A developmental perspective on marital closeness and health in older adulthood

Abstract

Although closeness is a necessary element in relationships, its role in later-life marriages is less clear, with some finding benefits of greater closeness and others finding costs. To begin to reconcile these findings, the current study sought to explore older couples’ experiences of both desired and actual closeness over time. Drawing upon both socioemotional selectivity theory and the dynamic goal theory of marital satisfaction, the current study examined the stability of marital closeness among older couples and the role of health in shaping both within-spouse and cross-spouse associations. We utilized a series of cross-lagged, mutual influence, actor–partner interdependence models to examine how higher-functioning older couples’ (N_T1 = 64, N_T2 = 55) actual and desired marital closeness were linked over a year and whether subjective and objective health moderated these pathways. Findings revealed that older spouses reported enjoying and desiring close relationships within their marriages, but these feelings of and desire for closeness were more stable for husbands. Moderational analyses, however, indicated that the desire for and experience of closeness were sensitive to spouses’ health. These findings suggest that both gender and health may be important considerations when understanding the stability of and cross-partner associations in marital closeness for older couples.

Keywords

Health marital closeness marriage older adults socioemotional selectivity theory

Establishing close connections is considered paramount for individual and relational well-being (Ben-Ari & Lavee, 2007; Erikson, 1959). Dating back to the 1950s, Erikson posited that forming close, romantic relationships is essential to the future well-being of young adults—those unsuccessful in doing so were thought to be disadvantaged moving into adulthood. Although scholars have found robust empirical support for the importance and benefits of closeness in early adult romantic relationships (Conger, Cui, Bryant, & Elder, 2000; Rauer, Pettit, Lansford, Dodge, & Bates, 2013), the role of closeness in later-life relationships is decidedly more mixed.

On the one hand, some find that closeness is strongly protective of aging adults’ physical and mental health, even their mortality risk (Mancini & Bonanno, 2006; Tower, Kasl, & Darefsky, 2002). However, Polenick, Martire, Hemphill, and Stephens (2015) uncovered potentially harmful outcomes of closeness among spouses dealing with health challenges, with greater closeness exacerbating the effects of an older adult’s illness severity on their caregiving spouses’ depressive symptoms.

Although these findings suggest that marital closeness and its links to health may change, particularly in older adulthood, it is unclear how stable marital closeness is when experienced within the context of poorer health. Theories suggest that marital closeness is prioritized in older adulthood (Carstensen, 1996; Li & Fung, 2011), but researchers have tended to focus on actual closeness while neglecting couples’ desire for closeness, perhaps working under the assumption that most couples want to maximize closeness later in life. However, there can be different levels of closeness desired in relationships, and both actual and desired closeness can be uniquely important for relationship quality (Mashek & Sherman, 2004). Furthermore, although couples’ desires and experiences of closeness are likely interdependent, this may potentially change in the context of poor health. For example, partners may seek to optimize closeness when confronting the mortality of a spouse or perhaps instead they desire less closeness as they contemplate an uncertain future. Therefore, we drew upon data from 64 older, married couples collected at two time points to examine the stability of older couples’ actual and desired marital closeness over a year and the role of both spouses’ health in potentially moderating these associations.

Theoretical frameworks for studying marital closeness in older adulthood

Several prominent theories have tried to explain how individuals desire and achieve closeness in their intimate relationships and to what extent it may depend on his or her developmental stage. To understand these frameworks, it is important to differentiate between how individuals desire and achieve closeness. Focusing solely on actual closeness is limited because the closeness that a couple would like to experience may be more or less than the level of actual closeness in their relationship. Considering both actual and desired closeness may have been overlooked because couples have historically been assumed to desire high levels of closeness (Mashek & Sherman, 2004). However, potential differences in the stability of actual and desired closeness may be especially crucial to explore among older adults in the light of the intersection between their developmental drive to maximize closeness and the practical challenges of dealing with both spouses’ declining health.

Developmental theories suggest that couples will desire increasing closeness later in life with conceivably minimal cost. Perhaps the most well-known of these theories is Carstensen’s (1992, 1996) socioemotional selectivity theory (SST), which proposes that as individuals’ age, they perceive their time as limited. This perception motivates them to seek emotional intimacy and companionship within their close relationships (Carstensen, Isaacowitz, & Charles, 1999). By deliberately focusing on their most intimate relationships, older adults are thought to maximize their emotional rewards, such as feeling close to one’s partner. Extending these tenets of SST to the marital relationship, Li and Fung (2011) proposed the dynamic goal theory of marital satisfaction. This theory proposes that spouses have multiple goals (i.e., personal growth, instrumental, companionship) to achieve throughout marriage and that couples prioritize these goals differently across the life span. Similar to SST (Carstensen, 1996), the dynamic goal theory suggests that couples in older adulthood focus more on companionship goals as opposed to personal growth or instrumental goals. Li and Fung (2011) suggest that older couples’ limited future time perspective explains why “the need to be bonded with the spouse is particularly salient in older adulthood, but less so in the younger age” (p. 249). Thus, from a theoretical perspective, older couples should mutually both desire and achieve closeness in their marriages as they age, not only despite the health challenges spouses may face but perhaps even due to a perception of limited time because of one’s own or one’s partner’s declining health.

However, it is important to consider the possibility raised by Polenick et al.’s (2015) work that the theoretically based developmental drive to increase closeness within relationships may operate differently in the context of health challenges. For example, if older adults are increasing their closeness with a spouse with poor or declining health during this developmental stage, they may also be considering their spouse’s mortality in conjunction with their own mortality. In turn, they may recognize the potential pain of losing a spouse, which could affect the amount of closeness they are experiencing and want to experience with their partner. From this perspective, the stability of actual and desired closeness may in fact diminish when one or both spouses are experiencing poor or declining health.

