The Social Context of Partnered Older Adults’ Insomnia Symptoms

Abstract

The current study included an examination of social factors that mitigate or exacerbate insomnia symptoms among older adults who are married or living with a partner. We first examined the unique effects of spousal support and strain on insomnia symptoms and then evaluated the degree to which extramarital social factors (e.g., friend support) moderated spousal influences. Data came from Waves 2 and 3 of the National Social Life, Health, and Aging Project. Our sample consisted of 495 participants who were either married or cohabitating with a partner (M age in years = 69.84, SD = 8.08). Spousal strain—but not support—predicted higher insomnia symptoms 5 years later. Spousal influences on sleep, however, were moderated by extramarital factors in nuanced ways. Findings highlight the importance of taking into account older adults’ wider social context when examining the ways in which sleep is sensitive to positive and negative aspects of marital quality.

Keywords

sleep insomnia symptoms social context spousal strain spousal support marital quality

Insomnia symptoms increase in prevalence as people get older; 15−20% of adults over the age of 65 experience difficulty initiating and maintaining sleep, or waking up too early in the morning (Kocevska et al., 2021). Insomnia symptoms render older adults more susceptible to both chronic diseases and depression (Jaussent et al., 2011). A growing area of research suggests that older adults’ sleep is sensitive to social influences and that spouses (or cohabitating partners) play an integral role in either promoting or hindering sleep quality (Chen et al., 2015; Kane et al., 2014). Existing research focused on sleep and marital quality, however, often does not take into account the larger social context. The current study therefore included a more comprehensive examination of social factors that mitigate or exacerbate insomnia symptoms among older adults who are partnered (meaning that they are married or living with a partner). We evaluated associations between marital quality and insomnia symptoms within the larger context of older adults’ social lives, thereby also considering the role of family members, friends, and social participation more broadly.

Spousal Influences on Sleep

The marital relationship is thought of as the “primary social relationship” that influences partnered adults’ sleep, especially given the likelihood that marital partners sleep in the same bed (Troxel et al., 2007, p. 398). Troxel et al.'s model of social relationships and sleep therefore focuses primarily on the impact of marital relationships, including the role of both positive (e.g., security) and negative (e.g., conflict) aspects of these relationships. In accordance with this model, spousal support has been shown to facilitate sleep, whereas spousal strain has been shown to interfere with it (Chen et al., 2015; Hasler & Troxel, 2010; Kane et al., 2014).

Research has even shown that changes in both spousal support and spousal strain are coordinated with changes in older adults’ insomnia symptoms (Lee et al., 2017). More specifically, decreases in insomnia symptoms coincided with both increases in spousal support (e.g., spousal understanding) and decreases in spousal strain (e.g., spousal criticism). Further, among older (but not younger) adults, poor marital quality has been shown to predict increased sleep disturbances four years later (Yang et al., 2013). Thus, relational stress that induces negative or intrusive thoughts may interfere with sleep, whereas relational support that equips individuals with a sense of emotional security and companionship may promote sleep (Chen et al., 2015; Holliday & Troxel, 2017).

Extramarital Influences on Sleep

Having both a higher quantity and quality of social network ties has been shown to be beneficial for one's sleep (see Gordon et al., 2021 for a review). Perceived support from family (other than one's spouse) and friends has been shown to promote sleep quality, whereas strain from family and friends has been shown to hinder it within samples of middle-aged and older adults (Chung, 2017; Mesas et al., 2020). These findings highlight the potential for older adults’ sleep to be sensitive to influences from social ties other than their spouse/partner (i.e., extramarital influences). In fact, even older adults’ social participation more broadly—that is their active engagement in social activities—yields benefits for sleep. More specifically, Chen et al. (2016) found that higher social participation was associated with less objective wake after sleep onset. The authors speculated that social participation may, in part, benefit sleep because it fosters a sense of belonging and social integration.

Although social factors outside of marriage may influence sleep, few (if any) studies have examined how they interact with spousal influences. For example, can extramarital support (e.g., friend support) buffer the negative effects of spousal strain on insomnia symptoms? In other words, can other supportive social ties help mitigate the negative effects of strain from spouses on sleep? Although Troxel et al.’s (2007) model is focused primarily on spousal influences on sleep, it also acknowledges the potential for factors external to this primary social relationship to influence “for whom such links are most likely to be evident” (p. 391). Thus, other social stressors (e.g., strain from family) and resources (e.g., support from friends) may render individuals more or less vulnerable to the effects of marital quality on sleep. Although this possibility has yet to be explored when examining sleep-related outcomes, other research has shown that perceived network support from friends buffered negative effects of marital conflict on daily diurnal cortisol slopes (Keneski et al., 2018). Such findings demonstrate the potential for extramarital relationships to moderate spousal influences on health.

The Current Study

In being guided by Troxel et al.’s (2007) model, we first evaluated the unique influences of spousal support and strain on insomnia symptoms. We hypothesized that, even after taking into account extramarital supports and strains, spousal support would be associated with lower subsequent insomnia symptoms (Hypothesis 1), whereas spousal strain would be associated with higher subsequent insomnia symptoms (Hypothesis 2). We then examined the potential for extramarital variables to moderate spousal influences on sleep. We expected extramarital supports (i.e., perceived support from friends and family, higher number of friends and close relatives, and social participation) to buffer the negative effects of spousal strain on insomnia symptoms (Hypothesis 3). We expected extramarital strains (i.e., strain from friends and family) to exacerbate the negative effects of spousal strain on insomnia symptoms (Hypothesis 4). Finally, we explored whether extramarital variables moderated the effects of spousal support (as opposed to strain) on insomnia symptoms (research questions 1 and 2). Due to a lack of existing research, we did not make hypotheses about how extramarital variables may moderate the effects of spousal support on insomnia symptoms.

