An unusual attraction to the eyes in Williams-Beuren syndrome: A manipulation of facial affect while measuring face scanpaths

Abstract

Introduction. This study aimed to investigate face scanpaths and emotion recognition in Williams-Beuren syndrome (WBS) and whether: (1) the eyes capture the attention of WBS individuals faster than typically developing mental age-matched controls; (2) WBS patients spend abnormally prolonged periods of time viewing the eye region; and (3) emotion recognition skills or eye gaze patterns change depending on the emotional valance of the face.

Methods. Visual scanpaths were recorded while 16 WBS patients and 16 controls passively viewed happy, angry, fearful, and neutral faces. Emotion recognition was subsequently measured.

Results. The eyes did not capture the attention of WBS patients faster than controls, but once WBS patients attended to the eyes, they spent significantly more time looking at this region. Unexpectedly, WBS patients showed an impaired ability to recognise angry faces, but face scanpaths were similar across the different facial expressions.

Conclusions. Findings suggest that face processing is atypical in WBS and that emotion recognition and eye gaze abnormalities in WBS are likely to be more complex than previously thought. Findings highlight the need to develop remediation programmes to teach WBS patients how to explore all facial features, enhancing their emotion recognition skills and “normalising” their social interactions.

Keywords:

• Face processing
• Face gaze
• Eye gaze
• Emotion processing
• Social salience

Acknowledgements

This study was funded by a Macquarie University Research Development Grant, awarded to MP. We would like to say a special thank you to participants and their families for their ongoing support in our research

Notes

¹Developmental age and IQ were established using the Woodcock-Johnson Test of Cognitive Ability–Revised (Woodcock & Johnson, 1989/1990). We chose to match WBS patients to typically developing controls using global rather than verbal or nonverbal abilities, because: (1) global ability scores (Broad Cognitive Ability and overall developmental age) are the most reliable estimate of cognitive ability from the WJ-R COG; (2) there was no significant difference between WJ-R COG Oral Language Domain or Visual Processing Domain scores for our WBS cohort, p>.1; and (3) there was no significant correlation between visual scanpath (or emotion recognition) data and either Oral Language Domain, p >.1 or Visual Processing Domain, p>.1. Verbal (Oral Language) domain scores on the WJ-R COG ranged from 37 to 80 (M=63, SD=12). Nonverbal (Visual Processing) domain scores ranged from 17 to 90 (M=62, SD=20). Broad Cognitive Ability is similar to Full-Scale IQ on Wechsler Intelligence tests (M=100, SD=15) on a standardised sample population).

²WBS and TD controls spent a very similar amount of time engaged in gaze on screen, with MDTP on screen being 8868 ms (SD=179 ms), on average, for WBS and 8406 ms (SD=175 ms), on average, for TD controls; there was no significant group difference, p>.1.

³Results for left eye, right eye, and brow region are available upon request.

⁴Results for MFP are available upon request.

⁵Comparing the data published in Marsh (2008) on typically developing 12–18 year olds to that of Loughland et al. (2002) on adults, children displayed significantly more fixations to the eye region of the face for happy and neutral expressions, t(96) = 7.762, p<.000 for happy faces, and t(96) = 4.651, p<.000 for neutral faces, and children displayed significantly more fixations to the salient feature areas of the face for happy and neutral expressions, t(89) = 5.660, p<.000 for happy faces, and t(96) = 5.451, p<.000 for neutral faces; no other expressions were available for analysis. Marsh and Loughland et al. used the same eyetracker equipment, the same face stimuli (Ekman faces), similar methodology, and the same software analysis programs.