The role of health in marital closeness later in life

Many older couples will unfortunately face numerous health challenges later in life, as older adults are now experiencing higher rates of chronic illness, especially heart disease, cancer, and diabetes (National Center for Health Statistics, 2011). When health challenges occur within a marriage, it often leads to increases in caregiving responsibilities for a spouse, as they tend to be the initial caregiver (Stoller & Cutler, 1992). To note, men tend to experience health concerns both more frequently and earlier than do women in older adulthood (World Health Organization, 2003), resulting in more wives taking on caregiving responsibilities. As to the effects of these increased health concerns and caregiving behaviors, cross-sectional and longitudinal studies reveal that patients and caregiving spouses are both likely to report mental health problems, such as depressive symptoms (Berg, Wiebe, & Butner, 2011; Lee, Martire, Zarit, & Rovine, 2017; Lyons et al., 2014). Perhaps more concerning is that higher quality marriages, which normally confer substantial health benefits (Kiecolt-Glaser & Newton, 2001; Robles, Slatcher, Trombello, & McGinn, 2014), appear to put caregiving spouses at greater risk of problems when their partner is in poorer health (Hoppmann & Gerstorf, 2009). For example, in a cross-sectional study of highly satisfied older couples, Rauer Sabey, and Jensen (2014) found that receiving greater compassionate love from a spouse was linked with poorer health for both spouses.

As to this seemingly paradoxical finding that higher quality marriages confer additional risks to caregivers, Kiecolt-Glaser and Wilson (2017) suggest that “greater spousal closeness when facing one partner’s health crisis may spread poorer health, particularly among older couples” (p. 77). Indeed, Polenick et al. (2015) found that greater marital closeness magnified the effects of an older adult’s illness severity on their caregiving spouses’ depressive symptoms. However, it is important to note that other studies find that closeness is strongly protective of aging adults’ physical and mental health (Mancini & Bonanno, 2006; Tower et al., 2002). To reconcile these findings, what is needed is research examining how marital closeness functions in the context of health. Although closeness can benefit couples (Mancini & Bonanno, 2006; Tower & Kranser, 2006), there may be a stage when spouses desire less closeness as a result of health concerns, and thus what was once beneficial may no longer be so. By utilizing longitudinal data, we hope to elucidate when this stage may occur and how health challenges may catalyze decreases in stability of actual and desired closeness for older spouses.

Current study

The inconsistent outcomes previously found regarding older adults’ marital closeness (Mancini & Bonanno, 2006; Tower et al., 2002) coupled with a paucity of research on the desire for closeness in this population underscore the importance of examining the experience of closeness for older couples over time. We further consider the potential moderating role of two related, yet distinct facets of well-being, namely, objective and subjective health (Cutler & Grams, 1988; Hodes & Suzman, 2007; Idler & Kasl, 1995), in the stability of actual and desired closeness. To accomplish these goals, the current study utilizes two waves of data from a sample of 64 higher functioning older couples to examine whether and how husbands’ and wives’ reports of actual and desired marital closeness are linked over a year and the role of couples’ subjective and objective health in these pathways.

We seek to answer three questions in the current study. First, how stable are older adults’ perceptions of and desires for closeness over a year? Second, to what extent are spouses’ perceptions and desires linked both concurrently and over time? Finally, how do husbands’ and wives’ subjective and objective health moderate the associations between couples’ actual and desired closeness over time? Drawing upon previous theoretical and empirical work, we hypothesize that the levels of closeness will be stable for both spouses. We also expect that spouses’ perceptions and desires of closeness will be linked both concurrently and over time due to the mutual influence spouses have on each other. Finally, we hypothesize that couples in which one or both spouses report poorer health will experience less stability in both their desired and actual experiences of closeness in comparison to couples in which both spouses report better subjective and objective health.

Method

Participants

Sixty-four heterosexual married couples were recruited as a part of a larger study investigating links between marriage and health in older adulthood. Participants were recruited locally through health-care agencies, newspaper advertisements, churches, and other organizations in the Southeast U.S. To be eligible to participate, couples had to meet the following criteria: (1) be married, (2) be retired or partially retired (working fewer than 40 hr a week), and (3) be able to drive to the research center to ensure they were reasonably healthy and high functioning. Approximately 1 year (M = 16.4 months) after the first data collection, participating couples were contacted to complete a second wave of questionnaires. Husbands and wives were, on average, approximately 71 years old (SD = 7.4) and 69 (SD = 7.0) years old, respectively, and were almost exclusively European American (n = 60 and n = 61). Couples had been married for 42 years, on average (SD = 15), and 50 (79%) couples were in their first marriage. Couples were, on average, very happily married (M_W = 6.0, M_H = 6.2 on a 7-point scale of marital happiness). Couples had an average of 2.6 children (SD = 1.3; range = 0–6). Couples were highly educated—60 (94%) of the husbands and 54 (84%) of the wives attained at least some higher education. The average household income was US$85,875 (SD = US$64,074) and average total wealth (i.e., property, pensions, Individual Retirement Accounts [IRAs], and income) for couples was US$1,082,547 (SD = US$1,277,611). Forty-seven (73%) couples were fully retired, and 17 couples were partially retired (e.g., at least one spouse currently working for pay).

At the second wave of data, 55 of the 64 original couples agreed to participate, yielding a retention rate of 86%. Reasons for attrition included failure to locate participants, death, health limitations, and declined ongoing participation. Attrition analyses revealed that spouses lost to attrition did not differ from the retained spouses on age, marital duration, interpersonal closeness, objective health, or subjective health at T1.

Procedures

Prior to participating in data collection, all participants read and signed an informed consent letter that was approved by the local institutional review board. For the first wave of data collection (T1), couples participated in a 2- to 3-hr marital interview at an on-campus research laboratory. At the end of the interview, each spouse was given a set of questionnaires, which assessed the aspects of both health and marriage, including the measures of interest for the current study. Upon completing and returning these questionnaires, couples were paid US$75. For the second wave of data collection (T2), recontacted couples who agreed to participate were sent questionnaires via mail, again including the measures of interest here. Once the completed questionnaires were returned, couples were compensated US$50.

