Patients with dementia experience a steady deterioration in cognitive function that increases mortality and impairments. Moreover, dementia is also anticipated to increase significantly in prevalence as the world's population ages, placing a strain on healthcare systems throughout the globe. Hence, early identification and prediction of dementia are essential due to timely treatments, enhanced patient care, and the potential for preventative measures. Therefore, the aim of this project is to construct a diagnostic system that leverages patient electronic medical data to predict dementia as well as dementia risk factors. We developed a novel variable selection method (VSM) based on data mining techniques to accomplish this goal by selecting the most relevant variables from the dataset that contribute to the onset of dementia in older people. We employed a random forest (RF) model to classify dementia, healthy subjects, and the hyperparameters of the selected RF model were adjusted using a random search approach. The proposed diagnostic system is based on two components that hybridize as a single system; therefore, we named it the VSM_RF model. We obtained the dataset from the Swedish National Study on Aging and Care (SNAC) to verify the reliability and accuracy of the proposed VSM_RF model. The three SNAC locations collectively yielded 8191 data observations, each including 75 variables. Numerous validation metrics, including accuracy, balance accuracy, sensitivity, specificity, and Matthew's correlation coefficient, were deployed to thoroughly assess the efficiency of the proposed VSM_RF model. Only six out of the 75 variables were used to achieve the maximum accuracy, along with balance accuracy of 98.00% and 97.29%, respectively. 

Introduction: Blood-based biomarkers for Alzheimer's disease (AD) have the potential to improve diagnostic accessibility, but their clinical interpretation requires understanding of variability and biological influences.

Methods: We repeatedly sampled blood from 57 adults referred for lumbar puncture as part of a cognitive evaluation at a memory clinic. We measured serum phosphorylated- tau-181 (s-p-tau181) and plasma amyloid beta (Aβ)42/40 ratio (p-Aβ42/Aβ40) and evaluated the impact of renal and blood–brain barrier (BBB) function.

Results: Test–retest analysis revealed large variability of s-p-tau181 and small for p-Aβ42/Aβ40. Markers of renal function and BBB integrity significantly influenced s-p-tau181 levels, whereas p-Aβ42/Aβ40 was not affected.

Discussion: This study emphasizes the need for caution when interpreting longitudinal changes in s-p-tau181. Inter-individual variability is to a large degree due to susceptibility to biological influences where a novel association with integrity of BBB function were identified. These results have implications for the clinical application of blood-based biomarkers in AD diagnostics and monitoring.

Highlights: Blood phosphorylated- tau-181 (p-tau181) shows high test–retest variability in memory clinic patients. Blood amyloid beta (Aβ)42/Aβ40 ratio is stable but has poor diagnostic accuracy. Renal function and blood–brain barrier (BBB) integrity affect blood p-tau181 levels. Caution is needed when interpreting longitudinal changes in blood p-tau181. Renal and BBB disorders should be considered when assessing blood p-tau181. 

Introduction: Exploring the association between technology use and sleep health in older adults is important as digital engagement becomes integrated into society.

Objective: This study aimed to examine sleep health and its association with technology use in a population-based cohort of 60 years and older.

Methods: This cross-sectional, population-based study (2023) included 436 older adults from the Swedish National Study on Aging and Care, Blekinge (SNAC-B) population. These participants were sent questionnaires about their sleep, internet usage, Digital Social Participation (DSP), Technology Anxiety (TA), Technology Enthusiasm (TE), and use of information and communication technology. We used a multidimensional instrument, SATED, to measure sleep health. In this study, we conducted statistical analyses using the chi2 test, T-test, Pearson correlation, and backward linear and logistic regression.

Results: Our study found that older adults (60 years+) have a mean sleep health score of 7.40 (SD = 2.03). TE (,) and DSP (,) were positively associated with better sleep health, while TA (,) was negatively associated. Frequent internet users(M = 7.6) and engaging with screens before bedtime (M = 7.7) had higher sleep health scores compared to non-frequent users (M = 6.90,) and none or seldom engagement with screens before bedtime (M = 7.10,) respectively. Linear regression showed TE positively associated (= 0.241,) while TA negatively associated (= -0.220,) with sleep health. DSP was found to be a predictor of better satisfaction (OR: 1.32,), efficiency (OR: 1.16,), and duration of sleep (OR:1.16,). Lower TA predicted better satisfaction (OR: 0.81,), timing (OR: 0.74,), and efficiency (OR:0.78,) of sleep. Older adults who use technology one hour before sleep have better sleep timing (OR: 3.003,), while those who do use mobile phones with a screen during the awake period after sleep onset have poor sleep timing (OR:0.016,).

