Archives of gerontology and geriatrics最新文献

Background: The objective of this study was to construct a predictive model using multiple machine learning algorithms to predict the risk of dementia superimposed on delirium (DSD) in dementia patients in the ICU.

Methods: The data for this study were sourced from the Medical Information Mart for Intensive Care IV database. The dataset was divided into a development set (70%) and a test set (30%). Feature selection was conducted using the Boruta algorithm to identify clinically relevant predictors, followed by model development using logistic regression, decision trees, random forest, XGBoost, glmnet, k-nearest neighbors (KNN), support vector machine (SVM), and neural network (NNET). The performance of the model was evaluated using several performance metrics, including the area under the curve (AUC), accuracy, precision, sensitivity, specificity, and the F-beta score.

Results: The findings demonstrated that the XGBoost model showed stable and acceptable discriminative performance, with an AUC of 0.7072 in the development set and 0.7258 in the test set, along with reasonable agreement between predicted and observed risks. The model achieved a balanced classification profile, with satisfactory specificity and sensitivity, and an F-beta score of 0.7594 in the test set. Decision curve analysis indicated that the XGBoost model provided net clinical benefit across a range of decision thresholds.

Conclusions: XGBoost-based machine learning models demonstrate moderate discriminative ability for predicting the occurrence of DSD in ICU patients with dementia and may support risk stratification in similar ICU settings, pending further external validation in independent cohorts.

Background: With the aging population, the demand for geriatric orthopedic surgery is increasing, and the rate of postoperative complications is escalating. This study aimed to construct a prediction model based on geriatric comprehensive assessment data and potential risk factors for surgery to assess the risk of complications in older patients within three months after various surgical procedures.

Methods: Using a prospective cohort design, we split a central dataset (7:3 ratio) into training and internal validation sets. We identified relevant variables using LASSO regression and built predictive models using logistic regression. External validation was performed using data from four other centers. Performance was evaluated by AUC, calibration plots, and DCA, and a Shiny web calculator was subsequently deployed.

Results: The study incorporated six variables related to frailty, nutrition, ADL, anesthesia modality, surgery complexity, and operation duration for modeling. To validate the model, we employed data from the training and validation sets independently. The AUC of the nomogram model based on the training set was 0.816 (95% CI: 0.769-0.863), and the AUC based on the internal validation set was 0.791 (95% CI: 0.716-0.866); the AUC for external validation was 0.789 (95% CI: 0.710-0.868). All calibration and DCA curves demonstrated good predictive performance.

Conclusions: The predictive model effectively identifies the risk of postoperative complications within three months in older individuals undergoing orthopedic surgery, aiding preoperative risk assessment and patient management to enhance clinical decision-making and improve patient outcomes.

Background: Population ageing and the increasing complexity of care pathways make transitions from hospital to residential healthcare facilities a critical phase for continuity of care. Inadequate management may result in clinical errors, repeated hospitalisations, and increased healthcare costs. The lack of standardised models and limited understanding of transitional care pathways highlight the need for further investigation of existing approaches.

Objectives: This scoping review aims to identify and map geriatric transitional care models described in the literature, focusing on their main characteristics, the professionals involved, and their potential applicability within the Italian National Health Service.

Methods: A scoping review was conducted according to the guidelines of the Joanna Briggs Institute and Prisma-ScR. The bibliographic search, carried out on Medline, ScienceDirect and Web of Science, identified 2107 articles. After removing duplicates and screening against predefined inclusion and exclusion criteria, seven studies were included.

Results: Across the seven studies, twelve transitional care projects were identified. Most models adopted a multidisciplinary, nurse-centred approach with active involvement of patients and family caregivers. Advanced Practice Nurses or equivalent expert nurses played a central role in care coordination, medication reconciliation, and post-discharge follow-up, contributing to reduced hospital readmissions and improved patient satisfaction.

Conclusions: Existing models suggest that multidisciplinary transitional care, supported by expert nurses, can enhance continuity and quality of care for older adults transferred to residential facilities. The conceptual framework derived from this mapping should be interpreted as a hypothesis‑generating synthesis to guide local adaptation and piloting in Italy, particularly by strengthening the role and competencies of Family and Community Nurses.

Background and objectives: We explored the association between a constructed multi-dimensional social frailty index and incidence of disability and all-cause mortality among community-dwelling older adults, examining its prognostic validity.

Research design and methods: We obtained data from the 2016 Japan Gerontological Evaluation Study (JAGES), surveying community-dwelling individuals aged ≥65 years, not certified as requiring long-term care, at baseline. The analytical sample comprised 73,472 participants (mean age, 73.7 years; 53.4% women) from 18 municipalities across 10 prefectures in Japan. The social frailty index comprised seven items across four domains: general resources (subjective financial conditions), social resources (living arrangements; social networks [visiting friends]), social behavior/activities (social participation; decreased frequency of going out), and basic social needs fulfillment (emotional social support; social cohesion [community attachment]). Incident disability and all-cause mortality were recorded over a mean 3.5-year follow-up. Associations were examined using Cox proportional hazards models adjusted for demographics, health status, and physical, cognitive, and depressive symptoms.

Results: Participants were classified as socially non-frail (51.6%), socially pre-frail (39.5%), and socially frail (9.0%). Higher social frailty scores were significantly associated with increased risk of disability and mortality. In the demographics and health status-adjusted model, incident disability hazard ratios were 1.41 for social pre-frailty and 1.77 for social frailty, whereas corresponding mortality hazard ratios were 1.35 and 1.65, respectively (all P < 0.001).

