Data harmonization and federated learning for multi-cohort dementia research using the OMOP common data model: A Netherlands consortium of dementia cohorts case study.

BACKGROUND: Establishing collaborations between cohort studies has been fundamental for progress in health research. However, such collaborations are hampered by heterogeneous data representations across cohorts and legal constraints to data sharing. The first arises from a lack of consensus in standards of data collection and representation across cohort studies and is usually tackled by applying data harmonization processes. The second is increasingly important due to raised awareness for privacy protection and stricter regulations, such as the GDPR. Federated learning has emerged as a privacy-preserving alternative to transferring data between institutions through analyzing data in a decentralized manner. METHODS: In this study, we set up a federated learning infrastructure for a consortium of nine Dutch cohorts with appropriate data available to the etiology of dementia, including an extract, transform, and load (ETL) pipeline for data harmonization. Additionally, we assessed the challenges of transforming and standardizing cohort data using the Observational Medical Outcomes Partnership (OMOP) common data model (CDM) and evaluated our tool in one of the cohorts employing federated algorithms. RESULTS: We successfully applied our ETL tool and observed a complete coverage of the cohorts' data by the OMOP CDM. The OMOP CDM facilitated the data representation and standardization, but we identified limitations for cohort-specific data fields and in the scope of the vocabularies available. Specific challenges arise in a multi-cohort federated collaboration due to technical constraints in local environments, data heterogeneity, and lack of direct access to the data. CONCLUSION: In this article, we describe the solutions to these challenges and limitations encountered in our study. Our study shows the potential of federated learning as a privacy-preserving solution for multi-cohort studies that enhance reproducibility and reuse of both data and analyses.

Visual associative learning to detect early episodic memory deficits and distinguish Alzheimer's disease from other types of dementia.

OBJECTIVE: We investigated how well a visual associative learning task discriminates Alzheimer's disease (AD) dementia from other types of dementia and how it relates to AD pathology. METHODS: 3,599 patients (63.9 ± 8.9 years old, 41% female) from the Amsterdam Dementia Cohort completed two sets of the Visual Association Test (VAT) in a single test session and underwent magnetic resonance imaging. We performed receiver operating curve analysis to investigate the VAT's discriminatory ability between AD dementia and other diagnoses and compared it to that of other episodic memory tests. We tested associations between VAT performance and medial temporal lobe atrophy (MTA), and amyloid status (n = 2,769, 77%). RESULTS: Patients with AD dementia performed worse on the VAT than all other patients. The VAT discriminated well between AD and other types of dementia (area under the curve range 0.70-0.86), better than other episodic memory tests. Six-hundred forty patients (17.8%) learned all associations on VAT-A, but not on VAT-B, and they were more likely to have higher MTA scores (odds ratios range 1.63 (MTA 0.5) through 5.13 for MTA ≥ 3, all p < .001) and to be amyloid positive (odds ratio = 3.38, 95%CI = [2.71, 4.22], p < .001) than patients who learned all associations on both sets. CONCLUSIONS: Performance on the VAT, especially on a second set administered immediately after the first, discriminates AD from other types of dementia and is associated with MTA and amyloid positivity. The VAT might be a useful, simple tool to assess early episodic memory deficits in the presence of AD pathology.

Contributions of amyloid beta and cerebral small vessel disease in clinical decline.

INTRODUCTION: We assessed whether co-morbid small vessel disease (SVD) has clinical predictive value in preclinical or prodromal Alzheimer's disease. METHODS: In 1090 non-demented participants (65.4 ± 10.7 years) SVD was assessed with magnetic resonance imaging and amyloid beta (Aß) with lumbar puncture and/or positron emission tomography scan (mean follow-up for cognitive function 3.1 ± 2.4 years). RESULTS: Thirty-nine percent had neither Aß nor SVD (A-V-), 21% had SVD only (A-V+), 23% Aß only (A+V-), and 17% had both (A+V+). Pooled cohort linear mixed model analyses demonstrated that compared to A-V- (reference), A+V- had a faster rate of cognitive decline. Co-morbid SVD (A+V+) did not further increase rate of decline. Cox regression showed that dementia risk was modestly increased in A-V+ (hazard ratio [95% confidence interval: 1.8 [1.0-3.2]) and most strongly in A+ groups. Also, mortality risk was increased in A+ groups. DISCUSSION: In non-demented persons Aß was predictive of cognitive decline, dementia, and mortality. SVD modestly predicts dementia in A-, but did not increase deleterious effects in A+. HIGHLIGHTS: Amyloid beta (Aß; A) was predictive for cognitive decline, dementia, and mortality. Small vessel disease (SVD) had no additional deleterious effects in A+. SVD modestly predicted dementia in A-. Aß should be assessed even when magnetic resonance imaging indicates vascular cognitive impairment.

Genetically identical twin-pair difference models support the amyloid cascade hypothesis.