Conclusions: DSP and TE support better sleep health, while TA negatively impacts sleep satisfaction, timing, and efficiency. Encouraging positive digital engagement and minimizing technology-related stress may promote healthier sleep in older adults. 

Background: In addition to impairing the lung function, chronic obstructive pulmonary disease (COPD) also affects phonatory characteristics. Recent research highlights the potential of voice as a digital biomarker to support clinical decision-making. While machine learning (ML) can detect disease patterns from acoustic features, clinical relevance requires understanding the relationship between the disorder and acoustic features.

Objective: This study investigates both statistical and clinical significance using Baseline Acoustic (BLA) and Mel-Frequency Cepstral Coefficient (MFCC) features with focusing on individuals with COPD and healthy controls (HC).

Method: Acoustic features derived from Swedish utterances of the vowel [a:], recorded via mobile phones from 48 age-matched participants (24 COPD, 24 HC; equal gender distribution), were analyzed. To reduce bias from varying recording counts, features were aggregated by averaging 10 randomly selected recordings per participant over 100 iterations. Vowel articulation was visualized in the vowel quadrilateral space using F1 (tongue height) and F2 (tongue advancement). Group differences were assessed using the Shapiro-Wilk test, Mann-Whitney U test (α = 0.05), Benjamini-Hochberg (BH) and Bonferroni corrections, Permutational Multivariate Analysis of Variance (PERMANOVA) test, and Cliff's Delta (δ).

Results: Of 101 features, 29 remained significant after BH correction and one after Bonferroni. Multivariate testing (p = 0.019) showed group separation. Additionally, 34 features demonstrated large effect sizes, suggesting potential as digital biomarkers.

Conclusion: Voice data recorded via mobile phones capture meaningful acoustic differences associated with COPD. These findings support the integration of voice-based assessments into eHealth platforms for noninvasive COPD screening and monitoring, which is pending further validation on larger populations.

Clinical Trial: NCT06705647

Inadequate sleep in older adults is linked to health issues such as frailty, cognitive impairment, and cardiovascular disorders. Maintaining regular sleep patterns is important for healthy aging, making effective sleep monitoring essential. While polysomnography (PSG) is the gold standard for diagnosing sleep disorders, its regular use in home settings is limited. Alternative objective monitoring methods in the home can offer insights into natural sleep patterns and factors affecting them without the limitations of PSG.

This scoping review aims to examine current technologies, sensors, and sleep parameters used for home-based sleep monitoring in older adults. It also aims to explore various predictors and outcomes associated with sleep to understand the factors of sleep monitoring at home. 

We identified 54 relevant articles using PubMed, Scopus, Web of Science, and an AI tool (Research Rabbit), with 48 studies using wearable technologies and eight studies using non-wearable technologies. Further, six types of sensors were utilized. The most common technology employed was actigraphy wearables, while ballistocardiography and electroencephalography were less common. The most frequent objective parameters of sleep measured were Total Sleep Time (TST), Wakeup After Sleep Onset (WASO), and Sleep Efficiency (SE), with only six studies evaluating sleep architecture in terms of sleep stages. Additionally, six categories of predictors and outcomes associated with sleep were analyzed, including Health-related, Environmental, Interventional, Behavioral, Time and Place, and Social associations. These associations correlate with TST, WASO, and SE and include in-bed behaviors, exterior housing conditions, aerobic exercise, living place, relationship status, and seasonal thermal environments.

Depression in older adults is a significant public health issue with broad impacts on both individuals and society. The multifaceted nature of depression underscores the complexity of identifying and predicting risk factors, necessitating a sophisticated and accurate approach based on new emerging technologies. Compared to traditional statistical methods, machine learning provides a more detailed and individualized understanding of risk variables by analyzing large datasets, identifying patterns, and building predictive models. This study presented a novel feature selection method based on the relief and lasso algorithms. The proposed feature selection method selected the ten most significant features from the dataset. A neural network (NN) with hyperparameters optimized by a grid search technique was used to categorize depression. The feature selection and classification modules work together as a single unit, namely as (Relief_Lasso_NN). Data from the Swedish National Study on Aging and Care (SNAC) was used for this study. The collected dataset consists of 726 samples with 75 features per sample. Four experiments were conducted to validate the performance of the proposed (Relief_Lasso_NN) framework. The proposed model achieved an accuracy of 90.34% in predicting depression using only ten features from the dataset. The top 10 features identified by the proposed feature selection method significantly impact depression in older adults. Furthermore, the performance of seven other state-of-the-art machine learning models was also compared with the proposed framework.

