Modeling of Functional Assessment Questionnaire (FAQ) as continuous bounded data from the ADNI database.

Area of Science:

• Neurology
• Biostatistics
• Gerontology

Background:

• Assessing independent daily functioning is crucial for Alzheimer's disease (AD) management.
• The Functional Assessment Questionnaire (FAQ) is a key clinical endpoint for AD patients and caregivers.
• Standard statistical methods may not adequately capture the full range of FAQ scores due to boundary data.

Purpose of the Study:

• To develop and validate a mathematical model for describing the natural history of FAQ score changes in AD.
• To compare a censored approach with a standard approach for handling boundary data in FAQ scores.
• To identify covariates that influence disease progression as measured by FAQ.

Main Methods:

• Utilized data from the Alzheimer's Disease Neuroimaging Initiative (ADNI) study, including normal elderly, mild cognitive impairment, and mild AD participants.
• Developed a mathematical model to describe FAQ score changes over time.
• Compared a censored approach (accounting for boundary data) with a standard approach (ignoring boundaries) for FAQ scores.

Main Results:

• The censored approach significantly improved the predictability of disease progression in FAQ scores compared to the standard approach.
• Baseline severity, ApoE4 genotype, age, sex, and imaging biomarkers were evaluated as covariates.
• The censored method demonstrated superior performance in modeling FAQ score changes.

Conclusions:

• A censored approach for handling boundary data in FAQ scores enhances the predictability of Alzheimer's disease progression.
• This modeling technique offers a more accurate way to assess functional decline in AD.
• The methodology is applicable to other bounded outcome measures in clinical research.