“Digital eye tracking and plasma biomarkers: Distinguishing functional cognitive impairment from Alzheimer's disease biology”

1.

Dear Editor,

We sincerely appreciate the insightful feedback from Dr. Julián Benito‐León and Dr. Carla María Benito‐Rodríguez on our recent study, “A novel eye‐tracking digital marker outperforms plasma biomarkers in detecting cognitive impairment,” ¹ published in Alzheimer's & Dementia. As the authors of the paper, we would like the opportunity to draw a clearer conclusion for our study.

We fully agree with the point that our study demonstrated that the novel eye‐tracking digital marker outperforms plasma biomarkers in detecting clinically defined cognitive impairment, not necessarily in detecting Alzheimer's disease (AD) biology. As clearly stated in our paper, ¹ our primary outcome was classification into cognitively unimpaired (CU) and cognitively impaired (CI) groups, with the latter comprising patients with mild cognitive impairment and dementia. We did not claim that eye‐tracking markers replace or surpass plasma biomarkers in identifying AD‐specific pathology. In fact, we explicitly emphasized that eye‐tracking markers capture functional deficits and cognitive decline, which are the clinical manifestations of underlying AD pathology.

We acknowledge that the eye‐panel marker includes two recall‐based variables—instantaneous recall task score (IRTS) and delayed recall task score (DRTS)—alongside three purely oculomotor variables (saccades times [ST], fixation time of cubes [FTcube], and sphere motion path length [SMPL]). As the letter rightly notes, these recall scores share criterion proximity with clinical cognitive status, which may contribute to the panel's strong syndromic discrimination. However, we wish to clarify that our study was designed to evaluate the utility of the eye‐panel marker for detecting cognitive impairment as a clinical syndrome, not for isolating AD pathology. The inclusion of IRTS and DRTS reflects the multi‐domain nature of cognitive decline and enhances the panel's ecological validity as a rapid, game‐based screening tool. Importantly, as shown in Figure 3A of our paper, even when considered alone, purely oculomotor measures such as ST (area under the curve [AUC] = 0.80) and SMPL (AUC = 0.75) demonstrated meaningful diagnostic performance. The eye‐panel marker integrated these oculomotor variables with recall scores to achieve the highest overall accuracy (AUC = 0.865).

Given that the primary goal of our study was to develop a practical, non‐invasive screening tool for cognitive impairment in community and clinical settings, rather than to isolate the biological signature of AD, we believe the inclusion of IRTS and DRTS is methodologically justified. We fully agree with the letter's perspective that eye‐tracking markers are complementary to plasma biomarkers, and we have clarified in our response that the comparison was performed at the syndromic level, not as a claim of biological superiority over AD‐specific plasma markers.

Regarding the suggestion to stratify analyses by amyloid and tau status, we fully agree that this would help clarify whether the eye‐panel marker detects AD pathology, downstream functional impairment, or both. However, as you noted, the number of participants with available positron emission tomography (PET) data in our study was relatively limited. Currently, it is insufficient to yield stable statistical estimates. We agree that future studies with larger PET‐imaged cohorts are needed to address this important question, and we are actively collecting additional data to enable such stratified analyses soon.

A key strength of our study lies in the evaluation of these novel digital eye‐tracking markers for detecting cognitive decline in both clinical and community settings. Moreover, as we reported in the paper, the efficacy of the eye‐panel marker appears to stem from its strong association with core AD pathogenesis, particularly tau pathology, neurodegeneration, and clinical cognitive function.

In summary, we appreciate for helping refine the interpretation of our findings. We fully agree that eye‐tracking should be viewed as a complementary functional biomarker for the syndromic detection of cognitive impairment, rather than a replacement for molecular AD pathology markers. The letter's comments will directly inform our future research directions, particularly regarding external validation and biology‐stratified analyses.

AUTHOR CONTRIBUTIONS

Yi Ling:Writing – original draft, Visualization. Pan Sun:Writing – original draft, Visualization. Tengfei Guo:Writing – review & editing. Benyan Luo:Writing – review & editing.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest. Author disclosures are available in the Supporting Information.

Supporting information

Supporting Information: alz71574‐sup‐0001‐ICMJE.pdf