Predictors of Dementia in the Oldest Old: A Novel Machine Learning Approach.
Ontology highlight
ABSTRACT: Background
Incidence of dementia increases exponentially with age; little is known about its risk factors in the ninth and 10th decades of life. We identified predictors of dementia with onset after age 85 years in a longitudinal population-based cohort.Methods
On the basis of annual assessments, incident cases of dementia were defined as those newly receiving Clinical Dementia Rating (CDR) ≥1. We used a machine learning method, Markov modeling with hybrid density-based and partition-based clustering, to identify variables associated with subsequent incident dementia.Results
Of 1439 participants, 641 reached age 85 years during 10 years of follow-up and 45 of these became incident dementia cases. Using hybrid density-based and partition-based, among those aged 85+ years, probability of incident dementia was associated with worse self-rated health, more prescription drugs, subjective memory complaints, heart disease, cardiac arrhythmia, thyroid disease, arthritis, reported hypertension, higher systolic and diastolic blood pressure, and hearing impairment. In the subgroup aged 85 to 89 years, risk of dementia was also associated with depression symptoms, not currently smoking, and lacking confidantes.Conclusions
An atheoretical machine learning method revealed several factors associated with increased probability of dementia after age 85 years in a population-based cohort. If independently validated in other cohorts, these findings could help identify the oldest-old at the highest risk of dementia.
SUBMITTER: Jia Y
PROVIDER: S-EPMC7677183 | biostudies-literature |
REPOSITORIES: biostudies-literature
Similar Datasets
Project description:ObjectiveTo examine the association between brain structural changes and β-amyloid deposition, and incident dementia in 183 elderly subjects without dementia (mean age 85.5 years) 2 years later.MethodsSubjects had a brain structural MRI scan and a PET scan with (11)C-labeled Pittsburgh compound B (PiB) in 2009, and were evaluated clinically in 2011.ResultsAt baseline evaluation, of the 183 participants (146 cognitively normal [CN]); 37 mild cognitive impairment [MCI]), 139 (76%) were PiB+, had small hippocampal volume (<25th percentile), or had high white matter lesion (WML) volume (>75th percentile). Two years later, 111 (61%) were classified as CN, 51 (28%) as MCI, and 21 (11%) as dementia. At baseline, 51% of the CN participants and 67.5% of the MCI cases were PiB+. Thirty percent of the CN and 51% of the MCI cases had small hippocampi, and 24% of the CN and 40.5% of the MCI cases had abnormal WMLs. Of the 21 participants who progressed to dementia, 20 (95%) had at least one imaging abnormality. Only 3 (14%) were only PiB+, 1 (5%) had only small hippocampi, 1 (5%) had only WMLs, 1 (5%) was biomarker negative, and the other 16 had various pairs of imaging abnormalities. Continuous variables of PiB retention, left and right hippocampal volume, and WML volume were independent predictors of dementia in a logistic regression analysis controlling for age, sex, education level, and Mini-Mental State Examination scores.ConclusionsThe prevalence of β-amyloid deposition, neurodegeneration (i.e., hippocampal atrophy), and small vessel disease (WMLs) is high in CN older individuals and in MCI. A combination of 2 or 3 of these factors is a powerful predictor of short-term incidence of dementia.
| S-EPMC4240431 | biostudies-literature
Project description:IntroductionLittle is understood about the socioeconomic predictors of tooth loss, a condition that can negatively impact individual's quality of life. The goal of this study is to develop a machine-learning algorithm to predict complete and incremental tooth loss among adults and to compare the predictive performance of these models.MethodsWe used data from the National Health and Nutrition Examination Survey from 2011 to 2014. We developed multiple machine-learning algorithms and assessed their predictive performances by examining the area under the receiver operating characteristic curve (AUC), accuracy, sensitivity, specificity, and positive and negative predictive values.ResultsThe extreme gradient boosting trees presented the highest performance in the prediction of edentulism (AUC = 88.7%; 95%CI: 87.1, 90.2), the absence of a functional dentition (AUC = 88.3% 95%CI: 87.3,89.3) and for predicting missing any tooth (AUC = 83.2%; 95%CI, 82.0, 84.4). Although, as expected, age and routine dental care emerged as strong predictors of tooth loss, the machine learning approach identified additional predictors, including socioeconomic conditions. Indeed, the performance of models incorporating socioeconomic characteristics was better at predicting tooth loss than those relying on clinical dental indicators alone.ConclusionsFuture application of machine-learning algorithm, with longitudinal cohorts, for identification of individuals at risk for tooth loss could assist clinicians to prioritize interventions directed toward the prevention of tooth loss.
