Skip to main content
NCBI home page
The Journals of Gerontology Series B: Psychological Sciences and Social Sciences logoLink to The Journals of Gerontology Series B: Psychological Sciences and Social Sciences
. 2020 Dec 30;76(10):1948–1959. doi: 10.1093/geronb/gbaa230

Rumination and Sleep Quality Among Older Adults: Examining the Role of Social Support

Christina M Marini 1,✉, Stephanie J Wilson 2, Suyoung Nah 3, Lynn M Martire 3, Martin J Sliwinski 3
Editor: Rodlescia Sneed
PMCID: PMC8598998  PMID: 33378473

Abstract

Objectives

Although the adverse link between rumination and sleep quality is well established, much of the literature neglects the role of social factors. This study examined the role of older adults’ perceived social support from spouses and from family/friends in modifying the association between trait rumination and sleep quality. Existing hypotheses suggest that social support may play 3 unique roles, each tested within the current study: (H1) support may act as a protective factor that buffers negative effects of rumination on sleep quality, (H2) support may curtail rumination and, in turn, promote sleep quality, and (H3) rumination may erode support and, in turn, undermine sleep quality.

Method

Data came from 86 partnered older adults in independent-living or retirement communities (Mage = 75.70 years). We utilized 3 waves of interview data collected annually between 2017 and 2019. The first hypothesis was tested using moderation in multilevel models; the second 2 hypotheses were evaluated with prospective associations using multilevel mediation.

Results

Negative effects of high-trait rumination on time-varying sleep quality were attenuated among those who reported high, stable levels of support from their spouses. Perceived family/friend support did not yield the same protective effect. There was no evidence that support preempted, or was eroded by, rumination.

Discussion

Perceived spousal support may act as a psychosocial resource that mitigates negative effects of trait rumination on older adults’ sleep quality. Interventions aimed at mitigating maladaptive outcomes of rumination on sleep quality for older adults should consider spousal support as a key target.

Keywords: Emotion/emotion regulation, Marriage, Stress

Changes that coalesce in later life, including the onset of chronic illness, transition to retirement, and weakening and desynchronization of circadian rhythms, make it increasingly difficult for many older adults to get a good night’s sleep (Neikrug & Ancoli-Israel, 2010). Although sleep is a pillar of health across the life span, it is especially important for older adults. Poor sleep quality significantly predicts morbidity and mortality later in life. It has further been associated with declines in mental health, cognitive function, balance, and sensorimotor abilities (see Neikrug & Ancoli-Israel, 2010 for a review).

Prominent models of sleep, such as the Cognitive Model of Insomnia (Harvey, 2002), point to the central role of cognitive processes in determining sleep outcomes. These models suggest that rumination, or unconstructive repetitive thought about past events (including unresolved issues from the day and ongoing concerns), naturally leads to increased arousal at bedtime, which degrades sleep quality (Morin et al., 2003). Accordingly, individuals who engage in high levels of rumination experience poorer sleep than those who ruminate less (Carney et al., 2010).

Although the link between rumination and sleep is well established, much of the existing literature neglects the role of social factors. The primary purpose of the current study is therefore to examine the impact of older adults’ perceived social support from spouses (or cohabitating partners) and friends/family on the association between rumination and sleep quality. Researchers have speculated that perceived social support may broadly benefit older adults’ sleep insofar as it promotes positive health behaviors and a sense of belonging (Troxel et al., 2010). Little existing research, however, has examined whether perceived support acts as a psychosocial resource that promotes older adults’ sleep specifically by curtailing rumination, or by mitigating negative consequences of rumination on sleep quality.

Rumination as a Perseverative Cognitive Coping Process

According to Response Styles Theory (Nolen-Hoeksema, 1991; Nolen-Hoeksema et al., 2008), rumination encompasses negative, self-focused thinking about the past. This neurotic self-attentiveness is a coping mechanism by which people fixate on (or ruminate about) stressors, which prolongs (and perhaps intensifies) distress and interferes with adaptive coping responses (Trapnell & Campbell, 1999). Rumination is therefore an example of a perseverative cognitive coping process, which creates and sustains acute stress responses (Smyth et al., 2013). By reliving past stressors, rumination can serve as its own source of stress. Moreover, individuals who tend to ruminate (i.e., high-trait ruminators) demonstrate higher levels of reactivity to stressors (Moberly & Watkins, 2008).