Method

Participants and Procedure

Data came from Wave 2 (2010−2011) and Wave 3 (2015–2016) of the National Social Life, Health, and Aging Project (NSHAP). The NSHAP is a nationally representative study of community-dwelling older adults intended to aid the understanding of social factors that contribute to later-life health and well-being. We excluded data from Wave 1 (2005−2006) because insomnia symptoms were added as part of a sleep module in Wave 2 (see Lauderdale et al., 2014 for more details). Our sample consisted of Wave 2 respondents who indicated that they were married or living with a partner and completed the sleep module (N = 495). Half (49.7%) of participants identified as male and the other half (50.3%) identified as female. Participants were, on average, 69.84 (SD = 8.08) years old. The majority (95.8%) of participants were married and the remainder (4.2%) were living with a partner (we therefore use the term spouse). Most (83.8%, n = 415) participants held a high school degree or its equivalent. Of these participants, 122 also held a bachelor's degree (or higher). Most (75.9%) participants identified as White, whereas 8.5% identified as Black, 11.7% identified as Hispanic non-Black, and 3.8% identified as other. The majority of the sample (68.5%) was retired. The median household income was $50,000 (SD = $72,193.50).

There was some attrition in our sample; 27.3% of respondents (n = 135) who participated in Wave 2 did not have valid insomnia symptoms data in Wave 3. A series of independent samples t-tests conducted with Wave 2 data revealed that participants who dropped out of the study between waves were older (t[207.55] = 6.85, p < .001) and had more chronic health conditions (t[181.08] = 3.33, p = .001) than those who stayed in the study. Further, participants who dropped out reported lower levels of social participation (t[493] = −2.94, p = .003) and family strain (t[491] = −2.02, p = .044) than those who stayed in the study.

Measures

Support and Strain

The support and strain items were originally derived from Schuster et al.’s (1990) scale of supportive and negative interactions. Two questions measured spousal support: (1) How often can you open up to your spouse/partner if you need to talk about your worries and (2) How often can you rely on your spouse/partner for help if you have a problem? Response options ranged from 0 (never) to 3 (often). In accordance with other studies (e.g., Hsieh & Hawkley, 2018), we combined the two items into a mean scale (α = .53). The same set of items was repeated for perceived support from family (other than one's spouse) and from friends. We created parallel mean scales of family support (α = .60) and friend support (α = .76) in Wave 2.

Two questions measured spousal strain: (1) How often does your spouse/partner make too many demands on you and (2) How often does your spouse/partner criticize you? Response options ranged from 0 (never) to 3 (often). In accordance with other studies (e.g., Hsieh & Hawkley, 2018), we combined the two items into a mean scale (α = 0.61). The same set of items was repeated for perceived strain from family and from friends. We created parallel mean scales of family strain (α = 0.55) and friend strain (α = 0.57) in Wave 2 (see Table 1).

Table 1.
Descriptive Statistics and Correlations.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16

1. Wave 2 insomnia symptoms –

2. Wave 3 insomnia symptoms .54* –

3. Wave 2 spousal support −.07 −.03 –

4. Wave 2 spousal strain .02 .05 −.24* –

5. Wave 2 number of friends −.05 −.11* .10* −.002 –

6. Wave 2 number of relatives −.05 −.10 .02 −.04 .33*** –

7. Wave 2 friend support −.02 −.10 .11* −.08 .34* .20* –

8. Wave 2 friend strain .04 −.03 .01 .23* .12 −.05 .14 –

9. Wave 2 family support −.05 −.05 .16* −.02 .16* .33* .33* .01 –

10. Wave 2 family strain .08 .02 −.06 .33* .04 .02 .10* .50* .01 –

11. Wave 2 social participation −.03 −.16 .01 .01 .27* .13 .16* .12 .19*** .09 −

12. Gender (% female) .05 .07 −.07 −.10* .01 .13 .20 −.06 .27* .08 .07 −

13. Age .03 .01 −.05 −.06 .02 .02 −.04 −.12 −.06 −.11* .04 −.23** –

14. Sleep with partner −.06 −.09 .11* −.13 .06 .13 −.03 −.10* .08 −.07 .08 .02 −.04 –

15. Sleep medications (% no) .22* .19* −.02 −.01 .04 −.06 .11* .04 −.05 .02 −.02 .11* −.02 −.01 −

16. Comorbidity index .15*** .12* −.11* .03 −.06 .05 −.03 −.01 −.07 −.04 −.08 −.04 .19** −.02 .10* −

M (%) 2.39 2.14 2.72 1.25 3.40 2.99 2.02 .52 2.40 .78 2.04 49.7 69.84 3.00 81.2 2.57

SD 1.51 1.69 .45 .80 1.22 1.05 .83 .54 .67 .66 1.50 − 8.08 1.54 − 2.01

N 495 360 495 495 488 493 488 488 493 493 495 495 495 409 495 495

Observed range 0–6 0–6 0– 3 0– 3 0–5 0–5 0–3 0–2.5 0–3 0–3 0–5 – 36–91 0–4 – 0–12

Note. p < .05; p < .01; p < .001.