Measures

Actual and desired marital closeness

Both actual and desired levels of marital closeness were measured using the Inclusion of Other in the Self (IOS) Scale (Aron, Aron, & Smollan, 1992). This pictorial measure was designed to assess both actual and desired objective (behaving close) and subjective (feeling close) perceptions of interpersonal connectedness with a spouse. Participants were first asked to choose from a set of seven increasingly overlapping Venn diagrams, the diagram that best described his or her actual relationship with their spouse. Participants were then asked to choose from a set of seven increasingly overlapping Venn diagrams, the diagram that best described his or her desired relationship with their spouse. The IOS Scale has been well validated using numerous indicators of relationship functioning and is widely regarded as optimal for assessing the centrality of closeness in intimate relationships (Agnew, Loving, Le, & Goodfriend, 2004; Agnew, Van Lange, Rusbult, & Langston, 1998).

Health

Participants reported on both their subjective and objective health. To assess subjective health, participants evaluated their current health on a 4-point scale by asking, “Overall, would you describe your health as poor, fair, good, or excellent?” As most individuals evaluated their health favorably, we dichotomized this indicator, such that individuals who evaluated their health as poor, fair, or good were coded (0), and individuals who evaluated their health as excellent were coded (1). Furthermore, some scholars suggest that dichotomizing subjective health may have greater practical utility (Hoerger et al., 2016). To assess objective health, participants were asked to indicate whether they had ever been diagnosed by their doctor with any of the following diseases/conditions: heart trouble, diabetes, cancer, arthritis, asthma, stroke, lung disease, stomach problems/ulcers, leg problems, back problems, and/or depression. Participants indicated either “yes” or “no” to each condition, and affirmative responses were summed to create a total doctor-diagnosed disease score that ranged from 0 to 11, with higher scores indicating poorer objective health. Similar to previous work on older adults (Hodes & Suzman, 2007), this measure was not continuously distributed, with many older adults reporting at least one chronic condition. Thus, we dichotomized this indicator, with individuals who had two or more chronic conditions coded (0) and those with one or no chronic conditions coded (1).

Plan of analysis

We first conducted descriptive and preliminary analyses to examine the nature of and relations among the study variables. We then conducted mutual influence (Kenny, 1996), actor–partner interdependence models (APIMs; Kashy & Kenny, 2000) in MPlus Version 6.0 (Muthén & Muthén, 2007) to examine the contributions of self-reported actual and desired marital closeness at T1 on both their own and their spouses’ reports of actual and desired closeness at T2. We fit two separate models, one examining actual marital closeness and the other desired marital closeness. According to the recommendations from Kenny, Kashy, and Cook (2006) for cross-lagged dyadic models, T1 variables were centered, and all pathways were originally allowed to covary. Missing data were handled using full information maximum likelihood. To evaluate the robustness of the significant effects, we tested two control variables in each of the models, marital duration and age, and conducted sensitivity analyses by removing nonsignificant effects. We then conducted a series of delta χ² tests to explore potential gender differences in the pathways. Finally, to determine whether there were any difference in the pathways based on spouses’ objective and subjective health, we conducted multigroup analyses comparing the pathways of actual and desired closeness over time for husbands’ and wives’ objective and subjective health. Delta χ² tests were used to test differences between groups for each model.

Results

Preliminary analyses

Correlations and descriptive statistics are presented in Table 1. Paired sample t-tests revealed that both actual and desired closeness were fairly stable from T1 to T2 for husbands but not for wives—wives reported that actual, t(50) = −3.23, p < .01, and desired closeness, t(50) = 3.54, p < .01, decreased from T1 to T2. Desired closeness was significantly higher than actual closeness for wives at T1, t(58) = −2.58, p < .05, and for husbands at both T1, t(63) = −2.65, p < .05, and T2, t(51) = −3.25, p < .01. Finally, husbands reported significantly higher actual, t(50) = −2.39, p < .05, and desired closeness, t(50) = −3.80, p < .01, than wives at T2 but not at T1.

Table 1.

Descriptive statistics and intercorrelations for study variables for husbands and wives.

	1	2	3	4	5	6	7	8
1. Wife actual closeness T1	—
2. Wife actual closeness T2	.62**	—
3. Husband actual closeness T1	.40**	.30*	—
4. Husband actual closeness T2	.32*	.17	.61**	—
5. Wife desired closeness T1	.75**	.56**	.48**	.54**	—
6. Wife desired closeness T2	.44**	.79**	.24	.39**	.55**	—
7. Husband desired closeness T1	.33*	.22	.51**	.89**	.59**	.49**	–
8. Husband desired closeness T2	.36**	.36**	.78**	.45**	.42**	.32*	.48**	—
M (SD)	5.72 (1.20)	5.20 (1.31)	5.77 (1.23)	5.74 (1.30)	5.95 (1.06)	5.37 (1.32)	6.03 (1.18)	6.01 (1.25)
Range	2–7	3−7	2−7	1–7	4–7	3–7	3–7	1–7

*p < .05; **p < .01.

Regarding health, 46 wives (72%) and 43 husbands (67%) reported having poor, fair, or good subjective health, whereas 16 wives (25%) and 20 husbands (31%) reported excellent health. For objective health, 21 wives (33%) and 22 husbands (34%) reported having either no chronic illnesses or only one, with 42 wives (66%) and 42 husbands (66%) reporting at least two. Correlational analyses suggested that spouses’ reports of subjective and objective health were related to both husbands and wives, r = −.31, p < .05 and r = −.43, p < .01, respectively, such that spouses reported better subjective health when they had fewer doctor-diagnosed diseases. To note, independent sample t-tests indicated that none of the closeness variables differed by either wives’ or husbands’ subjective or objective health groups.

