(PS2-52) Capturing Loneliness in Real-time: Differences in Momentary Loneliness Between Autistic and Non-autistic Adults

Abstract Text: While prevailing theories suggest autistic individuals prefer solitude (Chevallier et al., 2012), autistic people report a strong desire for social connection and higher loneliness compared to non-autistic peers (e.g., Jaswal & Akhtar, 2019). Existing research on loneliness in autistic adults relies on retrospective self-reports without specifying time frames, introducing biases and limiting insights into contextual variability. Ecological momentary assessment (EMA) offers a promising alternative that addresses these challenges through real-time reporting of events. Despite its feasibility among autistic adults (Chen et al., 2025), EMA has rarely been used to investigate loneliness. The current study addresses this gap by using EMA to examine momentary loneliness among autistic and non-autistic adults, and how it varies across social context. Participants included 214 young adults aged 18-24 (65% female assigned at birth) across two diagnostic groups: non-autistic (N=176) and autistic (N=38). Participants reported their momentary loneliness and current activity five times daily over 10 days for up to 50 total responses. Multilevel modeling (MLM) assessed the relationship between autism traits and momentary loneliness. Additional models examined the relationship between momentary loneliness and social contexts: alone, with family, friends, a romantic partner, and acquaintances. Diagnostic group (autistic vs. non-autistic) was included as a moderator to assess whether the relationship between social context and loneliness differed by group. There were no significant group differences in momentary loneliness. Both groups experienced increased loneliness when alone (β=0.17, p< .001), with a stronger effect in autistic individuals (β=0.26, p< .001) compared to non-autistic individuals (β=0.15, p< .001). Both groups reported less loneliness when with family (β=-0.08, p=.009), friends (β=-0.16, p< .001), or a romantic partner (β=-0.23, p< .001), with no significant group interactions. However, a significant interaction between group and being with an acquaintance emerged (β=-0.19, p=.03). Simple slopes analyses showed that autistic individuals reported lower loneliness with acquaintances (β=-0.18, p=.02), while non-autistic individuals did not. Our findings revealed that, in contrast to trait loneliness, momentary loneliness was not greater in autistic adults. Further, while both groups reported more loneliness when alone, this effect was stronger in autistic individuals. This suggests that autistic individuals may be more sensitive to solitude than non-autistic peers, making them particularly vulnerable to loneliness. Additionally, autistic individuals reported lower loneliness with acquaintances, which was not the case for non-autistic peers. This finding aligns well with recent research suggesting that social interactions with ‘weak-ties’ are uniquely meaningful for autistic people (Pellicano & Heyworth, 2025). Our findings demonstrate the value of social context when studying loneliness in autistic adults, highlighting the importance of interventions that go beyond fostering close friendships to support a range of meaningful social interactions for autistic individuals.