Number of Friends and Close Relatives

In Wave 2, participants indicated how many friends they have in accordance with the following response options: 0 (none), 1 (one), 2 (two–three), 3 (four–nine), 4 (10–20), and 5 (more than 20). Participants also indicated how many family members/relatives they feel close to according to the same set of response options (see Table 1).

Social Participation

Social participation was assessed by three items from Wave 2 that asked participants how frequently they did each of the following in the past 12 months: (1) attended religious services, (2) did volunteer work, and (3) attended meetings of an organized group. Response options included: 0 (never), 1 (once/twice a year), 2 (several times a year), 3 (once a month), 4 (every week), and 5 (several times a week). Research indicates that these items load onto a common factor within the NSHAP data (Chen et al., 2016). We therefore created a mean scale of social participation (see Table 1).

Insomnia Symptoms

Self-reported insomnia symptoms were measured in Waves 2 and 3. The three questions included in both waves were how often do you have trouble: (1) falling asleep, (2) waking up during the night, and (3) waking up too early and not being able to fall asleep again? Response options included: 0 (rarely/never), 1 (sometimes), and 2 (most of the time). In accordance with existing research (Chen et al., 2016), we combined these items into a sum scale that ranged from 0 to 6 (Wave 2 α = .52; Wave 3 α = .68; see Table 1).

Covariates

Older age, gender, taking sleep medications, and having more than one chronic disease are all significant risk factors for symptoms of insomnia (Jaussent et al., 2011). These variables were therefore included as baseline covariates. The majority (81.2%) of participants indicated that they did not take any medications or treatments in the last two weeks to help them sleep. To operationalize chronic disease, we calculated the NSHAP Comorbidity Index (NCI; Vasilopoulos et al., 2014). The NCI is a weighted sum of 15 different conditions (e.g., diabetes) and has a possible range of 0 to 21 points. The mean reported in our sample was 2.57 (SD = 2.01). Finally, we included how often participants sleep in the same bed as their partner as a covariate given that cosleeping may have negative implications for one's sleep quality (see Richter et al., 2016 for a review). This variable was operationalized from a single item that asked participants how often they sleep in the same bed as their partner (in Wave 2). Response options ranged from 0 (never) to 4 (all of the time). More than half (65%) of the sample reported sleeping in the same bed as their partner all of the time.

Data Analysis

Analyses were conducted using multiple regression models with complex survey and sampling designs in Mplus V8.4 (Muthén & Muthén, 2017). Full information maximum likelihood estimation was used to handle missing data, which has been identified as an appropriate technique for handling missing longitudinal data within NSHAP (Santini et al., 2020). All analyses took into account the clustering and stratification of NSHAP's sample design. NSHAP-supplied sampling weights from Wave 2 were used to account for differential probabilities of selection.

To test that spousal support (Hypothesis 1) and spousal strain (Hypothesis 2) would each predict subsequent insomnia symptoms, we estimated three models. Model 1 controlled only for insomnia symptoms in Wave 2 to account for stability in insomnia symptoms over time. In Model 2, demographics and sleep- and health-related variables were added as covariates. In Model 3, the extramarital supports (e.g., support from friends) and extramarital strains (e.g., strain from friends) were added as additional covariates.

To test Hypothesis 3, that extramarital supports would buffer the negative effect of spousal strain on subsequent insomnia symptoms, we added interaction terms between each of the extramarital support variables and spousal strain to Model 3. To test Hypothesis 4, that extramarital strains would exacerbate the negative effect of spousal strain on subsequent insomnia symptoms, we added interactions between each extramarital strain variable and spousal strain to Model 3. Each interaction was examined in a separate model. For each significant interaction effect (p < .05), additional analyses were conducted to test the region of significance. Finally, to examine whether extramarital variables moderate the effects of spousal support (as opposed to strain) on insomnia symptoms (research questions 1 and 2), we conducted a parallel set of analyses where we tested interactions between each extramarital variable and spousal support. Each interaction was added (one by one) to Model 3.

Results

We first examined bivariate correlations among all study variables (see Table 1). To test our first two hypotheses, we then estimated a series of three main effects models. Model 1 included only insomnia symptoms in Wave 2 as a covariate. In Model 2, we added demographic and sleep/health-related covariates. Finally, in Model 3, we added the extramarital supports and strains. The R² values for each of these models are reported in Table 2. We did not find support for Hypothesis 1; spousal support was not associated with subsequent insomnia symptoms (see Model 1, Table 2). Consistent with Hypothesis 2, however, spousal strain was positively associated with subsequent insomnia symptoms. This main effect remained significant upon adding additional covariates in Models 2 and 3 (see Table 2).

Table 2.
Main Effects of Spousal Support and Strain on Insomnia Symptoms (N = 495).