Couples’ experiences of actual and desired closeness over time

Two APIMs (Kashy & Kenny, 2000) were fit to examine the contributions of husbands’ and wives’ actual and desired marital closeness at T1 to the same variables at T2. According to the recommendations for analyzing APIMs with distinguishable dyads (Peugh, DiLillo, & Panuzio, 2013), both models were fully saturated (df = 0). We tested the control variables (marital duration and age) separately in both models. As there were no path differences between the uncontrolled and controlled models, we present the uncontrolled models per recommendations (Simmons, Nelson, & Simonsohn, 2011). Sensitivity analyses also revealed no changes in path differences when nonsignificant paths were removed.

As seen in Figure 1, the model examining husbands’ and wives’ self-reported actual marital closeness over time revealed significant actor effects for both spouses but no significant partner effects. Thus, we found evidence that both husbands and wives experienced significant stability in their feelings of actual closeness over a year but that husbands’ earlier closeness was not linked to wives’ later reports of actual closeness, and vice versa. Cross-sectionally, wives’ and husbands’ actual marital closeness was strongly correlated at T1 though not at T2. Overall, the model explained 38.7% of the variance of wives’ actual marital closeness at T2 and 37.2% of the variance of husbands’ actual marital closeness at T2.

Figure 1.

APIM of wives’ and husbands’ actual marital closeness from T1 to T2. Only significant pathways are shown with unstandardized path coefficients and standardized coefficients in parentheses. APIM = actor–partner interdependence model.

As seen in Figure 2, the model examining husbands’ and wives’ self-reported desired marital closeness over time revealed significant actor effects for both husbands and wives and a significant partner effect for the path from wife desired closeness at T1 to husband desired closeness at T2. Consistent with the model examining spouses’ reports of actual closeness, spouses’ reports of desired closeness demonstrated significant stability over time. In contrast to the findings on actual closeness, however, wives’ reports of desired closeness at T1 predicted greater desired closeness on the part of husbands at T2. Wives’ and husbands’ desired marital closeness was again correlated at T1 but not at T2. The model explained 28.2% of wives’ desired marital closeness at T2 and 38.1% of husbands’ desired marital closeness at T2.

Figure 2.

APIM of wives’ and husbands’ desired marital closeness from T1 to T2. Only significant pathways are shown with unstandardized path coefficients and standardized coefficients in parentheses. APIM = actor–partner interdependence model.

To capture whether husbands and wives differed in these pathways, we conducted a series of delta χ² tests. Analyses revealed no gender differences for actor, Δχ²(1) = .09, p = .76, or partner effects, Δχ²(1) = .01, p = .92, for actual marital closeness. For desired marital closeness, there were no gender differences for the actor effects, Δχ²(1) = 2.25, p = .13. The path from wives’ T1 desired closeness to husbands’ T2 desired closeness, however, was stronger than the path from husbands’ T1 desired closeness to wives’ T2 desired closeness, Δχ²(1) = 4.19, p = .04, suggesting spillover from wives to husbands was stronger than from husbands to wives.

Does health alter the links between couples’ actual and desired closeness over time?

To explore whether couples’ initial subjective and/or objective health moderated the links between couples’ closeness over time, a series of multigroup analyses were conducted. Models were fully saturated across health groups, df = 0. Examining wives’ objective health as a moderator of actual closeness (see Table 2), husband-and-wife actor effects were significant in both groups; however, the husband actor effect was stronger when wives had no chronic illness or only one than when wives had two or more chronic illnesses, Δχ²(1) = 5.97, p < .01. Although both husbands’ and wives’ actual closeness was stable over time across health groups, husbands reported greater stability in their actual closeness when wives reported being in better health. In contrast, we found no evidence of moderation for actual closeness based on husbands’ objective health.

Table 2.

Summary of multigroup analyses for objective health and significantly different pathways for actual and desired closeness.

	Wives’ objective health			Husbands’ objective health
	2+	0–1		2+	0–1
	β (SE)	β (SE)	χ² Difference	β (SE)	β (SE)	χ² Difference
Actual closeness
H_T1 → H_T2	.40* (.16)	.89** (.05)	Δχ²(1) = 5.97**	.57** (.13)	.74** (.11)
W_T1 → W_T2	.60** (.13)	.49** (.19)		.53** (.14)	.73** (.13)
W_T1 → H_T2	.16 (.18)	.03 (.11)		−.07 (.17)	.24 (.14)
H_T1 → W_T2	.12 (.15)	−.14 (.21)		.13 (.15)	−.18 (.19)
Desired closeness
H_T1 → H_T2	−.10 (.17)	.69** (.13)	Δχ²(1) = 8.45**	.12 (.16)	.71** (.11)	Δχ²(1) = 8.50**
W_T1 → W_T2	.71** (.15)	.37^† (.19)		.60** (.13)	.28 (.24)
W_T1 → H_T2	.74** (.13)	.31^† (.16)		.53** (.14)	.26^† (.13)
H_T1 → W_T2	−.20 (.17)	.33^† (.20)		.02 (.15)	.18 (.24)

Note. All β coefficients and SEs are standardized. SE = standard error; H = husband; W = wife.

^†p< .10; *p < .05; **p < .01.

For desired closeness, we again found wives’ objective health acted as a moderator. The husband actor effect was stronger when wives reported one or no chronic illnesses than when they reported two or more, Δχ²(1) = 8.45, p < .01, indicating that the stability of husbands’ desired closeness was greater when wives were in better health. For the role of husbands’ objective health in spouses’ desired closeness, we found evidence of moderation for the actor effects. Specifically, the husband actor effect was stronger when husbands were in better health, Δχ²(1) = 8.50, p < .01. Overall, the significance and strength of the links between husbands’ and wives’ actual and desired closeness over time across objective health groups were similar. However, a consistent exception to this pattern of similarity was that the stability of husbands’ closeness over time was stronger when wives reported better health.

