Understanding preclinical dementia : early detection of dementia through cognitive and biological markers
Author: Payton, Nicola Maria
Date: 2020-10-08
Location: Eva & Georg Klein, Solnavägen 9, Karolinska Institutet, Solna
Time: 13.30
Department: Inst för neurobiologi, vårdvetenskap och samhälle / Dept of Neurobiology, Care Sciences and Society
View/ Open:
Thesis (1.665Mb)
Abstract
Dementia is becoming a growing healthcare crisis, therefore identifying individuals at risk or in the earliest stages of dementia is essential if prevention or disease modification is to be achieved. The objective of this thesis was to examine cognitive performance and decline during the preclinical phase and explore the ability of cognitive and biological markers to identify those at risk of future dementia. Data from a population-based longitudinal study, SNAC-K, were used to investigate this aim.
Study I examined the ability of neuropsychological tests, genetics, and structural MRI volumes to predict dementia six years later. Models were systematically created to identify the best combinations for prediction. A model containing all three modalities: hippocampal volume, a task of category fluency, presence of an APOE ɛ4 allele, white-matter hyperintensities volume, and a task of general knowledge, displayed the most predictive value (AUC=.924; C.I=.883–.965). However, this model did not significantly improve predictive value over one containing only cognitive and genetic markers, suggesting that minor increases in predictivity should be weighed against the costs of additional tests.
Study II investigated the benefit of DTI, alongside neuropsychological tests, genetics, and brain volume markers in predicting future dementia. MD values for tracts CHC, CS, FMAJ, and IFOF (AUC=.837– .862) and the FA IFOF latent factor (AUC=.839) were significantly associated with dementia at six years. A final model consisting of a measure of perceptual speed, hippocampal volume, and MD of the FMAJ tract was created with the highest predictive value (AUC=.911). Assessment of microstructural white matter integrity via DTI was associated with future dementia but the additional benefit when combined with other markers was relatively small.
Study III narrowed its focus to the ability of cognitive markers alone and the effect of modifying factors (age, sex, education, the presence of an ɛ4 allele, AD–only dementia, and time to diagnosis) on identifying those at risk of dementia. The most predictive model, consisting of category fluency, word recall, and pattern comparison, achieved good prediction values (AUC=.913) for dementia six years later. Tests in the domains of category fluency, episodic memory, and perceptual speed were, in general, good predictors across all subgroups and up to 6 years before a dementia diagnosis. However, cognitive tests became increasingly unreliable at predicting dementia beyond that time.
Study IV explored the trajectories of cognitive decline over a 12-year period during the preclinical stage of dementia, before examining the ability of early cognitive decline in identifying those with increased likelihood of future dementia. Persons in the preclinical phase showed increased rate of decline in all cognitive domains compared to those who did not develop dementia (β:-.07 to -.11), this difference was particularly noticeable closer to diagnosis. Those classified as fast decliners for 3 or more cognitive tests demonstrated the highest risk of dementia (HR: 3.38, CI: 1.91-6.01). Although, changes in early rates of decline were small and rates of decline may be more predictive closer to diagnosis.
Collectively, these studies confirm a long preclinical period in dementia development, which allows for the use of a wide range of markers (cognitive, genetic, MRI, and DTI) capable of identifying those at high risk of dementia. The ability of these markers to predict future dementia is increased through combining within and between modalities.
Study I examined the ability of neuropsychological tests, genetics, and structural MRI volumes to predict dementia six years later. Models were systematically created to identify the best combinations for prediction. A model containing all three modalities: hippocampal volume, a task of category fluency, presence of an APOE ɛ4 allele, white-matter hyperintensities volume, and a task of general knowledge, displayed the most predictive value (AUC=.924; C.I=.883–.965). However, this model did not significantly improve predictive value over one containing only cognitive and genetic markers, suggesting that minor increases in predictivity should be weighed against the costs of additional tests.
Study II investigated the benefit of DTI, alongside neuropsychological tests, genetics, and brain volume markers in predicting future dementia. MD values for tracts CHC, CS, FMAJ, and IFOF (AUC=.837– .862) and the FA IFOF latent factor (AUC=.839) were significantly associated with dementia at six years. A final model consisting of a measure of perceptual speed, hippocampal volume, and MD of the FMAJ tract was created with the highest predictive value (AUC=.911). Assessment of microstructural white matter integrity via DTI was associated with future dementia but the additional benefit when combined with other markers was relatively small.
Study III narrowed its focus to the ability of cognitive markers alone and the effect of modifying factors (age, sex, education, the presence of an ɛ4 allele, AD–only dementia, and time to diagnosis) on identifying those at risk of dementia. The most predictive model, consisting of category fluency, word recall, and pattern comparison, achieved good prediction values (AUC=.913) for dementia six years later. Tests in the domains of category fluency, episodic memory, and perceptual speed were, in general, good predictors across all subgroups and up to 6 years before a dementia diagnosis. However, cognitive tests became increasingly unreliable at predicting dementia beyond that time.