Model 1 Model 2 Model 3

Est. SE Est. SE Est. SE

Spousal support .075 .153 .174 .150 .213 .153

Spousal strain .176 .088* .184 .091* .225 .091*

Baseline insomnia symptoms .615 .057* .581 .063* .584 .058***

Gender .248 .148 .323 .134*

Age .008 .015 .010 .013

Sleep with partner −.045 .064 −.037 .060

Sleep medications .256 .213 .318 .191

Comorbidity index .096 .053 .078 .053

Number of friends −.121 .084

Friend support −.099 .104

Friend strain −.075 .150

Number of close relatives −.022 .084

Family support .074 .128

Family strain −.148 .108

Social participation −.117 .059*

R ² .337 .058* .359 .056* .396 .050***

Note. All predictors were measured in wave 2, whereas the dependent variable was measured in wave 3; unstandardized estimates are shown; *p < .05; ***p < .001.

Extramarital Variables Moderating Effects of Spousal Strain

To test Hypothesis 3, we examined whether extramarital supports buffered the negative effect of spousal strain on subsequent insomnia symptoms. The only significant interaction we detected was between spousal strain and social participation (see Model 1, Table 3). However, contrary to our hypothesis, social participation exacerbated (rather than buffered) the negative effect of spousal strain on insomnia symptoms. Although the main effect of social participation indicates that more participation is associated with fewer insomnia symptoms, the interaction with spousal strain suggests that high social participation coupled with high spousal strain is associated with more insomnia symptoms. This pattern is illustrated in Figure 1 which depicts the slope of spousal strain predicting insomnia symptoms 5 years later across different levels of social participation (in Wave 2). The positive slope indicates that as spousal strain increases, so do insomnia symptoms. However, this association only reaches statistical significance when social participation is close to or above the mean (the gray-shaded area of Figure 1—or the portion of the slope where the confidence bands do not include zero).

Figure 1.
Moderating role of social participation on the effect of spousal strain on insomnia symptoms. Note. The center black line depicts the magnitude of the association between spousal strain in Wave 2 and insomnia symptoms in Wave 3 across different levels of the mean-centered moderating variable of social participation in Wave 2 (i.e., simple slopes). The shaded gray area indicates at which levels of social participation there is a statistically significant association between spousal strain and insomnia symptoms (i.e., where the confidence bands do not include zero). Spousal strain is therefore positively associated with subsequent insomnia symptoms at moderate and high (but not low) levels of social participation.

Table 3.
Moderating Role of Social Participation on the Effect of Spousal Strain on Insomnia Symptoms (N = 495).

Model 1 Model 2 Model 3 Model 4

Est. SE Est. SE Est. SE Est. SE

Spousal support .203 .148 .199 .154 .187 .151 .208 .143

Spousal strain −.018 .114 .064 .111 .043 .097 −.019 .107

Baseline insomnia symptoms .573 .060* .568 .061* .578 .060* .580 .059*

Gender .322 .131* .333 .131* .319 .129* .327 .129*

Age .011 .013 .012 .013 .008 .014 .010 .014

Sleep with partner −.029 .060 −.043 .061 −.030 .059 −.032 .060

Sleep medications .342 .185 .349 .186 .350 .187 .310 .185

Comorbidity index .071 .051 .070 .052 .077 .050 .072 .049

Number of friends −.124 .082 −.145 .085 −.129 .080 −.135 .080

Friend support −.098 .104 −.097 .101 −.102 .109 −.099 .112

Friend strain −.111 .150 −.093 .153 −.127 .154 −.091 .149

Number of close relatives −.007 .083 .002 .085 −.006 .082 .002 .081

Family support .079 .131 .067 .128 .055 .136 .055 .133

Family strain −.161 .101 −.165 .105 −.128 .095 −.180 .100

Social participation −.248 .086**

Religious services −.167 .078*

Volunteering −.188 .084*

Group meetings −.216 .076**

Social participation × spousal strain .118 .056*

Religious services × spousal strain .067 .050

Volunteering × spousal strain .109 .055*

Group meetings × spousal strain .126 .054*

R ² .400 .050* .397 .051* .398 .050* .400 .050*

Note. Unstandardized estimates are shown; Model 1 included the composite social participation variable; Models 2–4 included one social participation item (e.g., religious services in Model 2); p < .05; p < .01; p < .001.

To better understand this counterintuitive finding, we conducted a post-hoc analysis wherein we separately examined the moderating role of each item in the social participation scale. This approach is consistent with that of other studies that have utilized these data (e.g., Chen et al., 2016). These follow-up analyses are presented in Table 3. Models 2 through 4 each test the main effect of one of the three items included in the composite social participation scale, as well as its corresponding interaction with spousal strain. Each type of social participation was significantly negatively associated with subsequent insomnia symptoms. However, we only detected significant interactions between spousal strain and volunteering and attending social group meetings, but not religious services. The patterns of both of these interactions mimicked that of the interaction between spousal strain and the composite social participation variable. That is, high spousal strain coupled with moderate–high volunteer work and social meetings predicted higher subsequent levels of insomnia symptoms. Finally, to test Hypothesis 4, we examined whether extramarital strains exacerbated the negative effect of spousal strain on subsequent insomnia symptoms. We did not find support for Hypothesis 4; neither family strain nor friend strain significantly moderated the effect of spousal strain on insomnia symptoms.

Extramarital Variables Moderating Effects of Spousal Support

Lastly, we examined whether extramarital variables moderated the effect of spousal support (as opposed to spousal strain) on insomnia symptoms. The only significant interaction we detected was between spousal support and family support (unstandardized estimate = .454, p = .033). The pattern of the interaction suggests that high family support coupled with high spousal support is associated with higher subsequent insomnia symptoms (see Figure 2).

