Precision prevention of Alzheimer's and other dementias: Anticipating future needs in the control of risk factors and implementation of disease‐modifying therapies

Frisoni, Giovanni B.; Molinuevo, José Luís; Altomare, Daniele; Carrera, Emmanuel; Barkhof, Frederik; Berkhof, Johannes; Delrieu, Julien; Dubois, Bruno; Kivipelto, Miia; Nordberg, Agneta; Schott, Jonathan M.; Flier, Wiesje M. van der; Vellas, Bruno; Jessen, Frank; Scheltens, Philip; Ritchie, Craig W.

doi:10.1002/alz.12132

Cited by 56 publications

(67 citation statements)

References 44 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Therefore, any intervention in the emotional health of this participant chosen for illustrative purposes could further reduce their risk. This approach is exactly what is envisaged in the Brain Health Clinics being developed across Europe (Frisoni et al, 2020 ) based on a consensus led by our group in how to change clinical services for dementia prevention (Ritchie et al, 2017 ). This is based on collecting data from these Brain Health Clinics to support Real World machine learning approaches and using these algorithms to support the development of personalised prevention plans driven by early disease detection and comprehensive risk profiling.…”

Section: Discussionmentioning

confidence: 89%

“…Therefore, any intervention in the emotional health of this participant chosen for illustrative purposes could further reduce their risk. This approach is exactly what is envisaged in the Brain Health Clinics being developed across Europe (Frisoni et al, 2020) based on a consensus led by our group in how to change clinical services for dementia prevention (Ritchie et al,FIGURE 4 | Feature importance as ranked by the weights derived from SHARE_RF_pred (A) and SHARE_XGBoost_pred (B) prediction models that were trained using SHARE dataset. It also shows the ranking of features of PREVENT only (C) and PREVENT target (D).…”

Section: Discussionmentioning

confidence: 99%

See 1 more Smart Citation

Developing an Explainable Machine Learning-Based Personalised Dementia Risk Prediction Model: A Transfer Learning Approach With Ensemble Learning Algorithms

et al. 2021

View full text Add to dashboard Cite

Alzheimer's disease (AD) has its onset many decades before dementia develops, and work is ongoing to characterise individuals at risk of decline on the basis of early detection through biomarker and cognitive testing as well as the presence/absence of identified risk factors. Risk prediction models for AD based on various computational approaches, including machine learning, are being developed with promising results. However, these approaches have been criticised as they are unable to generalise due to over-reliance on one data source, poor internal and external validations, and lack of understanding of prediction models, thereby limiting the clinical utility of these prediction models. We propose a framework that employs a transfer-learning paradigm with ensemble learning algorithms to develop explainable personalised risk prediction models for dementia. Our prediction models, known as source models, are initially trained and tested using a publicly available dataset (n = 84,856, mean age = 69 years) with 14 years of follow-up samples to predict the individual risk of developing dementia. The decision boundaries of the best source model are further updated by using an alternative dataset from a different and much younger population (n = 473, mean age = 52 years) to obtain an additional prediction model known as the target model. We further apply the SHapely Additive exPlanation (SHAP) algorithm to visualise the risk factors responsible for the prediction at both population and individual levels. The best source model achieves a geometric accuracy of 87%, specificity of 99%, and sensitivity of 76%. In comparison to a baseline model, our target model achieves better performance across several performance metrics, within an increase in geometric accuracy of 16.9%, specificity of 2.7%, and sensitivity of 19.1%, an area under the receiver operating curve (AUROC) of 11% and a transfer learning efficacy rate of 20.6%. The strength of our approach is the large sample size used in training the source model, transferring and applying the “knowledge” to another dataset from a different and undiagnosed population for the early detection and prediction of dementia risk, and the ability to visualise the interaction of the risk factors that drive the prediction. This approach has direct clinical utility.

show abstract

Section: Discussionmentioning

confidence: 89%

“…Therefore, any intervention in the emotional health of this participant chosen for illustrative purposes could further reduce their risk. This approach is exactly what is envisaged in the Brain Health Clinics being developed across Europe (Frisoni et al, 2020) based on a consensus led by our group in how to change clinical services for dementia prevention (Ritchie et al,FIGURE 4 | Feature importance as ranked by the weights derived from SHARE_RF_pred (A) and SHARE_XGBoost_pred (B) prediction models that were trained using SHARE dataset. It also shows the ranking of features of PREVENT only (C) and PREVENT target (D).…”

