Predicting dementia risk in primary care: development and validation of the Dementia Risk Score using routinely collected data

Incremental Value of Multidomain Risk Factors for Dementia Prediction: A Machine Learning Approach

OBJECTIVE

The current evidence regarding how different predictor domains contributes to predicting incident dementia remains unclear. This study aims to assess the incremental value of five predictor domains when added to a simple dementia risk prediction model (DRPM) for predicting incident dementia in older adults.

DESIGN

Population-based, prospective cohort study.

SETTING

UK Biobank study.

PARTICIPANTS

Individuals aged 60 or older without dementia.

MEASUREMENTS

Fifty-five dementia-related predictors were gathered and categorized into clinical and medical history, questionnaire, cognition, polygenetic risk, and neuroimaging domains. Incident dementia (all-cause) and the subtypes, Alzheimer's disease (AD) and vascular dementia (VaD), were determined through hospital and death registries. Ensemble machine learning (ML) DRPMs were employed for prediction. The incremental values of risk predictors were assessed using the percent change in Area Under the Curve (∆AUC%) and the net reclassification index (NRI).

RESULTS

The simple DRPM which included age, body mass index, sex, education, diabetes, hyperlipidaemia, hypertension, depression, smoking, and alcohol consumption yielded an AUC of 0.711 (± 0.008 SD). The five predictor domains exhibited varying levels of incremental value over the basic model when predicting all-cause dementia and the two subtypes. Neuroimaging markers provided the highest incremental value in predicting all-cause dementia (∆AUC% +9.6%) and AD (∆AUC% +16.5%) while clinical and medical history data performed the best at predicting VaD (∆AUC% +12.2%). Combining clinical and medical history, and questionnaire data synergistically enhanced ML DRPM performance.

CONCLUSION

Combining predictors from different domains generally results in better predictive performance. Selecting predictors involves trade-offs, and while neuroimaging markers can significantly enhance predictive accuracy, they may pose challenges in terms of cost or accessibility.

Development and validation of a patient-level model to predict dementia across a network of observational databases

BACKGROUND

A prediction model can be a useful tool to quantify the risk of a patient developing dementia in the next years and take risk-factor-targeted intervention. Numerous dementia prediction models have been developed, but few have been externally validated, likely limiting their clinical uptake. In our previous work, we had limited success in externally validating some of these existing models due to inadequate reporting. As a result, we are compelled to develop and externally validate novel models to predict dementia in the general population across a network of observational databases. We assess regularization methods to obtain parsimonious models that are of lower complexity and easier to implement.

METHODS

Logistic regression models were developed across a network of five observational databases with electronic health records (EHRs) and claims data to predict 5-year dementia risk in persons aged 55-84. The regularization methods L1 and Broken Adaptive Ridge (BAR) as well as three candidate predictor sets to optimize prediction performance were assessed. The predictor sets include a baseline set using only age and sex, a full set including all available candidate predictors, and a phenotype set which includes a limited number of clinically relevant predictors.

RESULTS

BAR can be used for variable selection, outperforming L1 when a parsimonious model is desired. Adding candidate predictors for disease diagnosis and drug exposure generally improves the performance of baseline models using only age and sex. While a model trained on German EHR data saw an increase in AUROC from 0.74 to 0.83 with additional predictors, a model trained on US EHR data showed only minimal improvement from 0.79 to 0.81 AUROC. Nevertheless, the latter model developed using BAR regularization on the clinically relevant predictor set was ultimately chosen as best performing model as it demonstrated more consistent external validation performance and improved calibration.

CONCLUSIONS

We developed and externally validated patient-level models to predict dementia. Our results show that although dementia prediction is highly driven by demographic age, adding predictors based on condition diagnoses and drug exposures further improves prediction performance. BAR regularization outperforms L1 regularization to yield the most parsimonious yet still well-performing prediction model for dementia.

Rural-Urban mild cognitive impairment comparison in West Michigan through EHR

INTRODUCTION

Mild cognitive impairment (MCI) is a significant public health concern and a potential precursor to Alzheimer's disease (AD). This study leverages electronic health record (EHR) data to explore rural-urban differences in MCI incidence, risk factors, and healthcare navigation in West Michigan.

METHODS

Analysis was conducted on 1,528,464 patients from Corewell Health West, using face-to-face encounters between 1/1/2015 and 7/31/2022. MCI cases were identified using International Classification of Diseases (ICD) codes, focusing on patients aged 45+ without prior MCI, dementia, or AD diagnoses. Incidence rates, cumulative incidences, primary care physicians (PCPs), and neuropsychology referral outcomes were examined across rural and urban areas. Risk factors were evaluated through univariate and multivariate Cox regression analyses. The geographic distribution of patient counts, hospital locations, and neurology department referrals were examined.

RESULTS

Among 423,592 patients, a higher MCI incidence rate was observed in urban settings compared to rural settings (3.83 vs. 3.22 per 1,000 person-years). However, sensitivity analysis revealed higher incidence rates in rural areas when including patients who progressed directly to dementia. Urban patients demonstrated higher rates of referrals to and completion of neurological services. While the risk factors for MCI were largely similar across urban and rural populations, urban-specific factors for incident MCI are hearing loss, inflammatory bowel disease, obstructive sleep apnea, insomnia, being African American, and being underweight. Common risk factors include diabetes, intracranial injury, cerebrovascular disease, coronary artery disease, stroke, Parkinson's disease, epilepsy, chronic obstructive pulmonary disease, depression, and increased age. Lower risk was associated with being female, having a higher body mass index, and having a higher diastolic blood pressure.

DISCUSSION

This study highlights rural-urban differences in MCI incidence and access to care, suggesting potential underdiagnosis in rural areas likely due to reduced access to specialists. Future research should explore socioeconomic, environmental, and lifestyle determinants of MCI to refine prevention and management strategies across geographic settings.

Highlights

Leveraged EHRs to explore rural-urban differences in MCI in West Michigan.Revealed a significant underdiagnosis of MCI, especially in rural areas.Observed lower rates of neurological referrals and completions for rural patients.Identified risk factors specific to rural and urban populations.

Peripheral GFAP and NfL as early biomarkers for dementia: longitudinal insights from the UK Biobank

BACKGROUND

Peripheral glial fibrillary acidic protein (GFAP) and neurofilament light chain (NfL) are sensitive markers of neuroinflammation and neuronal damage. Previous studies with highly selected participants have shown that peripheral GFAP and NfL levels are elevated in the pre-clinical phase of Alzheimer's disease (AD) and dementia. However, the predictive value of GFAP and NfL for dementia requires more evidence from population-based cohorts.

METHODS

This was a prospective cohort study to evaluate UK Biobank participants enrolled from 2006 to 2010 using plasma GFAP and NfL measurements measured by Olink Target Platform and prospectively followed up for dementia diagnosis. Primary outcome was the risk of clinical diagnosed dementia. Secondary outcomes were cognition. Linear regression was used to assess the associations between peripheral GFAP and NfL with cognition. Cox proportional hazard models with cross-validations were used to estimate associations between elevated GFAP and NfL with risk of dementia. All models were adjusted for covariates.

RESULTS

A subsample of 48,542 participants in the UK Biobank with peripheral GFAP and NfL measurements were evaluated. With an average follow-up of 13.18 ± 2.42 years, 1312 new all-cause dementia cases were identified. Peripheral GFAP and NfL increased up to 15 years before dementia diagnosis was made. After strictly adjusting for confounders, increment in NfL was found to be associated with decreased numeric memory and prolonged reaction time. A greater annualized rate of change in GFAP was significantly associated with faster global cognitive decline. Elevation of GFAP (hazard ratio (HR) ranges from 2.25 to 3.15) and NfL (HR ranges from 1.98 to 4.23) increased the risk for several types of dementia. GFAP and NfL significantly improved the predictive values for dementia using previous models (area under the curve (AUC) ranges from 0.80 to 0.89, C-index ranges from 0.86 to 0.91). The AD genetic risk score and number of APOE*E4 alleles strongly correlated with GFAP and NfL levels.

CONCLUSIONS

These results suggest that peripheral GFAP and NfL are potential biomarkers for the early diagnosis of dementia. In addition, anti-inflammatory therapies in the initial stages of dementia may have potential benefits.

Identifying dementia from cognitive footprints in hospital records among Chinese older adults: a machine-learning study

Background

By combining theory-driven and data-driven methods, this study aimed to develop dementia predictive algorithms among Chinese older adults guided by the cognitive footprint theory.

Methods

Electronic medical records from the Clinical Data Analysis and Reporting System in Hong Kong were employed. We included patients with dementia diagnosed at 65+ between 2010 and 2018, and 1:1 matched dementia-free controls. We identified 51 features, comprising exposures to established modifiable factors and other factors before and after 65 years old. The performances of four machine learning models, including LASSO, Multilayer perceptron (MLP), XGBoost, and LightGBM, were compared with logistic regression models, for all patients and subgroups by age.

Findings

A total of 159,920 individuals (40.5% male; mean age [SD]: 83.97 [7.38]) were included. Compared with the model included established modifiable factors only (area under the curve [AUC] 0.689, 95% CI [0.684, 0.694]), the predictive accuracy substantially improved for models with all factors (0.774, [0.770, 0.778]). Machine learning and logistic regression models performed similarly, with AUC ranged between 0.773 (0.768, 0.777) for LASSO and 0.780 (0.776, 0.784) for MLP. Antipsychotics, education, antidepressants, head injury, and stroke were identified as the most important predictors in the total sample. Age-specific models identified different important features, with cardiovascular and infectious diseases becoming prominent in older ages.

Interpretation

The models showed satisfactory performances in identifying dementia. These algorithms can be used in clinical practice to assist decision making and allow timely interventions cost-effectively.

Funding

The Research Grants Council of Hong Kong under the Early Career Scheme 27110519.

Data-driven clustering approach to identify novel clusters of high cognitive impairment risk among Chinese community-dwelling elderly people with normal cognition: A national cohort study

Background

Cognitive impairment is a highly heterogeneous disorder that necessitates further investigation into the distinct characteristics of populations at varying risk levels of cognitive impairment. Using a large-scale registry cohort of elderly individuals, we applied a data-driven approach to identify novel clusters based on diverse sociodemographic features.