Rumination and Sleep Quality

High-trait ruminators are prone to experiencing frequent unconstructive repetitive thought throughout the day, as well as increased emotional and cognitive arousal at bedtime, which undermines sleep quality (Morin et al., 2003). Research suggests that even after accounting for negative mood, rumination predicts poorer subjective sleep quality (Kirkegaard Thomsen et al., 2003). Thus, independent of emotional arousal associated with negative affect, it appears that cognitive arousal associated with rumination threatens subjective sleep quality. Compared to low-trait ruminators, high-trait ruminators also have been shown to experience more intrusive thoughts during the presleep period and poorer subjective sleep quality (Guastella & Moulds, 2007).

The Role of Perceived Social Support

Disclosing negative emotions to others has been shown to alleviate intrusive thoughts (Lepore et al., 2000). Researchers have therefore suggested that people who frequently ruminate seek social support to disrupt cycles of negative thought (Zawadzki et al., 2018). In fact, classic research focused on spousal bereavement indicates that the more widows talked with friends about their spouse’s death, the less they ruminated about it (Pennebaker & O’Heeron, 1984). The availability of supportive others with whom individuals can discuss problems and concerns may therefore curtail rumination (DeLongis et al., 2010). “When others are supportive of emotional expression and help individuals understand their distressing situations in new ways, rumination may decline” (Nolen-Hoeksema et al., 2008, p. 411). Older adults who have high perceived support may therefore ruminate less, which, in turn, may promote their sleep quality. The mere perception that support is available may act as a source of comfort, much like having money in a savings account, that enhances sleep (Pow et al., 2017). The sense of security afforded by perceived support may not only reduce rumination, but also encourage more adaptive coping responses that promote sleep quality (Holliday & Troxel, 2017).

According to Response Styles Theory (Nolen-Hoeksema, 1991; Nolen-Hoeksema et al., 2008), however, chronic (or trait-level) rumination may actually erode social support because chronic ruminators behave in ways that challenge their social relationships, including aggressive and dependent (or clingy) interpersonal behaviors, and report higher levels of both depressed mood and social friction. In line with this support erosion hypothesis, research suggests that rumination is positively associated with marital tension (King & DeLongis, 2014). Support erosion due to rumination may then undermine sleep quality given that perceived support is protective for sleep quality (Chung, 2017).

Researchers have also suggested that social support may protect against associations between rumination and maladaptive outcomes (DeLongis et al., 2010). In other words, engaging in rumination may be less harmful for older adults who have supportive social ties with whom they can discuss concerns. Regardless of whether support changes the amount of trait-level rumination older adults engage in, it may render the effects of rumination less harmful. In line with this support-buffering hypothesis, research indicates that social support weakens negative effects of rumination on psychological outcomes (Nolen-Hoeksema & Davis, 1999; Puterman et al., 2010).

We found only one study that examined the degree to which social support buffers the association between rumination and sleep quality (i.e., Zawadzki et al., 2013). In their study of college students, Zawadzki and colleagues found that rumination only predicted poor sleep quality among those who had low (as opposed to moderate or high) levels of perceived social support. They speculated that social support might distract students from continued rumination, thereby allowing them to relax enough to fall asleep. It remains unknown whether such findings generalize to older adults. Although older adults may ruminate less than younger adults do (Robinette & Charles, 2016), research indicates that there is consistency in the relationship between perseverative cognitive processes, such as rumination, and psychological well-being across young (19–39 years old), middle (40–59 years old), and late (60–83 years old) adulthood (Zawadzki et al., 2018). In addition, rumination has been shown to have larger effects on blood pressure reactivity and recovery among older adults than younger adults (Robinette & Charles, 2016). Thus, even if older adults ruminate less than younger adults, the effects of rumination on sleep quality may be just as—if not more—detrimental later in life given strong ties between sleep quality, morbidity, and mortality among older adults (Neikrug & Ancoli-Israel, 2010).

