Umeå universitets logga

Background: This thesis takes as a predicate that humans are social animals and as such, that their social networks and health are intrinsically connected. A guiding heuristic of this thesis was the Berkman model which conceptualizes how social networks impact health. Considering social networks as a ‘mesosocial’ level between individuals’ health and larger social structures permits assessments of four processes, or pathways: macro↔meso↔micro. The overall aim of this thesis was to contribute to a better understanding of the role of social networks for the health of aging populations in the Southeast Asian context of Indonesia. This work was divided into four sub-studies (I-IV) aiming to i) review existing evidence on social networks and adult health; ii) assess the health effects of social networks; iii) examine how macro-level conditions affect health behavioral pathways through social networks; and iv) understand the macro-level determinants of social network formations.

Methods: This thesis employed a mixed methods design integrating evidence syntheses (I) with quantitative (II, III) and qualitative (IV) approaches. The four sub-studies included different conceptual angles on structural, functional and quality aspects of social networks, i.e., network diversity (II), dynamics (III), deficits and loneliness (IV). Following the PRISMA guidelines, a systematic literature review covering the years 2000 to 2015 was conducted. For the two quantitative studies, data on older adults aged 50+ years from the fourth (2007/08) and fifth (2014/15) waves of the Indonesian Family Life Surveys, were analyzed. First, employing an outcome-wide epidemiological approach, evidence for causal effects of social network diversity on 19 outcomes along the disablement process spectrum were evaluated using multivariable regression adjustments, propensity score matching, and instrumental variable analyses. Second, following a systems-approach to conceptualize social networks on four ecological levels, causal mediation analyses were performed to cross-sectionally examine the mediating role of social networks and their dynamics between a latent socioeconomic position indicator and two healthcare utilization outcomes. Also, the socioeconomic patterning of healthcare utilization and network dynamics were analyzed longitudinally, i.e., before and after Indonesia’s Universal Health Coverage implementation. The qualitative study draws on eight focus group discussions with 48 ‘older adult children’. Data were collected in four rural villages in the Gunung Kidul regency during 2016. The theoretical framework was inspired by symbolic interactionism and sensitizing concepts of systemic networks and network deficits, interpersonal emotions, and the social construction of risks. Data were analyzed using Grounded Theory. The final mixed-methods synthesis (I-IV) was based on notions of systems-thinking, ecological transitions theory, and a framework for studying social processes in aging.

Results: The results showed (i) that despite an increased focus on social determinants of health, in Indonesia, social networks remained an under- and unresearched social determinant of non-communicable diseases and inequalities therein, respectively; (ii) that social network diversity conferred a vast array of strong and long-term – yet heterogeneous and gender-specific – effects across the entire disablement process, i.e. affecting pulmonary health outcomes, proinflammatory processes, physical and cognitive functioning, and disability states; (iii) that both social networks on various ecological levels and their dynamics were socio-economically patterned and mediated the association between socioeconomic position and healthcare utilization outcomes while showing stronger mediating effects after Universal Health Coverage implementation; (iv) that ‘older adult children’ experienced their own aging as a process of ‘bargaining for a sense of security’ which reified how and why loneliness emerged amidst the challenges of social, demographic and epidemiological transitions and how in response to this, unconventional compromises, which affected both their networks of caretakers and the places of old-age care, were being made.

Conclusion: The results explicated several previously less understood pathways and particularly spoke to the so-called ‘downstream experiences’, ‘upstream questions’, and offered a holistic perspective. Notions of ‘transitions’ within different levels collectively conveyed an understanding that such transitions created risks, and that networks played a crucial interstitial role in both cushioning or exacerbating effects of and on these risks. The final mixed-methods synthesis allowed a comprehensive view on various dimensions of social networks and enabled a multi-contextual and multi-dimensional integration of findings. Collective insights were distilled into two main discussion points which were divided based on their ‘upstream’ and ‘downstream‘ foci. The first explicated the macro-social phenomena that appeared in this thesis i.e., epidemiological transitions, population aging, social change, and discussed – with a focus on network diversity, dynamics and deficits – the meso-social effects on and of such macro-level transitions. The later elucidated both the ‘social network-health’ nexus by explicating the ‘Janus face’ of social networks, as well as the transitions in the interplay of formal and informal social networks and their implications for old-age care. After presenting methodological considerations, the thesis concludes with selected implications for research, policy, and practice which emphasize on strengthening social network research in low and middle-income countries, addressing the precarious conditions of older adult women in rural economies, and promoting the role of the healthcare system.