AbstractBackgroundAlzheimer’s disease can influence gaze behavior, even in prodromal stages such as mild cognitive impairment (MCI). Alzheimer’s pathology impacts exploratory eye movements, and can be detected with eye tracking applied to novelty preference tasks that leverage visual short‐term memory. MCI due to Alzheimer’s disease is often associated with attentional deficits, with distracting stimuli capturing a greater amount of attention than seen in healthy controls (HC). In this pilot study, we examine the effects of distractors, including categorical, background, and “unexpected” distractors, in a novel naturalistic visual search task.MethodsUsing a novel visual search task, 45 participants (35 HC, 10 MCI) were asked to count the number of animals (either flying or land animals) in a natural scene while their eye movements were recorded. Scenes contained either no distractors (11 scenes), “expected” distractors of the opposite animal type (8 scenes), or “unexpected” distractors that appeared unusual, such as fish on land (2 scenes). Four of the scenes (1 with distractors) contained a vibrant or interesting background.ResultsNo difference was found between MCI and HC in gaze behavior towards expected distractors, however unexpected distractors captured significantly (p<.05) longer and more frequent visits in MCI participants (1.2 seconds and 7.0 visits) compared with HC participants (0.6 seconds and 4.5 visits). In contrast, duration and visits towards targets did not differ, however MCI participants took significantly longer to first attend targets (1.5 seconds) than HC participants (0.5 seconds) when unexpected distractors were present (p = .02). In MCI participants, interesting background stimuli captured significantly more gaze in the first 3 seconds of presentation (35.3% MCI vs 25.8% HC; p = .04).ConclusionsIn this preliminary study, we found that HC participants showed greater resilience to distractions compared to MCI participants, who showed less ability to maintain focus in the presence of unanticipated distractors. Attention was pulled from the counting task towards a distracting background as well as towards the unexpected distractors in MCI participants, but not when distractors appeared to fit naturally in the surrounding scene. Future research will examine effects of age‐related decline and biomarker status in unimpaired participants.