Article Text


Executive dysfunction and depressive symptoms in cerebrovascular disease
  1. G Bellelli1,
  2. E Lucchi1,
  3. G Cipriani2,
  4. G B Frisoni3,
  5. M Trabucchi4
  1. 1“Ancelle della Carità” Hospital, Cremona, and Geriatric Research Group, Brescia, Italy
  2. 2Geriatric Research Group, Brescia, Italy
  3. 3Laboratory of Epidemiology and Neuroimaging IRCCS “San Giovanni di Dio Fatebenefratelli”, Brescia, and Geriatric Research Group, Brescia, Italy
  4. 4Geriatric Research Group, Brescia, and University “Tor Vergata”, Roma, Italy
  1. Correspondence to:
 Dr G Bellelli, “Ancelle della Carità” Hospital, via Aselli 14, 26100 Cremona, Italy;
  1. J Kramer5
  1. 5University of California San Francisco, Department of Psychiatry and Langley Porter Psychiatric Institute, 401 Parnassus Avenue, Box PAC-0984, San Francisco, CA 94143–0984; kramer{at}

    Statistics from

    The article by Kramer et al1 suggests that subcortical ischaemic vascular disease is associated with subtle declines in executive functioning and visual memory, even in non-demented patients. The authors compared 27 control subjects and 12 non-demented patients, who were selected after exclusion of major depression, bipolar affective disorder, and other DSM-IV I axis disorders. We wish to contribute with personal data to this topic suggesting that, even in absence of a clinical diagnosis of depression, depressive symptoms may modulate executive dysfunctions in non-demented subjects with cerebrovascular disease.

    We examined 34 consecutive patients with cognitive impairment-no dementia (CI-ND) (mean (SD), age: 78.1 (6.3), range 65–90; years of education: 4.9 (1.7), range 3–10; Mini Mental State Examination score: 24.0 (2.4), range 18–27). The diagnosis of CI-ND was made on the basis of a standardised multidimensional protocol including history, clinical examination, detailed neuropsychological testing, and computed tomography. The presence and severity of cortical, white matter, and deep subcortical lesions and of leukoaraiosis were assessed on computed tomographic film with a standardised visual rating scale.2 With this method, the patients were quantitatively divided in two groups (50th centile) according to the severity of cerebrovascular disease: 17 patients had none or mild, and 17 had moderate or severe cerebrovascular disease. The two groups had similar age (mean (SD), age: 79.1 (6.5) and 76.9 (6.0); p=0.31, t test), educational level (mean (SD), years of schooling: 4.7 (1.5) and 5.3 (1.9); p=0.27 t test), cognitive impairment (mean (SD), Mini Mental State Examination score: 24.2 (2.7) and 23.8 (2.0); p=0.63 t test), functional status (mean (SD), Barthel Index: 85.3 (18.9) and 80.8 (21.4); p=0.57, t test), and comorbidity (mean (SD), Charlson Index: 2.1 (2.2) and 2.2 (1.6); p=0.95, t test). Comparing the neuropsychological tests, we found that the patients in the group with none or mild cerebrovascular disease performed better on Babcock (mean (SD), score: 11.3 (2.3) and 8.6 (3.0); p=0.01, t test), on digit symbol (mean (SD), associations in 90 seconds: 14.3 (6.8) and 11.7 (4.9); p=0.24, t test), on trail making A (mean (SD), seconds: 143.0 (82.6) and 228.2 (157.7); p=0.05, t test), but were significantly less depressed (mean (SD) number of symptoms on Center of Epidemiological Studies depression scale: CES-D, 10.8 (7.2) and 18.8 (6.6); p=0.002, t test). However, when the effect of cerebrovascular severity on all significant variables was weighted in a multivariate linear regression model, only depressive symptoms maintained the statistical significance (CES-D: β 0.82; 95% confidence intervals, 1.2 to 4.5; p=0.02).

    Several studies have shown that depressed patients have a lower performance than non-depressed ones on executive functions.3 Furthermore, the relation between cerebrovascular disease and depression has recently been established.4 In their study, Kramer and colleagues performed an extensive neuropsychological evaluation, but failed to take into consideration that depressive symptoms may be detected in elderly population even after exclusion of major depression. On the contrary, we suggest that depressive symptoms need to be considered in the interpretation of even subtle executive dysfunctions in cerebrovascular disease patients.


    Author’s reply

    We completely agree with Dr Bellelli and his colleagues about the importance of considering depressive symptoms when evaluating executive and other cognitive abilities in subjects with cerebrovascular disease. It is increasingly clear that subcortical-frontal ischaemic vascular disease, symptoms of depression, and executive impairment are related, although the nature of these relations is complex and not yet fully understood. Their data on patients with mild cognitive impairment amply illustrate the need to consider depression and other moderator variables when evaluating cognitive dysfunction.

    We would like to emphasise two points. Firstly, data on depression related symptoms (for example, depressed mood, anhedonia, guilt feelings, etc) were collected on all but two cases in our sample. Although we did not report these data in our paper, there were no correlations between any of our executive measures and depressive symptoms. In addition, when we used linear regression to simultaneously consider depression and subcortical lacunes, only the presence of subcortical lacunes predicted executive functioning.

    Our second point is to emphasise that while linear regression is a powerful technique for considering multivariate models, researchers must not confuse correlation with causation. For example, in the analyses by Bellelli et al, depression but not cerebrovascular severity remained in the model predicting performance on executive tasks. This cannot be interpreted to imply that the executive impairment in cerebrovascular patients is caused by their depression and not by their cerebrovascular disease. Cerebrovascular disease might cause both depression and executive impairment. Because cerebrovascular disease, depression, and executive impairment are all related, however, it would not be unusual for one of these variables to be excluded from a regression model.

    A causal relation between subcortical-frontal neuropathology and deficits on frontally mediated executive tasks makes sense neuroanatomically. Elucidation of the underlying mechanisms (for example, white matter disease versus lacunar volume versus strategic lacunes versus cortical microvascular change) and the precise behaviour consequences (for example, processing speed; working memory; set maintenance; inhibition) remains a high research priority.

    View Abstract

    Request permissions

    If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.