Marital Effects in Later Life: Dyadic Approaches and Gender Differences

Abstract Marriage is a dyadic system, within which the characteristics and experiences of each partner can have implications for both. Moreover, gender of both spouses may impact these dyadic influences. The five papers comprising this symposium all take a dyadic approach to studying midlife and older couples, and how their effects on one another may vary by gender. Donnelly examines the consequences of precarious work among midlife couples, finding heightened risks for marital strain and divorce, depending on which gender spouse is exposed to precarious work. Garcia also analyzes gender differences – in this case, how the gender of a woman’s spouse may affect associations between daily marital strain and sleep quality, with only women married to men showing adverse sleep outcomes. Polenick and colleagues study the long-term repercussions of chronic condition discordance, finding that both individual-level and couple-level discordance had impacts for husbands’ and wives’ physical activity. Gallagher and Stokes focus on cognitive functioning within dyads, revealing gendered effects: Wives’ poorer cognitive functioning was associated with their own (better) marital quality, while husbands’ poorer cognitive functioning was associated with wives’ (worse) marital quality. Lastly, Stokes and Barooah examine longitudinal dyadic associations between loneliness and vascular health, finding that own and partner’s baseline loneliness were associated with increased HbA1c levels only in the context of inferior marital support. Carr will assess the strengths and limitations of these papers, and discuss the contributions these studies can make to the field and to future research on marital effects and gender in later life.

Social networks are critical for end-of-life decision-making and hands on support and may also impact end-of-life outcomes including location of death. Yet we fail to consider these factors in the context of cultural values and variation in healthcare system financing and resources, e.g., availability of palliative care. Using SHARE, we examined the following factors for 6,391 decedents across 13 countries interviewed in the last year of life: family size, living alone, and size and characteristics of social networks. We compared these characteristics cross-nationally for persons dying at home as opposed to in hospital and other formal settings. While individuals with larger social networks are more likely to die at home, we find a cross-national gradient of higher unmet healthcare needs resulting in overall more homebased deaths. Our findings suggest that individual factors such as family availability must be considered in the context of country-level factors when examining quality end-of-life indicators.

KINLESSNESS, LONELINESS, AND END OF LIFE: A CROSS-NATIONAL COMPARISON OF 20 COUNTRIES
Esteban Calvo, 1 Christine Mair, 2 Katherine Ornstein, 3 Rosario Donoso, 1 and José Medina, 1 1. Universidad Mayor,Santiago,Region Metropolitana,Chile,2. University of Maryland,Baltimore County,Baltimore,Maryland,United States,3. Icahn School of Medicine at Mount Sinai,New York,New York,United States Countries across the globe are experiencing declining rates of fertility and marriage, which present a distinct challenge for older adults' social integration, well-being, and end-of-life care. However, older adults who are "alone" (e.g., no partner, no child) may not be lonely, and end-of-life risks faced by "kinless" older adults likely vary significantly by country context. Using harmonized, cross-national data from 20 countries (United States (HRS), England (ELSA), and European Union (SHARE)), we examine associations between family structure, loneliness, and end-of-life outcomes. Although "kinless" family structures are associated with greater loneliness in the pooled sample, the percent of "kinless" who report no signs of loneliness ranges from 7% (Greece) to 56% (Denmark). Family structure is associated with various end-of-life outcomes, and these associations vary by country-likely reflecting differences in healthcare structure. We discuss distinctions between "being alone," "being lonely," and "being without care" in light of cross-national variation.

MARITAL EFFECTS IN LATER LIFE: DYADIC APPROACHES AND GENDER DIFFERENCES Chair: Jeffrey Stokes Discussant: Deborah Carr
Marriage is a dyadic system, within which the characteristics and experiences of each partner can have implications for both. Moreover, gender of both spouses may impact these dyadic influences. The five papers comprising this symposium all take a dyadic approach to studying midlife and older couples, and how their effects on one another may vary by gender. Donnelly examines the consequences of precarious work among midlife couples, finding heightened risks for marital strain and divorce, depending on which gender spouse is exposed to precarious work. Garcia also analyzes gender differences -in this case, how the gender of a woman's spouse may affect associations between daily marital strain and sleep quality, with only women married to men showing adverse sleep outcomes. Polenick and colleagues study the long-term repercussions of chronic condition discordance, finding that both individual-level and couple-level discordance had impacts for husbands' and wives' physical activity. Gallagher and Stokes focus on cognitive functioning within dyads, revealing gendered effects: Wives' poorer cognitive functioning was associated with their own (better) marital quality, while husbands' poorer cognitive functioning was associated with wives' (worse) marital quality. Lastly, Stokes and Barooah examine longitudinal dyadic associations between loneliness and vascular health, finding that own and partner's baseline loneliness were associated with increased HbA1c levels only in the context of inferior marital support. Carr will assess the strengths and limitations of these papers, and discuss the contributions these studies can make to the field and to future research on marital effects and gender in later life. Loneliness is a contributor to later life declines in health, including vascular health. Importantly, loneliness is not restricted to those who lack close social ties: More than onethird of married U.S. older adults experience loneliness, and having a lonely spouse increases the likelihood of experiencing loneliness oneself. Thus, over time loneliness in either spouse may lead to worse health for both spouses. Using longitudinal dyadic data from the Health and Retirement Study (2008-2014), we estimated multilevel lagged dependent variable models to examine implications of both partners' loneliness at baseline for each spouse's HbA1c four years later. Findings revealed that effects of both partners' loneliness were contingent upon marital quality: Own and partner's loneliness led to increases in HbA1c when perceived marital support was low, but this was attenuated at higher levels of marital support. These results extend prior research concerning loneliness and vascular health, and loneliness as a relational experience.

