Abstract
Alzheimer’s disease (AD) is a prevalent neurodegenerative condition primarily impacting the elderly population. Despite its widespread occurrence, the precise etiological factors remain elusive, emphasizing the critical need for early detection and intervention to mitigate disease progression. This thesis endeavours to address this challenge by focusing on the identification of sociodemographic, lifestyle, vital signs, neuropsychological, and behavioural risk factors that potentially contribute to the development of AD.This thesis introduces the research, highlighting the urgency of addressing Alzheimer’s disease and the potential impact of early detection on preventative strategies. A thorough literature review provides a foundation for understanding the current state of knowledge on AD and sets the stage for the subsequent chapters.
In this thesis, Pearson's chi-squared test was used for categorical variables and one-way ANOVA for continuous variables to analyze the association of sociodemographic, lifestyle, and neuropsychological risk factors with the development of Alzheimer’s disease (AD) in two separate cohorts, as detailed in Chapter 4. The first cohort consisted of patients without a history of AD development, while all patients in the second cohort developed dementia within 36 months. Significant differences were observed in Marital Status, Occupation, APOE4 genotype, Geriatric Depression Scale, and Functional Questionnaire Assessment score between Normal control and AD patients in both cohorts. Logistic regression was employed to develop Alzheimer’s early prediction models, which exhibited high performance in ROC (receiver operating characteristic) measurements. These prediction models investigated the effects of socio-demographic and neuropsychological risk factors (Age, Occupation, APOE4 gene, Geriatric depression scale (GDTOTAL) Functional questionnaire assessment score (FAQTOTAL)) on AD development across both cohorts. The identified significant risk factors from both cohorts could potentially contribute to the development of a reliable and easily accessible Alzheimer's risk app.
The daily dietary habits represent one of the most crucial aspects of people's lifestyles, with significant implications for brain health. In this thesis, Chapter 5undertook a comprehensive examination of dietary patterns, comparing individuals with Alzheimer’s disease to those without Alzheimer’s disease using advanced statistical techniques like Multiple factor analysis (MFA) and Classification Modeling. Utilizing MFA, the research delved into dietary patterns, while employing Random forest (RF) classifier and Sparse logistic regression (SLR) for Variable Importance analysis to pinpoint food items significantly linked to AD. The MFA analysis revealed striking trends, particularly a robust correlation (Lg=.92, RV=0.65) between the consumption of processed foods and meats in Alzheimer's patients, whereas no significant relationships were observed within the Healthy control (HC) group. Notably, food items such as Meat pie, Hamburger, Ham, Sausages, Beef, and Vegetables such as Capsicum and Cabbage emerged as crucial variables associated with Alzheimer’s disease through RF and SLR analyses. The findings suggest that the diversity or balance of daily dietary intake may decrease the risk of development of AD. RF and SLR classifications underscore the association of processed foods, particularly deli meats and meat-based products, with AD, indicating a need for reducing reliance on such items in daily consumption, especially among younger generations.
This thesis (Chapter 6) also examines various lifestyle factors influencing the transition from Mild cognitive impairment (MCI) to Alzheimer’s disease (AD) using time-to-event data and longitudinal measures. Cox proportional hazard models and Aalen's additive regression models are employed to identify significant variables affecting AD development. A comparative analysis of survival models, including Cox proportional hazard regression, Time-dependent covariates Cox model, and Aalen's additive hazards (AH) model, is conducted utilizing both time-to-event data and longitudinal data in the context of Alzheimer’s progression. The primary focus of using survival analysis was on identifying and evaluating socio-demographic, neuropsychological, behavioural, and lifestyle-related risk factors associated with disease progression from MCI to AD. The study particularly emphasizes socio-demographic, lifestyle, behavioural, and vital sign factors, introducing two novel variables, Sleeping disorder and Eating disorder, have not previously utilized in survival analyses with ADNI data. Surprisingly, the research reveals that Daily functional activities score and Sleeping disorder emerges as a significant risk factor for AD, increasing the likelihood of its occurrence over time.
This thesis also entails the development of an Alzheimer’s disease (AD) risk index in chapter 7, which incorporates socio-demographic factors, lifestyle, medical history, behavioural patterns, and neuropsychological risk factors. Through the analysis of these variables and assigning scores to healthy individuals, those with mild cognitive impairment, and Alzheimer's patients, an application for early detection of Alzheimer’s disease or dementia risk could be developed.
Overall, this doctoral thesis represents an extensive exploration and proposes implications for early detection strategies and potential interventions. The research presented holds significant promise in advancing our understanding of Alzheimer's disease, providing insights that may inform preventative measures and strategies for effective early intervention.
| Date of Award | 28 Nov 2024 |
|---|---|
| Original language | English |
| Supervisor | Kuldeep Kumar (Supervisor) & Ping Zhang (Supervisor) |