Visual pursuit of a moving object signifies the transition from unawareness to awareness in prolonged disorders of consciousness (PDOC) – i.e. from vegetative state to minimally conscious state [1]. Accurate differential diagnosis is crucial in PDOC because treatment decisions may be influenced by differences in likelihood of functional recovery that are associated with each diagnostic group [2, 3].

Two recent studies found that visual pursuit is significantly more likely to be observed in PDOC when the stimulus to be pursued is the patient’s own face reflected in a mirror, relative to when the stimulus is an object or person. [4, 5]. The cause of this effect is unclear, although it has been hypothesised that the patient’s own face is more likely to attract the gaze as it is a salient and auto-referential stimulus [4, 5]. This hypothesis, however, presumes that reflected-own-face pursuit indexes consciousness to the same extent as object pursuit. In other words, a patient who only pursues their reflected-own-face, and not an object, is considered to be conscious to the same clinical level as a patient who pursues both. Any difference in pursuit behaviour between these two patients is considered to stem from an attention deficit that can be overcome with a sufficiently salient stimulus [4, 5].

However, extremely highly learned stimuli, such as one’s own name or own face, can elicit complex responses in the absence of awareness. For example, one’s own name is subjected to complex neural processing even during non-REM sleep — a period in which an individual is demonstrably unaware [6]. Furthermore, the brain’s electrophysiological response differentiates one’s own-face from faces of others even when the individual is unaware of having seen a face at all [7]. An alternate interpretation, therefore, is that processing of high relevance stimuli, such as pursuit of one’s own reflected face, may be less reliant on overt consciousness than processing of less salient stimuli, such as pursuit of an object. In other words, is reflected-own-face pursuit less demanding than object pursuit, thus allowing patients with lower cognitive abilities, or even lower levels of consciousness, to exhibit reliable pursuit? This interpretation has important implications for diagnoses achieved through responses to high relevance stimuli, such as own-face and own-name [8–10]. A further alternative physiological interpretation for the difference in pursuit behaviour is that conscious patients are actively attempting to pursue both stimuli, but that the eye movements during reflected-own-face pursuit are naturally smoother and therefore easier for an assessor to identify confidently.

To test these hypotheses, we asked a group of healthy participants to separately pursue an object and their reflection in a mirror while completing a challenging secondary task designed to reduce the cognitive resources available for smooth pursuit. If reflected-own-face pursuit does not draw on conscious resources to the same extent as object pursuit, we would predict differential performance on the secondary task and on measures of smooth pursuit, quantified via a head-mounted eye-tracker.

Our data indicate that pursuit of one’s own face reflected in a mirror is no less cognitively demanding than pursuit of an object, and therefore that both stimuli tap similar diagnostically important abilities in PDOC. Indeed, a Bayesian analysis indicates substantial evidence in favour of no differences in secondary task performance during reflected-own-face or object-pursuit.

Eye-tracking indicates that reflected-own-face pursuit is naturally smoother than object pursuit, and may therefore be less likely to be missed or confused with irrelevant staccato eye movements (e.g. nystagmus). Indeed, the data provide very strong evidence for significantly shorter mean eye fixations during reflected-own-face pursuit, which we interpret as a sign of smoother pursuit because the eyes do not take long pauses to fixate. Fixations also occur more uniformly across the space of object pursuit, consistent with less smooth pursuit.

While it is possible that the eye movements of healthy participants reported here do not generalise to PDOC, recent eye-tracking evidence suggests generalisability between patients and healthy controls [17]. We cannot rule out the possibility that a more demanding secondary task will have an effect on our measures of smooth pursuit, unlike the two-back task we employed here. However, even if this were the case, it is unlikely that the interaction between task performance and pursuit stimulus would be borne out, as indicated by the strong evidence against interaction models in our data (number of fixations: interaction JZS-BF₁₀ = .040; duration of fixations: interaction JZS-BF₁₀ = .404; relative to model with highest evidence). We therefore conclude that the cognitive demands of reflected-own-face pursuit and object pursuit are comparable.

These data provide further support for the use of a mirror in clinical assessments of awareness [4, 10]. It has been suggested that the greater sensitivity of the reflected-own-face stimulus stems from its higher salience and subsequent auto-referential processing – i.e. self-awareness [4, 5]. Indeed, there is evidence that autonomic responses to own-face photographs in the minimally conscious state are more similar to healthy individuals, while the autonomic response of patients in the vegetative state fails to differentiate between own-face and control stimuli [18]. Nevertheless, open questions remain regarding the subjective experience of patients who pursue their reflected own-face: does reflected-own-face pursuit index a simultaneous subjective experience of self-awareness on the part of the patient? Contemporary advances in brain imaging and physiological recordings may provide some insight into this question. Regardless, our data are consistent with the view that visual pursuit of reflected-own-face in PDOC is evidence of a similar level of consciousness and cognition to that identified by object pursuit, and therefore is a more appropriate method in clinical assessment.

Our data provide substantial evidence that a mirror provides greater sensitivity to detect visual pursuit in patients with PDOC not because reflected-own-face pursuit is a less complex ability, but because the mirror naturally elicits smoother eye movements that can be identified more confidently by the assessor. These smoother and more reliable eye movements may be driven by the self-referential nature of the reflected-own-face stimulus, although further research is needed to appropriately characterise the relationship with subjective self-awareness.