| S-EPMC8213149 | biostudies-literature
Project description:BackgroundEfforts to identify important risk factors for cognitive impairment and dementia have to date mostly relied on meta-analytic strategies. A comprehensive empirical evaluation of these risk factors within a single study is currently lacking.ObjectiveWe used a combined methodology of machine learning and semi-parametric survival analysis to estimate the relative importance of 52 predictors in forecasting cognitive impairment and dementia in a large, population-representative sample of older adults.MethodsParticipants from the Health and Retirement Study (N = 9,979; aged 50-98 years) were followed for up to 10 years (M = 6.85 for cognitive impairment; M = 7.67 for dementia). Using a split-sample methodology, we first estimated the relative importance of predictors using machine learning (random forest survival analysis), and we then used semi-parametric survival analysis (Cox proportional hazards) to estimate effect sizes for the most important variables.ResultsAfrican Americans and individuals who scored high on emotional distress were at relatively highest risk for developing cognitive impairment and dementia. Sociodemographic (lower education, Hispanic ethnicity) and health variables (worse subjective health, increasing BMI) were comparatively strong predictors for cognitive impairment. Cardiovascular factors (e.g., smoking, physical inactivity) and polygenic scores (with and without APOEɛ4) appeared less important than expected. Post-hoc sensitivity analyses underscored the robustness of these results.ConclusionsHigher-order factors (e.g., emotional distress, subjective health), which reflect complex interactions between various aspects of an individual, were more important than narrowly defined factors (e.g., clinical and behavioral indicators) when evaluated concurrently to predict cognitive impairment and dementia.
| S-EPMC7934087 | biostudies-literature
Project description:IntroductionThe study objective was to build a machine learning model to predict incident mild cognitive impairment, Alzheimer's Disease, and related dementias from structured data using administrative and electronic health record sources.MethodsA cohort of patients (n = 121,907) and controls (n = 5,307,045) was created for modeling using data within 2 years of patient's incident diagnosis date. Additional cohorts 3-8 years removed from index data are used for prediction. Training cohorts were matched on age, gender, index year, and utilization, and fit with a gradient boosting machine, lightGBM.ResultsIncident 2-year model quality on a held-out test set had a sensitivity of 47% and area-under-the-curve of 87%. In the 3-year model, the learned labels achieved 24% (71%), which dropped to 15% (72%) in year 8.DiscussionThe ability of the model to discriminate incident cases of dementia implies that it can be a worthwhile tool to screen patients for trial recruitment and patient management.
| S-EPMC6920083 | biostudies-literature
Project description:Aging is known to have deleterious effects on cerebral white matter, yet little is known about these white matter alterations in advanced age. In this study, 94 oldest-old adults without dementia (90-103 years) underwent diffusion tensor imaging to assess relationships between chronological age and multiple measures of integrity in 18 white matter regions across the brain. Results revealed significant age-related declines in integrity in regions previously identified as being sensitive to aging in younger-old adults (corpus callosum, fornix, cingulum, external capsule). For the corpus callosum, the effect of age on genu fractional anisotropy was significantly weaker than the relationship between age and splenium fractional anisotropy. Importantly, age-related declines in white matter integrity did not differ in cognitively normal and cognitively impaired not demented oldest-old, suggesting that they were not solely driven by cognitive dysfunction or preclinical dementia in this advanced age group. Instead, white matter in these regions appears to remain vulnerable to normal aging processes through the 10th decade of life.