Overview of the Current Study

The aforementioned research suggests the potential for perceived support to play three unique roles in determining the relationship between trait-level rumination and sleep quality, each of which informs the current study’s hypotheses. First, perceived support may buffer an expected negative association between trait rumination and time-varying sleep quality. We refer to this as the support-buffering hypothesis (see Figure 1). Second, perceived support may curtail rumination, which may in turn promote sleep quality (rumination-reduction hypothesis). Third and finally, rumination may erode social support, which may in turn undermine sleep quality (support-erosion hypothesis). We tested these hypotheses within a sample of partnered (i.e., married or cohabitating) older adults.

Figure 1.

Figure 1.

Open in a new tab

Study hypotheses.

In addition to building on existing research with a sample of older adults, we sought to examine the unique effects of support from specific social ties. Existing studies have measured only general perceptions of available social support. In the current study, we therefore examined the distinct roles of perceived support from partners and family/friends. Although partners are a key source of support, growing evidence suggests that peripheral social ties play a uniquely important role in predicting adjustment and well-being later in life (Huxhold et al., 2014). In fact, having a greater number of weaker social ties (e.g., friends/family) has been shown to be more protective against older adults’ depressed affect than having a greater number of close social ties (e.g., spouses/partners; Huxhold et al., 2020). This may be because more peripheral (or weaker) social ties require less time and energy to maintain and may even provide functions that close ties do not, such as novelty. In light of research suggesting that having a diverse social network made up of close and peripheral social ties is beneficial for older adults’ well-being (Fiori et al., 2007), we posit that perceived support from partners and from family/friends would each play a uniquely important role in the current study.

Method

Participants and Procedures

The current study used longitudinal data from a larger observational study of older adults designed to capture changes in close relationships and health. This study was approved by the Pennsylvania State University Institutional Review Board (IRB # MODCR00002495). Since 2016, participants in the larger study have completed annual in-person interviews and a follow-up online survey 6 months later. There are currently eight waves of available data that were collected between 2016 and 2019 (i.e., one in-person interview and one follow-up survey per year). In the current study, we utilized data from the interviews (odd waves) because rumination was not measured in the follow-up surveys. Further, we were unable to use Wave 1 because the full measure of social support was not included until Wave 3. We therefore utilized data from Wave 3 in 2017, Wave 5 in 2018, and Wave 7 in 2019, which we hereinafter refer to as Time 1 (T1), Time 2 (T2), and Time 3 (T3), respectively.

Participants lived in independent-living or retirement communities in central Pennsylvania and were recruited via e-mail, flyers, and community presentations by research staff. Eligible participants were those who did not have language, hearing, or vision problems that precluded their completion of interactive interviews in English, and those who were not cognitively impaired as indicated by a score ≤4 on the Memory Impairment Screen (Buschke et al., 1999). A total of 132 participants were screened for participation. Of these, one participant was not eligible due to cognitive impairment. A total of 131 older adults between the ages of 58 and 94 were therefore enrolled in the study.

Because we aimed to examine the unique effects of support from partners versus family/friends, we limited our analytical sample to those participants who were partnered (i.e., married or cohabitating) at T1, which resulted in a sample size of 86 participants. There was minimal attrition within this analytical sample between T1 and T3. Approximately 86% (N = 74) of the sample provided data at all three time points and 8% (N = 7) provided data at two time points. Only 6% of the sample (N = 5) provided data only at T1. Reasons for missing assessments included lack of time, health issues, and death (N = 2). Using a series of independent samples t tests, we found that there were no significant mean differences in demographics or scores on key study variables—that is, age (t(84) = 0.33, p = .741) or T1 sleep quality (t(84) = 1.01, p = .317)—among participants who had complete or missing data. Descriptive information for the current study’s sample is displayed in Supplementary Table 1.