Depression and cognitive impairment frequently co-occur in older adults, but their temporal relationship remains unclear. While depression is often considered a risk factor for cognitive decline, evidence is mixed, particularly in individuals with mild cognitive impairment or early dementia (MCI/ED). This study analyzed longitudinal data from 1086 participants (M = 74.49, SD = 7.24) in the SMART4MD clinical trial, conducted in Spain and Sweden over 18 months, with assessments every six months. Cognitive impairment was measured using the Mini-Mental State Examination, and depression was assessed with the Geriatric Depression Scale-15. Findings revealed a concurrent association between depressive symptoms and cognitive impairment. In regression mixed analysis, depression levels predicted increased cognitive decline over time, but no evidence was found for cognitive impairment predicting future depression. These associations were confirmed using a bivariate latent growth curve model with cross-lagged paths, which revealed early but attenuating bidirectional effects between depression and cognition. These results highlight depression as a medium-term risk factor for cognitive decline, emphasizing the importance of addressing depressive symptoms to mitigate cognitive deterioration in MCI/ED populations.

Vowel-based voice analysis is gaining attention as a potential non-invasive tool for COPD classification, offering insights into phonatory function. The growing need for voice data has necessitated the adoption of various techniques, including segmentation, to augment existing datasets for training comprehensive Machine Learning (ML) models

This study aims to investigate the possible effects of segmentation of the utterance of vowel "a" on the performance of ML classifiers CatBoost (CB), Random Forest (RF), and Support Vector Machine (SVM). This research involves training individual ML models using three distinct dataset constructions: full-sequence, segment-wise, and group-wise, derived from the utterance of the vowel "a" which consists of 1058 recordings belonging to 48 participants. This approach comprehensively analyzes how each data categorization impacts the model's performance and results.

A nested cross-validation (nCV) approach was implemented with grid search for hyperparameter optimization. This rigorous methodology was employed to minimize overfitting risks and maximize model performance. Compared to the full-sequence dataset, the findings indicate that the second segment yielded higher results within the four-segment category. Specifically, the CB model achieved superior accuracy, attaining 97.8% and 84.6% on the validation and test sets, respectively. The same category for the CB model also demonstrated the best balance regarding true positive rate (TPR) and true negative rate (TNR), making it the most clinically effective choice.

These findings suggest that time-sensitive properties in vowel production are important for COPD classification and that segmentation can aid in capturing these properties. Despite these promising results, the dataset size and demographic homogeneity limit generalizability, highlighting areas for future research. Trial registration The study is registered on clinicaltrials.gov with ID: NCT06160674. 

Introduction: Dementia is a condition (a collection of related signs and symptoms) that causes a continuing deterioration in cognitive function, and millions of people are impacted by dementia every year as the world population continues to rise. Conventional approaches for determining dementia rely primarily on clinical examinations, analyzing medical records, and administering cognitive and neuropsychological testing. However, these methods are time-consuming and costly in terms of treatment. Therefore, this study aims to present a noninvasive method for the early prediction of dementia so that preventive steps should be taken to avoid dementia.

Methods: We developed a hybrid diagnostic system based on statistical and machine learning (ML) methods that used patient electronic health records to predict dementia. The dataset used for this study was obtained from the Swedish National Study on Aging and Care (SNAC), with a sample size of 43040 and 75 features. The newly constructed diagnostic extracts a subset of useful features from the dataset through a statistical method (F-score). For the classification, we developed an ensemble voting classifier based on five different ML models: decision tree (DT), naive Bayes (NB), logistic regression (LR), support vector machines (SVM), and random forest (RF). To address the problem of ML model overfitting, we used a cross-validation approach to evaluate the performance of the proposed diagnostic system. Various assessment measures, such as accuracy, sensitivity, specificity, receiver operating characteristic (ROC) curve, and Matthew’s correlation coefficient (MCC), were used to thoroughly validate the devised diagnostic system’s efficiency.

Results: According to the experimental results, the proposed diagnostic method achieved the best accuracy of 98.25%, as well as sensitivity of 97.44%, specificity of 95.744%, and MCC of 0.7535.

Discussion: The effectiveness of the proposed diagnostic approach is compared to various cutting-edge feature selection techniques and baseline ML models. From experimental results, it is evident that the proposed diagnostic system outperformed the prior feature selection strategies and baseline ML models regarding accuracy.