Citations

There is a growing interest in the application of machine learning (ML) in Alzheimer’s disease (AD) research. However, neuropsychiatric symptoms (NPS), frequent in subjects with AD, mild cognitive impairment (MCI), and other related dementias have not been analyzed sufficiently using ML methods. To portray the landscape and potential of ML research in AD and NPS studies, we present a comprehensive literature review of existing ML approaches and commonly studied AD biomarkers. We conducted PubMed searches with keywords related to NPS, AD biomarkers, machine learning, and cognition. We included a total of 38 articles in this review after excluding some irrelevant studies from the search results and including 6 articles based on a snowball search from the bibliography of the relevant studies. We found a limited number of studies focused on NPS with or without AD biomarkers. In contrast, multiple statistical machine learning and deep learning methods have been used to build predictive diagnostic models using commonly known AD biomarkers. These mainly included multiple imaging biomarkers, cognitive scores, and various omics biomarkers. Deep learning approaches that combine these biomarkers or multi-modality datasets typically outperform single-modality datasets. We conclude ML may be leveraged to untangle the complex relationships of NPS and AD biomarkers with cognition. This may potentially help to predict the progression of MCI or dementia and develop more targeted early intervention approaches based on NPS.

AbstractBackgroundPhysical and cognitive activities are associated with decreased risk of mild cognitive impairment and dementia. However, the longitudinal associations between physical and cognitive activities with trajectories of Alzheimer’s disease (AD) neuroimaging biomarkers among older adults free of dementia remain unclear.MethodWe conducted a longitudinal study derived from the population‐based Mayo Clinic Study of Aging, including individuals aged ≥ 50 years who were free of dementia. Participants had information on physical and cognitive activity engagement in midlife (ages 30‐45 for participants aged 50‐69; ages 50‐65 for participants aged ≥ 70) and late‐life (12 months prior to baseline assessment), and AD biomarker assessments. We calculated linear mixed‐effect models to examine the association between baseline physical and cognitive activity composite scores for mid‐ and late‐life, and trajectories for individual yearly change in amyloid deposition (measured by PiB‐PET), tau burden (measured by Tau‐PET), regional glucose hypometabolism (measured by FDG‐PET), cortical thickness (measured by MRI), and white matter hyper intensities (WMH, measured by FLAIR‐MRI). The models were adjusted for age, sex, and APOE ɛ4 carrier status.ResultThe sample included 2755 persons (52% males; 2462 cognitively unimpaired, 293 with MCI). The mean [SD] age was 72.7 [9.8] years; 774 participants were APOE ɛ4 carriers. Although, overall, participants showed an increase in amyloid deposition, tau burden, WMH, and a decrease in glucose metabolism and cortical thickness over time, those with higher late‐life physical activity experienced less pronounced increase in Tau‐PET SUVR over time [estimate for late‐life physical activity with time interaction: ‐0.002; 95% CI ‐0.004, ‐0.000; p = 0.034]. In addition, participants with higher cognitive activities in midlife [est. for midlife cognitive activity with time interaction: 0.001; 95% CI 0.000, 0.003; p = 0.026] and late‐life [est. for late‐life cognitive activity with time interaction 0.002; 95% CI 0.001, 0.003; p = 0.005] had less pronounced decrease in FDG‐PET SUVR over time. There were no further significant interactions.ConclusionPreliminary findings suggest significant associations between 1) late‐life physical activity and less tauopathy, and 2) cognitive activities in mid‐ and late‐life with less synaptic dysfunction over time. Further research is needed to validate study findings.

ImportanceIndividuals with total joint arthroplasty (TJA) have long-term exposure to metal-containing implants; however, whether long-term exposure to artificial implants is associated with cognitive function is unknown.ObjectiveTo compare long-term cognitive trajectories in individuals with and without TJA.Design, Setting, and ParticipantsThis population-based cohort study assessed serial cognitive evaluations of 5550 participants (≥50 years of age) from the Mayo Clinic Study of Aging between November 1, 2004, and December 31, 2020.ExposuresTotal joint arthroplasty of the hip or the knee.Main Outcomes and MeasuresLinear mixed-effects models were used to compare the annualized rate of change in global and domain-specific cognitive scores in participants with and without TJA, adjusting for age, sex, educational level, apolipoprotein E ε4 carrier status, and cognitive test practice effects.ResultsA total of 5550 participants (mean [SD] age at baseline, 73.04 [10.02] years; 2830 [51.0%] male) were evaluated. A total of 952 participants had undergone at least 1 TJA of the hip (THA, n = 430) or the knee (TKA, n = 626) before or after entry into the cohort. Participants with TJA were older, more likely to be female, and had a higher body mass index than participants without TJA. No difference was observed in the rate of cognitive decline in participants with and without TJA until 80 years of age. A slightly faster cognitive decline at 80 years or older and more than 8 years from surgery was observed (b = −0.03; 95% CI, −0.04 to −0.02). In stratified analyses by surgery type, the faster decline was observed primarily among older participants with TKA (b = −0.04; 95% CI, −0.06 to −0.02).Conclusions and RelevanceIn this cohort study, long-term cognitive trajectories in individuals with and without TJA were largely similar except for a slightly faster decline among the oldest patients with TKA; however, the magnitude of difference was small and of unknown clinical significance.

