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.

Microvascular Dysfunction and Whole-Brain White Matter Connectivity: The Maastricht Study.

BACKGROUND: Microvascular dysfunction is involved in the development of various cerebral disorders. It may contribute to these disorders by disrupting white matter tracts and altering brain connectivity, but evidence is scarce. We investigated the association between multiple biomarkers of microvascular function and whole-brain white matter connectivity. METHODS AND RESULTS: Cross-sectional data from The Maastricht Study, a Dutch population-based cohort (n=4326; age, 59.4±8.6 years; 49.7% women). Measures of microvascular function included urinary albumin excretion, central retinal arteriolar and venular calibers, composite scores of flicker light-induced retinal arteriolar and venular dilation, and plasma biomarkers of endothelial dysfunction (intercellular adhesion molecule-1, vascular cell adhesion molecule-1, E-selectin, and von Willebrand factor). White matter connectivity was calculated from 3T diffusion magnetic resonance imaging to quantify the number (average node degree) and organization (characteristic path length, global efficiency, clustering coefficient, and local efficiency) of white matter connections. A higher plasma biomarkers of endothelial dysfunction composite score was associated with a longer characteristic path length (ß per SD, 0.066 [95% CI, 0.017-0.114]) after adjustment for sociodemographic, lifestyle, and cardiovascular factors but not with any of the other white matter connectivity measures. After multiple comparison correction, this association was nonsignificant. None of the other microvascular function measures were associated with any of the connectivity measures. CONCLUSIONS: These findings suggest that microvascular dysfunction as measured by indirect markers is not associated with whole-brain white matter connectivity.

Associations between HbA1c Reduction and Change in Depressive Symptoms following Glucose-lowering Treatment in Adults: A Systematic Review of Intervention Studies.

INTRODUCTION: Hyperglycemia constitutes a likely pathway linking diabetes and depressive symptoms; lowering glycemic levels may help reduce diabetes-comorbid depressive symptoms. Since randomized controlled trials can help understand temporal associations, we systematically reviewed the evidence regarding the potential association of hemoglobin HbA1c lowering interventions with depressive symptoms. METHODS: PubMed, PsycINFO, CINAHL, and EMBASE databases were searched for randomized controlled trials evaluating HbA1c-lowering interventions and including assessment of depressive symptoms published between 01/2000-09/2020. Study quality was evaluated using the Cochrane Risk of Bias tool. PROSPERO registration: CRD42020215541. RESULTS: We retrieved 1,642 studies of which twelve met our inclusion criteria. Nine studies had high and three unclear risks of bias. Baseline depressive symptom scores suggest elevated depressive symptoms in five studies. Baseline HbA1c was <8.0% (<64 mmol/mol) in two, 8.0-9.0% (64-75 mmol/mol) in eight, and ≥10.0% (≥86 mmol/mol) in two studies. Five studies found greater HbA1c reduction in the treatment group; three of these found greater depressive symptom reduction in the treatment group. Of four studies analyzing whether the change in HbA1c was associated with the change in depressive symptoms, none found a significant association. The main limitation of these studies was relatively low levels of depressive symptoms at baseline, limiting the ability to show a lowering in depressive symptoms after HbA1c reduction. CONCLUSIONS: We found insufficient available data to estimate the association between HbA1c reduction and depressive symptom change following glucose-lowering treatment. Our findings point to an important gap in the diabetes treatment literature. Future clinical trials testing interventions to improve glycemic outcomes might consider measuring depressive symptoms as an outcome to enable analyses of this association.

Bear in mind: the role of personal background in semantic animal fluency - The SMART-MR study.

Objectives: Semantic fluency is a prominent neuropsychological task, typically administered within the category 'animals'. With the increasing development of novel item-level metrics of semantic fluency, a concern around the validity of item-level analyses could be that personal background factors (e.g., hobbies like birdwatching or fishing) may disproportionally influence performance. We analyzed animal fluency performance at the item level and investigated the prevalence of individuals with abundant knowledge in specific classes of animals (e.g., birds, fish, insects) and the relationship of such knowledge with personal background factors and other cognitive tasks (episodic memory and executive functioning). Method: Participants included 736 Dutch middle-aged to older adults from the SMART-MR cohort (mean age 58 ± 9.4 years, 18% women). Individuals were asked to name as many animals as possible for 2 min. Number of people with abundant animal class knowledge was calculated for the ability to recall a series of minimum ≥5 and up to ≥15 animals within a specific class with at most one interruption by an animal from another class. Subsequent analyses to investigate relationships of abundant class knowledge with sociodemographic characteristics (t-tests and chi-square tests) and cognitive performance (linear regressions) were performed for a cut-off of ≥10 animals within a specific class (90th percentile), with a sensitivity analysis for ≥7 animals (67th percentile). Results: A total of 416 (56.2%) participants recalled a series of ≥5 animals from a specific class, 245 (33.3%) participants recalled ≥7, 78 (10.6%) participants recalled ≥10, and 8 (1.1%) participants recalled ≥15. Those who recalled a series of at least 10 animals within a class were older, more often men, and more often retired than those who did not. Moreover, they had a higher total score on animal fluency, letter fluency (i.e., executive functioning), and episodic memory tasks compared to those who did not. Discussion: Our results suggest that the benefit of abundant animal class knowledge gained by personal background does not disproportionally influence animal fluency performance as individuals with such knowledge also performed better on other cognitive tasks unrelated to abundant knowledge of animal classes.