Analyses revealed no evidence of moderation in actual closeness based on either spouses’ subjective health (see Table 3). Similarly, there was no significant moderation for desired closeness based on husbands’ subjective health. There was, however, evidence of moderation in desired closeness based on wives’ subjective health, with the effect of wives’ initial closeness on husbands’ later desired closeness was stronger for those reporting poor, fair, or good health than those reporting excellent health, Δχ²(1) = 4.18, p < .05. Specifically, wives’ initial desired closeness was more strongly predictive of husbands’ later desired closeness when wives themselves reported poorer health. Results remained robust even after conducting sensitivity checks in which nonsignificant paths were removed.

Table 3.

Summary of multigroup analyses for subjective health and significantly different pathways for actual and desired closeness.

	Wives’ subjective health			Husbands’ subjective health
	Poor, fair, or good	Excellent		Poor, fair, or good	Excellent
	β (SE)	β (SE)	χ² Difference	β (SE)	β (SE)	χ² Difference
Actual closeness
H_T1 → H_T2	.51** (.14)	.78** (.11)		.63** (.12)	.52** (.17)
W_T1 → W_T2	.64** (.12)	.54* (.17)		.59** (.13)	.75** (.14)
W_T1 → H_T2	.05 (.16)	.12 (.16)		−.09 (.16)	.27 (.19)
H_T1 → W_T2	−.09 (.15)	.35^† (.18)		.10 (.15)	−.15 (.18)
Desired closeness
H_T1 → H_T2	.22 (.15)	.65** (.13)		.15 (.16)	.49** (18)
W_T1 → W_T2	.61** (.13)	.33 (.24)		.52** (.13)	.38 (.24)
W_T1 → H_T2	.54** (.13)	.32^† (.16)	Δχ²(1) = 4.18*	.55** (.13)	.26 (.20)
H_T1 → W_T2	−.10 (.16)	.25 (.23)		.20 (.18)	−.11 (.25)

Note. All β coefficients and SEs are standardized. SE: standard error; H = husband; W = wife.

^†p < .10; *p < .05; **p < .01.

Discussion

Although closeness is a necessary element in relationship formation and maintenance (Ben-Ari & Lavee, 2007; Hess, Fannin, & Pollom, 2007), greater closeness has been found to be both beneficial and harmful for older adults’ physical and mental health (Mancini & Bonanno, 2006; Polenick et al., 2015). These contradictory findings underscore the importance of considering how closeness is experienced later in life, as this is a time when health becomes particularly problematic and thus potentially more relevant for couples’ closeness (National Center for Health Statistics, 2011). Thus, the current study drew upon Carstensen’s (1992) SST and Li and Fung’s (2011) dynamic goal theory to understand the experiences of marital closeness over time and the role of health in a sample of high-functioning older couples. To note, spouses in our sample were highly satisfied in their marriages and thus uniquely qualified to test the theoretical tenets of maximizing the emotional rewards of the marital relationship (Carstensen, 1992; Li & Fung, 2011), as opposed to a distressed sample of older couples that may be maximizing other relationships outside the marital relationship.

Our findings suggest that older adulthood is a time when marriages flourish and spouses enjoy feeling close and desire to be close to one another, supporting the central tenets of both SST and dynamic goal theory (Carstensen, 1992; Li & Fung, 2011). Although this sample of satisfied older couples enjoyed fairly stable marital closeness—achieved and desired—over time, it is noteworthy that the initial levels of closeness only explained moderate levels of variance in closeness a year later, with more variance explained for husbands’ desire for closeness than wives’. Additional evidence emerged to suggest that gender differences may become more marked as individuals age. In particular, the within-person analyses of closeness for husbands suggest that husbands are experiencing and desiring more stable levels of closeness with their wives over time. Further, husbands’ desires for closeness appeared to be more sensitive to wives’ earlier desires, whereas wives’ desires for closeness were not predicted by husbands’ desires. Finally, our moderational analyses also suggested that wives’ initial desire for closeness and the state of her health may be especially crucial for the experience of closeness for both spouses.

Is closeness in older adulthood stable and developmentally driven? Or does it depend?

Consistent with SST (Carstensen, 1992, 1996) and Li and Fung’s (2011) dynamic goal theory, older spouses in our modestly sized sample reported high levels of both actual and desired closeness and these were relatively stable over a 1-year period. Such findings seemingly contrast with longitudinal work showing uniform declines in marital satisfaction across the life span (Vaillant & Vaillant, 1993; VanLaningham, Johnson, & Amato, 2001), suggesting that marital satisfaction and closeness may become increasingly uncoupled at some point in the life course. Given that our maritally satisfied sample still demonstrated different patterns of stability in closeness, future research should explore the differences between closeness and marital satisfaction over time. Such an inquiry may help pinpoint when and why these key marital constructs begin to diverge in couples’ lives, if they in fact do at all given recent work suggesting that the oft-reported decline of marital satisfaction over the life span may be a methodological artifact (see Proulx, Ermer, & Kanter, 2017, for a review). Thus, longer time frames and larger samples with more diverse marital experiences may be necessary to accomplish this goal, as the current study only considered marital closeness over a 1-year time point in a sample of satisfied couples. Finally, future research should also examine any gender difference in this possible uncoupling, as our findings suggest that it could be less marked for wives as they reported significantly less desired and actual closeness over time.

Beyond wives’ closeness decreasing over time, wives’ initial desires for closeness appear to play a key role within the marital relationship, as husbands’ desires for closeness were predicted by wives’ initial desires for closeness but not the other way around. These findings are consistent with the work by Schwarzer and Gutiérrez-Doña’s (2005), suggesting that the importance of the marital relationship is especially intensified for older men as their social networks decrease in size. Furthermore, given that Tower and Kasl (1996) found that older husbands are more likely than their wives to name their spouse as a confidant, older husbands may take more relational cues, such as desires for closeness, from their wives than wives take from their husbands, as wives tend to have larger social networks with other close confidants.