Methods

A prospective cohort of 6398 elderly people from the Chinese Longitudinal Healthy Longevity Survey, followed between 2008-14, was used to develop and validate the model. Participants were aged ≥60 years, community-dwelling, and the Chinese version of the Mini-Mental State Examination (MMSE) score ≥18 were included. Sixty-nine sociodemographic features were included in the analysis. The total population was divided into two-thirds for the derivation cohort (n = 4265) and one-third for the validation cohort (n = 2133). In the derivation cohort, an unsupervised Gaussian mixture model was applied to categorise participants into distinct clusters. A classifier was developed based on the most important 10 factors and was applied to categorise participants into their corresponding clusters in a validation cohort. The difference in the three-year risk of cognitive impairment was compared across the clusters.

Results

We identified four clusters with distinct features in the derivation cohort. Cluster 1 was associated with the worst life independence, longest sleep duration, and the oldest age. Cluster 2 demonstrated the highest loneliness, characterised by non-marital status and living alone. Cluster 3 was characterised by the lowest sense of loneliness and the highest proportions in marital status and family co-residence. Cluster 4 demonstrated heightened engagement in exercise and leisure activity, along with independent decision-making, hygiene, and a diverse diet. In comparison to Cluster 4, Cluster 1 exhibited the highest three-year cognitive impairment risk (adjusted odds ratio (aOR) = 3.31; 95% confidence interval (CI) = 1.81-6.05), followed by Cluster 2 and Cluster 3 after adjustment for baseline MMSE, residence, sex, age, years of education, drinking, smoking, hypertension, diabetes, heart disease and stroke or cardiovascular diseases.

Conclusions

A data-driven approach can be instrumental in identifying individuals at high risk of cognitive impairment among cognitively normal elderly populations. Based on various sociodemographic features, these clusters can suggest individualised intervention plans.

Development and validation of cognitive ageing risk score (CARS) for early detection of subtle cognitive deficits in older people

BACKGROUND

Early cognitive deficits commonly seen in older people have not been well defined and managed in primary care. The objectives are (1) to develop and validate a new risk score to estimate the risk of dementia in Chinese older population; and (2) to evaluate the use of risk score in conjunction with cognitive screening in detecting early cognitive deficits in community older people.

METHODS

A development cohort of 306 cognitive healthy older adults aged 60 or above were followed for 6 years. A CARS was constructed using the estimated coefficients of risk factors associated with dementia at follow up. Validation was carried out in another five-year cohort of 383 older adults. The usefulness of CARS in detecting early cognitive deficits was evaluated.

RESULTS

Risk factors include older age, male gender, low level of education, poorly controlled diabetes, prolonged sleep latency, fewer mind body or light exercise, loneliness, and being apolipoprotein e4 carriers. A cutoff of CARS at -1.3 had a sensitivity of 83.9% and a specificity of 75.4% to predict dementia. The area under curve was 82.5% in the development cohort. Early cognitive deficits were characterized by impaired retention (p <.001, 95% CI 0.2-0.9) and attention (p =.012, 95% CI 0.1-0.8).

CONCLUSION

The CARS can be used as a standard risk assessment of dementia or in conjunction with a computerized cognitive screening to evaluate a full cognitive profile for detecting early cognitive deficits. The result put forward the integration of risk algorithm into smart healthcare system to provide personalized lifestyle interventions.

Artificial intelligence approaches for early detection of neurocognitive disorders among older adults

Introduction

Dementia is one of the major global health issues among the aging population, characterized clinically by a progressive decline in higher cognitive functions. This paper aims to apply various artificial intelligence (AI) approaches to detect patients with mild cognitive impairment (MCI) or dementia accurately.

Methods

Quantitative research was conducted to address the objective of this study using randomly selected 343 Saudi patients. The Chi-square test was conducted to determine the association of the patient's cognitive function with various features, including demographical and medical history. Two widely used AI algorithms, logistic regression and support vector machine (SVM), were used for detecting cognitive decline. This study also assessed patients' cognitive function based on gender and developed the predicting models for males and females separately.

Results

Fifty four percent of patients have normal cognitive function, 34% have MCI, and 12% have dementia. The prediction accuracies for all the developed models are greater than 71%, indicating good prediction capability. However, the developed SVM models performed the best, with an accuracy of 93.3% for all patients, 94.4% for males only, and 95.5% for females only. The top 10 significant predictors based on the developed SVM model are education, bedtime, taking pills for chronic pain, diabetes, stroke, gender, chronic pains, coronary artery diseases, and wake-up time.

Conclusion

The results of this study emphasize the higher accuracy and reliability of the proposed methods in cognitive decline prediction that health practitioners can use for the early detection of dementia. This research can also stipulate substantial direction and supportive intuitions for scholars to enhance their understanding of crucial research, emerging trends, and new developments in future cognitive decline studies.

Incorporating frailty to address the key challenges to geriatric economic evaluation

BACKGROUND

The multidimensional and dynamically complex process of ageing presents key challenges to economic evaluation of geriatric interventions, including: (1) accounting for indirect, long-term effects of a geriatric shock such as a fall; (2) incorporating a wide range of societal, non-health outcomes such as informal caregiver burden; and (3) accounting for heterogeneity within the demographic group. Measures of frailty aim to capture the multidimensional and syndromic nature of geriatric health. Using a case study of community-based falls prevention, this article explores how incorporating a multivariate frailty index in a decision model can help address the above key challenges.

METHODS

A conceptual structure of the relationship between geriatric shocks and frailty was developed. This included three key associations involving frailty: (A) the shock-frailty feedback loop; (B) the secondary effects of shock via frailty; and (C) association between frailty and intervention access. A case study of economic modelling of community-based falls prevention for older persons aged 60 + was used to show how parameterising these associations contributed to addressing the above three challenges. The English Longitudinal Study of Ageing (ELSA) was the main data source for parameterisation. A new 52-item multivariate frailty index was generated from ELSA. The main statistical methods were multivariate logistic and linear regressions. Estimated regression coefficients were inputted into a discrete individual simulation with annual cycles to calculate the continuous variable value or probability of binary event given individuals' characteristics.

RESULTS

All three conceptual associations, in their parameterised forms, contributed to addressing challenge (1). Specifically, by worsening the frailty progression, falls incidence in the model increased the risk of falling in subsequent cycles and indirectly impacted the trajectories and levels of EQ-5D-3 L, mortality risk, and comorbidity care costs. Intervention access was positively associated with frailty such that the greater access to falls prevention by frailer individuals dampened the falls-frailty feedback loop. Association (B) concerning the secondary effects of falls via frailty was central to addressing challenge (2). Using this association, the model was able to estimate how falls prevention generated via its impact on frailty paid and unpaid productivity gains, out-of-pocket care expenditure reduction, and informal caregiving cost reduction. For challenge (3), frailty captured the variations within demographic groups of key model outcomes including EQ-5D-3 L, QALY, and all-cause care costs. Frailty itself was shown to have a social gradient such that it mediated socially inequitable distributions of frailty-associated outcomes.

CONCLUSION

The frailty-based conceptual structure and parameterisation methods significantly improved upon the methods previously employed by falls prevention models to address the key challenges for geriatric economic evaluation. The conceptual structure is applicable to other geriatric and non-geriatric intervention areas and should inform the data selection and statistical methods to parameterise structurally valid economic models of geriatric interventions.

Incident dementia risk among patients with type 2 diabetes receiving metformin versus alternative oral glucose-lowering therapy: an observational cohort study using UK primary healthcare records

INTRODUCTION

4.2 million individuals in the UK have type 2 diabetes, a known risk factor for dementia and mild cognitive impairment (MCI). Diabetes treatment may modify this association, but existing evidence is conflicting. We therefore aimed to assess the association between metformin therapy and risk of incident all-cause dementia or MCI compared with other oral glucose-lowering therapies (GLTs).

RESEARCH DESIGN AND METHODS

We conducted an observational cohort study using the Clinical Practice Research Datalink among UK adults diagnosed with diabetes at ≥40 years between 1990 and 2019. We used an active comparator new user design to compare risks of dementia and MCI among individuals initially prescribed metformin versus an alternative oral GLT using Cox proportional hazards regression controlling for sociodemographic, lifestyle and clinical confounders. We assessed for interaction by age and sex. Sensitivity analyses included an as-treated analysis to mitigate potential exposure misclassification.

RESULTS

We included 211 396 individuals (median age 63 years; 42.8% female), of whom 179 333 (84.8%) initiated on metformin therapy. Over median follow-up of 5.4 years, metformin use was associated with a lower risk of dementia (adjusted HR (aHR) 0.86 (95% CI 0.79 to 0.94)) and MCI (aHR 0.92 (95% CI 0.86 to 0.99)). Metformin users aged under 80 years had a lower dementia risk (aHR 0.77 (95% CI 0.68 to 0.85)), which was not observed for those aged ≥80 years (aHR 0.95 (95% CI 0.87 to 1.05)). There was no interaction with sex. The as-treated analysis showed a reduced effect size compared with the main analysis (aHR 0.90 (95% CI 0.83 to 0.98)).

CONCLUSIONS

Metformin use was associated with lower risks of incident dementia and MCI compared with alternative GLT among UK adults with diabetes. While our findings are consistent with a neuroprotective effect of metformin against dementia, further research is needed to reduce risks of confounding by indication and assess causality.

What's New in Dementia Risk Prediction Modelling? An Updated Systematic Review

Introduction

Identifying individuals at high risk of dementia is critical to optimized clinical care, formulating effective preventative strategies, and determining eligibility for clinical trials. Since our previous systematic reviews in 2010 and 2015, there has been a surge in dementia risk prediction modelling. The aim of this study was to update our previous reviews to explore, and critically review, new developments in dementia risk modelling.

Methods

MEDLINE, Embase, Scopus, and Web of Science were searched from March 2014 to June 2022. Studies were included if they were population- or community-based cohorts (including electronic health record data), had developed a model for predicting late-life incident dementia, and included model performance indices such as discrimination, calibration, or external validation.