What Drives Task Performance in Animal Fluency in Individuals Without Dementia? The SMART-MR Study.

PURPOSE: In this study, we aim to understand whether and how performance in animal fluency (i.e., total correct word count) relates to linguistic levels and/or executive functions by looking at sequence information and item-level metrics (i.e., clusters, switches, and word properties). METHOD: Seven hundred thirty-one Dutch-speaking individuals without dementia from the Second Manifestations of ARTerial disease-Magnetic Resonance study responded to an animal fluency task (120 s). We obtained cluster size and number of switches for the task, and eight different word properties for each correct word produced. We detected variables that determine total word count with random forests, and used conditional inference trees to assess points along the scales of such variables, at which total word count changes significantly. RESULTS: Number of switches, average cluster size, lexical decision response times, word frequency, and concreteness determined total correct word count in animal fluency. People who produced more correct words produced more switches and bigger clusters. People who produced fewer words produced fewer switches and more frequent words. CONCLUSIONS: Concurrent with existing literature, individuals without dementia rely on language and executive functioning to produce words in animal fluency. The novelty of our work is that such results were shown based on a data-driven approach using sequence information and item-level metrics. SUPPLEMENTAL MATERIAL: https://doi.org/10.23641/asha.23713269.

The cross-sectional association between amyloid burden and white matter hyperintensities in older adults without cognitive impairment: A systematic review and meta-analysis.

Alzheimer's disease (AD) is the most common cause of dementia, characterized by the aggregation of amyloid-beta (Aß) proteins into plaques. Individuals with AD frequently show mixed pathologies, often caused by cerebral small vessel disease (CSVD), resulting in lesions such as white matter hyperintensities (WMH). The current systematic review and meta-analysis investigated the cross-sectional relationship between amyloid burden and WMH in older adults without objective cognitive impairment. A systematic search performed in PubMed, Embase, and PsycINFO yielded 13 eligible studies. Aß was assessed using PET, CSF, or plasma measurements. Two meta-analyses were performed: one on Cohen's d metrics and one on correlation coefficients. The meta-analyses revealed an overall weighted small-to-medium Cohen's d of 0.55 (95% CI: 0.31-0.78) in CSF, an overall correlation of 0.31 (0.09-0.50) in CSF, and a large Cohen's d of 0.96 (95% CI: 0.66-1.27) in PET. Only two studies assessed this relationship in plasma, with an effect size of - 0.20 (95% CI: -0.75 to 0.34). These findings indicate a relationship between both amyloid and vascular pathologies in cognitively normal adults in PET and CSF. Future studies should assess the possible relationship of blood amyloid-beta and WMH for broader identification of at risk individuals showing mixed pathology in preclinical stages.

The bidirectional longitudinal association between depressive symptoms and HbA1c : A systematic review and meta-analysis.

AIM: To investigate whether there is a bidirectional longitudinal association of depression with HbA1c . METHODS: We conducted a systematic literature search in PubMed, PsycINFO, CINAHL and EMBASE for observational, longitudinal studies published from January 2000 to September 2020, assessing the association between depression and HbA1c in adults. We assessed study quality with the Newcastle-Ottawa-Scale. Pooled effect estimates were reported as partial correlation coefficients (rp ) or odds ratios (OR). RESULTS: We retrieved 1642 studies; 26 studies were included in the systematic review and eleven in the meta-analysis. Most studies (16/26) focused on type 2 diabetes. Study quality was rated as good (n = 19), fair (n = 2) and poor (n = 5). Of the meta-analysed studies, six investigated the longitudinal association between self-reported depressive symptoms and HbA1c and five the reverse longitudinal association, with a combined sample size of n = 48,793 and a mean follow-up of 2 years. Higher levels of baseline depressive symptoms were associated with subsequent higher levels of HbA1c (partial r = 0.07; [95% CI 0.03, 0.12]; I2 38%). Higher baseline HbA1c values were also associated with 18% increased risk of (probable) depression (OR = 1.18; [95% CI 1.12,1.25]; I2 0.0%). CONCLUSIONS: Our findings support a bidirectional longitudinal association between depressive symptoms and HbA1c . However, the observed effect sizes were small and future research in large-scale longitudinal studies is needed to confirm this association. Future studies should investigate the role of type of diabetes and depression, diabetes distress and diabetes self-management behaviours. Our results may have clinical implications, as depressive symptoms and HbA1c levels could be targeted concurrently in the prevention and treatment of diabetes and depression. REGISTRATION: PROSPERO ID CRD42019147551.

Cross-sectional associations of tau protein biomarkers with semantic and episodic memory in older adults without dementia: A systematic review and meta-analysis.