Although our moderational analyses should be interpreted with caution due to sample size restrictions, they offer some insight into these gender differences. Perhaps because men experience health concerns more frequently and earlier than their female counterparts do (World Health Organization, 2003), wives may not be as sensitive to smaller variations in their husband’s health as they may already be primed to provide care in response to any and all of their husband’s health challenges. In fact, in a meta-analysis of caregiving, Pinquart and Sorensen (2006) found that women provided more caregiving hours and a higher number of caregiving tasks when compared with male caregivers. Wives’ chronic illnesses, however, appeared to act as a moderator of experiences within the marital relationship. When wives were in better health, husbands enjoying greater stability in their desired and achieved closeness over time. Perhaps wives in better health were able to partake in more activities and new experiences with their husbands. According to Aron and Aron (1986), these experiences would allow for self-expansion that would, in turn, foster more closeness to the benefit of both spouses. In contrast, when wives reported being in poorer health, husbands’ desires were linked to the wives’ initial desires for closeness. Our findings that husbands appeared to be guided by what their wives want when their wives were in poorer health are consistent with previous research. For example, maritally satisfied husbands experienced high levels of distress when their partners were suffering, regardless of the level of perceived suffering (Monin, Levy, & Kane, 2017). Although husbands may find caring for an ill spouse more stressful than their spouses do (England, 2005), husbands also appear to be highly aware of their wives’ desires in the midst of wives’ chronic illnesses.

Considerations and conclusions

Although our study included many notable strengths (i.e., longitudinal, dyadic examination of older couples), there are some limitations to consider when interpreting the results. First, our modestly sized sample was relatively well-off, higher functioning, maritally satisfied, and primarily European American. As economic hardship has been linked to marital distress (Conger, Rueter, & Elder, 1999) and race has been linked to differences in marital quality (Broman, 2005), our results may not generalize to all older adults. Furthermore, our sample includes older adults who were married during the middle of the 20th century, a time in which many marriages were defined as companionate—focusing on intimacy and the emotional bond between the spouses (Cherlin, 2009). These ideals map onto the tenets of SST (Carstensen et al., 1999) and Li and Fung’s (2011) dynamic goal theory, which suggest that this cohort of older spouses may be particularly interested in maximizing emotional rewards within marriage. In contrast, marriages in the 21st century are individualized—focusing on the self-actualization for each spouse (Cherlin, 2009). This historical shift in marital values and goals may result in newer cohorts holding to more stringent ideals of self-actualization in a marriage, instead of prioritizing companionship, and thus these findings may be more applicable to this cohort of older adults.

Although our study utilized measures of both objective and subjective health, future research should consider changes in health to assist with interpretation of findings, as we only looked at the role of initial health in these pathways. Also, the current study only measured self-reports of marital closeness, not distance or perceived partner closeness, despite the importance of both closeness and distance in the continuing and formation of relationships (Birtchnell, 2001) and the importance of perceived partner closeness in relationship satisfaction (Tomlinson & Aron, 2013). Finally, the current study utilized a single-item measure for marital closeness that—although widely used (e.g., Lehmiller, 2012; Morry, 2005; Polenick et al., 2015)—may have unintentionally caused couples who were already close to reach an asymptote, obscuring potential differences over time. Future research should consider utilizing open-ended interviews to provide a fuller picture of how older adults are experiencing and desiring closeness.

In conclusion, our findings add an important perspective to the study of marital closeness for both scholars and practitioners. First, our work has been largely guided by both SST and dynamic goal theory (Carstensen et al., 1999; Li & Fung, 2011), which suggest that older adults perceive their time as limited, which motivates them to optimize positive emotion within their close relationships. Although most of our findings are consistent with these theories, there are nuances to be considered. As couples are living longer and facing a new set of challenges than previous generations (e.g., more complex marital histories; Brown & Lin, 2012; Rauer & Albers, 2016), the central tenets of both theories may need to be updated. For example, as the goals and values of marriage evolve in our culture (Cherlin, 2009), these theories may not be as applicable for future cohorts whose individualized marriages may be more focused on spouses encouraging one another to recognize their full potential (Finkel, Hui, Carswell, & Larson, 2014). In a way, this new model of marriage hews more closely to self-expansion theory (Aron & Aron, 1986), as it would suggest that individuals enter relationships in order to make one’s life richer and more complex by acquiring the characteristics, resources, and experiences of a spouse. Integrating these theories may help scholars better understand the marital desires and experiences of more recent cohorts as they begin to enter older adulthood.

Second, practitioners should be aware that although older couples may be desiring and enjoying close relationships, wives’ desires for closeness and the role that their health plays in both spouses’ desires for closeness warrant further investigation. Although older couples have been largely overlooked in therapy due in part to their historical reticence toward seeking help (Goldstein & Preston, 1984), scholars suggest that there is going to be an increase in older adults attending therapy as the population becomes more affected by the challenges associated with aging (Zarit & Knight, 1996). Practitioners should be prepared to assess a couple’s current feelings of closeness and be especially cognizant of both spouses’ health problems in order to tailor interventions to promote closeness (Jacobson & Christensen, 1996; Pistole, 1994). However, if couples feel too close and desire less closeness, practitioners may also need to be adept at tailoring interventions to promote healthy boundaries and interests outside the relationship (Kreger, 2013). These tailored interventions, in turn, may promote the mental and physical well-being of spouses who are facing a variety of developmental and relational challenges (Mancini & Bonanno, 2006; Pruchno, Cartwight, & Wilson-Genderson, 2009), an important goal in the light of the rapidly growing number of individuals entering later life nowadays.

Footnotes

Authors’ note

Parts of this manuscript were presented at the 2015 Annual Quint State Conference in Athens, GA and at the 2017 Annual Conference for the National Council on Family Relations in Orlando, FL.

Acknowledgements

The authors would like to thank the couples who participated in the Marriage and Retirement Study.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the John E. Fetzer Institute and the Alabama Agricultural Experiment Station.

