Cognition in the Oldest Old Using the NIH Toolbox

AbstractBackgroundThe percentage of cognitively intact older adults living into their 80s and beyond is growing rapidly. This age group is not only one of the fastest growing segments of the population, but it is also the group most susceptible to developing dementia. Effective detection of early cognitive decline is paramount but is complicated by a dearth of normative data for older adults above the age of 85. It is well‐established that crystallized cognition remains stable into older adulthood, whereas fluid cognition tends to decrease; however, it is unknown whether the same holds true for the oldest old. We use the comprehensive NIH Toolbox Cognition Battery (NIHTB‐CB) to assess differences in cognition.MethodParticipants (age range 65 to 94) were diagnosed as having normal cognition and were either 65–85 years old (N = 175, Mage = 73.9, SDage = 6.06) or 86+ (N = 98, Mage = 88.6, SDage = 2.39). Participants completed the NIHTB‐CB, which assesses fluid reasoning, executive functioning, processing speed, episodic memory, and crystallized language abilities. A series of univariate linear models regressed age with covariates (education, gender, race/ethnicity, site) on each NIHTB‐CB domain.ResultAge was a significant predictor for all NIHTB‐CB fluid measures. The R2 for each fluid reason model ranged from 0.17 to 0.22, suggesting a medium effect size. Of the crystallized measures, Picture Vocabulary decreased with age, while oral reading ability remained constant. The effect for Picture Vocabulary appears to be at least partially driven by education levels, as more educated participants exhibited higher scores. For all statistically significant models, age remained a significant predictor after adjusting for demographic covariates and site. The fluid measure with the steepest decline was Picture Sequence Memory and Pattern Comparison, while the fluid measure with the least steep decline was Flanker.ConclusionAs expected, fluid reasoning ability decreased with age in the oldest old. Contrary to theoretical expectations of stability across crystallized abilities, scores on one crystallized measure decreased with age. This study comprehensively characterizes the oldest old across multiple facets of cognition and adds to the extant literature regarding cognitive performance in the oldest old.