Social participation and marital satisfaction in mid to late life marriage

Abstract

Objectives:

Research shows that social participation is beneficial for overall health and well-being. Yet, no research to our knowledge has examined whether social participation is associated with greater marital satisfaction in middle-aged and older couples. We hypothesized that middle-aged and older adults would have greater marital satisfaction when their spouse engaged in social groups because there would be greater opportunity for self-expansion and for social support from ties outside the marriage.

Methods:

We used background self-report data from a multi-method study of 98 middle-aged and older adult married couples (N = 196) with chronic conditions. As part of the study, spouses completed questionnaires that measured the frequency and intensity of involvement in social groups (e.g., church, business groups). Marital satisfaction was measured with the Locke Wallace Marital Adjustment Test.

Results:

Contrary to our hypotheses, results from actor–partner interdependence models provided no evidence that one’s own social participation was associated with one’s own marital satisfaction (actor effects). However, in line with our hypotheses regarding partner effects, one spouse’s (a) report of any social participation with church organizations, business groups, or social groups, (b) greater number of affiliations with different organizations, (c) greater frequency of participation, and (d) being an active officer in a social organization were significantly associated with the other spouse’s greater marital satisfaction.

Discussion:

Findings of this study suggest that having a spouse who participates in social groups is good for relationship satisfaction in mid to late life marriage.

Maintaining a happy marriage in mid to late life provides personal health benefits (Pienta et al., 2000). Individuals in happy marriages have better sleep than those in unhappy marriages (Lauderdale & Chen, 2017). They also have better cardiovascular health, endocrine systems, and immune systems (Robles & Kiecolt-Glaser, 2003). Although we know that having a satisfied marriage has important health benefits, less is known about what makes older adults happy in their marriages. The health benefits of satisfied marriages are likely to be strong in later life as older adults’ social networks become smaller and spouses increasingly rely on each other for social support and support for their health conditions (Carstensen, 1992).

Research on marriage across adulthood shows that having shared interests and sharing activities and experiences together is beneficial for marriage (Harasymchuk et al., 2020; Riekkola et al., 2019). However, spending time apart and indulging in separate interests has also been shown to have its benefits (Riekkola et al., 2019). Participation in social organizations, especially in which spouses take on leadership roles, may be one way that older adults maintain their separate interests that provide them with personal meaning (Menec, 2003). Thus, the aim of this study was to investigate the interpersonal associations between social participation and marital satisfaction in mid to late life marriage.

Although little is known about the benefits of social participation in the context of marriage, there is a substantial literature showing personal benefits of social participation for older adults outside the context of a specific relationship. Social participation, defined as involvement in religious organizations, social clubs, community organizations, and charity work, is associated with greater self-reported health, physical functioning, cognitive functioning, and health behavior (Sirven & Debrand, 2008). One reason is that social participation makes people feel less lonely and is a source of social support (Niedzwiedz et al., 2016). It may also increase feelings of mastery (Gadalla, 2009) and a sense of purpose, which have been associated with increased relationship satisfaction (Pfund et al., 2020).

But why would social participation affect relationship satisfaction of a spouse? One reason is that social participation of one spouse creates opportunities for self-expansion (Aron & Aron, 1997; Aron et al., 2013). Self-expansion is a central motivation in all humans that explains the need to grow or expand the self. Gaining and maintaining close relationships expands the self because one incorporates social and material resources, perspectives, and characteristics of the other in the self (Aron et al., 1991). Self-expansion, as measured with self-reported measures of self-expansion (Mattingly et al., 2014), has been linked to one’s own relationship satisfaction in younger couples (McIntyre et al., 2020). Further, it has been theorized that shared self-expanding activities contribute to a sense of security which leads to increased relationship satisfaction (Cortes et al., 2020). Studies have found that when partners report high self-expansion they feel more satisfied with their sex life and relationship (Raposo et al., 2019).

It is also possible that social participation of one spouse leads to expansion of the self in their partner, suggesting a potential mechanism through which greater social participation leads to greater relationship satisfaction in the partner. The partner might indirectly gain social resources and perspectives from a spouse that is involved in other activities, and this might have positive implications for both spouses’ personal and relational functioning. Spouses have the opportunity to grow and learn from their partner when their partner is involved in activities that they are passionate about (Aron & Aron, 1986; Mattingly et al., 2020). In addition, spouses can encourage one another to seek out opportunities for self-expansion, which has long-term implications for health and retirement satisfaction (Tomlinson et al., 2020).

