A new study published in the journal Autism Research indicates that empathy may operate quite differently in individuals with autism spectrum condition compared to those with social anxiety. Both groups tended to report elevated levels of emotional distress in social situations, but only individuals with autism showed lower levels of emotional concern for others. The researchers also introduced a new way to distinguish between these overlapping conditions: a ratio of self-focused distress to other-oriented concern. This metric, they argue, could help improve diagnosis and support strategies.
Autism and social anxiety are both linked to social difficulties, but for different reasons. People with autism spectrum condition often struggle with understanding and responding to social cues. Social anxiety, on the other hand, involves intense fear of being judged or rejected in social situations. These overlapping traits often make it difficult to distinguish between the two in clinical settings, especially since many autistic people also experience high levels of social anxiety. This overlap has created a need for better tools to tease apart the similarities and differences in how these groups perceive and respond to others.
One of the key traits involved in social functioning is empathy, which can be broadly divided into two components. Cognitive empathy involves understanding another person’s thoughts and feelings. Affective empathy relates to emotional responses to others, such as compassion or distress. Prior studies suggest that people with autism tend to show lower cognitive empathy and mixed patterns in emotional empathy, with some reports of heightened emotional distress. In contrast, individuals with social anxiety may show high emotional reactivity but impaired accuracy in interpreting others’ mental states. Yet few studies have directly compared these two groups on both forms of empathy, using tools that can separate the different aspects.
To address this gap, researchers at the Bruckner Autism Research Center at Ariel University recruited 105 university students and divided them into three groups. One group consisted of students with a confirmed autism diagnosis. A second group included students with high levels of social anxiety but no autism diagnosis. The third group served as a control sample, reporting neither autism nor elevated anxiety.
All participants completed several standardized assessments, including the Interpersonal Reactivity Index, which measures both cognitive and affective empathy, and the Reading the Mind in the Eyes Test, which evaluates the ability to infer emotions based on subtle facial cues.
The results suggested clear differences between the groups when it came to trait-based empathy. Individuals in the autism group scored lower on perspective taking and empathic concern compared to the control group. Their empathic concern scores were also lower than those in the social anxiety group. Both the autism and social anxiety groups reported significantly higher levels of personal distress in response to others’ emotions compared to the control group. This suggests that while both groups experience intense emotional reactions in social situations, only the autistic group showed a notable reduction in the ability to feel concern for others.
One of the study’s key contributions was the introduction of a novel metric: the personal distress-to-empathic concern ratio. This ratio compares how much emotional discomfort a person feels in response to others’ distress (personal distress) to how much they feel for the person experiencing it (empathic concern). A higher ratio indicates a more self-focused, emotionally overwhelmed response.
The researchers found that the autism group had the highest average ratio, followed by the social anxiety group, and then the control group. This ratio was also positively linked to the severity of autistic traits and anxiety symptoms, but only in the autism group. In other words, higher emotional distress relative to empathic concern was particularly associated with autism.
Importantly, the researchers tested whether this ratio could help distinguish between autism and social anxiety. Using a statistical approach called receiver operating characteristic analysis, they identified a cutoff value of 0.83 on the ratio scale. This value demonstrated good sensitivity, meaning it could correctly identify many individuals with autism, though its specificity was moderate, indicating a moderate ability to rule out autism in those without it. This finding suggests that the PD/EC ratio, while not a standalone diagnostic tool, could serve as a useful screening measure in conjunction with other assessments.
When it came to the Reading the Mind in the Eyes Test, which provides a snapshot of a person’s ability to recognize emotions from subtle visual cues, the groups performed similarly. All scored within the normal range, and no significant differences emerged. This suggests that state-based cognitive empathy, at least as measured by this task, may not differ meaningfully across these populations.
However, the researchers noted that the simplicity of the task might not fully capture the complexity of real-life social interactions, particularly for autistic individuals who may rely on compensatory strategies in structured testing environments.
“We observed a significant difference in empathy levels between the autism and social anxiety groups,” explained study author Esther Ben-Itzchak, the head of the Bruckner Autism Research Center. “Both groups experienced higher personal distress compared to the control group; however, only the participants with autism showed significantly lower levels of empathic concern. This difference was effectively illustrated by a new ratio we developed, called the personal distress-to-empathic concern ratio (PD/EC). This ratio successfully distinguished between autism and social anxiety, conditions that frequently overlap in clinical settings, which was both surprising and exciting.”
“We were also surprised that by using a well-known and widely applied measure, the Interpersonal Reactivity Index (IRI) in a new way, namely computing the ratio between personal distress and empathic concern, we could provide a fresh application for this classic tool.”
The findings support a growing view that empathy is not a single, unified trait but a multidimensional construct, with different profiles emerging across psychological conditions. In the case of autism, the study suggests a combination of reduced perspective taking and empathic concern, coupled with heightened personal distress. For those with social anxiety, the pattern includes elevated distress without a corresponding drop in empathic concern. These distinctions have important implications for how we assess and support individuals facing social challenges.
The study also aligns with existing theories about emotional regulation in autism. One model suggests that autistic individuals may experience high levels of emotional arousal in social situations but lack the cognitive tools to manage those emotions effectively. This imbalance may lead to withdrawal or self-focused responses, which can be misinterpreted as a lack of empathy. The current findings add nuance to this picture by showing that lower levels of empathic concern are not necessarily universal but may interact with how individuals experience and regulate emotional distress.
Like all research, there are limitations. Most notably, the social anxiety group was identified using a self-report scale rather than clinical interviews, which means some participants may not meet formal diagnostic criteria. The sample also included a relatively small number of women, which limits generalizability across genders.
Another limitation involves the overlapping nature of autism and social anxiety. In this study, about three-quarters of the autistic group also scored high on the social anxiety measure. Although the authors ran additional analyses to control for this overlap, it remains difficult to completely disentangle the effects of each condition. A follow-up study with larger groups could compare autistic individuals with and without social anxiety, alongside a group with social anxiety but no autism, to better isolate the unique contributions of each condition to empathy profiles.
“Incorporating experimental tasks or physiological measures of empathy in future work could enhance the validity and depth of the findings,” Ben-Itzchak said.
Still, the current findings offer a compelling direction for future research and clinical practice. The PD/EC ratio provides a fresh application of a well-established empathy questionnaire and shows promise as a simple, scalable measure to support diagnostic differentiation. It may also help guide targeted interventions by highlighting specific empathy-related challenges, such as managing distress or building perspective-taking skills.
“We see two main directions for extending this work,” Ben-Itzchak explained. “First, we plan to replicate the findings across different populations by including more women in similar young adult samples, as well as examining clinical populations and younger children. Second, we aim to integrate the behavioral questionnaires with physiological markers of emotional regulation (such as autonomic nervous system measures), which could enhance the clinical utility of empathy profiles.”
“This is the first study to directly compare detailed empathy subcomponents in autism and social anxiety, while carefully screening the social anxiety group for autistic traits,” she added. “By doing so, we were able to reveal unique empathy profiles for each condition. Clinically, these insights could support more accurate differential diagnosis, and theoretically, they advance our understanding of empathy as a multidimensional construct. Importantly, our work underscores that while both groups experience elevated personal distress in social situations, the presence or absence of empathic concern may be a key factor that sets them apart.”
The study, “Distinct Empathy Profiles in Autism and Social Anxiety: A Comparative Study,” was authored by Sigal Tikochinsky, Esther Ben‐Itzchak, and Gil Zukerman.