Section: Discussionmentioning

confidence: 99%

Developing an Explainable Machine Learning-Based Personalised Dementia Risk Prediction Model: A Transfer Learning Approach With Ensemble Learning Algorithms

et al. 2021

View full text Add to dashboard Cite

show abstract

“…All the studies stressed the importance of a specific feature of neurodegenerative diseases, which include a prodromal phase lasting up to 20 years before clinical diagnosis 10,34 . This makes it more difficult to separate possible early symptoms from causal risk factors when describing the natural course of the disease.…”

Section: Review Of Potential Risk Factors For Dementiamentioning

confidence: 99%

Data-driven identification of health conditions associated with incident Alzheimer’s disease dementia risk: a 15 years follow-up cohort from electronic health records in France and the United Kingdom

Nedelec

Couvy‐Duchesne

Monnet

et al. 2021

Preprint

View full text Add to dashboard Cite

Importance. The identification of modifiable risk factors for Alzheimer s disease (AD) is paramount for early prevention and the targeting of new interventions. Objective. To assess the associations between health conditions diagnosed in primary care and the risk of incident AD over time. Design, Setting, and Participants. Data for 20,214 AD patients from the United Kingdom and 19,458 AD patients from France were extracted from The Health Improvement Network (THIN) database. For each AD case, a control was randomly assigned after matching for sex and age at dementia diagnosis. We agnostically tested the associations between 123 different ICD10 diagnoses extracted from health records and AD dementia, by conditional logistic regression. We focused on two time periods: 2 to 10 years vs. 0 to 2 years before the diagnosis of AD, to separate risk factors from early symptoms/comorbidities. Exposures. We considered all health conditions that had been recorded in more than 0.1% of visits per 1000 person-years in both cohorts, corresponding to 123 potential types of exposure. Main Outcomes and Measures. Odds ratios (ORs) for the association of AD with the various health conditions were calculated after Bonferroni correction for multiple comparisons. Results. Ten health conditions were significantly associated with high odds ratios for AD when diagnosed 2 to 10 years before AD, in the British and French samples: major depressive disorder (OR 95% confidence interval (UK):1.23-1.46)), anxiety (1.25-1.47), reaction to severe stress (1.24-1.59), hearing loss (1.11-1.28), constipation (1.22-1.41), spondylosis (1.14-1.39), abnormal weight loss (1.33-1.63), malaise and fatigue (1.14-1.32), memory loss (6.65-8.76) and syncope and collapse (1.1-1.37). Depression was the first comorbid condition associated with AD, appearing at least nine years before the first clinical diagnosis of AD, followed by, anxiety, constipation and abnormal weight loss. Conclusions and Relevance. These results from two independent primary care databases provide new evidence on the temporality of risk factors and early signs of Alzheimer s disease. These results could guide new dementia prevention strategies.

show abstract

“…Knowledge regarding the mid-life onset of neurodegenerative disease has been accumulating over the past decade14 through programmes such as Prevent Dementia15 and the European Prevention of Alzheimer’s Dementia programme 16. These initiatives have informed the development of national policies (Brain Health Scotland, for example)17 and specialist brain health clinics founded to provide detailed risk profiling, detect early disease, and develop personalised prevention plans 1819. This approach provides an ideal pathway for former athletes to obtain independent specialist review and advice on how to improve brain health, which might reduce the long term consequences of previous traumatic brain injury.…”

Section: Act To Reduce Harmmentioning

confidence: 99%

“…14 These initiatives have informed the development of national policy (eg Brain Health Scotland) 15 and specialist Brain Health Clinics founded on three broad goals to: [1] provide detailed risk profiling; [2] detect early disease; and [3] provide personalised prevention plans. 16,17 This approach provides an ideal pathway for former contact sport athletes to obtain independent, specialist clinical review and advice on measures to improve brain health that might serve to reduce the long-term consequences of previous exposure to TBI.…”

Section: Second Act To Reduce Harmmentioning

confidence: 99%

Sport associated dementia

Stewart

2021

BMJ

View full text Add to dashboard Cite

show abstract

Precision prevention of Alzheimer's and other dementias: Anticipating future needs in the control of risk factors and implementation of disease‐modifying therapies

Cited by 56 publications

References 44 publications

Developing an Explainable Machine Learning-Based Personalised Dementia Risk Prediction Model: A Transfer Learning Approach With Ensemble Learning Algorithms

Developing an Explainable Machine Learning-Based Personalised Dementia Risk Prediction Model: A Transfer Learning Approach With Ensemble Learning Algorithms

Data-driven identification of health conditions associated with incident Alzheimer’s disease dementia risk: a 15 years follow-up cohort from electronic health records in France and the United Kingdom

Sport associated dementia

Contact Info

Product

Resources

About