amputations, further research is necessary to determine whether these findings are generalisable to the wider population of older adults.