Results

In total, 9,209 articles were identified from the electronic search, of which 74 met the inclusion criteria. We found a substantial increase in the number of new models published from 2014 (>50 new models), including an increase in the number of models developed using machine learning. Over 450 unique predictor (component) variables have been tested. Nineteen studies (26%) undertook external validation of newly developed or existing models, with mixed results. For the first time, models have also been developed in low- and middle-income countries (LMICs) and others validated in racial and ethnic minority groups.

Conclusion

The literature on dementia risk prediction modelling is rapidly evolving with new analytical developments and testing in LMICs. However, it is still challenging to make recommendations about which one model is the most suitable for routine use in a clinical setting. There is an urgent need to develop a suitable, robust, validated risk prediction model in the general population that can be widely implemented in clinical practice to improve dementia prevention.

Associations Between Longitudinal Loneliness, DNA Methylation Age Acceleration, and Cognitive Functioning

OBJECTIVES

Loneliness may influence aging biomarkers related to cognitive functioning, for example, through accelerated DNA methylation (DNAm) aging.

METHODS

In the present study, we tested whether six common DNAm age acceleration measures mediated the effects of baseline loneliness and five different longitudinal loneliness trajectories on general cognitive ability, immediate memory recall, delayed memory recall, and processing speed in 1,814 older adults in the Health and Retirement Study.

RESULTS

We found that baseline loneliness and individuals who belong to the highest loneliness trajectories had poorer general cognitive ability and memory scores. Only DNAm age acceleration measures that index physiological comorbidities, unhealthy lifestyle factors (e.g., smoking), and mortality risk-mediated effects of baseline loneliness on general cognitive ability and memory functioning but not processing speed. These same DNAm measures mediated effects of the moderate-but-declining loneliness trajectory on cognitive functioning. Additionally, immediate and delayed memory scores were mediated by GrimAge Accel in the lowest and two highest loneliness trajectory groups. Total and mediated effects of loneliness on cognitive functioning outcomes were mainly accounted for by demographic, social, psychological, and physiological covariates, most notably self-rated health, depressive symptomatology, objective social isolation, and body mass index.

DISCUSSION

Current findings suggest that DNAm biomarkers of aging, particularly GrimAge Accel, have promise for explaining the prospective association between loneliness and cognitive functioning outcomes.

Multipredictor risk models for predicting individual risk of Alzheimer's disease

BACKGROUND

Early prevention of Alzheimer's disease (AD) is a feasible way to delay AD onset and progression. Information on AD prediction at the individual patient level will be useful in AD prevention. In this study, we aim to develop risk models for predicting AD onset at individual level using optimal set of predictors from multiple features.

METHODS

A total of 487 cognitively normal (CN) individuals and 796 mild cognitive impairment (MCI) patients were included from Alzheimer's Disease Neuroimaging Initiative. All the participants were assessed for clinical, cognitive, magnetic resonance imaging and cerebrospinal fluid (CSF) markers and followed for mean periods of 5.6 years for CN individuals and 4.6 years for MCI patients to ascertain progression from CN to incident prodromal stage of AD or from MCI to AD dementia. Least Absolute Shrinkage and Selection Operator Cox regression was applied for predictors selection and model construction.

RESULTS

During the follow-up periods, 139 CN participants had progressed to prodromal AD (CDR ≥ 0.5) and 321 MCI patients had progressed to AD dementia. In the prediction of individual risk of incident prodromal stage of AD in CN individuals, the AUC of the final CN model was 0.81 within 5 years. The final MCI model predicted individual risk of AD dementia in MCI patients with an AUC of 0.92 within 5 years. The models were also associated with longitudinal change of Mini-Mental State Examination (p < 0.001 for CN and MCI models). An Alzheimer's continuum model was developed which could predict the Alzheimer's continuum for individuals with normal AD biomarkers within 3 years with high accuracy (AUC = 0.91).

CONCLUSIONS

The risk models were able to provide personalized risk for AD onset at each year after evaluation. The models may be useful for better prevention of AD.

Development and validation of a dementia risk score in the UK Biobank and Whitehall II cohorts

BACKGROUND

Current dementia risk scores have had limited success in consistently identifying at-risk individuals across different ages and geographical locations.

OBJECTIVE

We aimed to develop and validate a novel dementia risk score for a midlife UK population, using two cohorts: the UK Biobank, and UK Whitehall II study.

METHODS

We divided the UK Biobank cohort into a training (n=176 611, 80%) and test sample (n=44 151, 20%) and used the Whitehall II cohort (n=2934) for external validation. We used the Cox LASSO regression to select the strongest predictors of incident dementia from 28 candidate predictors and then developed the risk score using competing risk regression.

FINDINGS

Our risk score, termed the UK Biobank Dementia Risk Score (UKBDRS), consisted of age, education, parental history of dementia, material deprivation, a history of diabetes, stroke, depression, hypertension, high cholesterol, household occupancy, and sex. The score had a strong discrimination accuracy in the UK Biobank test sample (area under the curve (AUC) 0.8, 95% CI 0.78 to 0.82) and in the Whitehall cohort (AUC 0.77, 95% CI 0.72 to 0.81). The UKBDRS also significantly outperformed three other widely used dementia risk scores originally developed in cohorts in Australia (the Australian National University Alzheimer's Disease Risk Index), Finland (the Cardiovascular Risk Factors, Ageing, and Dementia score), and the UK (Dementia Risk Score).

CLINICAL IMPLICATIONS

Our risk score represents an easy-to-use tool to identify individuals at risk for dementia in the UK. Further research is required to determine the validity of this score in other populations.

Multi-domain prognostic models used in middle-aged adults without known cognitive impairment for predicting subsequent dementia

BACKGROUND

Dementia, a global health priority, has no current cure. Around 50 million people worldwide currently live with dementia, and this number is expected to treble by 2050. Some health conditions and lifestyle behaviours can increase or decrease the risk of dementia and are known as 'predictors'. Prognostic models combine such predictors to measure the risk of future dementia. Models that can accurately predict future dementia would help clinicians select high-risk adults in middle age and implement targeted risk reduction.

OBJECTIVES

Our primary objective was to identify multi-domain prognostic models used in middle-aged adults (aged 45 to 65 years) for predicting dementia or cognitive impairment. Eligible multi-domain prognostic models involved two or more of the modifiable dementia predictors identified in a 2020 Lancet Commission report and a 2019 World Health Organization (WHO) report (less education, hearing loss, traumatic brain injury, hypertension, excessive alcohol intake, obesity, smoking, depression, social isolation, physical inactivity, diabetes mellitus, air pollution, poor diet, and cognitive inactivity). Our secondary objectives were to summarise the prognostic models, to appraise their predictive accuracy (discrimination and calibration) as reported in the development and validation studies, and to identify the implications of using dementia prognostic models for the management of people at a higher risk for future dementia.

SEARCH METHODS

We searched MEDLINE, Embase, PsycINFO, CINAHL, and ISI Web of Science Core Collection from inception until 6 June 2022. We performed forwards and backwards citation tracking of included studies using the Web of Science platform.  SELECTION CRITERIA: We included development and validation studies of multi-domain prognostic models. The minimum eligible follow-up was five years. Our primary outcome was an incident clinical diagnosis of dementia based on validated diagnostic criteria, and our secondary outcome was dementia or cognitive impairment determined by any other method.

DATA COLLECTION AND ANALYSIS

Two review authors independently screened the references, extracted data using a template based on the CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies (CHARMS), and assessed risk of bias and applicability of included studies using the Prediction model Risk Of Bias ASsessment Tool (PROBAST). We synthesised the C-statistics of models that had been externally validated in at least three comparable studies.  MAIN RESULTS: We identified 20 eligible studies; eight were development studies and 12 were validation studies. There were 14 unique prognostic models: seven models with validation studies and seven models with development-only studies. The models included a median of nine predictors (range 6 to 34); the median number of modifiable predictors was five (range 2 to 11). The most common modifiable predictors in externally validated models were diabetes, hypertension, smoking, physical activity, and obesity. In development-only models, the most common modifiable predictors were obesity, diabetes, hypertension, and smoking. No models included hearing loss or air pollution as predictors. Nineteen studies had a high risk of bias according to the PROBAST assessment, mainly because of inappropriate analysis methods, particularly lack of reported calibration measures. Applicability concerns were low for 12 studies, as their population, predictors, and outcomes were consistent with those of interest for this review. Applicability concerns were high for nine studies, as they lacked baseline cognitive screening or excluded an age group within the range of 45 to 65 years. Only one model, Cardiovascular Risk Factors, Ageing, and Dementia (CAIDE), had been externally validated in multiple studies, allowing for meta-analysis. The CAIDE model included eight predictors (four modifiable predictors): age, education, sex, systolic blood pressure, body mass index (BMI), total cholesterol, physical activity and APOEƐ4 status. Overall, our confidence in the prediction accuracy of CAIDE was very low; our main reasons for downgrading the certainty of the evidence were high risk of bias across all the studies, high concern of applicability, non-overlapping confidence intervals (CIs), and a high degree of heterogeneity. The summary C-statistic was 0.71 (95% CI 0.66 to 0.76; 3 studies; very low-certainty evidence) for the incident clinical diagnosis of dementia, and 0.67 (95% CI 0.61 to 0.73; 3 studies; very low-certainty evidence) for dementia or cognitive impairment based on cognitive scores. Meta-analysis of calibration measures was not possible, as few studies provided these data.

AUTHORS' CONCLUSIONS

We identified 14 unique multi-domain prognostic models used in middle-aged adults for predicting subsequent dementia. Diabetes, hypertension, obesity, and smoking were the most common modifiable risk factors used as predictors in the models. We performed meta-analyses of C-statistics for one model (CAIDE), but the summary values were unreliable. Owing to lack of data, we were unable to meta-analyse the calibration measures of CAIDE. This review highlights the need for further robust external validations of multi-domain prognostic models for predicting future risk of dementia in middle-aged adults.