Open research statement

This research was not preregistered. The data and materials used in the research are available and can be obtained by e-mailing arauer@utk.edu.

References

Agnew

C. R.

Loving

T. J.

Goodfriend

(2004). Thinking close: Measuring relational closeness as perceived self-other inclusion. In Mashek

D. J.

Aron

(Eds.), Handbook of closeness and intimacy (pp. 103–115). Mahwah, NJ: Lawrence Erlbaum Associates Publishers.

Agnew

C. R.

Van Lange

P. A.

Rusbult

C. E.

Langston

C. A.

(1998). Cognitive interdependence: Commitment and the mental representation of close relationships. Journal of Personality and Social Psychology, 74, 939–954.

Aron

E. N.

(1986). Love and the expansion of self: Understanding attraction and satisfaction. New York, NY: Hemisphere Publishing Corp/Harper & Row Publishers.

Aron

E. N.

Smollan

(1992). Inclusion of other in the self scale and the structure of interpersonal closeness. Journal of Personality and Social Psychology, 63, 596–612.

Ben-Ari

Lavee

(2007). Dyadic closeness in marriage: From the inside story to a conceptual model. Journal of Social and Personal Relationships, 24, 627–644.

Berg

C. A.

Wiebe

D. J.

Butner

(2011). Affect covariation in marital couples dealing with stressors surrounding prostate cancer. Gerontology, 57, 167–172.

Birtchnell

(2001). Relating therapy with individuals, couples and families. Journal of Family Therapy, 23, 63–84.

Broman

C. L.

(2005). Marital quality in black and white marriages. Journal of Family Issues, 26, 431–441.

Brown

S. L.

Lin

I. F.

(2012). The gray divorce revolution: Rising divorce among middle-aged and older adults, 1990–2010. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 67, 731–741.

10.

Carstensen

L. L.

(1992). Social and emotional patterns in adulthood: Support for socioemotional selectivity theory. Psychology and Aging, 7, 331–338.

11.

Carstensen

L. L.

(1996). Socio-emotional selectivity: A life span developmental account of social behavior. In Merrens

M. R.

Brannigan

G. G.

(Eds.), The developmental psychologists: Research adventures across the life span (pp. 251–272). New York, NY: The McGraw-Hill Companies, Inc.

12.

Carstensen

L. L.

Isaacowitz

D. M.

Charles

S. T.

(1999). Taking time seriously: A theory of socioemotional selectivity. American Psychologist, 54, 165–181.

13.

Cherlin

A. J.

(2009). The marriage-go-round: The state of marriage and the family in America today. New York, NY: Knopf.

14.

Conger

R. D.

Cui

Bryant

C. M.

Elder

G. H.

(2000). Competence in early adult romantic relationships: A developmental perspective on family influences. Journal of Personality and Social Psychology, 79, 224–237.

15.

Conger

R. D.

Rueter

M. A.

Elder

G. H.

Jr . (1999). Couple resilience to economic pressure. Journal of Personality and Social Psychology, 76, 54–71.

16.

Cutler

S. J.

Grams

A. E.

(1988). Correlates of self-reported everyday memory problems. Journal of Gerontology, 43, 82–90.

17.

England

(2005). Emerging theories of care work. Annual Review of Sociology, 31, 381–399.

18.

Erikson

E. H.

(1959). Identity and the life cycle: Selected papers. Psychological Issues, 1, 1–171.

19.

Finkel

E. J.

Hui

C. M.

Carswell

K. L.

Larson

G. M.

(2014). The suffocation of marriage: climbing mount Maslow without enough oxygen. Psychological Inquiry, 25, 1–41.

20.

Goldstein

Preston

(1984). Marital therapy for the elderly. Canadian Medical Association Journal, 130, 1551–1553.

21.

Hess

J. A.

Fannin

A. D.

Pollom

L. H.

(2007). Creating closeness: Discerning and measuring strategies for fostering closer relationships. Personal Relationships, 14, 25–44.

22.

Hodes

R. J.

Suzman

(2007). Growing older in America: The health and retirement study. Bethesda, MD: US Department of Health and Human Services.

23.

Hoerger

Coletta

Sörensen

Chapman

B. P.

Kaukeinen

Duberstein

P. R.

(2016). Personality and perceived health in spousal caregivers of patients with lung cancer: the roles of neuroticism and extraversion. Journal of Aging Research, 2016, 5659793.

24.

Hoppmann

Gerstorf

(2009). Spousal interrelations in old age: A mini-review. Gerontology, 55, 449–459.

25.

Idler

E. L.

Kasl

S. V.

(1995). Self-ratings of health: Do they also predict change in functional ability? The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 50, 344–353.

26.

Jacobson

N. S.

Christensen

(1996). Integrative couple therapy: Promoting acceptance and change. New York, NY: Norton.

27.

Kashy

D. A.

Kenny

D. A.

(2000). The analysis of data from dyads and groups. In Reis

H. T.

Judd

C. M.

(Eds.), Handbook of research methods in social and personality psychology (pp. 451–477). New York, NY: Cambridge University Press.

28.

Kenny

D. A.

(1996). Models of non-independence in dyadic research. Journal of Social and Personal Relationships, 13, 279–294.

29.

Kenny

D. A.

Kashy

D. A.

Cook

W. L.

(2006). Dyadic data analysis. New York, NY: Guilford Press.

30.

Kiecolt-Glaser

J. K.

Newton

T. L.

(2001). Marriage and health: His and hers. Psychological Bulletin, 127, 472–503.

31.

Kiecolt-Glaser

J. K.

Wilson

S. J.

(2017). Lovesick: How couples’ relationships influence health. Annual Review of Clinical Psychology, 13, 421–443.

32.

Kreger

(2013). Enmeshment in Family Relationships: 1+1=1. Retrieved from https://www.psychologytoday.com/blog/stop-walking-eggshells/201301/enmeshment-infamily-relationships-111

33.