Discussion: The multi-dimensional social frailty index demonstrated robust prognostic validity for adverse health outcomes.

Implications: Integrating social dimensions into frailty assessments may facilitate early interventions, inform policy development, and improve health maintenance in aging populations.

Background: The interplay between frailty dynamics and the newly defined Cardiovascular-Kidney-Metabolic (CKM) syndrome remains poorly understood. We aimed to quantify the association between frailty transitions and the risk of incident advanced CKM syndrome and identify modifiable drivers of frailty progression.

Methods: Using data from the China Health and Retirement Longitudinal Study (CHARLS), we analyzed 8159 participants cross-sectionally and 3506 longitudinally over four years. Frailty was assessed using a deficit-accumulation index. The primary outcome was prevalent and incident advanced CKM syndrome (Stages 3-4). We employed multivariable regression models to evaluate associations and a machine learning pipeline to identify key predictors of frailty progression.

Results: Baseline frailty showed a robust dose-response relationship with prevalent advanced CKM (OR 1.44 per 0.1-unit FI increase; 95% CI 1.38-1.51). Longitudinally, individuals progressing to a frail state had a 40% increased risk of incident advanced CKM compared to stable non-frail peers (OR 1.40; 95% CI 1.01-1.94). Notably, this risk was sex-specific, observing a significant association in men (OR 2.20; 95% CI 1.33-3.64) but not in women. Machine learning identified life satisfaction, smoking status, and sleep duration as the top predictors of frailty progression.

Conclusions: Frailty progression acts as a potent, sex-specific risk amplifier for advanced CKM syndrome. Integrating frailty screening into CKM care and targeting psychosocial well-being-specifically life satisfaction-alongside lifestyle factors may be important strategies to preempt frailty and potentially mitigate cardiovascular-renal-metabolic risks.

This exploratory study examined how depth of processing, intention to learn, and age interact to influence free recall performance and the temporal contiguity effect (TCE). A total of 119 participants, being 67 younger adults (M = 19.2 years) and 52 older adults (M = 73.9 years), completed free recall tasks that varied in levels of processing (no-orienting, shallow, deep) and learning intention (intentional vs. incidental). Given modest cell sizes (n = 10-17 per condition), findings should be considered exploratory and hypothesis-generating. Patterns suggest that deeper processing improves overall recall across age groups, while the organization of recall may differ with age and learning intention, though definitive conclusions about interactions require replication with larger samples. Intentional learning strengthened temporal organization compared with incidental learning, but this pattern diverged by age. In younger adults, deeper processing disrupted the TCE during intentional learning, whereas older adults maintained stable temporal organization across processing levels, a pattern that may reflect strategic compensation, reduced encoding flexibility, or preserved temporal binding mechanisms. These findings contribute to understanding how encoding strategies and learning intention shape memory organization across the adult lifespan and highlight potential pathways for supporting episodic memory in aging.

Objectives: The current study explored how childhood conditions influence the trajectory of depression in older adults, with a focus on the mediating roles of health-related and social lifestyle factors.

Methods: Longitudinal data were drawn from the 2018, 2020, and 2022 waves of the China Family Panel Studies. It comprised 2, 878 older adults aged 60 or older. Latent growth modeling was used to assess the effects of childhood conditions on baseline levels and changes in depression over time, as well as the mediating roles of lifestyle.

Results: Poor childhood health and lower family status were associated with higher baseline levels of depressive symptoms. Health-promoting and social lifestyle factors, particularly connectedness with children and regular physical activity, were linked to lower initial depression levels, whereas smoking was associated with a faster increase in depressive symptoms over time. In addition, connectedness with children partially mediated the association between childhood health and baseline depression.

Conclusion: The findings address the long-term impact of childhood conditions on depression in later life and highlight the protective role of favorable social and health-promoting lifestyles. Suggestions for targeted policies and interventions to support the mental well-being of aging populations were discussed.

Objective: Early detection of mild cognitive impairment (MCI) is essential for dementia prevention. We developed the National Center for Geriatrics and Gerontology-Functional Assessment Tool (NCGG-FAT) using age- and education-adjusted norms with a 1.5 standard deviation (SD) cutoff. This study examined the associations between cognitive domains assessed by the NCGG-FAT and incident dementia using the existing and expanded databases.

Method: A 5-year prospective cohort of 2,441 participants without dementia at baseline was analyzed. Hazard ratios (HRs) and 95% confidence intervals (CIs) for individual cognitive tests and cognitive states (normal cognition, MCI, and global cognitive impairment) were estimated using Cox proportional hazards models. Competing risk of death was addressed using Fine-Gray models. Predictive models for dementia were also developed and validated.

Results: In the expanded database, declines ≥1.5 SD in word list memory, TMT-B, forward and backward digit span, and SDST were significantly associated with dementia onset (HRs 1.77-3.22). Fine-Gray analyses yielded similar results. For cognitive states, amnestic and non-amnestic MCI, particularly moderate subtypes, and global cognitive impairment showed elevated risks (HRs 1.55-2.92), although some associations lost significance after accounting for competing mortality. The NCGG-FAT composite score demonstrated high discrimination for incident dementia (AUC = 0.96; accuracy = 0.95).

Conclusion: The expanded NCGG-FAT database is a useful auxiliary tool for assessing future dementia risk in community-dwelling older adults.