Tuesday, July 16, 2024

Predictive Models for Alzheimer’s Disease Dementia in Older Adults with Mild Cognitive Impairment: An Overview

Similar articles

In light of the increasing prevalence of Alzheimer’s disease, the ability to accurately predict its onset in older adults with mild cognitive impairment has garnered significant interest. This review synthesizes existing research on predictive models, highlighting their strengths and weaknesses, ultimately aiming to enhance the identification and management of individuals at risk.

Mild cognitive impairment (MCI) is recognized as a precursor to Alzheimer’s disease, spurring numerous studies to develop predictive models. Despite the proliferation of research, the effectiveness of these models remains uncertain. This systematic review aims to evaluate and summarize the performance of prediction models for Alzheimer’s disease dementia in older adults with MCI, providing a comprehensive overview of their validity and applicability.

Research Methodology

The review involved extensive searches of PubMed, EMBASE, Web of Science, and MEDLINE databases up to October 2023. The inclusion criteria focused on cohort studies that developed or validated risk prediction models for Alzheimer’s disease dementia in older adults with MCI. The Predictive Model Risk of Bias Assessment Tool (PROBAST) was used to evaluate model bias and applicability. Random-effects models calculated the combined model AUCs and 95% prediction intervals, assessing heterogeneity with the I2 statistic. Funnel plot analysis was also conducted to detect publication bias.

Key Findings

The review incorporated 16 studies with a total of 9290 participants. Fourteen predictors, including age, functional activities questionnaire, and Mini-mental State Examination scores, were frequently identified. Despite the development of numerous models, only two underwent external validation. Machine learning techniques were employed in eleven studies, whereas traditional methods were utilized in four. However, many studies faced issues such as small sample sizes, missing crucial methodological details, and a lack of model presentation. All models were ultimately rated as having a high or unclear risk of bias, with the best-developed models achieving an average AUC of 0.87 (95% CI: 0.83, 0.90).

Practical Implications

– Future research must prioritize methodological rigor and external validation.
– Transparent reporting and adherence to the scientific method are crucial for improving model accuracy and generalizability.
– Researchers should ensure the inclusion of adequate sample sizes and comprehensive methodological information to enhance study reliability.

The review concludes that most predictive modelling studies in this area lack rigor, leading to a high risk of bias. Future studies should focus on improving methodological standards and conducting external validation to enhance the reliability and applicability of predictive models for Alzheimer’s disease dementia. Emphasizing transparent reporting and adherence to the scientific method will be essential in improving the accuracy and generalizability of future research findings.

Original Article: BMC Geriatr. 2024 Jun 19;24(1):531. doi: 10.1186/s12877-024-05044-8.

You can follow our news on our Telegram and LinkedIn accounts.

Subscribe to our newsletter

To be updated with all the latest news, offers and special announcements.

Latest article