Measures

Rumination

Rumination was measured with the Rumination-Reflection Questionnaire (RRQ; Trapnell & Campbell, 1999). The RRQ consists of two subscales, rumination and reflection. We utilized the rumination subscale, measured by 12 items rated on a scale from 1 (strongly disagree) to 5 (strongly agree). Example items include “You spend a great deal of time thinking back over embarrassing or disappointing moments” and “You tend to ‘ruminate’ or dwell over things that happen to you for a really long time afterward.” Items were averaged to create mean scores at T1 (Cronbach’s α = .91; M = 2.44, SD = 0.70), T2 (Cronbach’s α = .90; M = 2.38, SD = 0.68), and T3 (Cronbach’s α = .88; M = 2.38, SD = 0.62) with greater scores indicating higher levels of trait rumination.

Sleep quality

Participants’ sleep quality was assessed with the Patient-Reported Outcomes Information System (PROMIS) Sleep Disturbance Measure-Short Form version 2.1 (Buysse et al., 2010; Yu et al., 2012). The PROMIS is a National Institutes of Health-sponsored project to develop item banks and standard measures that assess physical, mental, and social health. The PROMIS Sleep Disturbance assessment reflects a global measure of sleep, rather than focusing on a particular disorder. Participants rated the overall quality of their sleep over the last week from 0 (very good) to 4 (very poor). Participants also reported the refreshing nature of sleep, problems with sleep, and difficulty in falling asleep over the last week from 0 (not at all) to 4 (very much). With the exception of the item about refreshing sleep, all items were reverse coded. Items were averaged to create mean scores at T1 (Cronbach’s α = .81; M = 2.89, SD = 0.74), T2 (Cronbach’s α = .82; M = 3.05, SD = 0.71), and T3 (Cronbach’s α = .84; M = 2.96, SD = 0.76) with greater scores indicating higher levels of sleep quality.

Social support

Perceived available social support was assessed using a series of items developed by Schuster and colleagues (1990) that were later modified for use in the third wave of the National Social Life, Health, and Aging Project (NSHAP). The items used to measure family/friend support (not including support from partners) were: (1) How often can you open up to members of your family and friends if you need to talk about your worries?, (2) How often can you rely on them for help if you have a problem?, (3) How much do your family and friends really care about you?, and (4) How much do they understand the way you feel about things? Items were reverse scored and then averaged to create mean scores of perceived support from family/friends at T1 (Cronbach’s α = .73; M = 3.69, SD = 0.38), T2 (Cronbach’s α = .75; M = 3.64, SD = 0.42), and T3 (Cronbach’s α = .79; M = 3.59, SD = 0.49) with greater scores indicating higher levels of support. These same items were then modified to ask about support from partners. Items were reverse scored and averaged to create mean scores of perceived support from partners at T1 (Cronbach’s α = .58; M = 3.72, SD = 0.35), T2 (Cronbach’s α = .66; M = 3.78, SD = 0.30), and T3 (Cronbach’s α = .73; M = 3.75, SD = 0.36) with higher scores indicating greater support.

Depressive symptoms

A key covariate in our models, participants’ depressive symptoms over the past week were assessed with the PROMIS Depression-Short Form (Pilkonis et al., 2011), which consists of four items measured on a five-point Likert scale that ranged from 0 (never) to 4 (always). Items were averaged to create mean scores at T1 (Cronbach’s α = .82; M = 0.22, SD = 0.42), T2 (Cronbach’s α = .71; M = 0.17, SD = 0.36), and T3 (Cronbach’s α = .82; M = 0.25, SD = 0.44) with higher scores indicating more severe depressive symptoms.

Comorbidities

Another covariate in our models, current physical health conditions, included up to 23 health problems (e.g., high cholesterol, heart disease, osteoarthritis, cancer, etc.; Martire & Scheier, 2000). Reports of physical health conditions were strongly correlated across time points (rT1,T2 = .84, p < .0001; rT1,T3 = .75, p < .0001; rT2,T3 = .84, p < .0001) and, thus, were averaged to create person-level means (M = 4.84, SD = 2.33).

