[Psychometric evaluation of depression in the elderly subject: which instruments? What are the future perspectives?].

Depression in the elderly is frequent but often unknown (in 30 to 50% of the cases) because of difficulties in detecting or diagnosing it. This is due to the clinical features and prognosis of depression in this kind of population but also to the non-existence of specific diagnostic tests. Most of the authors consider that the most useful diagnostic tests are screening assessments. Some are rating scales that have been validated in general population then secondarily in the elderly: Hamilton Rating Depression Scale (HDRS), Montgomery & Asberg Depression Rating Scale (MADRS), Zung Self Rating Depression Scale (Zung SDS), Beck Depression Inventory (BDI) or Center for Epidemiological Studies Depression Scale (CES-D). They usually involve biases linked to age and more particularly to somatic items; and the educational level required to answer is too high for this population. However, the MADRS is still interesting for measuring change under treatment and the CES-D for detection of depressive elderly. On the other hand, some screening scales are specific of depression in the elderly. The most commonly used is the GDS (Geriatric Depression Scale) with 30 items. Some points have been discussed to increase the achievement of these methods. For example, inventories are better than interviews and should be integrated into semi-standardized interviews which do not last more than 30 minutes. The quotation 'yes' or 'no' is preferable. The instrument have to be short but have to contain specific items for depression in the elderly. Several short forms are already validated or in progress such as GDS with 15 items and, recently, with 4 items, BASDEC, short Zung IDS, BDI with 13 items and DGDS. However, these screening scales loose a part of their validity in the moderate or severe demented elderly. Few instruments can screen depression in a demented population although depression and dementia syndromes are frequently associated. Some of the inventories used are not specific: they evaluate the general psychopathology in the elderly and contains subscales which screen depression or organic brain disease. Thus, GMS-AGECAT Package, CAMDEX, CARPER, BAS are often used by the Anglo-Saxons. At present, only one specific instrument has been validated: the Cornell Scale for Depression in Dementia. Recently, new screening instruments have been put forward: Dementia Mood Assessment Scale and Canberra Interview for the elderly which seem interesting but need further studies.