Figure 2.
Moderating role of family support on the effect of spousal support on insomnia symptoms. Note. The center black line depicts the magnitude of the association between spousal support in Wave 2 and insomnia symptoms in Wave 3 across different levels of the mean-centered moderating variable of family support in Wave 2 (i.e., simple slopes). The shaded gray area indicates at which levels of family support there is a statistically significant association between spousal support and insomnia symptoms (i.e., where the confidence bands do not include zero). Spousal support is therefore positively associated with subsequent insomnia symptoms at high (but not moderate or low) levels of family support.

Discussion

In line with Troxel et al.’s (2007) model of social relationships and sleep, findings from the current study suggest that spouses can exert a considerable influence on older adults’ insomnia symptoms, even after taking into account the role of their larger social context. In accordance with existing research (i.e., Chen et al., 2015), we found that spousal strain—but not spousal support—yielded a significant main effect on older adults’ insomnia symptoms. This may be because spousal strain promotes feelings of loneliness among partnered older adults (Marini et al., 2020), which is predictive of insomnia symptoms (Shankar, 2020).

The current study builds on existing research by examining spousal influences on sleep in the context of extramarital social factors. In doing so, we found that the effect of spousal strain on insomnia symptoms was qualified by individuals’ level of social participation. That is, spousal strain predicted higher subsequent insomnia symptoms in the context of moderate and high (but not low) social participation. This finding was contrary to our hypothesis that social participation would actually buffer the negative effect of spousal strain on insomnia symptoms. We speculated that social participation would afford a sense of belonging that would offset the negative effect of spousal strain on sleep. Our rationale was consistent with that of Lauderdale and Chen (2017) who expected social participation to be more beneficial for sleep among unmarried (compared to married) older adults in Wave 2 of the NSHAP data. They thought that social involvement would “substitute for the beneficial effects of marriage” (p. 112). Their hypothesis, however, was also not supported (marital status and social participation did not interact to predict sleep).

These findings beg the question of whether one spouse's social participation (and specifically attendance in social groups and volunteer work) is in and of itself a source of tension that either gives rise to—or exacerbates—spousal strain. Future research therefore ought to examine whether partnered older adults’ participation in social activities and groups is a joint or separate activity, and how spouses feel about each other's social participation (e.g., they see it as beneficial for their spouse versus they see it as consuming all of their spouse's time and energy). If an individual's spouse is critical of their social participation, but at the same time it brings them joy, they may experience dissonant and intrusive thoughts that disrupt their ability to fall and stay asleep (Morin et al., 2003).

Although there was no main effect of spousal support, our findings did suggest that high spousal support coupled with high (but not low or moderate) family support was associated with higher insomnia symptoms five years later. This finding may suggest that older adults who receive more support from both spouses and family are simply coping with higher levels of stress, which hinders their sleep quality. However, the measures of support utilized in the current study captured older adults’ perceived support availability, rather than received support. Thus, older adults who perceived that they could turn to both their spouse and other family members for support were not necessarily in need of higher support.

Future research is needed to replicate this unexpected finding, however, it begs the question of whether having multiple sources of support (i.e., spouses versus other family members) actually increases the likelihood of receiving conflicting advice, which in turn causes older adults to experience dissonant and intrusive thoughts that disrupt sleep (Morin et al., 2003). Conflicting advice from spouses versus other family members may also become a source of strain within the marriage that disrupts sleep. In fact, recent research conducted with NSHAP data has found that older adults experience increased anxiety as a result of their spouse's declining physical health, but only when their spouse discusses health-related concerns with other confidants in their social network in addition to their spouse (Martire et al., 2022). Other research utilizing NSHAP data has also found that older adults report lower levels of individual well-being when their spouse reports more frequent communication with members of their social network (Ermer & Proulx, 2020). These findings highlight the need to further examine older adults’ motivations for turning to other social ties (e.g., because they are questioning their spouse's advice, because they do not want to overburden their spouse), as well as the consequences of doing so for marital well-being (e.g., it threatens the spouse, it facilitates spousal communication) and for related health outcomes, including insomnia symptoms.

Limitations and Future Research Directions

Limitations of the current study pave the way for future research. First, some of our measures had low reliability—particularly support and strain—which was likely a function of there being only two items on each scale (Eisinga et al., 2013). Future research should utilize more reliable measures of specific positive (e.g., responsiveness) and negative (e.g., criticism) aspects of social relationships that may be particularly important predictors of sleep quality. Second, our measure of sleep relied solely on subjective (as opposed to objective) ratings of insomnia symptoms. Subjective and objective sleep data are only modestly correlated among older adults, thus suggesting that they tap into distinct aspects of sleep in later life (Landry et al., 2015). Future research is therefore needed to determine if findings from the current study generalize to objective markers of sleep quality. Although NSHAP collected objective sleep data via wrist actigraphy on a subset of participants, they did so in Wave 2 only, which precluded our ability to predict change in objective sleep from Wave 2 to Wave 3 as we did for insomnia symptoms. Third and finally, we evaluated social factors as predictors of change in insomnia symptoms across a 5-year period. Future research ought to zoom in and reexamine these associations at the daily level to determine how generalizable results are across timescales.

