Which Kinds of Social Networks Reduce Stress in Late Adulthood?

John Cacioppo, University of Chicago
Hsi-Yuan Chen, University of Chicago
Lea Ellwardt, University of Cologne
Louise Hawkley, University of Chicago
Rafael Wittek, University of Groningen

Background: Integration into social networks is commonly believed to benefit older adults’ wellbeing, since it reduces stress during adverse life-events and improves coping with morbidity and disability in late life. We argue, however, that stress buffering is conditional on the degree to which the relationships in a network are interconnected and balanced. Connectedness refers to presence of mutual strong ties, and balance refers to presence of mutual positive ties to third parties.

Objective: The aim is to investigate which kinds of social networks reduce stress. It is hypothesized that perceived stress is lowest for older adults embedded in highly connected and balanced networks. Previous research designs have almost exclusively focused on the relationships between an older adult and her direct contacts (ego-alter ties). These designs do not permit assessing connectedness and balance in networks. To address this shortcoming, the current study additionally includes the indirect relationships between an older adult’s contacts (alter-alter ties).

Data/Method: Unique panel data on egocentric networks in an older population stem from the population-based Chicago Health Aging and Social Relations Study (CHASRS). The sample contains five waves collected between 2002 and 2006, with 709 observations from 160 participants aged 50 to 68 years at baseline. Data include information on the participants’ direct and indirect social relationships, i.e. contact frequency and relationship quality for ego-alter ties and alter-alter ties, as well as participants’ perceived stress. Fixed-effects models test whether changes in a person’s network connectedness and balance are associated with changes in the same person’s perceived stress.

Results: Older adults reporting the highest number of balanced relationships (i.e., positive ties among alters) experience least stress. This effect holds independently of socio-demographic confounders, loneliness and network size. There is no stress-reducing effect for network connectedness, suggesting that balance matters more.

Presented in Poster Session 1