Depressive symptoms profile and dementia risk after spontaneous intracerebral haemorrhage.

Introduction: Depressive symptoms are commonly reported after spontaneous intracerebral haemorrhage (ICH) and frequently associated with cognitive decline. Using hierarchical clustering analysis (HCA), we aimed to identify different post-ICH depressive symptoms profiles and to evaluate their association with dementia risk.
Methods: We included consecutive patients from the prospective Prognosis of Intracerebral Haemorrhage (PITCH) study who survived 6 months after the ICH. We performed HCA using depressive symptoms severity (assessed using the Montgomery-Åsberg Depression Rating Scale (MADRS)), along with the presence of apathy and anxiety (screened using Neuropsychiatric Inventory questionnaire). Baseline clinical/neuroimaging characteristics and risk of incident dementia were compared between different profiles using univariate and multivariable models.
Results: Of 265 six-month ICH survivors, 221 (83%) underwent neuropsychiatric screening (mean age 65.5 years; 57% male). Using HCA, 3 profiles were identified: (1) without significant depressive symptoms ( n  = 152; median MADRS score = 2 [IQR 0-4]); (2) depressive symptoms with predominant apathy ( n  = 41; median MADRS score = 15 [IQR 5-20], 68% with apathy); (3) depressive symptoms profile with predominant anxiety ( n  = 28; median MADRS score = 17 [IQR 9-25]; 100% with anxiety). Compared to patients without depressive symptoms, patients with depressive symptoms and predominant apathy (but not those with predominant anxiety) were more likely to have cerebral atrophy (OR = 2.4, 95% CI = 1.4-4.2) and had significantly higher long-term new-onset dementia risk (adjusted hazard ratio = 2.2, 95% CI = 1.3-3.8).
Conclusion: Screening for apathy and anxiety on top of depressive symptoms might help identifying patients at risk for dementia. Future studies on treatment should account for different post-ICH depressive symptoms profiles that may impact on cognitive function.
Competing Interests: Declaration of conflicting interestsThe author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.