Conclusion

By taking into account the larger social landscape of older adults’ lives, we were able to identify nuanced linkages between positive and negative aspects of marital quality and insomnia symptoms. Findings suggest that extramarital social factors have the potential to modify effects of spousal support and strain on older adults’ insomnia symptoms. Older adults’ social participation, for example, appears to act as a risk factor that amplifies the negative effect of spousal strain on insomnia symptoms. Findings highlight the importance of examining marital influences on sleep within a broader social context.

	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16
1. Wave 2 insomnia symptoms	–
2. Wave 3 insomnia symptoms	.54***	–
3. Wave 2 spousal support	−.07	−.03	–
4. Wave 2 spousal strain	.02	.05	−.24***	–
5. Wave 2 number of friends	−.05	−.11*	.10*	−.002	–
6. Wave 2 number of relatives	−.05	−.10	.02	−.04	.33***	–
7. Wave 2 friend support	−.02	−.10	.11*	−.08	.34***	.20***	–
8. Wave 2 friend strain	.04	−.03	.01	.23***	.12**	−.05	.14**	–
9. Wave 2 family support	−.05	−.05	.16***	−.02	.16***	.33***	.33***	.01	–
10. Wave 2 family strain	.08	.02	−.06	.33***	.04	.02	.10*	.50***	.01	–
11. Wave 2 social participation	−.03	−.16**	.01	.01	.27***	.13**	.16***	.12**	.19***	.09	−
12. Gender (% female)	.05	.07	−.07	−.10*	.01	.13**	.20**	−.06	.27***	.08	.07	−
13. Age	.03	.01	−.05	−.06	.02	.02	−.04	−.12**	−.06	−.11*	.04	−.23**	–
14. Sleep with partner	−.06	−.09	.11*	−.13**	.06	.13**	−.03	−.10*	.08	−.07	.08	.02	−.04	–
15. Sleep medications (% no)	.22***	.19***	−.02	−.01	.04	−.06	.11*	.04	−.05	.02	−.02	.11*	−.02	−.01	−
16. Comorbidity index	.15***	.12*	−.11*	.03	−.06	.05	−.03	−.01	−.07	−.04	−.08	−.04	.19**	−.02	.10*	−
M (%)	2.39	2.14	2.72	1.25	3.40	2.99	2.02	.52	2.40	.78	2.04	49.7	69.84	3.00	81.2	2.57
SD	1.51	1.69	.45	.80	1.22	1.05	.83	.54	.67	.66	1.50	−	8.08	1.54	−	2.01
N	495	360	495	495	488	493	488	488	493	493	495	495	495	409	495	495
Observed range	0–6	0–6	0– 3	0– 3	0–5	0–5	0–3	0–2.5	0–3	0–3	0–5	–	36–91	0–4	–	0–12

	Model 1	Model 2	Model 3
Spousal support	.075	.153	.174	.150	.213	.153
Spousal strain	.176	.088*	.184	.091*	.225	.091*
Baseline insomnia symptoms	.615	.057***	.581	.063***	.584	.058***
Gender			.248	.148	.323	.134*
Age			.008	.015	.010	.013
Sleep with partner			−.045	.064	−.037	.060
Sleep medications			.256	.213	.318	.191
Comorbidity index			.096	.053	.078	.053
Number of friends					−.121	.084
Friend support					−.099	.104
Friend strain					−.075	.150
Number of close relatives					−.022	.084
Family support					.074	.128
Family strain					−.148	.108
Social participation					−.117	.059*
R ²	.337	.058***	.359	.056***	.396	.050***

	Model 1	Model 2	Model 3	Model 4
Spousal support	.203	.148	.199	.154	.187	.151	.208	.143
Spousal strain	−.018	.114	.064	.111	.043	.097	−.019	.107
Baseline insomnia symptoms	.573	.060***	.568	.061***	.578	.060***	.580	.059***
Gender	.322	.131*	.333	.131*	.319	.129*	.327	.129*
Age	.011	.013	.012	.013	.008	.014	.010	.014
Sleep with partner	−.029	.060	−.043	.061	−.030	.059	−.032	.060
Sleep medications	.342	.185	.349	.186	.350	.187	.310	.185
Comorbidity index	.071	.051	.070	.052	.077	.050	.072	.049
Number of friends	−.124	.082	−.145	.085	−.129	.080	−.135	.080
Friend support	−.098	.104	−.097	.101	−.102	.109	−.099	.112
Friend strain	−.111	.150	−.093	.153	−.127	.154	−.091	.149
Number of close relatives	−.007	.083	.002	.085	−.006	.082	.002	.081
Family support	.079	.131	.067	.128	.055	.136	.055	.133
Family strain	−.161	.101	−.165	.105	−.128	.095	−.180	.100
Social participation	−.248	.086**
Religious services			−.167	.078*
Volunteering					−.188	.084*
Group meetings							−.216	.076**
Social participation × spousal strain	.118	.056*
Religious services × spousal strain			.067	.050
Volunteering × spousal strain					.109	.055*
Group meetings × spousal strain							.126	.054*
R ²	.400	.050***	.397	.051***	.398	.050***	.400	.050***

Footnotes

Declaration of Conflicting Interests

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

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This work was supported by the National Institute on Aging (grant numbers R01 AG063241 and R03 AG064360).