Data Analysis

All multilevel models were fit in SAS 9.4 (Cary, NC) using PROC MIXED with the Kenward-Rogers degrees of freedom adjustment to control Type I error. Our sample of 86 participants included 33 couples and 20 married individuals whose spouse did not participate in the study. As such, we structured the random component of our models following recommendations for partially nested samples (Bauer et al., 2008). In this design, each dyad or singleton receives a unique value on a single grouping variable, which is used to distinguish between-group from within-group variance (i.e., within-dyad and over-time). Given this partially nested structure, and in keeping with recommendations for dyadic longitudinal data, we did not specify separate subject-level random effects (Ledermann & Kenny, 2017). We used all available data that were appropriate for each question, which differed between multilevel moderation and mediation. Indeed, mediation models tested lagged associations between rumination and support, whereas support-buffering models examined multilevel associations between trait rumination and time-varying sleep as moderated by concurrent (Level 1) and enduring (Level 2) support.

Moderation: support-buffering hypothesis

To evaluate moderation hypotheses, we tested two-way interactions between rumination and support in their associations with sleep quality. Spouse and family/friend support moderators were examined in separate models. Most prior studies have examined support-buffering of rumination with cross-sectional data; our longitudinal design afforded the opportunity to tease apart the unique effects of stable support, that is, the Level 2 person means, from that of time-varying support, that is, Level 1 person-mean centered values. Given the high degree of stability in trait rumination as evidenced by the strong correlations across time points (see Table 1), we opted to test the effect of baseline rumination on time-varying sleep as moderated by Level 1 and Level 2 support. This approach enabled us to capture the degree to which both stable and time-varying sources of support could modify the effect of trait-like rumination on fluctuating sleep quality. Given the skewed distribution of spouse and friend/family support variables, we probed statistically significant interactions at the first percentile, 25th percentile, and 100th percentile to characterize the link between rumination and sleep at meaningful values of support. Note that models met assumptions of homoscedasticity and normally distributed residuals and thus did not require logarithmic transformation.

Table 1.

Pearson Correlations Among Key Study Variables

Variables 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16.
1. Age —
2. Rumination T1 −.13 —
3. Rumination T2 −.18 .80*** —
4. Rumination T3 −.03 .73*** .69*** —
5. Sleep quality T1 −.04 −.40*** −.41*** −.39*** —
6. Sleep quality T2 −.03 −.40*** −.40*** −.34** .47*** —
7. Sleep quality T3 −.09 −.25* −.19 −.24* .56*** .56*** —
8. Spouse support T1 −.06 .01 .14 .13 .04 −.03 −.10 —
9. Spouse support T2 −.13 −.16 −.17 −.17 .12 .12 .03 .67*** —
10. Spouse support T3 .04 −.22 −.12 −.26* .05 −.02 .01 .65*** .85*** —
11. Family/friend support T1 −.21 −.09 −.10 −.15 .10 .06 .03 .41*** .29* .30* —
12. Family/friend support T2 −.16 −.17 −.21 −.13 .18 .05 −.03 .34** .46*** .39** .62*** —
13. Family/friend support T3 −.12 .03 −.02 −.04 .07 .06 .11 .18 .36** .31** .55*** .57*** —
14. Depressive symptoms T1 .17 .24* .20 .27* −.41*** −.17 −.20 −.26* −.34** −.23 −.33** −.30** −.14 —
15. Depressive symptoms T2 .09 .17 .16 .37** −.17 −.26* −.16 −.00 −.28* −.32** −.23* −.17 −.19 .38*** —
16. Depressive symptoms T3 .05 .06 .10 .33** −.31** −.32** −.34** .04 −.08 −.07 −.22 .03 .02 .49*** .66*** —
Open in a new tab

Note: *p < .05. **p < .01. ***p < .001.

