Chair: Rikke Lund

Poor social relations and loneliness are associated with adverse health outcomes among middle-aged and older people across gender and socioeconomic groups. Less is known about how aspects of social relations including the negative aspects are associated with gender and social class in contemporary middle-aged and older Nordic populations. The objective of this symposium is to provide knowledge about this in several Nordic countries as well as on the association with all-cause mortality based on large population based studies. Overall, we conclude that there is evidence of gender and social differences in poor social relations in Nordic countries.

S3.3.1 Gender and social class differences in social relations in middle-aged Danes

Rikke Lund

University of Copenhagen, Denmark

Background: Little is known about social class and gender differences in social relations among the middle-aged especially regarding the negative aspects of social relations. Methods: In Copenhagen Aging and Midlife Biobank (N~7.000) we investigated the association of gender and occupational social class combined on social relations. Results: Compared to women in high social class, women in low social class had poorer scores on all aspects of social relations (i.e. living alone, less contact frequency, less emotional support, and more worries and conflicts with close social relations). Compared to women in high social class, men in high social class significantly more often reported low emotional support from family and friends, but less often to have low support from partner and less often to experience serious worries and conflicts with children, family and friends. Compared to women in high social class, men in low social class were significantly more often living alone, having less contact with family and friends, more often to have low levels of emotional support from family and friends, and more frequent to have worries and conflicts with partner. In contrast, they were less often worried about children family and friends. Conclusion: Both gender and social differences in social relations are evident in this middle-aged cohort.

S3.3.2 Loneliness, social isolation, and all-cause mortality in older Swedish women and

men

Carin Lennartsson

Aging Research Centre, Karolinska Institutet, Stockholm, Sweden

Background: Socially isolated and lonely individuals have higher mortality risks, and the effects are comparable with well-established risk factors such as smoking and sedentary lifestyle. Social isolation and loneliness are less often analysed concurrently, which means that there is a lack of direct comparisons. Existing studies show mixed results and seldom have a gender perspective. We examine the relative effects of social isolation and loneliness as well as their combined effects on short- and long-term mortality in a representative national sample of older women and men. Finally, we examine if the mortality risk of loneliness and social isolation is linear, or if there is a threshold at which mortality risk becomes pronounced.  Methods: Baseline data derive from Swedish Panel Study of Living Conditions of the Oldest Old (SWEOLD), a nationally representative sample of persons aged 70 years and older living in Sweden in 2004, (n=1 180). Mortality information (date of occurrence) was obtained from the Swedish National Cause of Deaths Registry. We use Cox-regression to estimate the relative effects of social isolation and loneliness on two, four and 10-year mortality. Results: Our results show the importance of social isolation and/or loneliness for short- and long- term mortality for women and men and the findings will be discussed from a gender perspective.

S3.3.3 The social pattern of conflicting and low quality social relations

Marja Aartsen

NOVA, OsloMet – Oslo Metropolitan University, Norway

Background: Recent research found that people with high socio-economic positions (SEP) had larger networks with higher quality relationships compared to people with low SEP, which partly explains the well-known inequalities in health. Whether also low quality of relations or even conflicting relations differ across gender and SEP is less well known. Method: A logistic regression analysis of the quality of social relations (conflicts with partner, conflict with family, low quality with neighbors) on four exposure groups (high and low educated female, and high and ow educated men), controlling for age, was conducted in 834 women and 755 men living with a partner. All participated in the second wave of the Norwegian Life-course, Ageing and Generation (NorLAG) study. The age ranged between 67 and 84 years (Median age= 72). Results: Compared to higher educated women, higher and lower educated men had two to three times more often a conflicting relation with the partner, and more often low quality relations with neighbors. Lower educated women had more often low quality relations with neighbors than the higher educated women. Conclusion: Conflicting relations in the social network are patterned according to SEP and gender which may partly explain inequalities in health.

S3.3.4 Are old-age gender-differences in social relations modified by education in the

Nordic countries?

Charlotte Juul Nilsson

University of Copenhagen, Denmark

Background: Variations across countries in the social relations of older people have been suggested in older studies, also within the Nordic countries, The aim of this study was to describe differences by educational level in the association between gender and social relations in older age in three Nordic countries. Methods: Men and women aged 67-84 from the 2012/13-survey from the Danish Longitudinal Study of Aging (N=4289), the 2007/08-survey from the Norwegian Study on Life Course, Ageing and Generation (N=1937), and the 2014-survey from the Swedish Panel Study of Living Conditions of the Oldest Old (N=862) were included. In country-specific logistic regression the association between gender and education combined and social relations were studied. Results: Women with more than basic education were the reference group. In all countries, living alone and feeling lonely were less common among men, especially with high education. Low contact with children was less common among Danish and Swedish women with basic education, and more common for highly educated Danish men. Danish men and women and Norwegian women with basic education were more likely to lack emotional support. Conclusions: With some variation across countries, we found gender differences in social relations, which only appeared to be slightly moderated by education.

S3.3.5 Gender and educational differences in social relations among 60 to 69 years old

Finnish men and women

Johanna Surakka

University of Tampere, Finland

Background: People in their 60’s are between work and retirement and intergenerational relationships are important part of their life. We investigated 60 to 69 year old adult’s social relations by gender and educational class. Methods: Randomly picked, 60 to 69 year old Tampere residents answered a survey through mail or internet in 2017 (N=1495). Response rate was 70%, and 40% were male. Results: Almost all participants had children and/or grandchildren and friends. Taking part in children’s and grandchildren’s lives by supporting them economically was more usual among highly educated participants than among the less educated. Women (regardless of education class) spent more time with their children or grandchildren than men and highly educated men spent more time with children than lower educated men. Lower educated women discussed personal matters more often with their children than higher educated women. There was no such difference among men.  Women shared their life with friends more often than men did and the proportion of men who had no one to tell personal matters (16%) was higher than among women (4%). Both women and men with higher education discuss personal matters with friends more often than those with lower education. Conclusions. Both education and gender make a difference in the social relations of older adults.