A flexible modeling approach for biomarker-based computation of absolute risk of Alzheimer's disease dementia

INTRODUCTION

As Alzheimer's disease (AD) biomarkers rapidly develop, tools are needed that accurately and effectively communicate risk of AD dementia.

METHODS

We analyzed longitudinal data from >10,000 cognitively unimpaired older adults. Five-year risk of AD dementia was modeled using survival analysis.

RESULTS

A demographic model was developed and validated on independent data with area under the receiver operating characteristic curve (AUC) for 5-year prediction of AD dementia of 0.79. Clinical and cognitive variables (AUC = 0.79), and apolipoprotein E genotype (AUC = 0.76) were added to the demographic model. We then incorporated the risk computed from the demographic model with hazard ratios computed from independent data for amyloid positron emission tomography status and magnetic resonance imaging hippocampal volume (AUC = 0.84), and for plasma amyloid beta (Aβ)42/Aβ40 (AUC = 0.82).

DISCUSSION

An adaptive tool was developed and validated to compute absolute risks of AD dementia. This approach allows for improved accuracy and communication of AD risk among cognitively unimpaired older adults.

Can Patients with Dementia Be Identified in Primary Care Electronic Medical Records Using Natural Language Processing?

Dementia and mild cognitive impairment can be underrecognized in primary care practice and research. Free-text fields in electronic medical records (EMRs) are a rich source of information which might support increased detection and enable a better understanding of populations at risk of dementia. We used natural language processing (NLP) to identify dementia-related features in EMRs and compared the performance of supervised machine learning models to classify patients with dementia. We assembled a cohort of primary care patients aged 66 + years in Ontario, Canada, from EMR notes collected until December 2016: 526 with dementia and 44,148 without dementia. We identified dementia-related features by applying published lists, clinician input, and NLP with word embeddings to free-text progress and consult notes and organized features into thematic groups. Using machine learning models, we compared the performance of features to detect dementia, overall and during time periods relative to dementia case ascertainment in health administrative databases. Over 900 dementia-related features were identified and grouped into eight themes (including symptoms, social, function, cognition). Using notes from all time periods, LASSO had the best performance (F1 score: 77.2%, sensitivity: 71.5%, specificity: 99.8%). Model performance was poor when notes written before case ascertainment were included (F1 score: 14.4%, sensitivity: 8.3%, specificity 99.9%) but improved as later notes were added. While similar models may eventually improve recognition of cognitive issues and dementia in primary care EMRs, our findings suggest that further research is needed to identify which additional EMR components might be useful to promote early detection of dementia.

Supplementary Information

The online version contains supplementary material available at 10.1007/s41666-023-00125-6.

Global incidence of young-onset dementia: A systematic review and meta-analysis

INTRODUCTION

Reliable data on the incidence rates for young-onset dementia (YOD) are lacking, but are necessary for research on disease etiology and to raise awareness among health care professionals.

METHODS

We performed a systematic review and meta-analysis on population-based studies on the incidence of YOD, published between January 1, 1990 and February 1, 2022, according to Meta-analyses of Observational Studies in Epidemiology (MOOSE) guidelines. Data were analyzed using random-effects meta-analyses. Results were age-standardized, and heterogeneity was assessed by subgroup analyses and meta-regression.

RESULTS

Sixty-one articles were included. Global age-standardized incidence rates increased from 0.17/100,000 in age 30 to 34 years, to 5.14/100,000 in age 60 to 64 years, giving a global total age-standardized incidence rate of 11 per 100,000 in age 30 to 64. This corresponds to 370,000 new YOD cases annually worldwide. Heterogeneity was high and meta-regression showed geographic location significantly influenced this heterogeneity.

DISCUSSION

This meta-analysis shows the current best estimate of YOD incidence. New prospective cohort studies are needed.

Data-driven decision making for the screening of cognitive impairment in primary care: a machine learning approach using data from the ELSA-Brasil study

The systematic assessment of cognitive performance of older people without cognitive complaints is controversial and unfeasible. Identifying individuals at higher risk of cognitive impairment could optimize resource allocation. We aimed to develop and test machine learning models to predict cognitive impairment using variables obtainable in primary care settings. In this cross-sectional study, we included 8,291 participants of the baseline assessment of the ELSA-Brasil study, who were aged between 50 and 74 years and were free of dementia. Cognitive performance was assessed with a neuropsychological battery and cognitive impairment was defined as global cognitive z-score below 2 standard deviations. Variables used as input to the prediction models included demographics, social determinants, clinical conditions, family history, lifestyle, and laboratory tests. We developed machine learning models using logistic regression, neural networks, and gradient boosted trees. Participants' mean age was 58.3±6.2 years, 55% were female. Cognitive impairment was present in 328 individuals (4%). Machine learning algorithms presented fair to good discrimination (areas under the ROC curve between 0.801 and 0.873). Extreme Gradient Boosting presented the highest discrimination, high specificity (97%), and negative predictive value (97%). Seventy-six percent of the individuals with cognitive impairment were included among the highest ranked individuals by this algorithm. In conclusion, we developed and tested a machine learning model to predict cognitive impairment based on primary care data that presented good discrimination and high specificity. These characteristics could support the detection of patients who would not benefit from cognitive assessment, facilitating the allocation of human and economic resources.

External validation of existing dementia prediction models on observational health data

BACKGROUND

Many dementia prediction models have been developed, but only few have been externally validated, which hinders clinical uptake and may pose a risk if models are applied to actual patients regardless. Externally validating an existing prediction model is a difficult task, where we mostly rely on the completeness of model reporting in a published article. In this study, we aim to externally validate existing dementia prediction models. To that end, we define model reporting criteria, review published studies, and externally validate three well reported models using routinely collected health data from administrative claims and electronic health records.

METHODS

We identified dementia prediction models that were developed between 2011 and 2020 and assessed if they could be externally validated given a set of model criteria. In addition, we externally validated three of these models (Walters' Dementia Risk Score, Mehta's RxDx-Dementia Risk Index, and Nori's ADRD dementia prediction model) on a network of six observational health databases from the United States, United Kingdom, Germany and the Netherlands, including the original development databases of the models.

RESULTS

We reviewed 59 dementia prediction models. All models reported the prediction method, development database, and target and outcome definitions. Less frequently reported by these 59 prediction models were predictor definitions (52 models) including the time window in which a predictor is assessed (21 models), predictor coefficients (20 models), and the time-at-risk (42 models). The validation of the model by Walters (development c-statistic: 0.84) showed moderate transportability (0.67-0.76 c-statistic). The Mehta model (development c-statistic: 0.81) transported well to some of the external databases (0.69-0.79 c-statistic). The Nori model (development AUROC: 0.69) transported well (0.62-0.68 AUROC) but performed modestly overall. Recalibration showed improvements for the Walters and Nori models, while recalibration could not be assessed for the Mehta model due to unreported baseline hazard.

CONCLUSION

We observed that reporting is mostly insufficient to fully externally validate published dementia prediction models, and therefore, it is uncertain how well these models would work in other clinical settings. We emphasize the importance of following established guidelines for reporting clinical prediction models. We recommend that reporting should be more explicit and have external validation in mind if the model is meant to be applied in different settings.

Development of a novel dementia risk prediction model in the general population: A large, longitudinal, population-based machine-learning study

BACKGROUND

The existing dementia risk models are limited to known risk factors and traditional statistical methods. We aimed to employ machine learning (ML) to develop a novel dementia prediction model by leveraging a rich-phenotypic variable space of 366 features covering multiple domains of health-related data.

METHODS

In this longitudinal population-based cohort of the UK Biobank (UKB), 425,159 non-demented participants were enrolled from 22 recruitment centres across the UK between March 1, 2006 and October 31, 2010. We implemented a data-driven strategy to identify predictors from 366 candidate variables covering a comprehensive range of genetic and environmental factors and developed the ML model to predict incident dementia and Alzheimer's Disease (AD) within five, ten, and much longer years (median 11.9 [Interquartile range 11.2-12.5] years).

FINDINGS

During a follow-up of 5,023,337 person-years, 5287 and 2416 participants developed dementia and AD, respectively. A novel UKB dementia risk prediction (UKB-DRP) model comprising ten predictors including age, ApoE ε4, pairs matching time, leg fat percentage, number of medications taken, reaction time, peak expiratory flow, mother's age at death, long-standing illness, and mean corpuscular volume was established. Our prediction model was internally evaluated based on five-fold cross-validation on discrimination and calibration, and it was further compared with existing prediction scales. The UKB-DRP model can achieve high discriminative accuracy in dementia (AUC 0.848 ± 0.007) and even better in AD (AUC 0.862 ± 0.015). The model was well-calibrated (Hosmer-Lemeshow goodness-of-fit p-value = 0.92), and the predictive power was solid in different incidence time groups. More importantly, our model presented an apparent superiority over existing models like Cardiovascular Risk Factors, Aging, and Incidence of Dementia Risk Score (AUC 0.705 ± 0.008), the Dementia Risk Score (AUC 0.752 ± 0.007), and the Australian National University Alzheimer's Disease Risk Index (AUC 0.584 ± 0.017). The model was internally validated in the general population of European ancestry and White ethnicity; thus, further validation with independent datasets is necessary to confirm these findings.

INTERPRETATION

Our ML-based UKB-DRP model incorporated ten easily accessible predictors with solid predictive power for incident dementia and AD within five, ten, and much longer years, which can be used to identify individuals at high risk of dementia and AD in the general population.

FUNDING

This study was funded by grants from the Science and Technology Innovation 2030 Major Projects (2022ZD0211600), National Key R&D Program of China (2018YFC1312904, 2019YFA070950), National Natural Science Foundation of China (282071201, 81971032, 82071997), Shanghai Municipal Science and Technology Major Project (2018SHZDZX01), Research Start-up Fund of Huashan Hospital (2022QD002), Excellence 2025 Talent Cultivation Program at Fudan University (3030277001), Shanghai Rising-Star Program (21QA1408700), Medical Engineering Fund of Fudan University (yg2021-013), and the 111 Project (No. B18015).

