Impaired face recognition is associated with social inhibition

Abstract

Face recognition is fundamental to successful social interaction. Individuals with deficits in face recognition are likely to have social functioning impairments that may lead to heightened risk for social anxiety. A critical component of social interaction is how quickly a face is learned during initial exposure to a new individual. Here, we used a novel Repeated Faces task to assess how quickly memory for faces is established. Face recognition was measured over multiple exposures in 52 young adults ranging from low to high in social inhibition, a core dimension of social anxiety. High social inhibition was associated with a smaller slope of change in recognition memory over repeated face exposure, indicating participants with higher social inhibition showed smaller improvements in recognition memory after seeing faces multiple times. We propose that impaired face learning is an important mechanism underlying social inhibition and may contribute to, or maintain, social anxiety.

Keywords: Habituation; Individual differences; Novel faces; Social anxiety; Temperament.

Figure 1. Rate of improvement in face recognition memory correlates with social inhibition

The left panel shows the results of the main analysis. Higher social inhibition scores were correlated with smaller b′ slope values across the experiment (change from 1 to 7 exposures, p=0.04), indicating that higher social inhibition was associated with smaller improvement in recognition memory for faces seen seven times compared to faces seen once. The right panel shows the results of 1–7 face exposures exploratory post hoc analyses. Significant correlations are denoted with an asterisk (*). Post-hoc analyses showed a moderate association between higher social inhibition and smaller memory improvements during early face exposures (change from 1 to 3 exposures , p=0.03 uncorrected) and modest, nonsignificant (p>0.05) correlations across later exposure windows; however, the strength of correlations across adjacent windows were not significantly different, suggesting a similar overall impairment across discreet repetition windows.