Social participation can also strengthen other ties outside of the marriage. Consistent with this idea, a previous study found that wives whose husbands had close friends reported higher positive marital quality 5 years later (Zhaoyang & Martire, 2020). Social integration, a related concept to social participation but one that involves diversity in types of social roles and activities (e.g., engaging in both work and family roles), provides a sense of community, neighborhood attachment, and support from friends (Voydanoff, 2005). When people gain this sense of belonging, support, and community, they may feel more secure in their marriage.

Using data from a study of married older adults, the present study examined the association between an individual’s social participation and one’s own marital satisfaction (actor effect) and their partner’s marital satisfaction (partner effect). We hypothesized that any involvement in a church group, business group, or social club (Hypothesis 1), the greater number of social organizations one is a part of (Hypothesis 2), one’s own increased frequency of social participation (Hypothesis 3), and being an officer for a social organization (Hypothesis 4) would be associated with one’s own greater marital satisfaction. Additionally, we hypothesized that any organizational involvement, the greater number of social organizations one is a part of, one’s own increased frequency of social participation, and being an officer for a social organization would be associated with the other spouse’s greater marital satisfaction (Hypotheses 5–8). We did not make any hypotheses regarding gender differences; however, we explored differences by gender in our models.

Method

Participants

To examine our hypotheses, we analyzed data from a study of 98 middle-aged and older adult married couples (N = 196) from New Haven County, Connecticut. The APIM-PowerR was used for post hoc power analysis (Ackerman & Kenny, 2016). The aim of the original study was to experimentally manipulate mutual support between married couples in which both partners had at least one chronic health condition while they discussed their health concerns in the laboratory. Couples over the age of 50 were recruited because the presence of chronic conditions increases with older age (Ward et al., 2014). Heterosexual couples were recruited to increase homogeneity regarding the historical and developmental context of traditional gender roles. Newspaper advertisements and community bulletins were used to recruit participants. As part of the study, spouses completed separate questionnaires that were sent to their homes, assessing demographics, chronic conditions, social participation, and marital satisfaction. Participants received $40 each as compensation. Data collection took place between 2014 and 2016 (Monin et al., 2019). All variables in this study were taken from the background questionnaires that were sent in the mail.

Measures

Social participation.

Social participation was measured with the questions: “Are you involved in any organizations at the present time?” (yes or no); “How many organizations are you involved with currently?” (enter a numerical number); “Overall, about how often do you participate in some kind of organizational activity?” on a scale from 1 (less than once a month) to 5 (more than once a week); and “Are you currently an officer in any of these organizations?” (yes or no). We drew these questions from other studies measuring social participation (Douglas et al., 2017; van Hees et al., 2020; Zhang & Zhang, 2015).

Marital satisfaction.

Marital satisfaction was measured with the Locke Wallace Marital Adjustment Test, a 16-item questionaire measuring several aspects of marital quality (Locke & Wallace, 1959). Each spouses’ general level of marital happiness was measured on a scale from 1 (very unhappy) to 7 (perfectly happy). Perception of spousal agreement regarding issues such as handling family finances, matters of recreation, friends, and sexual relations were assessed on scales ranging from 1 (always disagree) to 6 (always agree). The reliability and validity of the Locke Wallace Marital Satisfaction Scale has been established among the general and spousal caregiving populations with a range of 0.72 to 0.90 (Jiang et al., 2013).

Covariates.

The following theoretically relevant covariates were measured. Socio-demographic factors included age (years), gender (male or female), education level, having any children (yes/no), marital length (in years), and income by increments of $10,000. The Physical Comorbidity Index was used to assess how many chronic conditions participants had including heart disease, stroke, and cancer (Katz et al., 1996). In addition, we calculated a variable indicating working status where full-time, part-time, and self-employed were coded as 1 (currently working) and retired, homemaker, and unemployed were coded as 0 (not working).

