April 2010
Reviewer: Jeff C. Huffman, MD
Support for the vascular depression hypothesis in late-life depression
Background: Vascular depression has been described as a subtype of major depression that is characterized by onset/presence in later life, presence of vascular lesions/white matter hyperintensities (WMH), and impaired cognitive function (especially executive dysfunction); this syndrome frequently does not respond to standard antidepressant treatment. It has been necessary to further characterize of the relationship between the imaging characteristics and cognitive findings in patients with apparent vascular depression to improve our understanding of this treatment-resistant condition, which is likely to grow in prevalence with the aging of society.
Methods: The authors recruited subjects (age inclusion criteria not reported) from two academic sites who had DSM-IV diagnosed major depression. Excluded were subjects with unstable/severe medical illness, those with dementia (ruled out using several measures, including the MMSE), and history of other Axis I diagnoses prior to depression onset. Subjects underwent a battery of neuropsychological tests evaluating executive function, processing speed, episodic memory, language processing, and working memory, in addition to depression severity ratings using the Montgomery-Asberg Depression Rating Scale (MADRS). Brain MRIs were also performed prior to treatment, with hyperintensities assessed using ‘Fazekas scores.’ Subjects received 12 weeks of sertraline, flexibly dosed up to 200 mg/day based on clinical response and tolerability.
Results: A total of 217 subjects were enrolled, with 190 completers. Mean age of subjects was 68.4 years, with mean age of depression onset 53.6. Mean MMSE was 27/30. A total of 72/190 completers reached remission (MADRS <8); mean sertraline dose among completers was 118mg. White matter hyperintensities (Fazeskas score) and a variety of cognitive deficits were closely associated with one another, and were both associated with MADRS score (i.e., more hyperintensities and greater cognitive deficits were associated with lack of treatment response). On mixed regression models accounting for baseline variables, specific cognitive deficits were independently associated with MADRS scores, while Fazekas scores were not. The cognitive deficits associated with depression non-response included processing speed, episodic memory, language processing, and executive function; among these, processing speed was most strongly associated with MADRS.
Commentary: This well-designed study found that both WMH and a variety of cognitive deficits appear to be associated with poor treatment response among persons with late life depression. These findings are consistent with the vascular depression hypothesis. Interestingly, when both were analyzed together, cognitive deficits remained predictive of depression scores, while WMH did not, suggesting that assessment for cognitive deficits may be more important (and more strongly associated with treatment nonresponse) in this cohort, at least via the WMH measurement methods used here. It may be that specific location of deficits may be as (or more) important than the total burden of lesions. Another interesting finding was that impaired processing speed, at least in this cohort, appeared to be more predictive of depression scores than executive function, and a variety of cognitive deficits (including language problems) correlated with less change in MADRS scores. This suggests that rapid tests for processing speed might be helpful in assessing likelihood of treatment response in later-life depression.
One major remaining question about vascular depression, not addressed in this study, remains: how do we treat these patients? Such patients typically do not respond well to standard antidepressant treatment. In our clinical experience, these patients do appear to respond relatively well to ECT, and it may make sense to consider this as an option for patients with severe depression with a vascular depression picture. Psychostimulants may be another option, given their effects on dopamine and some suggestion that they may rapidly help depression in the elderly. Other studies have used the calcium channel blocker nimodipine to augment antidepressants in vascular depression patients, with apparent benefit. Clearly, however, the jury is still out on this key clinical question.