Social Networks and Cognitive Function of Older Adults in Lebanon

Abstract

Social networks are a vital component of healthy aging, influencing emotional well-being, physical health, cognitive function, and mortality. Beyond these outcomes, cognitive function represents a critical outcome in later life that determines independence, functional ability, and overall quality of life. Despite this, research remains scarce in settings characterized by weak formal institutions and persistent instability, where social networks may be especially important for health outcomes. Social networks are complex and highly heterogeneous among older adults, reflecting the diverse social changes experienced across later life. However, their measurement has largely focused on isolated network characteristics. Therefore, the multidimensional nature of social networks remains poorly understood, especially in unstable, low-income settings. Moreover, although gender differences in social networks are well reported in the literature, a very limited number of studies have looked at the gender-stratified associations between social networks and health outcomes. In this dissertation, we used data from the Lebanon study on Aging and HeAlth (LSAHA), the first large scale, population-based study of Alzheimer’s Disease and Related Dementias (ADRD) and other aging-related health challenges in the Middle East and North Africa (MENA) region, to first examine the patterns and multidimensionality of social networks of older adults in Lebanon and then assess gender-stratified association between social networks and cognitive function. We used hierarchical clustering to identify social network types in our sample. We then used gender-stratified regression models to assess the associations between social networks and cognitive outcomes. We examined two levels of association: (1) the relationships between individual social network indicators, specifically network size, diversity, and perceived emotional support, and cognitive performance across multiple domains (memory, verbal fluency, executive function, and orientation); and (2) the associations between derived social network typologies, based on these indicators, and the same cognitive domains. We identified five social network types among older adults: “Small, homogeneous, supportive”, “Large, diverse, unsupportive”, “Small, homogeneous, unsupportive”, “Large, diverse, supportive”, “Very large, diverse, supportive”. Membership in these types differed significantly by gender, region, and age groups. Social networks were associated with cognitive outcomes at both the individual and typology levels, with typologies being more predictive. The strength and the nature of the associations differed between men and women. In men, compared to the “large, diverse, supportive”, the “Small, homogeneous, supportive”, “Large, diverse, unsupportive”, and “Small, homogeneous, unsupportive” networks were associated with lower performance across different cognitive domains. In women, the “Very large, diverse, supportive” network was associated with poorer memory scores compared to the “large, diverse, supportive” network. This work highlights the heterogeneity of social networks among older adults in Lebanon. It also underscores the importance of conducting gender-stratified analyses, as the associations were found to differ between men and women. The identified social network types can be used to understand the social lives of older adults and to inform the development of locally tailored interventions that account for gender differences and aim to enhance social relationships and protect against cognitive decline.