ORCID iD

Christina M. Marini

Author Biographies

Christina M. Marini is an assistant professor in the Gordon F. Derner School of Psychology at Adelphi University. Her research focuses on links between social relationships and health across the adult lifespan. She is an affiliated scientist of the Military Family Research Institute at Purdue University and a faculty affiliate of the Center for Healthy Aging at the Pennsylvania State University. She is also an external collaborator of the Stress, Health, and Aging Research Program at the National Center for PTSD, VA Boston Healthcare System.

Ruixue Zhaoyang has expertise and extensive experience in analyzing longitudinal and intensive repeated measurement data (e.g., daily diary, EMAs, complex longitudinal surveys). Her prior work identifies patterns of change in different features of daily social interactions over time and their influences on physical health.

Lynn M. Martire is professor of Human Development and Family Studies, and an affiliate of the Center for Healthy Aging, at The Pennsylvania State University. Her research examines the influence of close relationships on adults’ health and well-being, and the effects of chronic illness on the caregiving spouse.

Orfeu M. Buxton, PhD, serves as the Elizabeth Fenton Susman Professor of Biobehavioral Health, as a co-funded faculty affiliate of the Social Science Research Institute, and Associate Director of the Clinical and Translational Science Institute, Pennsylvania State University. He directs the Sleep, Health & Society Collaboratory at Penn State. His research primarily focuses on the causes and consequences of chronic sleep deficiency in the workplace, home, and society, including the physiologic and social mechanisms by which these outcomes arise. He is the Editor in Chief of Sleep Health.

References

Chen

J. H.

Lauderdale

D. S.

Waite

L. J

. (2016). Social participation and older adults’ sleep. Social Science & Medicine, 149, 164–173. https://doi.org/10.1016/j.socscimed.2015.11.045

Chen

J. H.

Waite

L. J.

Lauderdale

D. S

. (2015). Marriage, relationship quality, and sleep among US older adults. Journal of Health and Social Behavior, 56(3), 356–377. https://doi.org/10.1177/0022146515594631

Chung

. (2017). Social support, social strain, sleep quality, and actigraphic sleep characteristics: Evidence from a national survey of US adults. Sleep Health, 3(1), 22–27. https://doi.org/10.1016/j.sleh.2016.10.003

Eisinga

Grotenhuis

M. T.

Pelzer

. (2013). The reliability of a two-item scale: Pearson, Cronbach, or Spearman–Brown? International Journal of Public Health, 58, 637–642. https://doi.org/10.1007/s00038-012-0416-3

Ermer

A. E.

Proulx

C. M

. (2020). Social support and well-being among older adult married couples: A dyadic perspective. Journal of Social and Personal Relationships, 37(4), 1073–1091. https://doi.org/10.1177/0265407519886350

Gordon

A. M.

Carrillo

Barnes

C. M

. (2021). Sleep and social relationships in healthy populations: A systematic review. Sleep Medicine Reviews, 57, 101428. https://doi.org/10.1016/j.smrv.2021.101428

Hasler

B. P.

Troxel

W. M

. (2010). Couples’ nighttime sleep efficiency and concordance: Evidence for bidirectional associations with daytime relationship functioning. Psychosomatic Medicine, 72(8), 794–801. https://doi.org/10.1097/PSY.0b013e3181ecd08a

Holliday

S. B.

Troxel

(2017). Relationship quality: Implications for sleep quality and sleep disorders. In McHale

S. M.

King

Buxton

O. M.

(Eds.), Family contexts of sleep and health across the life course (pp. 53–84). Springer.

Hsieh

Hawkley

. (2018). Loneliness in the older adult marriage: Associations with dyadic aversion, indifference, and ambivalence. Journal of Social and Personal Relationships, 35(10), 1319–1339. https://doi.org/10.1177/0265407517712480

10.

Jaussent

Dauvilliers

Ancelin

M. L.

Dartigues

J. F.

Tavernier

Touchon

., & Besset, A. (2011). Insomnia symptoms in older adults: Associated factors and gender differences. The American Journal of Geriatric Psychiatry, 19(1), 88–97. https://doi.org/10.1097/JGP.0b013e3181e049b6

11.

Kane

H. S.

Slatcher

R. B.

Reynolds

B. M.

Repetti

R. L.

Robles

T. F

. (2014). Daily self-disclosure and sleep in couples. Health Psychology, 33(8), 813–822. https://doi.org/10.1037/hea0000077

12.

Keneski

Neff

L. A.

Loving

T. J

. (2018). The importance of a few good friends: Perceived network support moderates the association between daily marital conflict and diurnal cortisol. Social Psychological and Personality Science, 9(8), 962–971. https://doi.org/10.1177/1948550617731499

13.

Kocevska

Lysen

T. S.

Dotinga

Koopman-Verhoeff

M. E.

Luijk

M. P.

Antypa

, & Tiemeier, H. (2021). Sleep characteristics across the lifespan in 1.1 million people from the Netherlands, United Kingdom and United States: A systematic review and meta-analysis. Nature Human Behaviour, 5, 113–122. https://doi.org/10.1038/s41562-020-00965-x

14.

Landry

G. J.

Best

J. R.

Liu-Ambrose

. (2015). Measuring sleep quality in older adults: A comparison using subjective and objective methods. Frontiers in Aging Neuroscience, 7, 166. https://doi.org/10.3389/fnagi.2015.00166

15.

