Pathways to loneliness: a mediation analysis investigating the social gradient of loneliness in persons with disabilities in Switzerland

Background The experience of loneliness can have drastic consequences for health and quality of life. Given that loneliness is highly prevalent in persons with physical disabilities and that loneliness more profoundly affects persons of low socioeconomic status, more evidence is required in order to understand the mechanisms determining loneliness in this population. The objective of this study is therefore to investigate the potential pathways through which socioeconomic status influences loneliness in persons with spinal cord injury. Methods Mediation analysis utilising structural equation models and bias corrected and accelerated confidence intervals were used in order to test the mediation effects of health status, functioning, participation, social support and self-efficacy on the association between socioeconomic status and loneliness in persons with spinal cord injury. A latent construct was created for socioeconomic status with the indicators education, household income, financial hardship, subjective social status and engagement in paid work. Results This study found evidence to support the mediating role of psychosocial resources and of secondary health conditions in the association between socioeconomic status and loneliness. The study demonstrated robust associations between socioeconomic status and all potential mediators, whereby higher socioeconomic status was associated with better health, participation and psychosocial resources, however, not all potential mediators were associated with loneliness. The serial mediation model explained the interplay between socioeconomic status, mediators on different levels, and loneliness. For example, emotional support and self-efficacy were both positively associated with fewer restrictions to participation (0.08 (CI: 0.05, 0.12); 0.29 (CI: 0.24, 0.36) respectively), and fewer restrictions to participation were found to be a result of improved functional independence and fewer secondary health conditions (0.23 (CI: 0.15, 0.39); − 0.29 (CI: − 0.36, − 0.20) respectively). Conclusions Our findings highlight the vulnerability of persons with low socioeconomic status to loneliness in persons with spinal cord injury and identified potential mediating factors, such as health, functioning, participation and psychosocial resources, in the association between socioeconomic status and loneliness. This population-based evidence suggests potential targets of interventions on the pathway to loneliness, through which socioeconomic status influences loneliness. The complexity of the model shows the need for comprehensive interprofessional rehabilitation to identify and support people with lower socioeconomic status and concomitant risk factors for loneliness.