By Karen S. Rook
An older widowed woman appreciates the many ways her son helps her with important tasks and decisions, but recently she has noticed that he calls and stops by to visit less frequently. She knows he has a demanding job, but she cannot help feeling hurt that he does not seem to want to spend time with her. An older man in declining health has noticed that his daughter has become critical and “bossy” about what he should do to prevent further health problems. He knows she means well, but he feels annoyed and somewhat demeaned by her unwelcome advice.
Social relationships can be a source of mixed blessings. On the one hand, close relationships can provide emotional support and instrumental assistance that help us adjust to stressful life events, and they afford opportunities for companionship that fosters happiness and a sense of meaning in everyday life (Qualls, 2014).
Yet social relationships can also be a source of conflict, demands, and disappointments that detract from health and well-being. Negative exchanges with social network members typically happen less often than positive exchanges, but when they do occur, they have potent effects (Rook, 1998). Negative exchanges often take place in family relationships and friendships, rather than in peripheral relationships (Sorkin and Rook, 2004), which helps explain their strong impact. Such findings have stimulated research on the nature and effects of negative social exchanges. This article provides an overview of research on the association between negative social exchanges and physical health in later life, and highlights implications for future research and collaboration on interventions designed to limit the health-related toll of negative social exchanges.
Negative Social Exchanges: Definitions and Boundaries
Negative social exchanges have been defined in a number of ways, but definitions generally refer to behaviors by others that are likely to be viewed as misdeeds or violations of relationship norms and experienced as unpleasant, unwanted, or insensitive (Brooks and Dunkel-Schetter, 2011; Rook, 1990). Examples include others’ critical or insensitive comments, selfish actions, demands, intrusions, and interference. Negative social exchanges also include being let down by others when support was expected or being excluded by others from social activities. Such omissions can cause considerable distress. In one study, women who had experienced a significant stressful life event were most vulnerable to depression when the support they had anticipated from a partner was not forthcoming (Brown and Harris, 2001).
Negative social exchanges are widely regarded as multidimensional. Rook (1998) proposed four categories of negative interaction: insensitive or unsympathetic behavior by others; unwanted or unsound advice provided by others; failure by others to provide instrumental support in times of need; and rejection or neglect by others. These categories correspond to four categories of positive exchanges that have been found to affect health and well-being: emotional support, informational support, instrumental support, and companionship. These different types of positive and negative social exchanges have been found to have distinctive effects on older adults’ health and well-being (Newsom et al., 2005). Reliable access to emotional and instrumental support appears to be especially important in helping people recover from disruptive life events and transitions, but may be less important in boosting day-to-day well-being. Experiencing companionship with others, in contrast, can kindle positive emotions and provide a respite from minor stressors and hassles, thereby contributing to daily well-being (Rook, August, and Sorkin, 2011). Interventions designed to help older adults form new social ties or strengthen existing ties could benefit from considering these different categories of social exchanges, as they may suggest different intervention targets and strategies (Heller and Rook, 1997).
The research discussed in this overview emphasizes relatively commonplace affronts and disappointments occurring in social relationships, generally excluding more extreme forms of negativity such as marital infidelity or elder abuse. The literature discussed also excludes analyses of the health effects of exposure to others’ interpersonal stress, such as an adult child’s contentious divorce or a friend’s loneliness. Worries about family members’ or friends’ difficulties may take a toll on physical or psychological health (Fingerman et al., 2012), but this article focuses on directly experienced interpersonal stress. This article emphasizes the effects of negative social exchanges on physical health, but it should be noted that strong effects on psychological health have also been documented (e.g., see Finch et al., 1999; Rook, 1998).
Evidence Linking Negative Social Exchanges to Physical Health
Aspects of older adults’ health that have been found to be affected by negative social exchanges include self-rated health, disease and disability, cognitive functioning, and mortality. Key findings that have emerged from these lines of research are discussed below. Studies that have included middle-aged adults, as well as those that focus on older adults, are included in the discussion below because many of the major chronic conditions of later life begin or accelerate in middle adulthood.