Models included the following covariates: lag-1 sleep quality, concurrent depressive symptoms, age, gender, comorbidities, as well as Level 1 and Level 2 of the other source of support (i.e., family/friend support in the spousal support model and vice versa). Controlling for sleep quality at the previous time point allowed us to examine residualized change in sleep. Depressive symptoms were controlled because sleep problems are a cardinal symptom of depression (Smith & Quartana, 2010). Also, this covariate allowed us to tease apart the effects of ruminative thought patterns from mood disturbance, a common companion to rumination (McLaughlin & Nolen-Hoeksema, 2011). In addition to gender, we accounted for age, as sleep quality tends to decline with older age (Madrid-Valero et al., 2017). Comorbidities were included as a covariate to account for the known association between physical health conditions and poor sleep quality (Neikrug & Ancoli-Israel, 2010). We also accounted for family/friend support in spousal support moderation models, and vice versa, to test the unique contributions of each source.

Mediation

To evaluate rumination-reduction and support-erosion hypotheses, we employed Bauer and colleagues’ (2006) method for multilevel mediation. In this technique, the mediator and outcome are stacked into the same dependent variable so that paths A (predictor to mediator), B (mediator to outcome), and C′ (effect of predictor controlling for mediator, i.e., the direct effect) can be estimated simultaneously. Dummy variables are used to estimate separate fixed and random effects for the two “dependent variables,” that is, the mediator and the outcome. Separate residual variances were also specified for the mediator and the outcome. Random effects were specified for paths A and B to obtain the random covariance between the two paths, a necessary component to estimate the indirect effect (Bauer et al., 2006).

Bauer’s accompanying macro was used to calculate the indirect effect of each predictor on the outcome (i.e., rumination on sleep through support, or support on sleep through rumination) by multiplying coefficients for paths A and B. It calculated the total effect of each predictor on sleep by summing the indirect effect and direct effect. We generated standard errors using 10,000 bootstrapped samples and 95% confidence intervals (CIs). The indirect and total effects of spouse and family/friend support were evaluated in separate models. Because the key distinction between the two mediation hypotheses is the temporal ordering of rumination and support, we tested lagged associations for path A (i.e., lag-1 support predicting rumination, lag-1 rumination predicting support). We tested concurrent associations for path B to maximize the use of available data.

Support-erosion hypothesis

In the support-erosion models, we first examined lag-1 rumination as a predictor of spousal support (path A), controlling for lag-1 spousal support, lag-1 depressive symptoms, and gender. Controlling for spousal support at the previous time point accounted for its autocorrelation and allowed us to draw conclusions about how lag-1 rumination was associated with prospective, residualized change in spousal support. Depression served as a covariate to account for accompanying relationship strain and biases in support perceptions (Ibarra-Rovillard & Kuiper, 2011). Gender was controlled because in prior work, men reported receiving higher-quality spousal support than did women (Neff & Karney, 2005). In this model, we had two available time points for path A: rumination at Time 1 predicting Time 2 support, and rumination at Time 2 predicting Time 3 support. In turn, path B was also tested with two time points, with spousal support at Times 2 and 3 predicting concurrent sleep quality, controlling for lag-1 sleep quality, lag-1 depressive symptoms, gender, age, comorbidities, lag-1 family/friend support, and lag-1 rumination (to estimate path C′). A separate mediation model examining family/friend support as the mediator mirrored the structure of the spousal support model.

Rumination-reduction hypothesis

In the rumination-reduction models, we examined whether lag-1 spousal support (at Times 1 and 2) predicted rumination at Times 2 and 3 (path A), controlling for lag-1 rumination, lag-1 depressive symptoms, lag-1 family/friend support, and gender. Lag-1 rumination was included as a covariate to aid conclusions about the sequential ordering. Models covaried gender because women report higher rumination levels than men (Johnson & Whisman, 2013). As aforementioned, depressive symptoms share systematic associations with both support perceptions and rumination, and controlling family/friend support allowed us to tease apart the independent role of spousal support on rumination. In turn, rumination at Times 2 and 3 was examined as a predictor of sleep at Times 2 and 3 (path B), with the model also including lag-1 sleep quality, lag-1 depressive symptoms, gender, age, comorbidities, and lag-1 support (to estimate path C’). A separate mediation model examining family/friend support as the primary predictor mirrored the structure of the spousal support model.

