Identification of high likelihood of dementia in population-based surveys using unsupervised clustering: a longitudinal analysis.

BACKGROUND: Dementia is defined as a cognitive decline that affects functional status. Longitudinal ageing surveys often lack a clinical diagnosis of dementia though measure cognition and daily function over time. We used unsupervised machine learning and longitudinal data to identify transition to probable dementia. METHODS: Multiple Factor Analysis was applied to longitudinal function and cognitive data of 15,278 baseline participants (aged 50 years and more) from the Survey of Health, Ageing, and Retirement in Europe (SHARE) (waves 1, 2 and 4-7, between 2004 and 2017). Hierarchical Clustering on Principal Components discriminated three clusters at each wave. We estimated probable or "Likely Dementia" prevalence by sex and age, and assessed whether dementia risk factors increased the risk of being assigned probable dementia status using multistate models. Next, we compared the "Likely Dementia" cluster with self-reported dementia status and replicated our findings in the English Longitudinal Study of Ageing (ELSA) cohort (waves 1-9, between 2002 and 2019, 7840 participants at baseline). RESULTS: Our algorithm identified a higher number of probable dementia cases compared with self-reported cases and showed good discriminative power across all waves (AUC ranged from 0.754 [0.722-0.787] to 0.830 [0.800-0.861]). "Likely Dementia" status was more prevalent in older people, displayed a 2:1 female/male ratio, and was associated with nine factors that increased risk of transition to dementia: low education, hearing loss, hypertension, drinking, smoking, depression, social isolation, physical inactivity, diabetes, and obesity. Results were replicated in ELSA cohort with good accuracy. CONCLUSIONS: Machine learning clustering can be used to study dementia determinants and outcomes in longitudinal population ageing surveys in which dementia clinical diagnosis is lacking.

Identification of High Likelihood of Dementia in Population-Based Surveys using Unsupervised Clustering: a Longitudinal Analysis.

Background: Dementia is defined by cognitive decline that affects functional status. Longitudinal ageing surveys often lack a clinical diagnosis of dementia though measure cognitive and function over time. We used unsupervised machine learning and longitudinal data to identify transition to probable dementia. Methods: Multiple Factor Analysis was applied to longitudinal function and cognitive data of 15,278 baseline participants (aged 50 years and more) from the Survey of Health, Ageing, and Retirement in Europe (SHARE) (waves 1, 2 and 4-7, between 2004 and 2017). Hierarchical Clustering on Principal Components discriminated three clusters at each wave. We estimated probable or "Likely Dementia" prevalence by sex and age, and assessed whether dementia risk factors increased the risk of being assigned probable dementia status using multistate models. Next, we compared the "Likely Dementia" cluster with self-reported dementia status and replicated our findings in the English Longitudinal Study of Ageing (ELSA) cohort (waves 1-9, between 2002 and 2019, 7,840 participants at baseline). Findings: Our algorithm identified a higher number of probable dementia cases compared with self-reported cases and showed good discriminative power across all waves (AUC ranged from 0.754 [0.722-0.787] to 0.830 [0.800-0.861]). "Likely Dementia" status was more prevalent in older people, displayed a 2:1 female/male ratio and was associated with nine factors that increased risk of transition to dementia: low education, hearing loss, hypertension, drinking, smoking, depression, social isolation, physical inactivity, diabetes, and obesity. Results were replicated in ELSA cohort with good accuracy. Interpretation: Machine learning clustering can be used to study dementia determinants and outcomes in longitudinal population ageing surveys in which dementia clinical diagnosis is lacking.

A new approach to digitized cognitive monitoring: validity of the SelfCog in Huntington's disease.

Cognitive deficits represent a hallmark of neurodegenerative diseases, but evaluating their progression is complex. Most current evaluations involve lengthy paper-and-pencil tasks which are subject to learning effects dependent on the mode of response (motor or verbal), the countries' language or the examiners. To address these limitations, we hypothesized that applying neuroscience principles may offer a fruitful alternative. We thus developed the SelfCog, a digitized battery that tests motor, executive, visuospatial, language and memory functions in 15 min. All cognitive functions are tested according to the same paradigm, and a randomization algorithm provides a new test at each assessment with a constant level of difficulty. Here, we assessed its validity, reliability and sensitivity to detect decline in early-stage Huntington's disease in a prospective and international multilingual study (France, the UK and Germany). Fifty-one out of 85 participants with Huntington's disease and 40 of 52 healthy controls included at baseline were followed up for 1 year. Assessments included a comprehensive clinical assessment battery including currently standard cognitive assessments alongside the SelfCog. We estimated associations between each of the clinical assessments and SelfCog using Spearman's correlation and proneness to retest effects and sensitivity to decline through linear mixed models. Longitudinal effect sizes were estimated for each cognitive score. Voxel-based morphometry and tract-based spatial statistics analyses were conducted to assess the consistency between performance on the SelfCog and MRI 3D-T1 and diffusion-weighted imaging in a subgroup that underwent MRI at baseline and after 12 months. The SelfCog detected the decline of patients with Huntington's disease in a 1-year follow-up period with satisfactory psychometric properties. Huntington's disease patients are correctly differentiated from controls. The SelfCog showed larger effect sizes than the classical cognitive assessments. Its scores were associated with grey and white matter damage at baseline and over 1 year. Given its good performance in longitudinal analyses of the Huntington's disease cohort, it should likely become a very useful tool for measuring cognition in Huntington's disease in the future. It highlights the value of moving the field along the neuroscience principles and eventually applying them to the evaluation of all neurodegenerative diseases.

