Processed food intake assortativity in the personal networks of older adults

Abstract

Coalescing evidence shows that food habits circulate through social relationships. However, seniors' social networks and unhealthy dietary choices remain underexplored, particularly within rural Eastern Europe's unique socioeconomic and cultural contexts. We investigate how social networks affect the consumption of processed foods high in salt among community-dwelling older adults (64 years of age and older) from a Romanian rural community. We use an observational cross-sectional quantitative approach (a personal network research design) to collect self-reported information about the study participants (egos; n = 83), their social contacts (alters), and the relationships among them (ego-alter and alter-alter ties). We fit multi-level, mixed-effects models, clustering alters by their egos, and show that older adults eating processed food high in salt are more likely to be connected to alters with similar eating habits (assortativity). This result is of particular relevance when considering the nutritional health of older adults, as it implies that interventions targeting this demographic should extend beyond the individual to encompass their social environment. Despite the cross-sectional approach limitations, our work may enrich the understanding of dietary patterns among rural Eastern European seniors and provide an empirical foundation for designing targeted interventions that can mitigate health risks associated with high-salt diets in older populations.

Online media use and COVID-19 vaccination in real-world personal networks

“…Specifically, alters who had a higher proportion of vaccinated direct network neighbors, relative to the vaccination rate of the entire network, were more likely to be vaccinated themselves. This finding aligns with other research employing PNA to examine assortative mixing concerning attitudes towards COVID-19 vaccination [31] and processed food consumption [62]. Such alignment underlines the potential for further exploratory and comparative studies.…”

“…Given that an alter’s vaccination status is measured via a binary variable (1=vaccinated; 0=unvaccinated), a positive correlation coefficient indicates assortative mixing (i.e., the alter’s vaccination status is related to the vaccination status of its direct neighbors in the network). The computation method was previously advanced to detect assortative mixing of opinions related to COVID- 19 vaccination [31] and processed food intake [62].…”