Lee

J. E.

Martire

L. M.

Zarit

S. H.

Rovine

M. J.

(2017). Activity restriction and depressive symptoms in older couples. Journal of Aging and Health, 29, 1251–1267.

34.

Lehmiller

J. J.

(2012). Perceived marginalization and its association with physical and psychological health. Journal of Social and Personal Relationships, 29, 451–469.

35.

Fung

H. H.

(2011). The dynamic goal theory of marital satisfaction. Review of General Psychology, 15, 246–254.

36.

Lyons

K. S.

Bennett

J. A.

Nail

L. M.

Fromme

E. K.

Dieckmann

Sayer

A. G.

(2014). The role of patient pain and physical function on depressive symptoms in couples with lung cancer: A longitudinal dyadic analysis. Journal of Family Psychology, 28, 692–700.

37.

Mancini

A. D.

Bonanno

G. A.

(2006). Marital closeness, functional disability, and adjustment in late life. Psychology and Aging, 21, 600–610.

38.

Mashek

D. J.

Sherman

M. D.

(2004). Desiring less closeness with intimate others. In Mashek

D. J.

Aron

(Eds.), Handbook of closeness and intimacy (pp. 343–356). Mahwah, NJ: Lawrence Erlbaum Associates Publishers.

39.

Monin

J. K.

Levy

B. R.

Kane

H. S.

(2017). To love is to suffer: Older adults’ daily emotional contagion to perceived spousal suffering. The Journals of Gerontology: Series B, 72, 383–387.

40.

Morry

M. M.

(2005). Relationship satisfaction as a predictor of similarity ratings: A test of the attraction-similarity hypothesis. Journal of Social and Personal Relationships, 22, 561–584.

41.

Muthén

L. K.

Muthén

B. O.

(2007). Mplus user’s guide (5th ed.). Los Angeles, CA: Muthén & Muthén.

42.

National Center for Health Statistics. (2011). Health, United States, 2010: With Special Feature on Death and Dying. Washington, DC: US. Government Printing Office.

43.

Peugh

J. L.

DiLillo

Panuzio

(2013). Analyzing mixed-dyadic data using structural equation models. Structural Equation Modeling; A Multidisciplinary Journal, 20, 314–337.

44.

Pinquart

Sörensen

(2006). Gender differences in caregiver stressors, social resources, and health: An updated meta-analysis. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 61, P33–P45.

45.

Pistole

C. M.

(1994). Adult attachment styles: Some thoughts on closeness-distance struggles. Family Process, 33, 147–159.

46.

Polenick

C. A.

Martire

L. M.

Hemphill

R. C.

Stephens

M. A. P.

(2015). Effects of change in arthritis severity on spouse well-being: The moderating role of relationship closeness. Journal of Family Psychology, 29, 331–338.

47.

Proulx

C. M.

Ermer

A. E.

Kanter

J. B.

(2017). Group-based trajectory modeling of marital quality: A critical review. Journal of Family Theory and Review, 9, 307–327.

48.

Pruchno

R. A.

Cartwright

F. P.

Wilson-Genderson

(2009). Effects of marital closeness on the transition from caregiving to widowhood. Aging and Mental Health, 13, 808–817.

49.

Rauer

Albers

(2016). Marital happiness. In Whitbourne

S. K.

(Ed.), Encyclopedia of adulthood and aging. London, UK: Wiley-Blackwell.

50.

Rauer

Pettit

Lansford

Bates

Dodge

(2013). Romantic relationship patterns in young adulthood and their developmental antecedents. Developmental Psychology, 49, 2159–2171.

51.

Rauer

A. J.

Sabey

Jensen

J. F.

(2014). Growing old together: Compassionate love and health in older adulthood. Journal of Social and Personal Relationships, 31, 677–696.

52.

Robles

T. F.

Slatcher

R. B.

Trombello

J. M.

McGinn

M. M.

(2014). Marital quality and health: A meta-analytic review. Psychological Bulletin, 140, 140–187.

53.

Schwarzer

Gutiérrez-Doña

(2005). More spousal support for men than for women: A comparison of sources and types of support. Sex Roles, 52, 523–532.

54.

Simmons

J. P.

Nelson

L. D.

Simonsohn

(2011). False-positive psychology: Undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychological Science, 22, 1359–1366.

55.

Stoller

E. P.

Cutler

S. J.

(1992). The impact of gender on configurations of care among married elderly couples. Research on Aging, 14, 313–330.

56.

Tomlinson

J. M.

Aron

(2013). The path to closeness: A mediational model for overcoming the risks of increasing closeness. Journal of Social and Personal Relationships, 30, 805–812.

57.

Tower

R. B.

Kasl

S. V.

(1996). Gender, marital closeness, and depressive symptoms in elderly couples. The Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 51, 115–129.

58.

Tower

R. B.

Kasl

S. V.

Darefsky

A. S.

(2002). Types of marital closeness and mortality risk in older couples. Psychosomatic Medicine, 64, 644–659.

59.

Tower

R. B.

Krasner

(2006). Marital closeness, autonomy, mastery, and depressive symptoms in a US Internet sample. Personal Relationships, 13, 429–449.

60.

Vaillant

C. O.

Vaillant

G. E.

(1993). Is the U-curve of marital satisfaction an illusion? A 40 year study of marriage. Journal of Marriage and the Family, 55, 230–239.

61.

VanLaningham

Johnson

D. R.

Amato

(2001). Marital happiness, marital duration, and the U-shaped curve: Evidence from a five-wave panel study. Social Forces, 79, 1313–1341.

62.

World Health Organization. (2003). Gender, health and ageing. Geneva, Switzerland: WHO, 2003. http://www.who.int/gender/documents/en/Gender_Aging.pdf

63.

Zarit

S. H.

Knight

B. G.

(1996). A guide to psychotherapy and aging: Effective clinical interventions in a life-stage model. Washington, DC: American Psychological Association.