Analysis

First, we tested for potential covariates (using correlations, t-tests, ANOVAs, and chi-square analysis) and included the variables that were significantly associated with both marital satisfaction and one of the social participation variables in our preliminary models. We used a threshold of p < 0.10 when choosing what coviariates to include in the preliminary models. Then we included only the covariates that remained significant at a p < 0.10 threshold in the final models to make the models as parsimonious as possible. We included both the actor and partner versions of the covariate variables in the preliminary models. To test the main hypotheses, we used a dyadic data analytic technique, the Actor–Partner Interdependence Model (APIM; Kenny et al., 2006) using the mixed procedure in SPSS, which not only deals appropriately with the non-independence of the data but also takes advantage of it to address questions of mutual influence. It is a dual-intercept model that entails calculating “actor effects” and “partner effects.” An actor effect represents the influence that an individual’s score on a predictor variable has on one’s own score on a dependent variable (e.g., the effect of a person’s social participation on their own marital satisfaction), and a partner effect represents the influence that an individual’s score on a predictor has on one’s partner’s score on the dependent variable (e.g., the effect of a person’s social participation on their partner’s marital satisfaction). For each model, the independent variables were both partner’s social participation variables (e.g., any participation in social organizations, number of organizations involved with, frequency of participation, and officer status). We ran separate models for each of the four social participation variables predicting relationship satisfaction. Preliminary models adjusted for all covariates that were significantly related to both the independent and dependent variables in each model. Final models included only significant covariates, of which there were none. Although we did not have hypotheses concerning gender roles, we explored whether actor and partner effects were modified by role [wife (1) versus husband (2)] by adding interaction terms between role and each of the social participation variables to the preliminary and final models.

Results

Participant characteristics and descriptive statistics

As shown in Table 1, the mean age of wives was 67 years (min = 51, max = 89) and husbands was 70 years (min = 56, max = 90). Couples in our sample were mainly White and highly educated with more than 60% of husbands and wives having earned a bachelor’s degree or higher. A third of the couples had an annual household income of $100,000 or more. On average, couples had been married for 36 years (sd = 14.7) and 78% had children. Fifty-eight percent of husbands were not working, and 39% were currently working. Sixty-two percent of wives were not working, while 35% of wives were currently working.

Table 1.

Participant characteristics.

	Wives		Husbands
	N (%)*	Range	N (%)*	Range
Social participation
Involved in any organizations	64 (65)		64 (66)
Number of organizations involved in	1.81 ± (1.58)	0–7	1.48 ± (1.37)	0–5
Frequency of participation in organization
Less than once a month	24.8 (27.5)		27.7 (29.8)
Once a month	7.9 (8.8)		5 (5.3)
Less than once a week, but more than once a month	9.9 (11)		9.9 (10.6)
Once a week	21.8 (24.2)		21.8 (23.4)
More than once a week	25.7 (28.6)		28.7 (30.9)
Officer of an organization	23 (24)		17 (18)
Marital satisfaction	93.19 ± (16.06)	47–115	94.27 ± (15.64)	26–116
Demographic characteristics
Age (in years)	67.40 ± (7.24)	51–89	70.03 ± (7.66)	56–90
Race
White	95 (97)		91 (92)
Black	0 (0)		1 (1)
Asian	0 (0)		1 (1)
No primary group	3 (3)		5 (6)
Education
Less than high school	0 (0)		1 (1)
High school	13 (13)		10 (10)
Some college credit	13 (13)		15 (15)
Associate’s degree	8 (8)		7 (7)
Bachelor’s degree	15 (15)		21 (21)
Some graduate school	17 (17)		13 (13)
Professional degree	32 (33)		31 (32)
Household income
<$10,000			3 (3)
$10,000–$39,999			16 (16)
$40,000–$69,999			27 (28)
$70,000–$99,999			21 (21)
$100,000 or more			29 (30)
Employment status
Currently working	63 (62.4)		59 (58.4)
Not working	35 (34.7)		39 (38.6)
Months married			429.44 ± (176.03)	28–743
Have children			76 (78)
Number of children			1.93 ± (1.3)

^*Means and standard deviations are presented for continuous variables.

Data on income, months married, and children were collected for the couple and are consistent across husbands and wives.

Covariates and intraclass correlations

We considered husband’s and wive’s income and whether or not they had children as covariates in our preliminary models because these characteristics were both associated with marital satisfaction and at least one of the social participation variables. Marital satisfaction was negatively associated with having children (t(95) = − 1.95, p = 0.06) and associated with greater income (r(94) = 0.19, p = 0.07). Frequency of participation differed by whether the participant had children or not (t(91) = 2.00, p = 0.05). Whether they participated in organizations or not was also related to having children (χ2(1) = 1.86, p = 0.20).

See Supplemental Tables 1 and 3 for the rest of the non-significant associations with the covariates and Supplemental Table 4 for the non-significant gender difference tests of the variables. The intraclass correlation between each partners’ marital satisfaction scores was 0.37. A post-hoc power analysis was conducted for all four of the models. The range of the power analysis values was from 0.66 to 0.99.

