Abstract
XpressO is a digital cognitive prescreening tool developed by Montreal Cognition (MoCA Test Inc.). In this study, we evaluated real-world online data collected through XpressO in 2024 and early 2025, based on over 50,000 self-enrolled online participants. In line with expectations, we found that the XpressO score-which predicts screening positive for (mild) cognitive impairment-is associated with sex, age, and education. The results indicated that women have a 5.8% [5.3%, 6.3%] lower relative risk of prescreening positive for cognitive impairment, each additional year of age increases the relative risk by 0.59% [0.57%, 0.60%], whereas each year of education decreases the relative risk by 0.99% [0.94%, 1.06%]. We also visualized and quantified interaction effects among these demographic variables as predictors of the XpressO score. While the interaction effect between sex and age was not statistically significant, all other interaction terms, including the three-way interaction between sex, age, and education, were significantly associated with the XpressO score. Additionally, adjusting for demographic factors reduced the observed effect of potential confounders, the language and the platform used. However, when we evaluated a score adjusted for demographics in a clinical cohort of 101 participants, we found that this adjustment slightly but significantly reduced the discriminatory power of the XpressO tool in identifying individuals with cognitive impairment from 0.86 to 0.81. These findings from a real-world online cohort offer novel insights into the complex influence of demographic factors on digital cognitive prescreening. Moreover, they demonstrate that XpressO is a viable tool for online prescreening and can help streamline the diagnostic pathway for individuals who may be eligible for disease-modifying treatments for Alzheimer's disease.