Effect of trazodone on cognitive function in people with dementia: Cohort study using UK routinely collected data

Andrew Sommerlad, Nomi Werbeloff, Gayan Perera, Tanya Smith, Harry Costello, Christoph Mueller, Andrey Kormilitzin, Matthew Broadbent, Alejo J. Nevado-Holgado, Simon Lovestone, Robert Stewart, Gill Livingstone

Research output: Contribution to journalArticlepeer-review


The antidepressant trazodone has been found in recent evidence from mouse neurodegenerative disease models to be neuroprotective, leading to interest in whether trazodone could be repurposed to improve cognition in people with dementia. We aimed to examine trazodone’s effect using routinely collected clinical data, by comparing cognitive decline of people with dementia taking trazodone with those taking other antidepressants.

We conducted three identical naturalistic pharmaco‐epidemiological cohort studies using large secondary mental healthcare databases in North London, South London, and Oxford, covering around 2% of the UK’s population. We included all people with dementia who were recorded between 2008‐16 taking trazodone, citalopram or mirtazapine for at least 6 weeks and had cognition assessed by mini‐mental state examination (MMSE). Linear mixed models examined age, time and sex‐adjusted MMSE change in people with all‐cause dementia taking trazodone compared with those taking citalopram and mirtazapine. Sensitivity analyses examined those with non‐vascular dementia; mild dementia; and adjusted results for neuropsychiatric symptoms. Results from the three study sites were combined using random‐effects meta‐analysis.

We included 2,199 people with dementia, including 406 taking trazodone, with mean 2.2 years follow‐up. There was no difference in adjusted cognitive decline in people with all‐cause or non‐vascular dementia taking trazodone, citalopram or mirtazapine in any of the three study sites. Meta‐analysed results from the three sites indicated greater mean MMSE decline in people with all‐cause dementia taking trazodone compared to those taking citalopram (0.26 points per successive MMSE measurement, 95% CI 0.03 to 0.49; p=0.03). Pooled MMSE change was not different for those taking trazodone v mirtazapine. The main limitation of our study design was the potential for confounding by indication, but results in sensitivity analyses adjusting for neuropsychiatric symptoms and analysing only those with mild dementia were consistent with the primary analysis, suggesting that confounding by indication did not explain results.

There was no evidence of cognitive benefit from taking trazodone compared to citalopram or mirtazapine in people with established dementia in these three separate naturalistic cohort studies or when results were combined. Despite the preclinical evidence, trazodone should not be advocated for cognition in dementia.
Original languageEnglish
Pages (from-to)e039090
Number of pages2
JournalAlzheimer's & Dementia
Volume16 (Supple.10)
Publication statusPublished - 2020


Dive into the research topics of 'Effect of trazodone on cognitive function in people with dementia: Cohort study using UK routinely collected data'. Together they form a unique fingerprint.

Cite this