Weight loss in a cardiovascular trial population identifies people at future risk of dementia

Introduction

Populations at increased risk of dementia need to be identified for well-powered trials of preventive interventions. Weight loss, which often occurs in pre-clinical dementia, could identify a population at sufficiently high dementia risk.

Methods

In 12,975 survivors in the Heart Protection Study statin trial of people with, or at high risk of, cardiovascular disease, the association of weight change over 5 years during the trial with post-trial dementia recorded in electronic hospital admission and death records (n = 784) was assessed, after adjustment for age, sex, treatment allocation, and deprivation measures.

Results

Among the 60% without substantial weight gain (≤2 kg weight gain), each 1 kg weight loss was associated with a risk ratio for dementia of 1.04 (95% confidence interval, 1.02-1.07). Weight loss ≥4 kg and cognitive function below the mean identified participants aged ≥67 years with a 13% 10-year dementia risk.

Discussion

The combination of weight loss and high vascular risk identified individuals at high risk of dementia who could be recruited to dementia prevention trials.

Development and validation of a risk model for cognitive impairment in the older Chinese inpatients: An analysis based on a 5-year database

Early diagnosis of cognitive impairment is important but difficult. Prediction models may work as an efficient way to identify high risk individuals for this disease. This study aimed to develop a simple and convenient model to identify high-risk individuals of cognitive impairment in the older Chinese inpatients. We enrolled 1300 inpatients aged 60 years or above from the department of geriatrics of the First Affiliated Hospital of Chongqing Medical University during 2013 to 2017. The model for cognitive impairment was established in the developing cohort of 1100 participants and tested in another validating cohort of 200 participants. Logistic regression analyses were used to identify the candidate variables of cognitive impairment. Receiver operating curve was adopted to validate the model. Logistic regression analyses showed that increasing age, diabetes, depression and low educational level were independently associated with cognitive impairment. The model was generated in the following way: P_model = e^y/(1 + e^y), where y = -6.874 + 0.088 * age + 0.317 * diabetes + 0.647 * depression + 0.345 * education level. The value of P_model indicates the probability of cognitive impairment for each patient. The present model proved to be a reliable marker for identifying people at high risk of cognitive impairment (area under curve = 0.790, 95% CI = 0.728-0.852, p < 0.001). It had a high sensitivity (86.2%) but a relatively low specificity (59.4%). It may be helpful to "recognize" those at high risk of cognitive impairment rather than "rule out" those at low risk of this disease.

Risk factors for incidence of dementia in primary care practice: a retrospective cohort study in older adults

BACKGROUND

The dementias are long-term, chronic conditions caused by progressive neurological degeneration. Current literature suggests that cardiovascular disease risk factors may contribute to the onset of dementia; however, evidence of these associations is inconsistent.

OBJECTIVES

This study aimed to examine the impact of risk factors on dementia onset in older adults diagnosed and managed in Canadian primary care settings.

METHODS

A retrospective cohort study was employed utilizing electronic medical records data in the Canadian Primary Care Sentinel Surveillance Network (CPCSSN). Patients aged 65+ years with no dementia diagnosis at baseline who were followed from 2009 to 2017 with a run-in year to exclude existing undiagnosed dementia cases. Multivariate Cox proportional hazard models were used to estimate risk.

RESULTS

Age was associated with an increased incidence risk of dementia in both examined age groups: 65-79 years (13%) and 80+ years (5%). History of depression increased dementia risk by 38% and 34% in the age groups. There were significant associations with lower social deprivation area quintile, smoking history, osteoarthritis, and diabetes mellitus in patients aged 65-79 years but not in those aged 80+ years. Sex, hypertension, obesity, dyslipidemia, and the use of antihypertensive medications and statins were not associated with risk of incident dementia diagnosis.

CONCLUSIONS

The association between chronic health conditions and dementia onset is complicated. Primary care electronic medical record data might be useful for research in this topic, though follow-up time is still relatively short to observe a clear causal relationship. Future studies with more complete data may provide evidence for dementia preventive strategies within primary care practice.

AI-Based Predictive Modelling of the Onset and Progression of Dementia

Dementia, the most severe expression of cognitive impairment, is among the main causes of disability in older adults and currently affects over 55 million individuals. Dementia prevention is a global public health priority, and recent studies have shown that dementia risk can be reduced through non-pharmacological interventions targeting different lifestyle areas. The FINnish GERiatric Intervention Study to Prevent Cognitive Impairment and Disability (FINGER) has shown a positive effect on cognition in older adults at risk of dementia through a 2-year multidomain intervention targeting lifestyle and vascular risk factors. The LETHE project builds on these findings and will provide a digital-enabled FINGER intervention model for delaying or preventing the onset of cognitive decline. An individualised ICT-based multidomain, preventive lifestyle intervention program will be implemented utilising behaviour and intervention data through passive and active data collection. Artificial intelligence and machine learning methods will be used for data-driven risk factor prediction models. An initial model based on large multinational datasets will be validated and integrated into an 18-month trial integrating digital biomarkers to further improve the model. Furthermore, the LETHE project will investigate the concept of federated learning to, on the one hand, protect the privacy of the health and behaviour data and, on the other hand, to provide the opportunity to enhance the data model easily by integrating additional clinical centres.

Derivation and Validation of the Cognitive Impairment Prediction Model in Older Adults: A National Cohort Study

Objective

This prediction model quantifies the risk of cognitive impairment. This aim of this study was to develop and validate a prediction model to calculate the 6-year risk of cognitive impairment.

Methods

Participants from the Chinese Longitudinal Healthy Longevity Survey (CLHLS) 2008-2014 and 2011-2018 surveys were included for developing the cognitive impairment prediction model. The least absolute shrinkage and selection operator, clinical knowledge, and previous experience were performed to select predictors. The Cox proportional hazard model and Fine-Gray analysis adjusting for death were conducted to construct the model. The discriminative ability was measured using C-statistics. The model was evaluated externally using the temporal validation method via the CLHLS 2002-2008 survey. A nomogram was conducted to enhance the practical use. The population attributable fraction was calculated.

Results

A total of 10,053 older adults were included for model development. During a median of 5.68 years, 1,750 (17.4%) participants experienced cognitive impairment. Eight easy-to-obtain predictors were used to develop the model. The overall proportion of death was 43.3%. The effect of age on cognitive impairment reduced after adjusting the competing risk of death. The Cox and Fine-Gray models showed a similar discriminative ability, with average C-statistics of 0.71 and 0.69 in development and external validation datasets, respectively. The model performed better in younger older adults (65-74 years). The proportion of 6-year cognitive impairment due to modifiable risk factors was 47.7%.

Conclusion

This model could be used to identify older adults aged 65 years and above at high risk of cognitive impairment and initiate timely interventions on modifiable factors to prevent nearly half of dementia.

Risk Models Based on Non-Cognitive Measures May Identify Presymptomatic Alzheimer's Disease

BACKGROUND

Alzheimer's disease (AD) is a progressive disorder without a cure. Develop risk prediction models for detecting presymptomatic AD using non-cognitive measures is necessary to enable early interventions.

OBJECTIVE

Examine if non-cognitive metrics alone can be used to construct risk models to identify adults at risk for AD dementia and cognitive impairment.

METHODS

Clinical data from older adults without dementia from the Memory and Aging Project (MAP, n = 1,179) and Religious Orders Study (ROS, n = 1,103) were analyzed using Cox proportional hazard models to develop risk prediction models for AD dementia and cognitive impairment. Models using only non-cognitive covariates were compared to models that added cognitive covariates. All models were trained in MAP, tested in ROS, and evaluated by the AUC of ROC curve.

RESULTS

Models based on non-cognitive covariates alone achieved AUC (0.800,0.785) for predicting AD dementia (3.5) years from baseline. Including additional cognitive covariates improved AUC to (0.916,0.881). A model with a single covariate of composite cognition score achieved AUC (0.905,0.863). Models based on non-cognitive covariates alone achieved AUC (0.717,0.714) for predicting cognitive impairment (3.5) years from baseline. Including additional cognitive covariates improved AUC to (0.783,0.770). A model with a single covariate of composite cognition score achieved AUC (0.754,0.730).

CONCLUSION

Risk models based on non-cognitive metrics predict both AD dementia and cognitive impairment. However, non-cognitive covariates do not provide incremental predictivity for models that include cognitive metrics in predicting AD dementia, but do in models predicting cognitive impairment. Further improved risk prediction models for cognitive impairment are needed.

Detecting Cognitive Impairment Status Using Keystroke Patterns and Physical Activity Data among the Older Adults: A Machine Learning Approach

Cognitive impairment has a significantly negative impact on global healthcare and the community. Holding a person's cognition and mental retention among older adults is improbable with aging. Early detection of cognitive impairment will decline the most significant impact of extended disease to permanent mental damage. This paper aims to develop a machine learning model to detect and differentiate cognitive impairment categories like severe, moderate, mild, and normal by analyzing neurophysical and physical data. Keystroke and smartwatch have been used to extract individuals' neurophysical and physical data, respectively. An advanced ensemble learning algorithm named Gradient Boosting Machine (GBM) is proposed to classify the cognitive severity level (absence, mild, moderate, and severe) based on the Standardised Mini-Mental State Examination (SMMSE) questionnaire scores. The statistical method "Pearson's correlation" and the wrapper feature selection technique have been used to analyze and select the best features. Then, we have conducted our proposed algorithm GBM on those features. And the result has shown an accuracy of more than 94%. This paper has added a new dimension to the state-of-the-art to predict cognitive impairment by implementing neurophysical data and physical data together.

Atrial Fibrillation Is Associated with Cognitive Impairment, All-Cause Dementia, Vascular Dementia, and Alzheimer's Disease: a Systematic Review and Meta-Analysis

BACKGROUND

Atrial fibrillation (AF) is a risk factor for cognitive impairment and dementia in patients with stroke history. However, the association between AF and cognitive impairment in broader populations is less clear.

OBJECTIVE

To systematically review and quantitatively synthesize the existing evidence regarding the association of AF with cognitive impairment of any severity and etiology and dementia.