Background: Observational studies suggested that dementia risk in patients with rheumatoid arthritis (RA) is higher than in the general population. Objective: To examine the associations of RA with cognitive decline and dementia, and neuroimaging biomarkers of aging, Alzheimer’s disease, and vascular pathology in adult participants in the Mayo Clinic Study of Aging (MCSA). Methods: Participants with RA were matched 1:3 on age, sex, education, and baseline cognitive diagnosis to participants without RA. RA cases with MRI were also matched with non-cases with available MRI. All available imaging studies (i.e., amyloid and FDG PET, sMRI, and FLAIR) were included. The study included 104 participants with RA and 312 without RA (mean age (standard deviation, SD) 75.0 (10.4) years, 33% male and average follow-up (SD) 4.2 (3.8) years). Results: Groups were similar in cognitive decline and risk of incident dementia. Among participants with neuroimaging, participants with RA (n = 33) and without RA (n = 98) had similar amyloid burden and neurodegeneration measures, including regions sensitive to aging and dementia, but greater mean white matter hyperintensity volume relative to the total intracranial volume (mean (SD)% : 1.12 (0.57)% versus 0.76 (0.69)% of TIV, p = 0.01), and had higher mean (SD) number of cortical infarctions (0.24 (0.44) versus 0.05 (0.33), p = 0.02). Conclusion: Although cognitive decline and dementia risk were similar in participants with and without RA, participants with RA had more abnormal cerebrovascular pathology on neuroimaging. Future studies should examine the mechanisms underlying these changes and potential implications for prognostication and prevention of cognitive decline in RA.

Background: Considering the world’s rapidly increasing life expectancy, with people working and maintaining active lifestyles longer than ever before, addressing the effects of aging on cognition is of utmost importance. A greater understanding of cognitive aging may also be critical in distinguishing natural cognitive aging from pre-clinical stages of Alzheimer’s disease and related cognitive disorders. Objective: To systematically examine the association between aging and cognitive performance in a cognitively and otherwise healthy probability population-based sample using a computer-based method. Methods: This cross-sectional study enrolled 673 cognitively and otherwise healthy participants aged 25–89 years (mean age 52.3±14.2 years, 52.5% of whom were female) from the Kardiovize study cohort. Mild cognitive impairment and dementia cases were excluded, followed by measurement of cognitive performance with the computer-administered Cogstate Brief Battery. We used ANCOVA and Modified Signed-Likelihood Ratio tests to examine patterns of cognition across age groups. Results: We found a gradual decrease in cognitive performance across the lifespan, which required two decades to demonstrate significant changes. In contrast to attention and learning, psychomotor speed and working memory showed the most significant age-related decrease and variability in performance. The established pattern of cognitive aging was not altered by sex or education. Conclusion: These findings corroborate, validate, and extend the current understanding of natural cognitive aging and pinpoint specific cognitive domains with the most extensive age-related interindividual differences. This will contribute to the development of strategies to preserve cognition with aging and may also serve to improve early diagnostics of cognitive disorders using computer-based methods.

Dementia is one of the most prevalent health problems in the aging population. Despite the significant number of people affected, dementia diagnoses are often significantly delayed, missing opportunities to maximize life quality. Early identification of older adults at high risk for dementia may help to maximize current quality of life and to improve planning for future health needs in dementia patients. However, most existing risk prediction models predominantly use static variables, not considering temporal patterns of health status. This study used an attention-based time-aware model to predict incident dementia that incorporated longitudinal temporal health conditions. The predictive performance of the time-aware model was compared with three traditional models using static variables and demonstrated higher predictive power.