Large cross-sectional (Walen and Lachman, 2000) and longitudinal (Croezen et al., 2012; Newsom et al., 2008) studies have found that people who experience more negative social exchanges report worse self-rated health, controlling for other potential influences on health such as age, sex, race, education, income, marital status, and mental health. The longitudinal studies also control for participants’ health at the beginning of the study to rule out the possibility that poor health leads people to experience more negative social exchanges. It is important to acknowledge, however, that the association may be reciprocal for some people, with poor health leading to strained social relationships and strained social relationships leading to worsening health.
Some evidence suggests that negative social exchanges that persist or recur over time are most strongly related to poor self-rated health (Newsom et al., 2008). The possibility that chronic negative social exchanges are especially damaging to health is a theme that appears in other research, as discussed below.
Disease and disability
Negative social exchanges have been linked to worse self-rated health, but are they related to more objective aspects of health, such as disease and disability? Existing research suggests the answer is yes. Negative social exchanges have been linked in longitudinal studies to an increased risk of heart disease (De Vogli, Chandola, and Marmot, 2007), greater functional impairment and disability (Seeman and Chen, 2002), and worse recovery following medical procedures such as surgery (Stephens, Druley, and Zautra, 2002). One study examined trajectories of disability because some older adults decline rapidly, whereas others exhibit a relatively stable pattern of disability, and still others may experience improvements in functioning over time (Mavandadi, Rook, and Newsom, 2007). The key question was whether negative social exchanges played a role in downward trajectories, and here, too, the research suggests they do. Older adults who experienced more negative social exchanges were more likely to experience worsening disability and were also less likely to experience disability remission.
These studies controlled for participants’ sociodemographic characteristics, co-occurring health conditions, and initial health status, and some studies also controlled for biological risk factors (such as hypertension, elevated cholesterol) and health behaviors (such as smoking, lack of physical activity) (De Vogli, Chandola, and Marmot, 2007). The inclusion of such concontrols means that the adverse effects of negative social exchanges are unlikely to be due to lower income, poor health behaviors, or other risk factors. Instead, it seems more likely that experiencing frequent or recurring negative exchanges with members of one’s social network sets physiological processes in motion that damage health over time. Clues about these processes come from research, discussed later, that has linked negative social exchanges to physiological functioning.
Recent research suggests that negative social exchanges may contribute to declines in cognitive functioning. Middle-aged adults who experience more frequent negative social exchanges have been found to perform worse on tests of executive function, controlling for sociodemographic characteristics, chronic health conditions, functional health limitations, health behaviors, and depression (Seeman et al., 2010; Tun et al., 2013). Frequent conflict, criticism, or demands from others may make the neuroendocrine and cardiovascular systems more reactive to stress, and such increased reactivity, in turn, may cause negative changes in the brain that impair cognitive functioning (Tun et al., 2013).
In view of the evidence linking negative social exchanges to poor physical and cognitive health, we might ask whether such exchanges are related to mortality. This possibility is just beginning to be explored, and while some studies have found no association (Croezen et al., 2010), others have reported a link with mortality, controlling for sociodemographic characteristics, health conditions, biological risk factors, and initial health status.
In one large study, middle-aged Israeli men who reported more difficulties with their spouses and children experienced an increased risk of death due to stroke over a twenty-three-year period (Tanne, Goldbourt, and Medalie, 2004). In a large study of women in the United States, those who had previously been diagnosed with breast cancer and who experienced more demanding and intrusive interactions with their social network members were more likely to die (from any cause) over a fourteenyear period (Kroenke et al., 2012). In another large study of American adults, those who experienced more frequent criticism, demands, and lack of emotional support in their relationships with family members and friends were more likely to die over a nineteen-year period (Birditt and Antonucci, 2008).
Puzzlingly, the latter study also revealed that criticism and demands from network members were associated with a lower, rather than higher, risk of mortality among chronically ill individuals. A similar counterintuitive finding emerged in another study in which greater negativity in participants’ relationships with their children and friends predicted a lower risk of mortality over a thirteen-year period (Antonucci, Birditt, and Webster, 2010).
One possible explanation for these counterintuitive findings is that people may pressure chronically ill family members to improve their health behaviors; if such pressure fosters sustained improvements in health behaviors, it might contribute to greater longevity even though it is experienced as aversive (Antonucci, Birditt, and Webster, 2010). Another possibility is that disagreements, if not too abrasive, can encourage a dialogue that promotes mutual understanding. Such increased understanding, in turn, might contribute to increased relationship closeness that benefits health (Fung et al., 2009).