METHODS

Medline, Scopus, and Cochrane Central were searched in order to identify studies investigating the association between AF and cognitive impairment (or dementia) cross-sectionally and longitudinally. Studies encompassing and analyzing exclusively patients with stroke history were excluded. A random-effects model meta-analysis was conducted. Potential sources of between-study heterogeneity were investigated via subgroup and meta-regression analyses. Sensitivity analyses including only studies reporting data on stroke-free patients, vascular dementia, and Alzheimer's disease were performed.

RESULTS

In total, 43 studies were included. In the pooled analysis, AF was significantly associated with dementia (adjusted OR, 1.6; 95% CI, 1.3 to 2.1; I², 31%) and the combined endpoint of cognitive impairment or dementia (pooled adjusted OR, 1.5; 95% CI, 1.4 to 1.8; I², 34%). The results were significant, even when studies including only stroke-free patients were pooled together (unadjusted OR, 2.2; 95% CI, 1.4 to 3.5; I², 96%), but the heterogeneity rates were high. AF was significantly associated with increased risk of both vascular (adjusted OR, 1.7; 95% CI, 1.2 to 2.3; I², 43%) and Alzheimer's dementia (adjusted HR, 1.4; 95% CI, 1.2 to 1.6; I², 42%).

CONCLUSION

AF increases the risk of cognitive impairment, all-cause dementia, vascular dementia, and Alzheimer's disease. Future studies should employ interventions that may delay or even prevent cognitive decline in AF patients.

Clinical judgment of GPs for the diagnosis of dementia: a diagnostic test accuracy study

BACKGROUND

GPs often report using clinical judgment to diagnose dementia.

AIM

To investigate the accuracy of GPs' clinical judgment for the diagnosis of dementia.

DESIGN & SETTING

Diagnostic test accuracy study, recruiting from 21 practices around Bristol, UK.

METHOD

The clinical judgment of the treating GP (index test) was based on the information immediately available at their initial consultation with a person aged ≥70 years who had cognitive symptoms. The reference standard was an assessment by a specialist clinician, based on a standardised clinical examination and made according to the 10th revision of the International Classification of Diseases (ICD-10) criteria for dementia.

RESULTS

A total of 240 people were recruited, with a median age of 80 years (interquartile range [IQR] 75-84 years), of whom 126 (53%) were men and 132 (55%) had dementia. The median duration of symptoms was 24 months (IQR 12-36 months) and the median Addenbrooke's Cognitive Examination III (ACE-III) score was 75 (IQR 65-87). GP clinical judgment had sensitivity 56% (95% confidence interval [CI] = 47% to 65%) and specificity 89% (95% CI = 81% to 94%). Positive likelihood ratio was higher in people aged 70-79 years (6.5, 95% CI = 2.9 to 15) compared with people aged ≥80 years (3.6, 95% CI = 1.7 to 7.6), and in women (10.4, 95% CI = 3.4 to 31.7) compared with men (3.2, 95% CI = 1.7 to 6.2), whereas the negative likelihood ratio was similar in all groups.

CONCLUSION

A GP clinical judgment of dementia is specific, but confirmatory testing is needed to exclude dementia in symptomatic people whom GPs judge as not having dementia.

A Multimodal Risk Network Predicts Executive Function Trajectories in Non-demented Aging

Background: Multiple modalities of Alzheimer's disease (AD) risk factors may operate through interacting networks to predict differential cognitive trajectories in asymptomatic aging. We test such a network in a series of three analytic steps. First, we test independent associations between three risk scores (functional-health, lifestyle-reserve, and a combined multimodal risk score) and cognitive [executive function (EF)] trajectories. Second, we test whether all three associations are moderated by the most penetrant AD genetic risk [Apolipoprotein E (APOE) ε4+ allele]. Third, we test whether a non-APOE AD genetic risk score further moderates these APOE × multimodal risk score associations. Methods: We assembled a longitudinal data set (spanning a 40-year band of aging, 53-95 years) with non-demented older adults (baseline n = 602; Mage = 70.63(8.70) years; 66% female) from the Victoria Longitudinal Study (VLS). The measures included for each modifiable risk score were: (1) functional-health [pulse pressure (PP), grip strength, and body mass index], (2) lifestyle-reserve (physical, social, cognitive-integrative, cognitive-novel activities, and education), and (3) the combination of functional-health and lifestyle-reserve risk scores. Two AD genetic risk markers included (1) APOE and (2) a combined AD-genetic risk score (AD-GRS) comprised of three single nucleotide polymorphisms (SNPs; Clusterin[rs11136000], Complement receptor 1[rs6656401], Phosphatidylinositol binding clathrin assembly protein[rs3851179]). The analytics included confirmatory factor analysis (CFA), longitudinal invariance testing, and latent growth curve modeling. Structural path analyses were deployed to test and compare prediction models for EF performance and change. Results: First, separate analyses showed that higher functional-health risk scores, lifestyle-reserve risk scores, and the combined score, predicted poorer EF performance and steeper decline. Second, APOE and AD-GRS moderated the association between functional-health risk score and the combined risk score, on EF performance and change. Specifically, only older adults in the APOEε4- group showed steeper EF decline with high risk scores on both functional-health and combined risk score. Both associations were further magnified for adults with high AD-GRS. Conclusion: The present multimodal AD risk network approach incorporated both modifiable and genetic risk scores to predict EF trajectories. The results add an additional degree of precision to risk profile calculations for asymptomatic aging populations.

Markers of dementia-related health in primary care electronic health records

OBJECTIVES

Identifying routinely recorded markers of poor health in patients with dementia may help treatment decisions and evaluation of earlier outcomes in research. Our objective was to determine whether a set of credible markers of dementia-related health could be identified from primary care electronic health records (EHR).

METHODS

The study consisted of (i) rapid review of potential measures of dementia-related health used in EHR studies; (ii) consensus exercise to assess feasibility of identifying these markers in UK primary care EHR; (iii) development of UK EHR code lists for markers; (iv) analysis of a regional primary care EHR database to determine further potential markers; (v) consensus exercise to finalise markers and pool into higher domains; (vi) determination of 12-month prevalence of domains in EHR of 2328 patients with dementia compared to matched patients without dementia.

RESULTS

Sixty-three markers were identified and mapped to 13 domains: Care; Home Pressures; Severe Neuropsychiatric; Neuropsychiatric; Cognitive Function; Daily Functioning; Safety; Comorbidity; Symptoms; Diet/Nutrition; Imaging; Increased Multimorbidity; Change in Dementia Drug. Comorbidity was the most prevalent recorded domain in dementia (69%). Home Pressures were the least prevalent domain (1%). Ten domains had a statistically significant higher prevalence in dementia patients, one (Comorbidity) was higher in non-dementia patients, and two (Home Pressures, Diet/Nutrition) showed no association with dementia.

CONCLUSIONS

EHR captures important markers of dementia-related health. Further research should assess if they indicate dementia progression. These markers could provide the basis for identifying individuals at risk of faster progression and outcome measures for use in research.

Development of a dementia prediction model for primary care: The Hisayama Study

INTRODUCTION

We aimed to develop a risk prediction model for incident dementia using predictors that are available in primary-care settings.

METHODS

A total of 795 subjects aged 65 years or over were prospectively followed-up from 1988 to 2012. A Cox proportional-hazards regression was used to develop a multivariable prediction model. The developed model was translated into a simplified scoring system based on the beta-coefficient. The discrimination of the model was assessed by Harrell's C statistic, and the calibration was assessed by a calibration plot.

RESULTS

During the follow-up period, 364 subjects developed dementia. In the multivariable model, age, female sex, low education, leanness, hypertension, diabetes, history of stroke, current smoking, and sedentariness were selected as predictors. The developed model and simplified score showed good discrimination and calibration.

DISCUSSION

The developed risk prediction model is feasible and practically useful in primary-care settings to identify individuals at high risk for future dementia.

Development and validation of a predictive algorithm for risk of dementia in the community setting

Background

Most dementia algorithms are unsuitable for population-level assessment and planning as they are designed for use in the clinical setting. A predictive risk algorithm to estimate 5-year dementia risk in the community setting was developed.

Methods

The Dementia Population Risk Tool (DemPoRT) was derived using Ontario respondents to the Canadian Community Health Survey (survey years 2001 to 2012). Five-year incidence of physician-diagnosed dementia was ascertained by individual linkage to administrative healthcare databases and using a validated case ascertainment definition with follow-up to March 2017. Sex-specific proportional hazards regression models considering competing risk of death were developed using self-reported risk factors including information on socio-demographic characteristics, general and chronic health conditions, health behaviours and physical function.

Results

Among 75 460 respondents included in the combined derivation and validation cohorts, there were 8448 cases of incident dementia in 348 677 person-years of follow-up (5-year cumulative incidence, men: 0.044, 95% CI: 0.042 to 0.047; women: 0.057, 95% CI: 0.055 to 0.060). The final full models each include 90 df (65 main effects and 25 interactions) and 28 predictors (8 continuous). The DemPoRT algorithm is discriminating (C-statistic in validation data: men 0.83 (95% CI: 0.81 to 0.85); women 0.83 (95% CI: 0.81 to 0.85)) and well-calibrated in a wide range of subgroups including behavioural risk exposure categories, socio-demographic groups and by diabetes and hypertension status.

Conclusions

This algorithm will support the development and evaluation of population-level dementia prevention strategies, support decision-making for population health and can be used by individuals or their clinicians for individual risk assessment.

Barriers and facilitators to the adoption of electronic clinical decision support systems: a qualitative interview study with UK general practitioners

Background

Well-established electronic data capture in UK general practice means that algorithms, developed on patient data, can be used for automated clinical decision support systems (CDSSs). These can predict patient risk, help with prescribing safety, improve diagnosis and prompt clinicians to record extra data. However, there is persistent evidence of low uptake of CDSSs in the clinic. We interviewed UK General Practitioners (GPs) to understand what features of CDSSs, and the contexts of their use, facilitate or present barriers to their use.