Pathological tau is suggested to play a role in cognitive deterioration in the preclinical phase of Alzheimer's disease. We investigated cross-sectional associations of tau burden with episodic and semantic memory performance in older adults without dementia. A systematic search in MEDLINE (via PubMed), PsychINFO, and Embase resulted in 24 eligible studies for meta-analysis. Tau burden was assessed using CSF, PET, or histopathological measures. All studies evaluated associations of tau with episodic memory: weighted effect sizes were -0.46 (95 % CI [-0.73; -0.20], p < .001) for episodic composite scores, -0.19 ([-0.36; -0.03], p = .024) for delayed word list recall, and -0.05 ([-0.14; 0.04], p = .257) for logical memory. Fourteen studies evaluated associations of tau with semantic memory: weighted effect sizes were -0.28 ([-0.52; -0.04], p = .023) for semantic composite scores, -0.06 ([-0.16; 0.03], p = .194) for semantic fluency, and 0.06 ([-0.06; 0.18], p = .319) for picture naming. Our findings indicate that tau burden related to both episodic and semantic memory impairment in older individuals without a diagnosis of mild cognitive impairment or manifest dementia, with episodic composite scores showing the strongest association with tau burden. Future potential lies in developing more sensitive scores to detect this subtle cognitive impairment, which could contribute to early identification of individuals in the preclinical phase of Alzheimer's disease, thereby improving early diagnosis and timely intervention.

Association between social network characteristics and prevalent and incident depression: The Maastricht Study.

AIMS: Social network characteristics may provide a novel non-pharmaceutical target for the prevention of depression. We investigated the temporal association of a broad range of structural and functional social network characteristics with incident depressive symptoms over 5 years of follow-up. METHODS: We used data from The Maastricht Study, a population-based prospective cohort study (n=2,465, mean age 59.8±8.1 years, 49.1% women, 11,585 person-years of follow-up). Social network characteristics were assessed through a name generator questionnaire. Clinically relevant depressive symptoms (9-item Patient Health Questionnaire score≥10) were assessed at baseline and annually. We used multivariable logistic and Cox regression analyses, adjusted for sociodemographic, lifestyle and cardiovascular risk factors. RESULTS: In cross-sectional analyses less emotional support for discomfort and with important decisions, and less informational support were associated with prevalent depressive symptoms (OR[95%CI] 1.19 [1.01-1.40]; 1.22 [1.05-1.43], and 1.20 [1.04-1.39], respectively). Every fewer 10% of family members was associated with prevalent depressive symptoms (1.11 [1.01-1.23]). In longitudinal analyses, less emotional support on important decisions was also associated with higher risk of incident depressive symptoms (HR[95%CI] 1.13 [1.03-1.25]). In addition, every fewer 10% of the network that was a family member was associated with a higher hazard of incident depressive symptoms (1.07 [1.01-1.13]). CONCLUSIONS: This study shows that less emotional support and fewer family members in the network were associated with higher risk of both prevalent and incident depression. The importance of emotional support and the role that family plays should be considered in treatment and prevention of depression.

Quality control strategies for brain MRI segmentation and parcellation: Practical approaches and recommendations - insights from the Maastricht study.

Quality control of brain segmentation is a fundamental step to ensure data quality. Manual quality control strategies are the current gold standard, although these may be unfeasible for large neuroimaging samples. Several options for automated quality control have been proposed, providing potential time efficient and reproducible alternatives. However, those have never been compared side to side, which prevents consensus in the appropriate quality control strategy to use. This study aimed to elucidate the changes manual editing of brain segmentations produce in morphological estimates, and to analyze and compare the effects of different quality control strategies on the reduction of the measurement error. Structural brain MRI from 259 participants of The Maastricht Study were used. Morphological estimates were automatically extracted using FreeSurfer 6.0. Segmentations with inaccuracies were manually edited, and morphological estimates were compared before and after editing. In parallel, 12 quality control strategies were applied to the full sample. Those included: two manual strategies, in which images were visually inspected and either excluded or manually edited; five automated strategies, where outliers were excluded based on the tools "MRIQC" and "Qoala-T", and the metrics "morphological global measures", "Euler numbers" and "Contrast-to-Noise ratio"; and five semi-automated strategies, where the outliers detected through the mentioned tools and metrics were not excluded, but visually inspected and manually edited. In order to quantify the effects of each quality control strategy, the proportion of unexplained variance relative to the total variance was extracted after the application of each strategy, and the resulting differences compared. Manually editing brain surfaces produced particularly large changes in subcortical brain volumes and moderate changes in cortical surface area, thickness and hippocampal volumes. The performance of the quality control strategies depended on the morphological measure of interest. Overall, manual quality control strategies yielded the largest reduction in relative unexplained variance. The best performing automated alternatives were those based on Euler numbers and MRIQC scores. The exclusion of outliers based on global morphological measures produced an increase of relative unexplained variance. Manual quality control strategies are the most reliable solution for quality control of brain segmentation and parcellation. However, measures must be taken to prevent the subjectivity associated with these strategies. The detection of inaccurate segmentations based on Euler numbers or MRIQC provides a time efficient and reproducible alternative. The exclusion of outliers based on global morphological estimates must be avoided.