This study examined the interrelationships between giving and receiving emotional support and depressive symptomatology in older couples, guided by an Actor-Partner Interdependence Model. This model used the dyad as the unit of analysis and allowed for modeling of the interdependence between spouses’ own reports of support and depressive symptomatology. In accord with our hypothesis, perception of receiving support partially mediated the relationship between giving emotional support and depressive symptomatology for both husbands and wives. There was no gender difference in the mediation effect. In addition, our findings reflected the interdependence between spouses’ reported experiences of giving each other emotional support, receiving emotional support, and experiencing depressive symptomatology.
The findings from this study show that the perception of receiving emotional support can be an important component of the positive effect of giving emotional support on depressive symptomatology. Husbands and wives who perceived receiving support from their spouses were less depressed. Some other studies, however, have reported that receiving emotional support can lead to negative health outcomes (Bolger et al., 2000
; Barbee, Cunningham, Winstead, Derlega, Gulley, Yankeelov, & Druen, 1993
). Bolger and colleagues (2000)
found that when support recipients were aware of receiving emotional support, there was an emotional cost or burden which could negatively affect health, but when support recipients did not perceive receiving support, it was beneficial. Hence, the researchers concluded that social support is most beneficial when it is “invisible,” i.e., when the support is unnoticed by the support recipient.
At first glance, the results of this study appear to contradict these findings. A closer examination, however, may help clarify how the emotional social support process operates differently among couples experiencing different life events. In the study conducted by Bolger and colleagues (2000)
, support recipients were experiencing a stressful event (i.e., preparation for a bar exam) which demanded a great deal of time and energy. Their support providers may have been empathetic and supportive, but recipients may have experienced guilt and a responsibility to reciprocate support, while being stretched for time and availability (Bolger et al., 2000
). In the current study, however, the cost of receiving emotional support may have been negligible, as husbands and wives may not have been experiencing circumstances in which one partner was receiving more support than the other. In fact, husbands and wives reported giving support as much as they were receiving it. Similar findings have been reported in the literature, showing that receiving support is beneficial when the support recipient gives support back to his or her partner (Liang et al., 2001
; Gleason, Iida, Bolger, & Shrout, 2003). Reciprocity is a fundamental function of close on-going relationships (Ingersoll-Dayton & Antonucci, 1988
). Our findings support this notion among older couples.
There were other methodological factors that may have contributed to the differential findings between our study and Bolger et al. (2000)
. Bolger and colleagues examined far younger couples (mean age <30), and many had been together for only about three years. In contrast, the mean age of our study participants was 82 years old, and they had been married on average 52 years. The age of the participants and their length of marriage may have affected how they perceived the emotional support given and received in their relationship, and these older couples may have developed more successful strategies to communicate their emotional support to one another. Future studies that go beyond interview and survey methodology could examine how older couples communicate and enact support, to better understand which strategies are more or less successful. Our conclusions are necessarily conditioned by the fact we have conducted a secondary analysis of survey data which was not originally collected to study the exchange of emotional support among the older couples that participated in the CLOC study.
The findings from this study may also be a reflection of changes in the social networks of older couples and a resulting increase in the interdependence between husbands and wives as they grow older. Carstensen (1992)
reported that as people age, they actively increase their emphasis on emotionally close and significant relationships. For the participants in the CLOC study, social networks may have become smaller, and partners may have drawn closer to each other (Carstensen, 1992
; Lockenhoff & Carstensen, 2004
), thereby increasing the interdependence of emotional social support between spouses. In addition, this interdependence may be a reflection of length of marriage. Perhaps these couples had remained married for an average of 52 years because they were skilled at mutually giving and receiving emotional support.
This study also found interdependence in husbands’ and wives’ reports of depressive symptomatology. This phenomenon has been widely reported in the literature, i.e., an emotional experience of a family member has been shown to affect other members (Broderick, 1993
). In fact, in close relationships, emotions have been found to be “contagious,” and individuals often unconsciously reflect the emotions that their partners experience (Hatfield, Cacioppo, & Rapson, 1994
; Tower & Kasl, 1996
). Alternatively, older couples may have been experiencing similar negative life events within and outside their marital relationships. For example, older husbands and wives may be struggling with their own and each others’ health and financial concerns (Bruce, 2002
; Glass et al., 2006
), and they may be losing friends due to death (Bruce, 2002
; Glass et al., 2006
). The interdependence between couples’ reports of giving support and experiencing depressive symptomatology highlights the importance of collecting information on both couple members and capturing interdependence with APIM. These interpretations suggest that when depressive symptomatology is experienced by one spouse in late life marriage, the partner may be at risk as well, due to either intra-couple dynamics or external factors that impinge on the couple. Future studies that retain the older couple as the unit of analysis and follow them over time, measuring both intra-couple processes and external social stresses, can help determine if the interdependence evidenced in this sample holds true. Because of the cross-sectional nature of our analyses, we cannot rule out any cohort effects that may have been present. Research indicates, however, that emotional support does not decrease with age when cohort designs are used (Due, Holstein, Lund, Modvig, & Avlund, 1999
), suggesting that cohort effects may not account for the present findings.