Whether negative social exchanges increase the risk of suicide is not yet known, although greater family conflict and loneliness predicted an increased risk of suicide among people ages 75 and older in a Swedish study (Waern, Rubenowitz, and Wilhelmson, 2003). Negative social exchanges also have been linked to an increased risk of suicidal thoughts and feelings (Lincoln et al., 2012; Mavandadi et al., 2013), controlling for sociodemographic characteristics, psychological symptoms or psychiatric disorders, and substance abuse. One study distinguished between different kinds of negative social exchanges to ascertain whether a particular type was most strongly related to suicidal ideation (Mavandadi et al., 2013). It revealed that social exclusion (being left out of social activities by others), but not lack of instrumental support or emotional support, was related to greater suicidal ideation. Other research, as well, has underscored the need to consider social exclusion (and not simply the lack of social support) for understanding how social relationships affect health and well-being (Eisenberger and Cole, 2012; Sorkin, Rook, and Lu, 2002).
How Negative Social Exchanges May Damage Health
How do negative social exchanges “get under the skin” to damage health? The answer to this question is still being investigated, but promising clues come from studies of physiological processes associated with negative social exchanges. Negative social exchanges can be thought of as a type of stressor and, like other stressors, can trigger the release of stress hormones in the bloodstream. These hormones activate or deactivate bodily systems to prepare the body to deal with the stressor. Typically, if a stressor is short-lived, the release of stress hormones slows and ordinary bodily functions resume, but when the stressors are prolonged or recurring, stress hormones remain elevated, which can damage bodily systems (Cacioppo and Berntson, 2006). For this reason, social relationships that are prone to recurring conflicts or disappointments may contribute to the dysregulation of important bodily systems, such as the cardiovascular, neuroendocrine, and immune systems (Brooks and Dunkel-Schetter, 2011; Seeman and McEwen, 1996). In one large study, middle-aged adults who reported more negative interaction exhibited worse cortisol regulation (Friedman et al., 2012). Notably, this association was strongest among participants who had experienced more frequent negative interactions over a ten-year period.
Among people who experience chronic or recurring stressors over time, the repeated demands on the body of responding to stress can lead to dysregulation across multiple physiological systems. This accumulated wear and tear on multiple bodily systems is referred to as allostatic load (McEwen and Stellar, 1993), and it underlies a broad range of chronic health conditions. Evidence derived from large epidemiological studies (Seeman et al., 2002) indicates, moreover, that frequent negative social interaction is associated with greater allostatic load in older adults.
Similar evidence has emerged from laboratory studies, offering the opportunity to observe interaction patterns in close relationships that may compromise health. In a typical study, married couples are brought into a laboratory and asked to discuss a current area of disagreement in the relationship. As these discussions unfold, researchers observe the spouses’ verbal and nonverbal behaviors, and record their physiology. Negative and hostile behaviors during such discussions have been found to be associated with increased blood pressure, release of stress hormones, and declines in immune function (Robles and Kiecolt-Glaser, 2003).
These patterns have been found in older and in middle-aged couples (Kiecolt-Glaser et al., 1997), even though older spouses generally exhibit greater warmth and less hostility toward their partners during conflict discussions (Levenson, Carstensen, and Gottman, 1994). The physiological processes triggered by abrasive marital conflicts could increase the risk of chronic disease onset or progression if the conflicts recur frequently. In one study of late middle-aged married couples, marital discord was already related to asymptomatic coronary artery disease (Smith et al., 2012).
In taking stock of the evidence linking negative social exchanges to health-damaging physiological processes and poor health outcomes, it is worth noting that these exchanges can occur in relationships that are ambivalent rather than exclusively problematic. Ambivalent social ties serve as sources of positive and negative exchanges (Fingerman, Hay, and Birditt, 2004), and such ties have been found to be related to poor healthrelated outcomes, including functional impairment (Rook et al., 2012), elevated blood pressure (Holt-Lunstad et al., 2003), and shorter telomere length (an indicator of cellular aging) (Uchino et al., 2012). Thus, it is not only “bad” relationships that may compromise health, but also those that arouse mixed feelings.