Lauderdale

D. S.

Chen

J. H.

(2017). Sleep, marriage and social engagement among older adults. In McHale

S. M.

King

Buxton

O. M.

(Eds.), Family contexts of sleep and health across the life course (pp. 105–120). Springer International Publishing.

16.

Lauderdale

D. S.

Philip Schumm

Kurina

L. M.

McClintock

Thisted

R. A.

Chen

J. H.

Waite

. (2014). Assessment of sleep in the national social life, health, and aging project. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 69(Suppl_2), S125–S133. https://doi.org/10.1093/geronb/gbu092

17.

Lee

J. H.

Chopik

W. J.

Schiamberg

L. B

. (2017). Longitudinal associations between marital quality and sleep quality in older adulthood. Journal of Behavioral Medicine, 40, 821–831. https://doi.org/10.1007/s10865-017-9850-2

18.

Marini

C. M.

Ermer

A. E.

Fiori

K. L.

Rauer

A. J.

Proulx

C. M

. (2020). Marital quality, loneliness, and depressive symptoms later in life: The moderating role of self and spousal functional limitations. Research in Human Development, 17(4), 211–234. https://doi.org/10.1080/15427609.2020.1837598

19.

Martire

L. M.

Zhaoyang

Marini

C. M.

Nah

. (2022). Dyadic links between health changes and well-being: The role of non-spousal confidants. Journal of Social and Personal Relationships, 39(8), 2617–2638. https://doi.org/10.1177/02654075221086509

20.

Mesas

A. E.

Peppard

P. E.

Hale

Friedman

E. M.

Nieto

F. J.

Hagen

E. W

. (2020). Individuals’ perceptions of social support from family and friends are associated with lower risk of sleep complaints and short sleep duration. Sleep Health, 6(1), 110–116. https://doi.org/10.1016/j.sleh.2019.08.013

21.

Morin

C. M.

Rodrigue

Ivers

. (2003). Role of stress, arousal, and coping skills in primary insomnia. Psychosomatic Medicine, 65(2), 259–267. https://doi.org/10.1097/01.PSY.0000030391.09558.A3

22.

Muthén

L. K.

Muthén

(2017). Mplus user’s guide: Statistical analysis with latent variables, user’s guide. Muthén & Muthén.

23.

Richter

Adam

Geiss

Peter

Niklewski

. (2016). Two in a bed: The influence of couple sleeping and chronotypes on relationship and sleep. An overview. Chronobiology International, 33(10), 1464–1472. https://doi.org/10.1080/07420528.2016.1220388

24.

Santini

Z. I.

Jose

P. E.

Cornwell

E. Y.

Koyanagi

Nielsen

Hinrichsen

., & Koushede, V. (2020). Social disconnectedness, perceived isolation, and symptoms of depression and anxiety among older Americans (NSHAP): A longitudinal mediation analysis. The Lancet Public Health, 5(1), e62–e70. https://doi.org/10.1016/S2468-2667(19)30230-0

25.

Schuster

T. L.

Kessler

R. C.

Aseltine

R. H

. (1990). Supportive interactions, negative interactions, and depressed mood. American Journal of Community Psychology, 18(3), 423–438. https://doi.org/10.1007/BF00938116

26.

Shankar

. (2020). Loneliness and sleep in older adults. Social Psychiatry and Psychiatric Epidemiology, 55, 269–272. https://doi.org/10.1007/s00127-019-01805-8

27.

Troxel

W. M.

Robles

T. F.

Hall

Buysse

D. J

. (2007). Marital quality and the marital bed: Examining the covariation between relationship quality and sleep. Sleep Medicine Reviews, 11(5), 389–404. https://doi.org/10.1016/j.smrv.2007.05.002

28.

Vasilopoulos

Kotwal

Huisingh-Scheetz

M. J.

Waite

L. J.

McClintock

M. K.

Dale

. (2014). Comorbidity and chronic conditions in the National Social Life, Health and Aging Project (NSHAP), wave 2. Journals of Gerontology Series B: Psychological Sciences and Social Sciences, 69(Suppl 2), S154–S165. https://doi.org/10.1093/geronb/gbu025

29.

Yang

H. C.

Suh

Kim

Cho

E. R.

Lee

S. K.

Shin

. (2013). Testing bidirectional relationships between marital quality and sleep disturbances: A 4-year follow-up study in a Korean cohort. Journal of Psychosomatic Research, 74(5), 401–406. https://doi.org/10.1016/j.jpsychores.2013.01.005

	Model 1		Model 2		Model 3
	Est.	SE	Est.	SE	Est.	SE
Spousal support	.075	.153	.174	.150	.213	.153
Spousal strain	.176	.088*	.184	.091*	.225	.091*
Baseline insomnia symptoms	.615	.057***	.581	.063***	.584	.058***
Gender			.248	.148	.323	.134*
Age			.008	.015	.010	.013
Sleep with partner			−.045	.064	−.037	.060
Sleep medications			.256	.213	.318	.191
Comorbidity index			.096	.053	.078	.053
Number of friends					−.121	.084
Friend support					−.099	.104
Friend strain					−.075	.150
Number of close relatives					−.022	.084
Family support					.074	.128
Family strain					−.148	.108
Social participation					−.117	.059*
R ²	.337	.058***	.359	.056***	.396	.050***