The severe depressive disorders of late life are associated with high rates of medical morbidity and mortality, cognitive impairment, suicide, disability, complex treatment regimens, institutionalization and high costs to the community (Murphy, 1983; Murphy et al. 1988; Bruce & Leaf, 1989; NIH Consensus Development Panel, 1992; Alexopoulos et al. 1993a, b; Brodaty et al. 1993; Bruce et al. 1994; Forsell et al. 1994; Hickie et al. 1995; Blazer, 1996). Those disorders that are accompanied by cognitive impairment and/or concurrent medical morbidity have a particularly poor outcome (Bruce & Leaf, 1989; Alexopoulos et al. 1993b; Hickie et al. 1995, 1997a). Although psychosocial models of late-life depression place considerable importance on age-related psychological and social risk factors, those who survive into later life may actually be characterized by psychological resilience (Henderson, 1994; Blazer, 1997).

Current aetiological research in late-life depression, therefore, places particular emphasis on the potential role of biological risk factors. The potential importance of vascular risk factors is receiving renewed attention and may provide opportunities for specific prevention and intervention strategies in high-risk populations. This emphasis on possible vascular risk factors, and the wider importance of vascular pathologies in late-life neuropsychiatric disorders, mirrors the emphasis of much earlier clinico-pathological studies (Binswanger, 1894; Alzheimer, 1895). The specific focus on the importance of small progressive changes within the subcortical white matter, as distinct from more discrete cortical infarcts (Olszewski, 1962), is now supported by the emerging neuroimaging literature and theoretical constructs in late-life depression (Krishnan, 1991, 1993; Hickie et al. 1996, 1997b; Krishnan et al. 1997).