Alzheimer’s disease disrupts domain-specific and domain-general processes in numerosity estimation

ABSTRACT

Introduction

This study investigated how Alzheimer’s Disease (AD) affects numerosity estimation abilities (e.g., finding the approximate number of items in a collection).

Method

Across two experiments, performance from HOA (i.e., Healthy Older Adults; N = 48) and AD patients (N = 50) was compared on dot comparison tasks. Participants were presented with two dot arrays and had to select the more numerous dot array in comparison tasks. They also took a Simon task and a number-line tasks (i.e., number-line tasks in which they had to indicate the position of a number on a line 0 to 100 or on a line 0 to 1,000 in the number-line task).

Results

In Experiment 1, (a) AD patients obtained significantly poorer performance while comparing collections of dots, especially harder (small-ratio) collections, (b) these deficits correlated with poorer performance on the number-line task for larger numerosities (i.e., 0 to 1,000), and (c) AD patients showed poorer performance on incongruent (where numerosity and area occupied by dots mismatched) than on congruent items (where both features matched), while HOA showed no congruency effects. Experiment 2 showed (a) congruency effects in both groups when convex hull was tested as an incongruent feature, and (b) comparable sequential modulations of congruency effects in both groups.

Conclusions

Our findings showed that numerosity abilities decline in AD patients, and that this decline results from impaired domain-specific processes (i.e., numerosity processing) and domain-general processes (i.e., inhibition). These findings have important implications to further our understanding of how specific and general cognitive processes contribute to numerosity estimation/comparison performance, and how such contributions change during Alzheimer’s disease.

KEYWORDS:

• Alzheimer’s disease
• dot comparison
• numerosity processing
• inhibition
• executive control

Data availability

Data will be made publicly available on data sharing website (e.g., OSF), once this article is accepted for publication or during review process upon request.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1. Item-based correlations between speed and accuracy in each group were significantly positive in HOA (r=.46 &.34, in Expts. 1 & 2, respectively), and non-significant in AD (r=.15 &.04, in Expts. 1 & 2). These correlations showed that increased RTs associated with increased percentages of errors were significant in HOA, and that speed and accuracy were not related in AD. This was also seen when examining means. For example, in Expt. 1, both HOA and AD patients were faster on large-ratio items than on small-ratio items and both groups were more accurate on large-ratio items than on small-ratio items; the ratio effects were larger in AD patients than in HOA on both speed and accuracy. As another example, in Expt. 2, poorer performance on incongruent relative to congruent items was found on both speed and accuracy in AD patients and in HOA.

Additional information

Funding

This research was supported by a grant to PL from the Agence Nationale de la Recherche (Grant # ANR- 17-CE28-0003).