Social Networks and Inequalities in Health (EUSN)

Social Networks and Inequalities in Health (OS_2)
Room: P205 (02 445P205)
Topic: Social Networks and Inequalities in Health
Form of presentation: Oral presentation
Duration: 105 Minutes

03:35 pm
A network approach to integrate theories of cognition, society, and inequality

Mark McCann | University of Glasgow | United Kingdom

This presentation applies four theoretical perspectives to the emergence of inequalities during the transition from adolescence to emerging adulthood, and will discuss how network analysis could provide opportunities to study the development of inequality.
Lin’s (1999) overview of the social capital model outlines how initial (e.g. parental) status provides access to individual resources such as education, and – via greater extensity of ties – access to network resources. Secondly, mobilisation of social capital – mediated by the strength of ties and the status of alters – determines ego’s attained status.
Two elements of Tilly’s theory of durable inequalities are likely to operate through social network structures: opportunity hoarding, whereby groups of (often subordinate) individuals form an in-group that becomes overrepresented in relation to a resource or social position; and adaptation, formation of social structures and patterns which act to exclude subordinate groups. Operationalising these elements in a network analytic framework could help understand these processes.
Furlong et al’s theory of rationalised individualisation conceptualises successful youth transitions as part of a dynamic process, where an individual’s pattern of secured or frustrated outcomes iteratively determines how, at successive transitions, they rationalise actions towards or against behaviours which may influence outcomes.
A fourth perspective based on Galesic et al’s empirical social cognition research outlines how bias phenomena – e.g. self-enhancement, false consensus, ideological bias – can be explained by a social sampling model; the accurate estimation of alters’ traits, in the presence of homophily for that trait, leads to inaccurate estimations of the population distribution and one’s own position on the distribution.
These ideas allow us to conceptualise an adolescent developmental perspective on inequality. Initial status may confer differential academic orientation for young people based on whether education secured or frustrated their parents’ employment outcomes, health risk behaviours may be inherited from parents, and may also serve as a signal to their peers of school disengagement. Within schools, patterns of friendship may form around academic orientation: adaptation processes may act to exclude low-orientation individuals from the positive norms of high orientation cliques; while opportunity hoarding could lead low-orientation students to coalesce into groups and take up risk behaviour. Symmetric homophily would suggest both processes are at play, while differential formation of social ties at high versus low levels of orientation (asymmetric homophily) would suggest one process dominates the other. In the presence of homophily, social sampling may explain ‘biased’ rationalisations of the success or failure of high versus low academic orientation, amplifying such rationalisations and further embedding in-group norms over time.
Understanding the transition from relative equality of health in youth, to divergent risk behaviour in adolescence and inequality in life course outcomes is a key societal concern. This paper will explore how social network methods can be applied: firstly, to develop hypotheses derived from integrated theories; second, to test these against network data; and ultimately, to develop robust, theory-informed interventions.

03:55 pm
Social networks of older people in the UK: are there positives as well as negatives?

David Barron | University of Oxford | United Kingdom
Elizabeth West | University of Greenwich

There is clear evidence that social isolation among older people in the UK is a significant problem. It is problematic both because loneliness is an unpleasant and undesirable state to be in, and also because it is associated with increased risk of mortality and morbidity. However, recent research in the US has suggested that there are positive as well as negative impacts on social networks as people age. This is rooted in activity theory (Lemon, Bengtson and Peterson 1972) and continuity theory (Thoits 1992), which portray older people as resilient in the face of life events that might be thought to increase social isolation, such as retirement and bereavement.

Cornwell, Laumann and Schumm (2008) recently investigated the characteristics of people aged 57 to 85 in the United States. They found that age is indeed negatively related to network size and closeness to network members. On the other hand, they also found that age is positively associated with frequency of contact with neighbours and engagement with community activities such as volunteering. Using data from the English Longitudinal Study of Ageing, the British Household Panel Survey, and Understanding Society datasets, we build on Cornwell et al.’s work by investigating how a number of indicators of social networks vary with age in the UK. We also make use of the longitudinal nature of these three data sources to investigate how they are impacted by life events such as retirement and bereavement. We aim to show how changes in the social networks of older people can contribute to successful or positive ageing.

04:15 pm
Body-mass index and social networks among adolescents: Evidence from four European countries

Thomas Grund | University College Dublin | Ireland
Travis Tatum | University College Dublin | Ireland

Obesity is related with various illnesses and remains a major threat to public health. Previous research stresses the importance of social networks for health outcomes. We draw on network data (N = 18133) from the Children for Immigrants Longitudinal Study in four European Countries (CILS4EU) study for adolescents (mean age = 14.99, std = 0.81) and investigate the relationship between individuals’ body-mass index (BMI) and the BMI of their friends. Our study reveals strong evidence for BMI clustering in England, Germany, Netherlands and Sweden; adolescents tend to be friends with others who have a similar BMI. We also find evidence that BMI clustering is more pronounced for strong relationships and between adolescents of the same biological sex. We provide new insights into BMI clustering among adolescents in Europe and argue that targeted health interventions for weight loss among adolescents should take social networks into account.