Contextual influence of reinforcement learning performance of depression: evidence for a negativity bias?

BACKGROUNDS: Value-based decision-making impairment in depression is a complex phenomenon: while some studies did find evidence of blunted reward learning and reward-related signals in the brain, others indicate no effect. Here we test whether such reward sensitivity deficits are dependent on the overall value of the decision problem. METHODS: We used a two-armed bandit task with two different contexts: one 'rich', one 'poor' where both options were associated with an overall positive, negative expected value, respectively. We tested patients (N = 30) undergoing a major depressive episode and age, gender and socio-economically matched controls (N = 26). Learning performance followed by a transfer phase, without feedback, were analyzed to distangle between a decision or a value-update process mechanism. Finally, we used computational model simulation and fitting to link behavioral patterns to learning biases. RESULTS: Control subjects showed similar learning performance in the 'rich' and the 'poor' contexts, while patients displayed reduced learning in the 'poor' context. Analysis of the transfer phase showed that the context-dependent impairment in patients generalized, suggesting that the effect of depression has to be traced to the outcome encoding. Computational model-based results showed that patients displayed a higher learning rate for negative compared to positive outcomes (the opposite was true in controls). CONCLUSIONS: Our results illustrate that reinforcement learning performances in depression depend on the value of the context. We show that depressive patients have a specific trouble in contexts with an overall negative state value, which in our task is consistent with a negativity bias at the learning rates level.

The burden of Huntington's disease: A prospective longitudinal study of patient/caregiver pairs.

BACKGROUND: Caregiver burden is widely recognized in Huntington's disease, but little is known about the factors determining its evolution over time in the absence of longitudinal studies. Our objective was to identify typical patterns of caregiver burden level and evolution using both patients' and caregivers' characteristics over a one-year period to identify potential levers for alleviation. METHODS: We conducted a prospective multicenter longitudinal study in caregiver/patient pairs in Huntington's disease (NCT02876445) between March 2011 and May 2015. Caregiver data were derived from two questionnaires at one-year interval on perceived burden (Zarit Burden Interview), social environment and support. Caregiver data were linked to clinical and demographic data from patients included in the Biomarker study (NCT01590589). Unsupervised clustering analysis was performed using self-organizing maps. RESULTS: 105 caregiver/patient pairs were included in the analysis. We identified four clusters. Of the two clusters of patients with advanced disease, cluster A was characterized by high levels of irritability and obsessive-compulsive behaviors, with high and increasing burden (N = 30; 29%), cluster B, the more apathetic group, with low and decreasing burden (N = 22; 21%). Clusters C (N = 27; 26%) and D (N = 26; 25%) were composed of patients in earlier stages, associated with a stable burden in group C but a notably increasing one in group D driven by patients' depression scores increase. CONCLUSIONS: Our results revealed the dynamics of caregiver burden over time in Huntington's disease, combining the stage of the disease, the severity of the patients' decline, psychiatric and behavioral disorders, and their evolution over time.

Information about action outcomes differentially affects learning from self-determined versus imposed choices.

The valence of new information influences learning rates in humans: good news tends to receive more weight than bad news. We investigated this learning bias in four experiments, by systematically manipulating the source of required action (free versus forced choices), outcome contingencies (low versus high reward) and motor requirements (go versus no-go choices). Analysis of model-estimated learning rates showed that the confirmation bias in learning rates was specific to free choices, but was independent of outcome contingencies. The bias was also unaffected by the motor requirements, thus suggesting that it operates in the representational space of decisions, rather than motoric actions. Finally, model simulations revealed that learning rates estimated from the choice-confirmation model had the effect of maximizing performance across low- and high-reward environments. We therefore suggest that choice-confirmation bias may be adaptive for efficient learning of action-outcome contingencies, above and beyond fostering person-level dispositions such as self-esteem.