Background: The objectives of this study are 1) to examine the frequencies of neuropsychiatric symptom clusters in patients with stroke or transient ischemic attack (TIA) by cognitive level and stroke subtype; and 2) to evaluate effect of demographic, clinical, and neuroimaging measures of chronic brain changes and amyloid upon neuropsychiatric symptom clusters.
Methods: Hospital-based, cross-sectional study. 518 patients were administered the Neuropsychiatric Inventory (NPI) 3-6 months post index admission. NPI symptoms were classified into four symptom clusters (Behavioral Problems, Psychosis, Mood Disturbance & Euphoria) derived from a confirmatory factor analysis of the 12 NPI items. Multivariable logistic regression was used to determine independent associations between demographic, clinical and neuroimaging measures of chronic brain changes (white matter changes, old infarcts, whole brain atrophy, medial temporal lobe atrophy [MTLA] and frontal lobe atrophy [FLA]) with the presence of NPI symptoms and all symptom clusters except euphoria. 11C-Pittsburg Compound B Positron Emission Tomography (11C-PiB PET) was performed in 24 patients to measure amyloid retention for Alzheimer's Disease (AD) pathology.
Results: 50.6% of the whole sample, including 28.7% cognitively normal and 66.7% of patients with mild cognitive symptoms, had ≥1 NPI symptoms. Frequencies of symptom clusters were largely similar between stroke subtypes. Compared to patients with cardioembolic stroke and intracranial haemorrhage, those with TIA had less frequent mood disturbance. Stroke severity at admission and MTLA were the most robust correlates of symptoms. FLA was associated with behavioral problems cluster only. Frequency of symptom clusters did not differ between patients with and without significant amyloid retention.
Conclusion: Frequency of neuropsychiatric symptoms increased with level of cognitive impairment but was largely similar between stroke subtypes. Stroke severity and MTLA were associated with neuropsychiatric symptoms. AD pathology appeared to be unrelated to neuropsychiatric manifestations but further studies with larger sample size are required to substantiate this finding.