| S-EPMC5647141 | biostudies-literature
Project description:Background and purposePatients with ischemic stroke frequently develop hemorrhagic transformation (HT), which could potentially worsen the prognosis. The objectives of the current study were to determine the incidence and predictors of HT, to evaluate predictor interaction, and to identify the optimal predicting models.MethodsA prospective study included 360 patients with ischemic stroke, of whom 354 successfully continued the study. Patients were subjected to thorough general and neurological examination and T2 diffusion-weighted MRI, at admission and 1 week later to determine the incidence of HT. HT predictors were selected by a filter-based minimum redundancy maximum relevance (mRMR) algorithm independent of model performance. Several machine learning algorithms including multivariable logistic regression classifier (LRC), support vector classifier (SVC), random forest classifier (RFC), gradient boosting classifier (GBC), and multilayer perceptron classifier (MLPC) were optimized for HT prediction in a randomly selected half of the sample (training set) and tested in the other half of the sample (testing set). The model predictive performance was evaluated using receiver operator characteristic (ROC) and visualized by observing case distribution relative to the models' predicted three-dimensional (3D) hypothesis spaces within the testing dataset true feature space. The interaction between predictors was investigated using generalized additive modeling (GAM).ResultsThe incidence of HT in patients with ischemic stroke was 19.8%. Infarction size, cerebral microbleeds (CMB), and the National Institute of Health stroke scale (NIHSS) were identified as the best HT predictors. RFC (AUC: 0.91, 95% CI: 0.85-0.95) and GBC (AUC: 0.91, 95% CI: 0.86-0.95) demonstrated significantly superior performance compared to LRC (AUC: 0.85, 95% CI: 0.79-0.91) and MLPC (AUC: 0.85, 95% CI: 0.78-0.92). SVC (AUC: 0.90, 95% CI: 0.85-0.94) outperformed LRC and MLPC but did not reach statistical significance. LRC and MLPC did not show significant differences. The best models' 3D hypothesis spaces demonstrated non-linear decision boundaries suggesting an interaction between predictor variables. GAM analysis demonstrated a linear and non-linear significant interaction between NIHSS and CMB and between NIHSS and infarction size, respectively.ConclusionCerebral microbleeds, NIHSS, and infarction size were identified as HT predictors. The best predicting models were RFC and GBC capable of capturing nonlinear interaction between predictors. Predictor interaction suggests a dynamic, rather than, fixed cutoff risk value for any of these predictors.
| S-EPMC9731336 | biostudies-literature
Project description:IntroductionSubstantial gaps in research remain across oldest-old ethnic populations while the burden of dementia increases exponentially with age among Mexican and Mexican American older adults.MethodsPrevalence and correlates of dementia among individuals ≥82 years of age were examined using two population-based cohort studies: The Mexican Health and Aging Study (MHAS, n = 1078, 2012) and the Hispanic Established Populations for the Epidemiologic Study of the Elderly (HEPESE, n = 735, 2012-2013). The analytic MHAS and HEPESE samples had an average age of 86.4 and 88.0 years, 1.2 and 1.8 women to men, and 2.7 and 5.1 average years of education, respectively.ResultsWe identified 316 (29.2%) and 267 (36.3%) cases of likely dementia in the MHAS and HEPESE cohorts, respectively. For Mexicans but not Mexican Americans, age-adjusted prevalence rates of likely dementia were higher in women than men. For both populations prevalence rates increased with age and decreased with education for Mexican Americans but not for Mexicans. In both populations, odds of likely dementia increased with age. Health insurance for the low-income was significantly associated with higher odds of likely dementia for Mexican American men and women and Mexican women but not men. Living in extended households increased the odds of likely dementia in women, but not in men for both studies. Multiple cardiovascular conditions increased the odds of likely dementia for Mexicans but not for Mexican Americans.DiscussionOur study provides evidence of the high burden of dementia among oldest-old Mexicans and Mexican Americans and its association with health and social vulnerabilities.
