![[Adobe Stock]](https://sp-ao.shortpixel.ai/client/to_webp,q_glossy,ret_img,w_750,h_375/https://www.psypost.org/wp-content/uploads/2025/04/happy-sad-faces-750x375.jpg)
A new study published in the journal Development and Psychopathology offers a new perspective on why some individuals with elevated autistic traits struggle to recognize emotions in faces. Researchers found that while autistic traits were linked to poorer recognition of emotions in human faces, this association disappeared when controlling for alexithymia—a trait characterized by difficulty identifying and describing one’s own emotions. In fact, alexithymia alone predicted lower emotion recognition scores for both human and anime faces, suggesting it may play a more central role than autism-related traits in shaping emotional processing challenges.
Previous research has shown that many individuals on the autism spectrum have difficulty interpreting emotional expressions in human faces. At the same time, a strong preference for anime—an animated art form with exaggerated emotional cues—is often observed in these individuals. Some researchers have hypothesized that the more dramatic and stylized facial expressions typical in anime might be easier for autistic individuals to interpret. Others have questioned whether emotion recognition challenges are driven by autism itself or by co-occurring alexithymia.
“My interest stems from my younger years when I would frequent forums dedicated to the discussion of anime and manga. I believe people tend to be more open about themselves when they are anonymous or at least semi-anonymous, and one thing I noticed is that on these sites users would mention that they have autism more often than on other parts of the internet,” said study author Bridger Standiford, who conducted the research while an undergraduate at Penn State Abington.
“As I progressed through my college studies and learned more about autism in a formal setting, something clicked when I learned about reduced facial emotion recognition in people with autism. I wondered if the exaggerated expressions often made by anime characters are what in part attracted those with autism to the medium seemingly more often than others.”
To investigate, the researchers recruited 247 adults with varying levels of autistic traits. Participants completed standardized measures of autistic traits (the AQ-10) and alexithymia (the Revised Toronto Alexithymia Scale), along with a facial emotion recognition task. The task included 12 human faces and 12 anime-style faces, each showing one of six basic emotions such as happiness or sadness. Participants viewed each image for three seconds before identifying the emotion they believed was being expressed.
The results showed that individuals with higher autistic trait scores performed significantly worse at identifying emotions in human faces. However, when it came to anime faces, their performance was similar to that of individuals with lower autistic traits. This finding supports the idea that anime expressions may be more accessible to those high in autistic traits, possibly because their exaggerated style compensates for difficulties with subtle emotional cues.
Yet the most revealing finding came from the regression and mediation analyses. When researchers statistically controlled for alexithymia, the link between autistic traits and poor emotion recognition vanished. Alexithymia, on the other hand, consistently predicted worse performance on both the human and anime emotion tasks, and it fully explained the relationship between autistic traits and facial emotion recognition. In other words, it wasn’t the autistic traits per se that drove difficulty in recognizing emotions—it was alexithymia.
These findings build on a growing body of research suggesting that alexithymia may be a key factor underlying emotional processing challenges in autistic individuals. Earlier studies have shown that alexithymia is much more common among people on the autism spectrum than in the general population, with some estimates placing its prevalence between 50 and 85 percent. Unlike autism, which is typically associated with broader social communication challenges, alexithymia specifically affects how people identify and verbalize emotions—both in themselves and in others.
“There are two key takeaways from our study,” Standiford told PsyPost. “The first is alexithymia’s role in emotion recognition. We found that difficulties in recognizing facial emotions are more closely linked to alexithymia – a condition characterized by challenges in identifying and describing one’s own emotions – than to autistic traits alone.”
“The second involves the impact of human versus anime faces on emotion recognition. Participants with higher levels of autistic traits showed poorer performance in recognizing emotions on human faces, but not on anime faces. However, when accounting for alexithymia, this association diminished, indicating that alexithymia is a likely source of the emotion recognition struggles experienced by many individuals high in autistic traits. In other words, autistic traits negatively influence emotion recognition via alexithymic traits.”
“The results of our study also have some important implications for interventions,” Standiford continued. “First, interventions aimed at improving facial emotion recognition, especially for individuals on the autism spectrum, might be more effective if they also focus on and address alexithymic traits. Additionally, incorporating anime characters, which present emotions in a more stylized and exaggerated manner, into interventions designed to improve emotion recognition may serve to increase their effectiveness.”
The study is not without limitations. The participants were not clinically diagnosed with autism but instead represented a range of autistic traits as measured by a brief screening tool, the AQ-10. While this allowed the researchers to investigate these traits in a community sample, it limits the ability to generalize the findings to individuals formally diagnosed with autism spectrum conditions.
“A second caveat involves the anime faces used as targets in our study,” Standiford noted. “While the human faces we used in our study had been rigorously tested and validated in previous research, the anime faces were selected and validated solely by me and my co-author, Dr. Kevin Hsu, as no validated collection of anime facial expressions existed. Validating a set of anime faces with differing emotional expressions will be the focus of future research.”
The study, “Autistic traits, alexithymia, and emotion recognition of human and anime faces,” was published March 20, 2025.