Methods

We interviewed 11 practicing GPs in London and South England using a semi-structured interview schedule and discussed a hypothetical CDSS that could detect early signs of dementia. We applied thematic analysis to the anonymised interview transcripts.

Results

We identified three overarching themes: trust in individual CDSSs; usability of individual CDSSs; and usability of CDSSs in the broader practice context, to which nine subthemes contributed. Trust was affected by CDSS provenance, perceived threat to autonomy and clear management guidance. Usability was influenced by sensitivity to the patient context, CDSS flexibility, ease of control, and non-intrusiveness. CDSSs were more likely to be used by GPs if they did not contribute to alert proliferation and subsequent fatigue, or if GPs were provided with training in their use.

Conclusions

Building on these findings we make a number of recommendations for CDSS developers to consider when bringing a new CDSS into GP patient records systems. These include co-producing CDSS with GPs to improve fit within clinic workflow and wider practice systems, ensuring a high level of accuracy and a clear clinical pathway, and providing CDSS training for practice staff. These recommendations may reduce the proliferation of unhelpful alerts that can result in important decision-support being ignored.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12911-021-01557-z.

Detection and Prediction of Incident Alzheimer Dementia over a 10-Year or Longer Medical History: A Population-Based Study in Primary Care

BACKGROUND

Despite the crucial role played by general practitioners in the identification and care of people with cognitive impairment, few data are available on how they may improve the early recognition of patients with Alzheimer dementia (AD), especially those with long (i.e., 10 years and longer) medical history.

AIMS

To investigate the occurrence and the predictors of AD during a 10-year or longer period prior AD diagnosis in primary care patients aged 60 years or older.

MATERIALS AND METHODS

A cohort study with a nested case-control analysis has been conducted. Data were extracted from the Italian Health Search Database (HSD), an Italian database with primary care data. AD cases have been defined in accordance with the International Classification of Diseases, ninth edition (ICD-9-CM) codes and coupled with the use of anti-dementia drugs. Prevalence and incidence rates of AD have been calculated. To test the association between candidate predictors, being identified in a minimum period of 10 years, and incident cases of AD, we used a multivariate conditional logistic regression model.

RESULTS

As recorded in the primary care database, AD prevalence among patients aged 60 years or older was 0.8% during 2016, reaching 2.4% among nonagenarians. Overall, 1,889 incident cases of AD have been identified, with an incidence rate as high as 0.09% person-year. Compared with 18,890 matched controls, history of hallucinations, agitation, anxiety, aberrant motor behavior, and memory deficits were positively associated with higher odds of AD (p < 0.001 for all) diagnosis. A previous diagnosis of depression and diabetes and the use of low-dose aspirin and non-steroidal anti-inflammatory drugs were associated with higher odds of AD (p < 0.05 for all).

CONCLUSION

Our findings show that, in accordance with primary care records, 1% of patients aged 60 years and older have a diagnosis of AD, with an incident AD diagnosis of 0.1% per year. AD is often under-reported in primary care settings; yet, several predictors identified in this study may support general practitioners to early identify patients at risk of AD.

Longitudinal Associations of Stroke With Cognitive Impairment Among Older Adults in the United States: A Population-Based Study

Objective: The aim of this study was to explore the longitudinal associations of stroke with cognitive impairment in older US adults.

Method: The data used in this longitudinal analysis were extracted from the National Health and Aging Trends Study (NHATS) from 2011 to 2019. Univariate and multivariable Cox proportional hazards regression models were used to estimate the longitudinal association of stroke with cognitive impairment. The multivariable model was adjusted by demographic, physical, and mental characteristics, and the complex survey design of NHATS was taken into consideration.

Results: A total of 7,052 participants with complete data were included. At the baseline, the weighted proportion of cognitive impairment was 19.37% (95% CI, 17.92–20.81%), and the weighted proportion of the history of stroke was 9.81% (95% CI, 8.90–10.72%). In univariate analysis, baseline stroke history was significantly associated with cognitive impairment in the future (hazard ratio, 1.746; 95% CI, 1.461–2.088), and the baseline cognitive impairment was significantly associated with future report of stroke (hazard ratio, 1.436; 95% CI, 1.088–1.896). In multivariable model, stroke was also significantly associated with cognitive impairment (hazard ratio, 1.241; 95% CI, 1.011–1.522); however, the reverse association was not significant (hazard ratio, 1.068; 95% CI, 0.788–1.447). After the data from proxy respondents were excluded, in the sensitive analyses, the results remained unchanged.

Conclusion: Older adults in the United States who suffered strokes are more likely to develop cognitive impairment as a result in the future than those who have not had strokes. However, the reverse association did not hold. Furthermore, the study suggests that it is necessary to screen and take early intervention for cognitive impairment in stroke survivors and prevent the incidence of stroke by modifying risk factors in the general population with rapidly growing older US adults.

Automated detection of patients with dementia whose symptoms have been identified in primary care but have no formal diagnosis: a retrospective case-control study using electronic primary care records

OBJECTIVES

UK statistics suggest only two-thirds of patients with dementia get a diagnosis recorded in primary care. General practitioners (GPs) report barriers to formally diagnosing dementia, so some patients may be known by GPs to have dementia but may be missing a diagnosis in their patient record. We aimed to produce a method to identify these 'known but unlabelled' patients with dementia using data from primary care patient records.

DESIGN

Retrospective case-control study using routinely collected primary care patient records from Clinical Practice Research Datalink.

SETTING

UK general practice.

PARTICIPANTS

English patients aged >65 years, with a coded diagnosis of dementia recorded in 2000-2012 (cases), matched 1:1 with patients with no diagnosis code for dementia (controls).

INTERVENTIONS

Eight coded and nine keyword concepts indicating symptoms, screening tests, referrals and care for dementia recorded in the 5 years before diagnosis. We trialled machine learning classifiers to discriminate between cases and controls (logistic regression, naïve Bayes, random forest).

PRIMARY AND SECONDARY OUTCOMES

The outcome variable was dementia diagnosis code; the accuracy of classifiers was assessed using area under the receiver operating characteristic curve (AUC); the order of features contributing to discrimination was examined.

RESULTS

93 426 patients were included; the median age was 83 years (64.8% women). Three classifiers achieved high discrimination and performed very similarly. AUCs were 0.87-0.90 with coded variables, rising to 0.90-0.94 with keywords added. Feature prioritisation was different for each classifier; commonly prioritised features were Alzheimer's prescription, dementia annual review, memory loss and dementia keywords.

CONCLUSIONS

It is possible to detect patients with dementia who are known to GPs but unlabelled with a diagnostic code, with a high degree of accuracy in electronic primary care record data. Using keywords from clinic notes and letters improves accuracy compared with coded data alone. This approach could improve identification of dementia cases for record-keeping, service planning and delivery of good quality care.

Indicators of dementia disease progression in primary care: An electronic health record cohort study

BACKGROUND AND PURPOSE

The objectives were to assess the feasibility and validity of using markers of dementia-related health as indicators of dementia progression in primary care, by assessing the frequency with which they are recorded and by testing the hypothesis that they are associated with recognised outcomes of dementia. The markers, in 13 domains, were derived previously through literature review, expert consensus, and analysis of regional primary care records.

METHODS

The study population consisted of patients with a recorded dementia diagnosis in the Clinical Practice Research Datalink, a UK primary care database linked to secondary care records. Incidence of recorded domains in the 36 months after diagnosis was determined. Associations of recording of domains with future hospital admission, palliative care, and mortality were derived.

RESULTS

There were 30,463 people with diagnosed dementia. Incidence of domains ranged from 469/1000 person-years (Increased Multimorbidity) to 11/1000 (Home Pressures). An increasing number of domains in which a new marker was recorded in the first year after diagnosis was associated with hospital admission (hazard ratio for ≥4 domains vs. no domains = 1.24; 95% confidence interval = 1.15-1.33), palliative care (1.87; 1.62-2.15), and mortality (1.57; 1.47-1.67). Individual domains were associated with outcomes with varying strengths of association.

CONCLUSIONS

Feasibility and validity of potential indicators of progression of dementia derived from primary care records are supported by their frequency of recording and associations with recognised outcomes. Further research should assess whether these markers can help identify patients with poorer prognosis to improve outcomes through stratified care and targeted support.

Predictors of New Dementia Diagnoses in Elderly Individuals: A Retrospective Cohort Study Based on Prefecture-Wide Claims Data in Japan

Preventing dementia in elderly individuals is an important public health challenge. While early identification and modification of predictors are crucial, predictors of dementia based on routinely collected healthcare data are not fully understood. We aimed to examine potential predictors of dementia diagnosis using routinely collected claims data. In this retrospective cohort study, claims data from fiscal years 2012 (baseline) and 2016 (follow-up), recorded in an administrative claims database of the medical care system for the elderly (75 years or older) in Niigata prefecture, Japan, were used. Data on baseline characteristics including age, sex, diagnosis, and prescriptions were collected, and the relationship between subsequent new diagnoses of dementia and potential predictors was examined using multivariable logistic regression models. A total of 226,738 people without a diagnosis of dementia at baseline were followed. Of these, 26,092 incident dementia cases were detected during the study period. After adjusting for confounding factors, cerebrovascular disease (odds ratio, 1.15; 95% confidence interval, 1.11-1.18), depression (1.38; 1.31-1.44), antipsychotic use (1.40; 1.31-1.49), and hypnotic use (1.17; 1.11-1.24) were significantly associated with subsequent diagnosis of dementia. Analyses of routinely collected claims data revealed neuropsychiatric symptoms including depression, antipsychotic use, hypnotic use, and cerebrovascular disease to be predictors of new dementia diagnoses.

