Testing a Theoretical Model of Loneliness in Adults on the Autism Spectrum

Background:

Loneliness is an experience of negative affect as a result of a discrepancy between the relationships one desires and what one perceives to have and, feeling a lack of intimacy. This is the first study to assess the experience of loneliness in adults on the autism spectrum in comparison to adults not on the spectrum. It is also the first study to examine a theoretical model of loneliness for adults on the spectrum based on a systematic literature review and an adaptation of the 1987 De Jong-Gierveld loneliness model.

Objectives:

The aims of this study were to assess the extent of self-reported loneliness in adults on the autism spectrum compared to other adults and to investigate the determinants of loneliness in adults on the spectrum based on a theoretical model of loneliness.

Methods:

Data was obtained from the Australian Longitudinal Study of Adults with Autism (ALSAA). Participants were recruited nation-wide to complete a self-report questionnaire. A total of 264 participants were included from the ALSAA: 156 adults on the spectrum aged 25 to 80 years (Mean (M)=42.8, Standard deviation (SD)=12.7), and 108 control participants not on the spectrum aged 25 to 77 years (M=42.8, SD=13.0). Variables including demographics, autistic traits, social support characteristics, depression, anxiety, and self-efficacy were assessed through regression models for their impact on loneliness, measured by the 8-item UCLA Loneliness Scale (ULS-8). The study was grounded in an inclusive research approach that involved a research advisory panel of adults on the spectrum providing feedback on questionnaire development and the interpretation of findings.

Results:

Adults on the spectrum scored significantly higher on the ULS-8 (M=23) than adults not on the spectrum (M=15, p<0.001). Adults on the spectrum had lower education and employment, and higher depression and anxiety compared to control participants. Having a diagnosis of autism spectrum disorder accounted for the greatest variance in loneliness in univariate regression analyses (B = 8.37, 95% CI [7.15, 9.59], p <0.001, R²=0.41). In multiple regression analyses, social skill deficit and dissatisfaction with social support were associated with greater loneliness in both groups (p<0.05). Reduced frequency of contact with family was associated with loneliness only in the group on the spectrum (p=0.051), whereas female gender (p=0.02) was associated with loneliness in only the control group.

Conclusions:

Adults on the autism spectrum were significantly lonelier than adults not on the spectrum. While the determinants of loneliness in both populations were similar (social skill deficits and dissatisfaction with social support), their influence on loneliness was much greater in adults on the spectrum. Adults on the spectrum may rely less on friends for social support; support from and contact with family should be encouraged. Strategies to assist adults on the spectrum to improve their social skills should be developed. Future research could further explore the phenomenon of loneliness in adults on the spectrum, including adults with co-existing intellectual disability, to fill gaps in knowledge around social anxiety, depression, conceptual understanding of loneliness, and the change in these factors over time through longitudinal analysis.