Social Connectedness for Individuals with Serious Mental Illnesses - Pre and Post COVID

Significance and Background: Social connectedness is a fundamental human need, and having social connections is positively linked to physical, mental, and cognitive health, positive health behaviors, and longer lifespan. For individuals with SMIs, who die approximately 10-30 years earlier than those without a mental illness, the costs of social isolation, smaller social networks (SNs), and low-quality support can be particularly dire. There is also a growing recognition of the effects of COVID and changes in mental health services on social connectedness for individuals with SMIs. With that in mind, the aims of the current study are: 1. To get an in-depth understanding of social connectedness for individuals with SMIs since 2025; 2. To compare social connectedness dimensions pre- and post-COVID to get a better understanding of these shifts.

Methods: Egocentric SN, social cognition assessments, and self-report survey data from individuals with SMIs (N=240, collected in 2025-2026) were analyzed and compared with SN pilot data (N=102, collected in 2017-2018) using free-recall SN mapping. Surveys included demographics, clinical history, social cognition, social self-efficacy, social motivation, and experiences of stigma, homelessness, incarceration and hospitalization.

Results: Among 234 participants, the mean SN size was 12.62 alters (SD=6.79). In a linear regression model predicting social network size, age at which mental health symptoms were first experienced (b=-.12, p=.008), being male (b=- 3.73, p<.001), social amotivation (b=-.11, p=.03), social cognition (ER-40: b=-.002, p<.001), and a history of homelessness (b=-4.19, p<.001) were all inversely associated with social network size and scores on the hinting task were positively associated with social network size (b=.31, p=.006). Age, race, social self-efficacy, history of hospitalization or incarceration, and everyday and major discrimination were all not significantly associated with social network size. In a series of linear regressions predicting the dysfunctionality of relationships with social network members only their experiences of major discrimination were positively associated with the proportion of SN members who were perceived as hypercritical b=.005, p=.03; expect too much (b=.006, p=.006), interfere in personal affairs (b=.005 , p=.016), and takes advantage of them (b=.005, p<.001). For aim 2, mean SN size for participants pre-COVID (N=102, 18.02, sd=10.02) was significantly higher than post COVID (N=234, 12.62, sd=6.79, t=4.97, p<.01). Participants in the post-covid sample also had significantly higher proportions of family members (t=-4.99, p<0.001) and providers (t=-4.91, p<0.01) in their SNs and fewer acquaintances (t=6.36, p<0.01).

Discussion and Implications: Our findings showcase a marked shift in SNs of individuals with SMIs pre and post covid with respect to both the structure and function of these networks. Within the post covid data, SN structure (size) was predicted both by individual level factors (e.g. demographics, social cognition and social motivation) as well as relational factors that indicated the experiential aspects of social connectedness.