The amyloid cascade hypothesis has strongly impacted the Alzheimer's disease research agenda and clinical trial designs over the past decades, but precisely how amyloid-ß pathology initiates the aggregation of neocortical tau remains unclear. We cannot exclude the possibility of a shared upstream process driving both amyloid-ß and tau in an independent manner instead of there being a causal relationship between amyloid-ß and tau. Here, we tested the premise that if a causal relationship exists, then exposure should be associated with outcome both at the individual level as well as within identical twin-pairs, who are strongly matched on genetic, demographic and shared environmental background. Specifically, we tested associations between longitudinal amyloid-ß PET and cross-sectional tau PET, neurodegeneration and cognitive decline using genetically identical twin-pair difference models, which provide the unique opportunity of ruling out genetic and shared environmental effects as potential confounders in an association. We included 78 cognitively unimpaired identical twins with [18F]flutemetamol (amyloid-ß)-PET, [18F]flortaucipir (tau)-PET, MRI (hippocampal volume) and cognitive data (composite memory). Associations between each modality were tested at the individual level using generalized estimating equation models, and within identical twin-pairs using within-pair difference models. Mediation analyses were performed to test for directionality in the associations as suggested by the amyloid cascade hypothesis. At the individual level, we observed moderate-to-strong associations between amyloid-ß, tau, neurodegeneration and cognition. The within-pair difference models replicated results observed at the individual level with comparably strong effect sizes. Within-pair differences in amyloid-ß were strongly associated with within-pair differences in tau (ß = 0.68, P < 0.001), and moderately associated with within-pair differences in hippocampal volume (ß = -0.37, P = 0.03) and memory functioning (ß = -0.57, P < 0.001). Within-pair differences in tau were moderately associated with within-pair differences in hippocampal volume (ß = -0.53, P < 0.001) and strongly associated with within-pair differences in memory functioning (ß = -0.68, P < 0.001). Mediation analyses showed that of the total twin-difference effect of amyloid-ß on memory functioning, the proportion mediated through pathways including tau and hippocampal volume was 69.9%, which was largely attributable to the pathway leading from amyloid-ß to tau to memory functioning (proportion mediated, 51.6%). Our results indicate that associations between amyloid-ß, tau, neurodegeneration and cognition are unbiased by (genetic) confounding. Furthermore, effects of amyloid-ß on neurodegeneration and cognitive decline were fully mediated by tau. These novel findings in this unique sample of identical twins are compatible with the amyloid cascade hypothesis and thereby provide important new knowledge for clinical trial designs.

Abnormal cerebrospinal fluid levels of amyloid and tau are associated with cognitive decline over time in cognitively normal older adults: A monozygotic twin study.

Introduction: The contribution of genetic and environmental factors to the relation between cerebrospinal fluid (CSF) biomarkers and cognitive decline in preclinical Alzheimer's disease remains unclear. We studied this in initially cognitively normal monozygotic twins. Methods: We included 122 cognitively normal monozygotic twins (51 pairs) with a follow-up of 4.3 ± 0.4 years. We first tested associations of baseline CSF Aß1-42/1-40 ratio, total tau (t-tau), and 181-phosphorylated-tau (p-tau) status with subsequent cognitive decline using linear mixed models, and then performed twin specific analyses. Results: Baseline abnormal amyloid-ß and tau CSF markers predicted steeper decline on memory (p ≤ .003) and language (p ≤ 0.04). Amyloid-ß and p-tau markers in one twin predicted decline in memory in the co-twin and tau markers in one twin predicted decline in language in the co-twin (r range -0.26,0.39; p's ≤ .02). Discussion: These results suggest that memory and language decline are early features of AD that are in part determined by the same genetic factors that influence amyloid-ß and tau regulation.

Genetically identical twins show comparable tau PET load and spatial distribution.

Tau accumulation starts during the preclinical phase of Alzheimer's disease and is closely associated with cognitive decline. For preventive purposes, it is important to identify factors associated with tau accumulation and spread. Studying genetically identical twin-pairs may give insight into genetic and environmental contributions to tau pathology, as similarities in identical twin-pairs largely result from genetic factors, while differences in identical twin-pairs can largely be attributed to non-shared, environmental factors. This study aimed to examine similarities and dissimilarities in a cohort of genetically identical older twin-pairs in (i) tau load; and (ii) spatial distribution of tau, measured with 18F-flortaucipir PET. We selected 78 genetically identical twins (39 pairs; average age 73 ± 6 years), enriched for amyloid-ß pathology and APOE ε4 carriership, who underwent dynamic 18F-flortaucipir PET. We extracted binding potentials (BPND) in entorhinal, temporal, widespread neocortical and global regions, and examined within-pair similarities in BPND using age and sex corrected intra-class correlations. Furthermore, we tested whether twin-pairs showed a more similar spatial 18F-flortaucipir distribution compared to non-twin pairs, and whether the participant's co-twin could be identified solely based on the spatial 18F-flortaucipir distribution. Last, we explored whether environmental (e.g. physical activity, obesity) factors could explain observed differences in twins of a pair in 18F-flortaucipir BPND. On visual inspection, Alzheimer's disease-like 18F-flortaucipir PET patterns were observed, and although we mainly identified similarities in twin-pairs, some pairs showed strong dissimilarities. 18F-flortaucipir BPND was correlated in twins in the entorhinal (r = 0.40; P = 0.01), neocortical (r = 0.59; P < 0.01) and global (r = 0.56; P < 0.01) regions, but not in the temporal region (r = 0.20; P = 0.10). The 18F-flortaucipir distribution pattern was significantly more similar between twins of the same pair [mean r = 0.27; standard deviation (SD) = 0.09] than between non-twin pairings of participants (mean r = 0.01; SD = 0.10) (P < 0.01), also after correcting for proxies of off-target binding. Based on the spatial 18F-flortaucipir distribution, we could identify with an accuracy of 86% which twins belonged to the same pair. Finally, within-pair differences in 18F-flortaucipir BPND were associated with within-pair differences in depressive symptoms (0.37 < ß < 0.56), physical activity (-0.41 < ß < -0.42) and social activity (-0.32 < ß < -0.36) (all P < 0.05). Overall, identical twin-pairs were comparable in tau load and spatial distribution, highlighting the important role of genetic factors in the accumulation and spreading of tau pathology. Considering also the presence of dissimilarities in tau pathology in identical twin-pairs, our results additionally support a role for (potentially modifiable) environmental factors in the onset of Alzheimer's disease pathological processes, which may be of interest for future prevention strategies.