• stroke;
  • depression;
  • post-stroke depression;
  • phenomenology;
  • sensitivity;
  • specificity



Somatic and neurocognitive symptoms of depression may overlap with the physical symptoms of stroke, and thus make the diagnosis of post-stroke depression difficult.


To assess the sensitivity of individual depressive symptoms and their contribution to the diagnosis of post-stroke depression.

Patients and methods

Two hundred and six patients with first-ever stroke, participating in a longitudinal study, were administered the Structured Clinical Interview for DSM-IV and the Hamilton Depression Rating Scale (HAM-D). In a discriminant analysis the relative contribution of the individual HAM-D items to the diagnosis of major depressive disorder was evaluated.


The cumulative incidence of post-stroke major depressive disorder was 32%. The discriminant model based on HAM-D item scores was highly significant (p<0.001) and classified 88.3% of patients correctly as depressed or nondepressed. As expected, ‘depressed mood’ discriminated best between depressed and non-depressed stroke patients. ‘Reduced interests’ had a relatively low sensitivity and may in part reflect ‘apathy’, which often is considered a separate construct. With the exception of ‘suicidal thoughts’, most psychological symptoms, such as ‘hypochondriasis’, ‘lack of insight’ and ‘feelings of guilt’, were not very sensitive. Some somatic symptoms, such as ‘reduced appetite’, ‘psychomotor retardation’, and ‘fatigue’ had high discriminative properties.


Psychological, neurocognitive and somatic symptoms of depression differ among themselves in terms of diagnostic sensitivity, and should be considered individually. Some somatic symptoms are highly sensitive for depression and should not be neglected by following an ‘exclusive’ or ‘attributional’ approach to the diagnosis of PSD. Copyright © 2005 John Wiley & Sons, Ltd.