Novel PET Biomarkers to Disentangle Molecular Pathways across Age-Related Neurodegenerative Diseases

There is a need to disentangle the etiological puzzle of age-related neurodegenerative diseases, whose clinical phenotypes arise from known, and as yet unknown, pathways that can act distinctly or in concert. Enhanced sub-phenotyping and the identification of in vivo biomarker-driven signature profiles could improve the stratification of patients into clinical trials and, potentially, help to drive the treatment landscape towards the precision medicine paradigm. The rapidly growing field of neuroimaging offers valuable tools to investigate disease pathophysiology and molecular pathways in humans, with the potential to capture the whole disease course starting from preclinical stages. Positron emission tomography (PET) combines the advantages of a versatile imaging technique with the ability to quantify, to nanomolar sensitivity, molecular targets in vivo. This review will discuss current research and available imaging biomarkers evaluating dysregulation of the main molecular pathways across age-related neurodegenerative diseases. The molecular pathways focused on in this review involve mitochondrial dysfunction and energy dysregulation; neuroinflammation; protein misfolding; aggregation and the concepts of pathobiology, synaptic dysfunction, neurotransmitter dysregulation and dysfunction of the glymphatic system. The use of PET imaging to dissect these molecular pathways and the potential to aid sub-phenotyping will be discussed, with a focus on novel PET biomarkers.

Alzheimer-type dementia prediction by sparse logistic regression using claim data

This study aimed to predict the risk of Alzheimer-type dementia for persons aged over 75 years old without receiving long-term care services using regularly collected claim data. A refined dataset including 48,123 persons was prepared from claim data of health insurance and long-term care insurance in a large city in the metropolitan area in Japan. The utilized features include the age and sex of subjects, 502 diseases based on ICD-10 diagnosis codes, and 107 prescription drugs based on therapeutic classes. The most important challenge in this work was feature selection form a large number of features. We adopted sparse logistic regression models with L0 regularization (SLR-L0) and L1 regularization (SLR-L1) as classification models based on machine learning. These regularizations enable feature selection by estimating sparse solution of non-zero coefficients in the model optimization. Predictions were performed by integrating 100 predictors trained by bootstrap samples. As a result, the area under the ROC curves (AUCs) were 0.663 for SLR-L0 and 0.660 for SLR-L1. These performances were similar, however, the average numbers of selected features were 13 out of a total of 611 for SLR-L0 and 253 for SLR-R1. The results indicate that SLR-L1 tended to include less useful features, whereas SLR-L0 narrowed down influential features. SLR-L0 might be more useful than SLR-L1 for practical use or the discussion of risk factors with medical experts.

The Impact of Including Costs and Outcomes of Dementia in a Health Economic Model to Evaluate Lifestyle Interventions to Prevent Diabetes and Cardiovascular Disease

Objectives

Economic evaluations of lifestyle interventions, which aim to prevent diabetes/cardiovascular disease (CVD), have not included dementia. Lifestyle interventions decrease dementia risk and extend life expectancy, leading to competing effects on health care costs. We aim to demonstrate the feasibility of including dementia in a public health cost-effectiveness analysis and quantify the overall impacts accounting for these competing effects.

Methods

The School for Public Health Research (SPHR) diabetes prevention model describes individuals’ risk of type 2 diabetes, microvascular outcomes, CVD, congestive heart failure, cancer, osteoarthritis, depression, and mortality in England. In version 3.1, we adapted the model to include dementia using published data from primary care databases, health surveys, and trials of dementia to describe dementia incidence, diagnosis, and disease progression. We estimate the impact of dementia on lifetime costs and quality-adjusted life years (QALYs) gained of the National Health Service diabetes prevention program (NHS DPP) from an NHS/personal social services perspective with 3 scenarios: 1) no dementia, 2) dementia only, and 3) reduced dementia risk. Subgroup, parameter, and probabilistic sensitivity analyses were conducted.

Results

The lifetime cost savings of the NHS DPP per patient were £145 in the no-dementia scenario, £121 in the dementia-only scenario, and £167 in the reduced dementia risk scenario. The QALY gains increased by 0.0006 in dementia only and 0.0134 in reduced dementia risk. Dementia did not alter the recommendation that the NHS/DPP is cost-effective.

Conclusions

Including dementia into a model of lifestyle interventions was feasible but did not change policy recommendations or modify health economic outcomes. The impact on health economic outcomes was largest where a direct impact on dementia incidence was assumed, particularly in elderly populations.

Dementia risk scores as surrogate outcomes for lifestyle-based multidomain prevention trials-rationale, preliminary evidence and challenges

INTRODUCTION

Although not designed as such, dementia risk scores might be useful surrogate outcomes for dementia prevention trials. Their suitability may be improved by using continuous scoring systems, taking into account all changes in risk factors, not only those crossing cut-off values.

METHODS

In three large multidomain dementia prevention trials with 1.5 to 2 years of follow-up (Multidomain Alzheimer Preventive Trial, Prevention of Dementia by Intensive Vascular Care and Healthy Ageing Through Internet Counselling in the Elderly) we assessed (1) responsiveness (sensitivity to change) and (2) actual and simulated intervention effects of the original and crude/weighted z-score versions of the cardiovascular risk factors, aging and incidence of dementia, and Lifestyle for Brain Health scores.

RESULTS

All versions of the risk scores were generally responsive, and able to detect small though statistically significant between-group differences after multidomain interventions. Simulated intervention effects were well detected in z-score versions as well as in the original scores.

DISCUSSION

Dementia risk scores and their z-score versions show potential as surrogate outcomes. How changes in risk scores affect dementia remains to be determined.

APPLE-Tree (Active Prevention in People at risk of dementia: Lifestyle, bEhaviour change and Technology to REducE cognitive and functional decline) programme: Protocol

BACKGROUND

Observational studies indicate that approximately a third of dementia cases are attributable to modifiable cardiometabolic, physical and mental health, and social and lifestyle risk factors. There is evidence that intensive behaviour change interventions targeting these factors can reduce cognitive decline. [Figure: see text] METHODS AND ANALYSIS: We will design and test a low intensity, secondary dementia-prevention programme (Active Prevention in People at risk of dementia: Lifestyle, bEhaviour change and Technology to REducE cognitive and functional decline, "APPLE-Tree") to slow cognitive decline in people with subjective cognitive decline with or without objective cognitive impairment. We will embed our work within social science research to understand how dementia prevention is currently delivered and structured. We will carry out systematic reviews and around 50 qualitative interviews with stakeholders, using findings to coproduce the APPLE-Tree intervention. We plan a 10-session group intervention, involving personalised goal-setting, with individual sessions for those unable or unwilling to attend groups, delivered by psychology assistants who will be trained and supervised by clinical psychologists. The coproduction group (including public and patient involvement [PPI], academic and clinical/third-sector professional representatives) will use the Behaviour Change Wheel theoretical framework to develop it. We will recruit and randomly allocate 704 participants, 1:1 to the intervention: informational control group. This sample size is sufficient to detect a between-group difference at 2 years of 0.15 on the primary outcome (cognition: modified neuropsychological test battery; 90% power, 5% significance, effect size 0.25, SD 0.6).

DISSEMINATION

We will work with Public Health England and third-sector partners to produce an effective national implementation approach, so that if our intervention works, it is used in practice.

Validity and reliability of the Brain Health Self-Efficacy Scale for the elderly

BACKGROUND

With the rapid increase in ageing population, China is confronted with the daunting challenge of a growing number of patients with neurocognitive disorders (NCDs). This trend makes the maintenance of self-health and early intervention essential, highlighting the need for a tool that assesses self-efficacy of older adults in maintaining brain health or cognitive function.

AIM

This study aimed to design the Brain Health Self-Efficacy Scale (BHSES) to measure elderly individuals' attitudes to NCDs, motivations and future plans for controlling risks. The psychometric properties of BHSES have been validated.

METHODS

Based on the current literature and relevant models, a 19-item scale was created during the first stage. A total of 660 older adults in the Yinhang community of Shanghai were included. The statistical approaches of item analysis, exploratory factor analysis (EFA), confirmatory factor analysis (CFA), criterion-related validity and reliability test were used to evaluate the quality of BHSES. In addition, the Geriatric Depression Scale (GDS) and the Self-Rating Anxiety Scale (SAS) were used as criteria to test the criterion-related validity.

RESULTS

To test item differentiation, the study adopted item analysis and excluded item 8. Selecting a random half of the sample for EFA, the BHSES was refined to 16 items, which were categorised into the following three dimensions: 'memory belief efficacy', 'self-care efficacy' and 'future planning efficacy'. These were highly consistent with the hypothesis model. Its cumulative variance contribution rate reached 61.14%, with factor loads of all items above 0.5. The three-factor model was confirmed by the remaining data through CFA. All fit indices reached the acceptable level (χ2=3.045, Goodness of Fit Index=0.898, adjusted Goodness of Fit Index=0.863, Comparative Fit Index=0.916, Incremental Fit Index=0.917, Tucker-Lewis Index=0.900, root mean square error of approximation=0.079 and root mean residual=0.068). The GDS and SAS scores revealed significant correlations with the BHSES score, indicating a high criterion-related validity. The overall Cronbach's α coefficient was 0.793, with the α coefficients' distribution of subdimensions ranging from 0.748 to 0.883.

CONCLUSIONS

The 16-item, self-compiled BHSES is a reliable and valid measurement. It could help identify older adults with potential risks for developing NCDs or with high suspicion of cognitive impairment onset in recent periods and also offer insight into tracking brain health self-efficacy in association with cognition status.

Cohort discovery and risk stratification for Alzheimer's disease: an electronic health record-based approach

BACKGROUND

We sought to leverage data routinely collected in electronic health records (EHRs), with the goal of developing patient risk stratification tools for predicting risk of developing Alzheimer's disease (AD).

METHOD

Using EHR data from the University of Michigan (UM) hospitals and consensus-based diagnoses from the Michigan Alzheimer's Disease Research Center, we developed and validated a cohort discovery tool for identifying patients with AD. Applied to all UM patients, these labels were used to train an EHR-based machine learning model for predicting AD onset within 10 years.

RESULTS

Applied to a test cohort of 1697 UM patients, the model achieved an area under the receiver operating characteristics curve of 0.70 (95% confidence interval = 0.63-0.77). Important predictive factors included cardiovascular factors and laboratory blood testing.

CONCLUSION

Routinely collected EHR data can be used to predict AD onset with modest accuracy. Mining routinely collected data could shed light on early indicators of AD appearance and progression.