Study IV explored the trajectories of cognitive decline over a 12-year period during the preclinical stage of dementia, before examining the ability of early cognitive decline in identifying those with increased likelihood of future dementia. Persons in the preclinical phase showed increased rate of decline in all cognitive domains compared to those who did not develop dementia (β:-.07 to -.11), this difference was particularly noticeable closer to diagnosis. Those classified as fast decliners for 3 or more cognitive tests demonstrated the highest risk of dementia (HR: 3.38, CI: 1.91-6.01). Although, changes in early rates of decline were small and rates of decline may be more predictive closer to diagnosis.
Collectively, these studies confirm a long preclinical period in dementia development, which allows for the use of a wide range of markers (cognitive, genetic, MRI, and DTI) capable of identifying those at high risk of dementia. The ability of these markers to predict future dementia is increased through combining within and between modalities.
List of papers:
I. Payton, N. M., Kalpouzos, G., Rizzuto, D., Fratiglioni, L., Kivipelto, M., Bäckman, L., & Laukka, E. J. (2018). Combining Cognitive, Genetic, and Structural Neuroimaging Markers to Identify Individuals with Increased Dementia Risk. Journal of Alzheimer's disease. 64(2), 533–542.
Fulltext (DOI)
Pubmed
View record in Web of Science®
II. Müller, T., Payton, N. M., Kalpouzos, G., Jessen, F., Grande, G., Bäckman, L., & Laukka, E. J. (2020). Cognitive, genetic, brain volume, and diffusion tensor imaging markers as early indicators of dementia. [Accepted]
Fulltext (DOI)
Pubmed
III. Payton, N. M., Rizzuto, D., Fratiglioni, L., Kivipelto, M., Bäckman, L., & Laukka, E. J. (2020). Combining Cognitive Markers to Identify Individuals at Increased Dementia Risk: Influence of Modifying Factors and Time to Diagnosis. Journal of the International Neuropsychological Society. 26(8), 785–797.
Fulltext (DOI)
Pubmed
IV. Payton, N. M., Marseglia, A., Grande, G., Fratiglioni, L., Kivipelto, M., Bäckman, L., & Laukka, E. J. Trajectories of preclinical cognitive decline and dementia development: A 12-year longitudinal study. [Manuscript]
I. Payton, N. M., Kalpouzos, G., Rizzuto, D., Fratiglioni, L., Kivipelto, M., Bäckman, L., & Laukka, E. J. (2018). Combining Cognitive, Genetic, and Structural Neuroimaging Markers to Identify Individuals with Increased Dementia Risk. Journal of Alzheimer's disease. 64(2), 533–542.
Fulltext (DOI)
Pubmed
View record in Web of Science®
II. Müller, T., Payton, N. M., Kalpouzos, G., Jessen, F., Grande, G., Bäckman, L., & Laukka, E. J. (2020). Cognitive, genetic, brain volume, and diffusion tensor imaging markers as early indicators of dementia. [Accepted]
Fulltext (DOI)
Pubmed
III. Payton, N. M., Rizzuto, D., Fratiglioni, L., Kivipelto, M., Bäckman, L., & Laukka, E. J. (2020). Combining Cognitive Markers to Identify Individuals at Increased Dementia Risk: Influence of Modifying Factors and Time to Diagnosis. Journal of the International Neuropsychological Society. 26(8), 785–797.
Fulltext (DOI)
Pubmed
IV. Payton, N. M., Marseglia, A., Grande, G., Fratiglioni, L., Kivipelto, M., Bäckman, L., & Laukka, E. J. Trajectories of preclinical cognitive decline and dementia development: A 12-year longitudinal study. [Manuscript]
Institution: Karolinska Institutet
Supervisor: Laukka, Erika Jonsson
Co-supervisor: Bäckman, Lars; Fratiglioni, Laura; Kivipelto, Miia
Issue date: 2020-09-14
Rights:
Publication year: 2020
ISBN: 978-91-7831-912-1
Statistics
Total Visits
Views | |
---|---|
Understanding ... | 874 |
Total Visits Per Month
September 2023 | October 2023 | November 2023 | December 2023 | January 2024 | February 2024 | March 2024 | |
---|---|---|---|---|---|---|---|
Understanding ... | 22 | 35 | 17 | 18 | 20 | 18 | 23 |
File Visits
Views | |
---|---|
Thesis_Nicola_Maria_Payton.pdf | 543 |
Thesis_Nicola_Maria_Payton.pdf | 2 |
Top country views
Views | |
---|---|
Sweden | 222 |
United States | 141 |
Denmark | 90 |
Ireland | 84 |
United Kingdom | 53 |
China | 47 |
Germany | 28 |
France | 26 |
Netherlands | 16 |
Canada | 11 |
Top cities views
Views | |
---|---|
Dublin | 82 |
Stockholm | 33 |
Ashburn | 19 |
Hangzhou | 19 |
Huddinge | 16 |
San Mateo | 16 |
Lund | 14 |
Sundbyberg | 11 |
Amsterdam | 8 |
Malmo | 8 |