Main hypotheses

Hypotheses 1 through 4 (actor effects): We did not find significant actor effects. Results showed that one’s own social participation did not significantly relate to one’s own marital satisfaction (see Table 2). There was no significant association between the number of social organizations one is a part of and one’s own marital satisfaction. Similarly, frequency of participation or being an acting officer of a club was not significantly associated with one’s own marital satisfaction.

Table 2.

Final actor–partner interdependence models predicting marital satisfaction.

	β	SE	df	t	p
Model 1 (Are you involved in any organizations at the present time?)
Intercept	90.57	3.75	151.47	24.17	0.00
Role (1 = Wife, 2 = Husband)	−1.20	1.80	95.00	−0.66	0.51
Actor social participation	2.00	2.34	188.92	0.86	0.39
Partner social participation	5.45	2.34	188.96	2.33	0.02
Model 2 (How many organizations are you involved with currently?)
Intercept	91.22	3.34	132.05	27.35	0.00
Role (1 = Wife, 2 = Husband)	−1.30	1.67	89.05	−0.78	0.44
Actor social participation	0.47	0.75	169.23	0.63	0.53
Partner social participation	2.40	0.74	160.85	3.26	0.00
Model 3 (Overall, about how often do you participate in some kind of organizational activity?)
Intercept	84.33	4.45	142.79	18.93	0.00
Role (1 = Wife, 2 = Husband)	−0.43	1.80	87.00	−0.24	0.81
Actor social participation	0.24	0.70	171.50	0.34	0.73
Partner social participation	2.73	0.70	172.36	3.88	0.00
Model 4 (Are you currently an officer in any of these organizations?)
Intercept	92.87	3.25	103.17	28.59	0.00
Role (1 = Wife, 2 = Husband)	−1.39	1.89	90.00	−0.73	0.47
Actor social participation	5.31	2.81	162.67	1.89	0.06
Partner social participation	6.91	2.82	165.08	2.45	0.02

Hypothesis 5 through 8 (partner effects): In line with our hypothesis, one spouse’s social participation was significantly associated with the other spouse’s greater marital satisfaction. Specifically, the number of social organizations one is involved with and the frequency of participation in social activities were positively associated with marital satisfaction of the partner. Being an officer of an organization also had a significant association with the partner’s marital satisfaction. Any involvement with an organization was significantly associated with higher marital satisfaction for the partner in the final models.

Our analyses exploring differences in effects for husbands and wives, revealed no significant interactions with role (husband/wife) for actor and partner effects of social participation variables in models predicting marital satisfaction. See Table 3 for the non-significant results of these models.

Table 3.

Final models including gender interaction terms.

	β	SE	df	t	p
Model 1 (Are you involved in any organizations at the present time?)
Intercept	95.06	6.17	104.26	15.41	0.00
Role (1 = Wife, 2 = Husband)	−3.79	3.58	94.00	−1.06	0.29
Actor social participation	−3.37	8.96	118.21	−0.38	0.71
Partner social participation	2.23	8.85	118.14	0.25	0.80
Role * Actor social participation	2.71	5.26	137.29	0.52	0.61
Role * Partner social participation	1.26	5.26	137.31	0.24	0.81
Model 2 (How many organizations are you involved with currently?)
Intercept	90.47	4.98	101.24	18.15	0.00
Role (1 = Wife, 2 = Husband)	−1.24	2.87	88.00	−0.43	0.67
Actor social participation	1.02	3.04	104.37	0.34	0.74
Partner social participation	2.83	2.66	111.90	1.06	0.29
Role * Actor social participation	−0.06	1.72	121.44	−0.03	0.97
Role * Partner social participation	0.02	1.74	119.70	0.01	0.99
Model 3 (Overall, about how often do you participate in some kind of organizational activity?)
Intercept	87.65	8.94	119.53	9.80	0.00
Role (1 = Wife, 2 = Husband)	−0.24	4.73	86.00	−0.05	0.96
Actor social participation	−0.74	2.93	104.43	−0.25	0.80
Partner social participation	1.15	2.98	103.67	0.39	0.70
Role * Actor social participation	−0.23	1.61	121.14	−0.15	0.89
Role * Partner social participation	0.17	1.61	121.37	0.11	0.91
Model 4 (Are you currently an officer in any of these organizations?)
Intercept	91.14	4.71	88.00	19.34	0.00
Role (1 = Wife, 2 = Husband)	−1.24	2.87	88.00	−0.43	0.67
Actor social participation	0.44	2.71	107.43	0.16	0.87
Partner social participation	2.48	2.53	115.63	0.98	0.33
Role * Actor social participation	0.02	1.71	123.52	0.01	0.99
Role * Partner social participation	−0.06	1.72	121.48	−0.03	0.97