| S-EPMC7744026 | biostudies-literature
Project description:ObjectiveThe aim of this study was to analyze brain pathologies which cause dementia in the oldest old population.MethodsAll 601 persons aged ≥85 years living in the city of Vantaa (Finland), on April 1st, 1991 formed the study population of the Vantaa85 + study, 300 of whom were autopsied during follow-up (79.5% females, mean age-at-death 92 ± 3.7 years). Alzheimer's disease (AD) pathology (tau and beta-amyloid [Aβ]), cerebral amyloid angiopathy (CAA) and Lewy-related pathologies were analyzed. Brain infarcts were categorized by size (<2 mm, 2-15 mm, >15 mm) and by location. Brain hemorrhages were classified as microscopic (<2 mm) and macroscopic.Results195/300 (65%) were demented. 194/195 (99%) of the demented had at least one neuropathology. Three independent contributors to dementia were identified: AD-type tau-pathology (Braak stage V-VI), neocortical Lewy-related pathology, and cortical anterior 2-15 mm infarcts. These were found in 34%, 21%, and 21% of the demented, respectively, with the multivariate odds ratios (OR) for dementia 5.5, 4.5, and 3.4. Factor analysis investigating the relationships between different pathologies identified three separate factors: (1) AD-spectrum, which included neurofibrillary tau, Aβ plaque, and neocortical Lewy-related pathologies and CAA (2) >2 mm cortical and subcortical infarcts, and (3) <2 mm cortical microinfarcts and microhemorrhages. Multipathology was common and increased the risk of dementia significantly.InterpretationThese results indicate that AD-type neurodegenerative processes play the most prominent role in twilight cognitive decline. The high prevalence of both neurodegenerative and vascular pathologies indicates that multiple preventive and therapeutic approaches are needed to protect the brains of the oldest old.
| S-EPMC5338150 | biostudies-literature
Project description:Interpersonal firearm violence is a leading cause of death and injuries in the United States. Identifying community characteristics associated with firearm violence is important to improve confounder selection and control in health research, to better understand community-level factors that are associated with firearm violence, and to enhance community surveillance and control of firearm violence. The objective of this research was to use machine learning to identify an optimal set of predictors for urban interpersonal firearm violence rates using a broad set of community characteristics. The final list of 18 predictive covariates explain 77.8% of the variance in firearm violence rates, and are publicly available, facilitating their inclusion in analyses relating violence and health. This list includes the black isolation and segregation indices, rates of educational attainment, marital status, indicators of wealth and poverty, longitude, latitude, and temperature.
| S-EPMC5985152 | biostudies-literature
Project description:ObjectiveTo examine the incidence of dementia among the oldest-old people with normal cognition and different types of cognitive impairment.MethodsThis study included 395 participants without dementia (mean age 93.3 years) from The 90+ Study, a prospective, population-based study of aging and dementia in people aged 90 years and older. The participants had evaluations for dementia every 6 months, and their average follow-up was 2.5 years. We examined the incidence of all-cause dementia in participants stratified into 4 cognitive groups: normal, amnestic mild cognitive impairment (aMCI), nonamnestic mild cognitive impairment (naMCI), and other cognitive impairment (OCI).ResultsDementia incidence was highest for participants with aMCI (31.4% per year) and OCI (39.9% per year). Participants with naMCI had an incidence of 14.1% per year, and participants with normal cognition had an incidence of 8.4% per year. Dementia incidence was associated with increasing age in both normal and cognitively impaired participants; however, an APOE4 allele was associated with a higher dementia incidence only in participants with baseline cognitive impairment.ConclusionsThe risk of developing dementia in the oldest-old is high and increases to higher rates when cognitive impairment is present. Similar to results of studies in younger elderly individuals, cognitive impairment and increasing age were related to increased dementia incidence. High dementia incidence rates in the oldest-old individuals, particularly when cognitively impaired, emphasize the need to further study cognitive impairment and dementia in this rapidly expanding age group.
| S-EPMC3233189 | biostudies-literature