The ‘vascular depression’ hypothesis has attracted considerable interest, but its basis is by no means clear. Kaimal & Nair (Reference Kaimal and Nair2005) in their recent correspondence mention vascular comorbidity in late-onset depression, citing a high prevalence rate of hypertension in particular. However, despite reasonably consistent findings of neuroradiological abnormalities associated with late-life depression, there is actually little evidence that hypertension or other ‘traditional’ vascular risk factors are responsible. What evidence there is for higher comorbidity comes from case-control studies comparing people with late-life depression who are known to clinical services with community controls; these studies carry a high risk of selection bias. Studies carried out in community populations have not generally found any associations between hypertension and depression (Reference Kim, Stewart and ShinKim et al, 2004), even in high-risk samples (Reference Stewart, Prince and MannStewart et al, 2001).

The direction of causation between vascular disorders and depression is also unclear (Reference BaldwinBaldwin, 2005). Evidence for neuroradiological abnormalities associated with depression is derived almost entirely from cross-sectional studies, and there is currently much stronger prospective evidence that depression is a risk factor for vascular disorders than vice versa. The same appears to be the case for sub-syndromal depressive symptoms that may not be recalled or reported in later life and might well have been present in people whose depression is classified as ‘late-onset’. Neuropathological studies of late-life depression do not find the microvascular abnormalities that would be expected if hypertension were responsible, but instead find increased large vessel disease (Reference Thomas, Ferrier and KalariaThomas et al, 2001), which could equally be a consequence as a cause of depression.

The uncertainty surrounding specific vascular processes in late-life depression is not just of academic interest, since there are important public health implications. Depression is strongly predicted by poor physical health and associated disability and there is little evidence at present to justify distinguishing specific disorders as responsible. The relationship between vascular processes and affective disorder is interesting, but there is a danger that more obvious risk factors for late-life depression (disability, poverty, loneliness) are ignored because they do not fall within the exciting world of organic psychiatry.