COGNITIVE FUNCTIONING, GENDER, AND MARITAL QUALITY AMONG OLDER MARRIED COUPLES: A DYADIC APPROACH Elizabeth Gallagher, 1 and Jeffrey Stokes, 2 1. University of Massachusetts Boston, Kingston, Massachusetts, United States, 2. University of Massachusetts Boston, Boston, Massachusetts, United States
Older spouses influence one another in myriad ways, and dyadic effects of marital quality on health and well-being have been well-established. However, little attention has been paid to dyadic implications of cognitive functioning, including for spouses' perceptions of the relationship itself. This study examines associations of older husbands' and wives' cognitive functioning with both partners' reports of four marital quality outcomes. Structural equation modeling analyzed data from 1,414 opposite-sex couples drawn from the 2016 wave of the Health and Retirement Study. Findings revealed that (a) wives' poorer cognitive functioning was associated with wives' reporting greater closeness and higher ratings of enjoying time with a spouse, whereas (b) husbands' poorer cognitive functioning was associated with wives' reporting greater marital strain, lower marital support, lower closeness, and lower ratings of enjoying time with a spouse. This suggests that cognitive functioning/impairment has dyadic consequences for marital quality, which are highly gendered.

PRECARIOUS WORK, MARITAL QUALITY, AND DIVORCE: A GENDERED DYADIC ANALYSIS OF AGING COUPLES
Rachel Donnelly, Vanderbilt University, Nashville, Tennessee, United States Precarious work -work that is unstable and insecure -is often stressful and may contribute to marital strain and dissolution among midlife adults. However, prior research has not considered how precarious work spills over to spouses. Using longitudinal dyadic data of midlife couples from the Health and Retirement Study, I examine whether having a spouse in precarious work is associated with marital strain and dissolution, with attention to differences by gender. I find that indicators of precarious work (job insecurity, schedule variability) are associated with a heightened risk of marital strain and divorce in midlife. These patterns depend on the gender of the spouse experiencing precarious work. Understanding the implications of precarious work for marriage is important because poor marital quality and divorce hasten health declines at older ages. Thus, this study suggests that precarious work may be a risk factor for divorce and poor health among aging adults.

CHRONIC CONDITION DISCORDANCE AND PHYSICAL ACTIVITY: LONGITUDINAL ASSOCIATIONS AMONG OLDER COUPLES
Courtney Polenick, 1 Kira Birditt, 1 Angela Turkelson, 1 and Helen Kales, 2 1. University of Michigan,Ann Arbor,Michigan,United States,2. UC Davis,Sacramento,California,United States Chronic condition discordance (i.e., the extent that two or more conditions have non-overlapping self-management requirements) is detrimental for functional health but little is known about mechanisms accounting for these associations. We examined links between chronic condition discordance at both the individual level and the couple level (i.e., between spouses) and physical activity over time. Participants included 1,095 couples from five waves (2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014) of the Health and Retirement Study. Dyadic growth curve models showed that greater individual-level discordance was associated with lower baseline physical activity among individuals and their partners. When husbands had greater individuallevel discordance, wives and husbands had faster declines in physical activity. The findings highlight the importance of considering both members of a couple when examining the implications of chronic illness for physical activity in middle and later life.

DAILY MARITAL STRAIN AND SLEEP IN SAME-SEX AND DIFFERENT-SEX COUPLES Michael Garcia, University of Texas at Austin, Austin, Texas, United States
Marital strain has consistently been linked to many indicators of daily health and well-being, including sleep. Prior studies show that, on days when marital strain is higher, women in different-sex couples experience poorer sleep outcomes. However, this work has not yet considered whether and how these relationships differ for men and women in same-sex couples. Using 10 days of dyadic diary data from 756 midlife U.S. men and women in 378 gay, lesbian, and heterosexual marriages, we examine the associations of daily marital strain with sleep quality and duration and consider whether these relationships differ across union type. Results suggest that increased marital strain is associated with poorer sleep quality and shorter sleep duration, but only for women married to men. These findings underscore the importance