04:35 pm
Network, School and country variations in Adolescents’ Health Behaviour: A Multilevel and Multiple Classification Social Network Analysis of Binary Response Variables in six European cities.
Vincent Lorant | Université catholique de Louvain | Belgium

Vincent Lorant | Université catholique de Louvain | Belgium
Mark Tranmer | University of Glasgow | United Kingdom


Adolescents’ health behaviours result from their social connections to peers having similar health behaviour. Better knowledge of this peer similarity of health behaviours in school-aged adolescents will be useful in terms of network interventions. Yet, it is unknown whether this similarity applies more to some behaviour and whether this similarity is vulnerable to compositional and situational factors. This paper describes the magnitude of health behaviour similarity and the role of compositional and situational factors on it.

A social network survey (SILNE) was carried out in 50 schools from six medium-size comparable European cities (Namur-Belgium, Tampere-Finland, Hanover-Germany, Latina-Italy, Amersfoort-Netherlands, and Coimbra-Portugal). Adolescents in the 9th and 10th grades were recruited (n=11.015, participation rate = 79.4%). A method of estimating the extent of peer and school variations in health behaviour is described and applied, through the use of the Multiple Membership Multiple Classification (MMMC) Model. It is possible to assess the extent of similarity of a particular health behaviour within schools and within networks.

Friendship network similarity was an important variance component for the three risky behaviours, smoking, drinking, and cannabis use; it was absent for the only positive behaviour, physical activity. The school bears very little on these behaviours, even before controlling for the school composition. Country is a key variance component for cannabis use but is not an important component for smoking. Network similarity was not very much affect by compositional or positional factors, with the exception of cannabis use. Cannabis friendship similarity relies heavily on popular individuals. Finally, we found that school are generally doing well or poorly in the same way for smoking, drinking and cannabis.

In conclusion, smoking and drinking are best addressed at the network level, cannabis consumption is better addressed both at the network and country-level whereas physical activity depends more on the country-level. Our approach provides an tool to identify vulnerable schools.

04:55 pm
Social spaces and social networks of homosexual men in Zurich and Lausanne. An explorative inquiry into their function and significance with respect to quality of life and health

Riccarda Neff | University of Applied Sciences and Arts Northwestern Switzerland | Switzerland
Nadine Käser
Carlo Fabian
Tobias Drilling

Research in the area of social network support and health outcomes shows that social spaces and social net-works are important areas to promote health and quality of life. There is hardly any empirical evidence in this area for homosexual men in Switzerland. Simultaneously, findings of recent studies refer to the fact that the objectively measurable as well as the subjectively experienced health and quality of life of homosexual men in Switzerland seem to be poorer compared to the whole population. With the planned research project, the lack of evidence-based knowledge concerning the situation of homosexual men in Switzerland shall be reduced. The research project aims to contribute to the discussion about quality of life and health of homosexual men more broadly within the fields of health promotion, politics, professional work and the scientific discourse by discovering, describing and understanding the meaning of social spaces, social networks and social support of homosexual men within this context.

This knowledge will be explored for the cities Zurich and Lausanne with an exploratory mixed method de-sign. The sampling will focus on men, identifying themselves as homosexual and living in an urban context in Switzerland. There will be a two-step research procedure. The first step consists of a qualitative exploration of the research field, based on semi-structured interviews including egocentric network analysis. Building on these findings, the second step consists of the quantitative exploration, by means of a broadly supported online survey. Methodically the study is oriented towards an interpretative paradigm and is characterized by an exploratory and participative research approach.

The egocentric network is generated in a participative manner as part of the qualitative, semi-structured interview in the form of a network map. We use the method of the concentric circles according to Kahn/Antonucci (1980), in which the name of the interviewee is in the innermost circle, and the interviewee is asked to identify persons with whom he feels emotionally connected with descending emotional intensity in the outer radii. The instrument is consciously kept open so that respondents can develop their relevance and meaning as unaffected as possible. In this use, the map primarily is a means of gathering information, acts as a narrative stimulus, cognitive support and is also the goal of the survey itself.

The conference presentation provides an overview of the methodological approach and first outcomes of the qualitative exploration focusing on the egocentric network analysis as part of the semi-structured interviews.