Early detection of cognitive impairment in the general population through a digital tool targeting the prospective...

Based on Prospective Memory (PM), we developed a digitalized tool (Chrono) aimed at the detection of cognitive dysfunction. It incorporated event-based tasks signaled by focal/non-focal cues.

The validation group consisted of 67 individuals (${\overline{x}}_{\text{age}}$ = 61.24; range 50-74 years) assigned, according to MoCA's cut-off, to Probable impaired (n = 25) vs. Not impaired (n = 42) groups. Participants completed the Dementia Apraxia Subtest, FCRST, Prose Memory, Rey's 15 words, Digit Span, Corsi Block-Tapping, and Chrono PM tasks. Founded on focal trials, we defined a Chrono score (max = 5). Relationships between PM scores and Updating, Shifting, and Inhibition were also explored.

Chrono score achieved an AUC = 0.813. According to sensitivity (0.83) and specificity (0.64), the optimal cut-off of 2.5 allowed discriminating between groups (p < 0.001) for both focal (2.80 vs. 4.12; $t_{31.25}=-3.83$) and non-focal (2.72 vs. 4.19; $t_{32.49}=-4.56)$ performance. A positive correlation between IFR and the adjusted overall sum of the focal trials confirmed convergent validity. A stepwise logistic regression identified updating, PM task, shifting switch trials, and shifting non-switch trials as significant predictors of membership in the probable impaired group. Adding the PM task to the Chrono's cognitive tests model significantly improved its fit (AIC = 64.411 versus AIC = 62.255; LRT: 𝜒22 = 6.156, p = 0.131). The covariates effects were not significant: age (OR = 0.976, 95% CI [0.305,1.553]; z = -.401, p = 0.688), gender (OR = 0.377, 95% CI [0.089,1.501]; z = -1.375, p = 0.169) and schooling (LRT: 𝜒22 = 1.544, p = 0.819).

As an exploratory study, our findings partially support the potential application of Chrono as a digital screening device for cognitive impairment.