Brain‐age predicts subsequent dementia in memory clinic patients: Neuroimaging / Optimal neuroimaging measures for early detection. (7th December 2020)

• Record Type: Journal Article
• Authors: Biondo, Francesca; Jewell, Amelia; Pritchard, Megan; Mueller, Christoph; Steves, Claire J; Cole, James
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.037378
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: The lack of disease‐modifying therapies in the current dementia epidemic emphasises the importance of targeting earlier disease stages. Cognitive measures alone do not confer sufficient sensitivity and specificity for this, motivating research into neuroimaging biomarkers that might detect more subtle early changes. One promising measure, so‐called 'brain age', has previously been associated with neurological and psychiatric disorders, and predictive of future health outcomes. This includes the development of dementia in observational research cohorts (Gaser et al., 2013; Wang et al., 2019). Here we present results from a community sample of people referred to memory clinics (the Brain Ageing and Risk of Cognitive Decline [BARCODE] project), determining whether baseline brain‐age in people with subject or mild cognitive impairment predicted subsequent dementia diagnosis. Method: We used T1‐weighted MRI n=318 memory clinic patients with subjective or mild cognitive impairment (192 female and mean [SD] age of 76.3 [0.5] years). Using in‐house software ( https://github.com/james‐cole/brainageR) we calculated brain‐predicted age difference (brain‐PAD) for each patients. We determined which participants went on to develop dementia via linkage to electronic health records. Logistic regression analysis tested whether brain‐PAD at time of scan predicted a subsequent dementia diagnosis, controlling for age and sex. Result: Out of 318 participants, n=248 subsequently received a dementia diagnosis. Logistic regression (McFadden's R 2 =0.19) indicated that brain‐PAD scores at time of scan were significantly higher in patients who subsequently received a dementia diagnosis, B ( SE )=0.04(0.02), p <0.05, OR 1.04 (95% CI 1.00‐1.08). Conclusion: Structural brain ageing is detectable in older adults with memory impairments. Critically, memory clinic patients with older‐looking brains are more likely to receive a subsequent dementia diagnosis. This illustrates the value of using brain age as a sensitive biomarker that has the potential to be used early‐on in memory clinics to detect patients at high‐risk of developing dementia.