Psychometric approaches to defining cognitive phenotypes in the Old Order Amish

Area of Science:

• Neuroscience
• Genetics
• Gerontology

Background:

• Alzheimer's disease (AD) diagnosis relies on memory and cognitive deficits.
• Identifying genetic variants linked to AD is crucial.
• Current psychometric phenotyping often excludes cognitively unimpaired individuals.

Purpose of the Study:

• To compare psychometric methods for creating granular cognitive phenotypes.
• To identify phenotypes applicable to both affected and unaffected individuals.
• To facilitate the search for genetic variants associated with AD and healthy aging.

Main Methods:

• Analysis of 682 Old Order Amish individuals.
• Development of four cognitive models using adjusted Z-scores: global/memory thresholds and global/memory clusters.
• Ordinal regression and ANOVA to assess model coherence with clinical status, APOE-e4, sex, and age.

Main Results:

• The memory cluster model demonstrated the best fit, identifying four distinct phenotypes.
• Individuals in poorer performing memory phenotypes were more likely to be cognitively impaired (CI) or mildly impaired (MI) and carry the APOE-e4 allele.
• Significant differences in memory, language, and executive functioning were observed across the four memory phenotypes.

Conclusions:

• Memory cluster stratification effectively identified cognitive phenotypes aligning with clinical status and genetic factors.
• These refined phenotypes are predicted to aid in discovering protective genetic variants.
• This approach enhances the study of genetic influences on cognitive aging and AD.