BACKGROUND: As investigations of disease modifying drugs aim to slow down progression of Alzheimer' disease (AD) biomarkers to reliably track disease progression gain more importance. This is especially important as clinical symptoms, including psychometric measures, are only modestly associated with the underlying disease pathology, in particular at the pre-dementia stages. The decision which biomarkers to choose in clinical trials is crucial and depends on effect size. However, longitudinal studies of multiple biomarkers in parallel that allow direct comparison on effect size are scarce. METHODS: We calculated effect size and minimal sample size for three common imaging biomarkers of AD, namely amyloid deposition measured with PiB-PET, neuronal dysfunction measured with FDG-PET and cortical thickness measured with MRI in a prospective 24-month follow-up study in a monocentric cohort of early AD. RESULTS: Post hoc power calculation revealed large effect sizes of Cohen's d for PiB-PET and cortical thickness and a small effect size for FDG-PET (1.315, 0.914, and 0.341, respectively). Accordingly, sample sizes for PiB-PET and cortical thickness required significantly smaller sample sizes than FDG-PET to reliably detect statistically significant changes after 24 months in early AD (n = 7, n = 12, and n = 70, respectively). CONCLUSION: Amyloid imaging with PET and measuring cortical thickness with MRI are suitable biomarkers to detect disease progression in early AD within a small sample.      «

BACKGROUND: As investigations of disease modifying drugs aim to slow down progression of Alzheimer' disease (AD) biomarkers to reliably track disease progression gain more importance. This is especially important as clinical symptoms, including psychometric measures, are only modestly associated with the underlying disease pathology, in particular at the pre-dementia stages. The decision which biomarkers to choose in clinical trials is crucial and depends on effect size. However, longitudinal st...     »