This study did not find gender differences in the mediation effect. Further, the perception of receiving emotional support was a mediator for both husbands and wives. Although husbands’ and wives’ experiences of giving and receiving emotional support may be different on a day-today basis, the spouses in this study may have evaluated their exchange of emotional support globally, rather than keeping an account of each time they gave and received support (Franks, Wendorf, Gonzalez, & Ketterer, 2004
). Hence, the support that husbands and wives perceived may have been a reflection of the past 50 years of giving and receiving support, including what they give and receive today, have given and received in the past, and expect to give and receive in the future (Franks et al., 2004
). Additionally, over time, older couples may have developed strategies to show more affection towards one another and avoid escalation of negative emotions (Carstensen, Gottman, & Levenson, 1995
). Husbands and wives reports of giving and receiving emotional support may be a reflection of an increase in emotional closeness in their marital relationship as their social network shrinks and they increasingly emphasize this most significant relationship (Cartensen, 1992
; Lockenhoff & Cartensen, 2004
This study has a number of important limitations. Findings should be considered in light of its cross-sectional nature. Our analyses examined only the CLOC study baseline data; replications with longitudinal data are required to account for temporality between reports of giving emotional support, receiving emotional support, and depressive symptomatology. However, we were able to demonstrate an association between giving emotional support, receiving emotional support, and depressive symptomatology, a relationship suggested in the past but not previously tested in older married couples (Acitelli & Antonucci, 1994
; Lockenhoff & Carstensen, 2004
). Additionally, an alternative model may be plausible. For instance, married individuals who are depressed may view their social world more negatively and thus perceive less support as being available from their spouses. Future studies may include test of alternative models to expand our understanding of social support and depressive symptomatology among older couples.
We chose to conduct a secondary analysis of the CLOC data set because it is unique in that a number of social, health, and behavior variables were collected on a probability sample of older couples. Due to the sample size, however, many measures were shortened versions of longer measures, possibly limiting comparisons to other studies of social support in late life. Although we compared our findings with those of Bolger et al. (2000)
, our study differed from their study in several methodological ways. As Bolger and colleagues did, future studies of older couples could consider using daily diary methods to record reports of giving and receiving support and depressive symptomatology near the time they occur, to capture the interactions on a day-to-day basis, minimize retrospective recall bias, and decrease global assessment of the variables under study.
Overall, our study findings underscore the importance of dyadic research approaches to understanding the relationship between giving and receiving support and depressive symptomatology among older couples, using a mediation model. Our findings revealed that the effect of spouses’ perception of receiving emotional support on their depressive symptoms was influenced not only by the behavior of their partners, but also by their own engagement in giving support to their partners. We also found interdependence among older couples in exchange of emotional support and in depressive symptomatology. Our findings underscore the need expressed by others for more couple-based studies to investigate dyadic interactions, particularly of older couples and late life marriages who may be studied less frequently (e.g, Acitelli & Antonucci, 1994
; Bolger et al., 2000
; Franks et al., 2004
Finally, our findings can inform future intervention research. Interventions focused on enhancing older adults’ social, emotional, and physical health might be improved by efforts focused on both members of couples (Lewis, DeVellis, & Sleath, 2002
), thereby increasing effectiveness and reach. Programs might focus on increasing the social participation of couples, or involving them in health promotion activities together.
As the study of social support among older couples continues to evolve, special attention should be paid to the exchange of other types of support, such as the effect of caregiving on depressive symptomatology for both couple members, or the exchange of tangible instrumental support, between husbands and wives. Using a dyadic framework may yield a more comprehensive understanding of who benefits from what type of support and why, as well as the potential influence of exchanging social support on mental and physical health. Our findings suggest that continued study of the support exchanged in older couples is warranted and could extend our understanding of support mechanisms as related to health and well-being.