Implications for Research and Collaboration
Negative social exchanges, particularly those that persist or recur over time, have the potential to detract substantially from older adults’ health and well-being. This conclusion is supported by evidence derived from carefully controlled studies using diverse methods and samples and examining multiple facets of health. Epidemiological and survey studies generally have drawn upon large representative samples, and most of these studies have used longitudinal designs that allow for prospective analyses of health outcomes and transitions. Laboratory studies of negative interaction use smaller and less representative samples, but they offer the advantages of experimental control, observational assessments of negative behavior, and the opportunity to conduct physiological assessments during real-time social interaction. Research in this literature typically includes controls for the influence of sociodemographic characteristics, comorbid health conditions, biological risk factors, and health behaviors on participants’ health status, and the longitudinal studies routinely control for baseline health. Although more work needs to be done to resolve remaining uncertainties about causality, the striking convergence of findings across these different studies suggests that continued attention to persistent negative social exchanges is warranted in order to develop a comprehensive understanding of the role of social networks in older adults’ health and well-being.
A particularly important question that emerges from this research—and one of potential significance to researchers and practitioners—asks what makes some older adults more vulnerable than others to negative social exchanges. Studies of age differences suggest that older adults generally are less likely than younger age groups to experience negative interactions with others (Charles and Carstensen, 2010), but it is clear that some older adults experience troubling interactions, and identifying the factors that increase their likelihood of experiencing or being adversely affected by such interactions is an important challenge for any future research.
The limited research conducted to date on such factors has focused on stressful life circumstances (such as financial strains or declining health) that create substantial needs for support and overwhelm existing support providers (Krause and Shaw, 2002; Krause, Newsom, and Rook, 2008); intergenerational or cultural differences in expectations for support and companionship that kindle hurt feelings or resentments (Pillemer et al., 2007; Treas and Mazumdar, 2002); impediments to the use of avoidance as a strategy for limiting aversive interactions (such as difficulty avoiding conflict with a noxious family member or neighbor) (Charles, 2010); personal characteristics that may contribute to relationship tensions (such as neuroticism or limited social skills) (Lahey, 2009); and the lingering effects of childhood adversity that can impair social relationships in adulthood and increase physiological reactivity to conflicts (Fagundes, Glaser, and Kiecolt-Glaser, 2012). Forging a greater understanding of how such factors operate in affecting vulnerability to persistent negative social exchanges is crucial to building a strong scientific foundation for designing interventions to strengthen social ties between older adults.
This goal most likely will be achieved through collaboration between researchers and practitioners, in which practitioners are able to express their most pressing needs as they consider possible strategies and interventions to benefit older adults with whom they work. Practitioners might want to know whether a strategy of helping older adults avoid or resolve problems in key social relationships is more or less beneficial than a strategy of reducing the health-damaging stress associated with such problems. One could either foster positive interactions in alternative relationships or facilitate other forms of stress reduction. They also might want to know whether problems that occur in particular kinds of social relationships or settings pose unique challenges.
In an intensive ethnographic study of older adults living in congregate housing, Hochschild (1978) observed that the residents formed an informal pecking order based not only on health status but also on the quality of their relationships with family members. In such an environment, it might be difficult for an older person to disclose family difficulties to others. Hochschild also observed that older adults whose family relationships were not in good order seemed to have a harder time than others deriving pleasure from interactions with fellow residents. These interesting ideas, which have considerable practical importance, emerged from rich knowledge of a particular population. It is exactly such first-hand knowledge of particular client populations and their life circumstances that practitioners could bring to a collaboration with researchers in order to shape future research that would address targeted questions about designing interventions. In turn, researchers would derive new hypotheses for investigation from practitioners, highlighting the mutual benefits that would stem from greater collaboration between researchers and practitioners.
Karen S. Rook, Ph.D., is a professor of psychology and social behavior in the School of Social Ecology at the University of California, Irvine, in Irvine, California.
Editor’s Note: This article is taken from the Spring 2014 issue of ASA’s quarterly journal, Generations, an issue devoted to the topic “Relationships, Health, and Well-Being in Later Life.” ASA members receive Generations as a membership benefit; non-members may purchase subscriptions or single copies of issues at our online store. Full digital access to current and back issues of Generations is also available to ASA members and Generations subscribers at Ingenta Connect. For details, click here.