Discussion

Results of this study provide evidence to suggest that engaging in more social participation has interpersonal implications for marital satisfaction in middle-aged and older adult couples; however, we did not find evidence for intrapersonal associations. Specifically, we found that one spouse’s greater participation in social organizations was associated with greater marital satisfaction in the other spouse. In addition to the presence or absence of involvement, the intensity and level of involvement, and being an officer of an organization was significantly associated with the partner’s marital satisfaction.

The partner effect findings are in line with the self-expansion model which argues that the partner benefits from their spouse’s experiences and in the context of the present study, the potential knowledge or enrichment gained from participation in social organizations. The findings are also consistent with the idea that people may experience less emotional burden when they know their spouses have other emotional confidants who can meet their needs and they are more socially integrated within the larger community (Monin & Clark, 2011).

Contrary to our hypotheses, we did not find a significant association between one’s own social participation and own marital satisfaction. This lack of significant actor effects aligns with a previous study showing that social participation, measured in hours of volunteer work and contact with friends and neighbors, was not related to one’s own marital satisfaction (Voydanoff, 2005). One’s own social participation may not affect one’s own perception of their marriage. Rather, it may improve one’s own wellness or happiness. As shown by previous studies, spending time in leisure activities after retirement has been shown to increase happiness and fulfillment after retirement (Lewis & Hill, 2020), which may be equally likely in both happy and troubled marriages. For example, some individuals may try to escape spending time with their partner by participating in other social activities; whereas others may be participating in social organizations along with their spouse as a way to spend more time together.

A major strength of the present study is its dyadic focus that accounts for the influence of one’s immediate social environment on an individual’s outcomes. Additionally, the study inquired about multiple aspects of social participation including confirmation of participation, number of organizations one is involved in, frequency of participation, and leadership role within the organization. The different components that make up our definition of social participation allow us to capture the type of social participation as well as the intensity or level of involvement. For instance, we included a measure of whether an individual was an officer of an organization; being an officer indicates a higher level of involvement in the organization which could create more potential for self-expansion.

Despite these strengths there are some limitations. Our sample size was relatively small and may have reduced the study’s statistical power to reject the null hypothesis. Furthermore, our findings may not be generalizable to the larger population of older adults because we restricted our analysis to heterosexual couples living together due to the aims of the parent study (Monin et al., 2019). Additionally, individuals were excluded from the parent study for taking beta blockers as one of the main outcomes of the parent study was heart rate. The sample is also predominately White, educated, and wealthy, making results generalizable to only part of the population. Examining the same hypothesis with a more diverse sample would be more telling of the broader implications of self-expanding activities on the population as a whole. We also recognize the items used to define and capture social participation and leadership in organizations could be improved upon. Questions could be reworded to capture positions in social groups more accurately in terms of whether the individual sees themselves in a leadership role in the different types of organizations (i.e., not just officers). Lastly, due to the cross-sectional design of the study we cannot make any causal inference regarding the association between social participation and marital satisfaction.

It will be important in future research to investigate whether couples are in social organizations together and the extent to which they participate in social activities together or separately. Couples who engage in exciting activities together also report high marital satisfaction, and this is attributed to self-expansion theory (Reissman et al., 1993). Encouraging a spouse to participate in social organizations may also be an important way to help one another thrive (Feeney & Collins, 2015), and measuring encouragement for participating in activities may also help elucidate these findings (Tomlinson et al., 2020). Future research might also consider curvilinear relationships between social participation and marital satisfaction. If a spouse is involved in too many activities leaving little time for the marriage it could negatively impact their partner’s marital satisfaction.

Taken together, this study extended past research that shows social participation has benefits for individuals’ well-being and examined whether social participation also has benefits for marital satisfaction for the self and spouses in midlife and older adult marriages. We found that social participation in church groups, social clubs, or business groups is associated with one’s spouse’s marital satisfaction, but not one’s own marital satisfaction. Next we need to replicate these findings in larger and more diverse samples and examine the mechanisms and processes that lead to interpersonal but not intra-personal benefits. This evidence will be useful for informing interventions to increase relationship satisfaction in midlife and older adult marriages.