Results

Support Buffering

As depicted in Figure 2, mean levels of spousal support significantly buffered the association between baseline rumination and time-varying sleep (Table 2, estimate = 0.449, SE = 0.175, p = .011), such that baseline rumination was a weaker predictor of sleep quality among people with greater spousal support (estimatemax support = −0.176, SE = 0.100, p = .080; estimateQ1 support = −0.400, SE = 0.088, p < .0001; estimate1st percentile support = −0.962, SE = 0.264, p = .0004). In contrast, year-to-year fluctuations in spousal support did not buffer the link between rumination and sleep quality (p > .250). Unlike spousal support, neither mean levels (p = .106) nor fluctuations in family/friend support (p = .996) significantly buffered the association between rumination and sleep quality (Table 3).

Figure 2.

Figure 2.

Open in a new tab

Association between baseline rumination and time-varying sleep quality as moderated by spousal support. The solid line represents the effect of rumination on sleep at maximum levels of support (estimatemax support = −0.176, SE = 0.100, p = .080). The dashed line represents the slope of rumination on sleep at low (first quartile) support (estimateQ1 support = −0.400, SE = 0.088, p < .0001). The dotted line shows the association between rumination and sleep at very low levels (first percentile) of spousal support (estimatefirst percentile support = −0.962, SE = 0.264, p = .0004).

Table 2.

Spousal Support as a Buffer of the Association Between Rumination and Sleep Quality

Fixed effects Estimate SE p 95% CI
Intercept 3.041 0.086 <.0001 2.870, 3.212
Lag-1 sleep −0.170 0.075 .025 −0.319, −0.022
Depression −0.337 0.161 .038 −0.654, −0.019
Age −0.003 0.008 .756 −0.019, 0.014
Comorbidities −0.039 0.024 .102 −0.086, 0.008
Gender −0.150 0.100 .134 −0.337, 0.046
L1 family/friend support 0.133 0.167 .428 −0.197, 0.463
L2 family/friend support 0.268 0.171 .119 −0.069, 0.606
Baseline rumination −0.307 0.082 .0003 −0.470, −0.145
L1 spouse support 0.113 0.309 .716 −0.497, 0.723
Rumination × L1 spouse support −0.374 0.373 .317 −1.110, 0.362
L2 spouse support −0.080 0.186 .667 −0.449, 0.288
Rumination × L2 spouse support 0.449 0.175 .011 0.103, 0.796
Open in a new tab

Note: L1 = Level 1 person-centered values; L2 = Level 2 person means. Nobservations = 224.

Table 3.

Family/Friend Support as a Buffer of the Association Between Rumination and Sleep Quality

Fixed effects Estimate SE p 95% CI
Intercept 3.058 0.087 <.0001 2.886, 3.230
Lag-1 sleep −0.180 0.077 .020 −0.331, −0.029
Depression −0.316 0.164 .057 −0.640, 0.009
Age 0.002 0.008 .842 −0.015, 0.018
Comorbidities −0.041 0.024 .092 −0.089, 0.007
Gender −0.201 0.100 .045 −0.399, −0.004
L1 spousal support 0.003 0.303 .992 −0.595, 0.601
L2 spousal support −0.123 0.194 .528 −0.506, 0.261
Baseline rumination −0.267 0.088 .003 −0.440, −0.094
L1 family/friend support 0.068 0.184 .714 −0.296, 0.432
Rumination × L1 family/friend support 0.001 0.215 .996 −0.424, 0.426
L2 family/friend support 0.433 0.172 .013 0.093, 0.773
Rumination × L2 family/friend support 0.291 0.179 .106 −0.063, 0.645
Open in a new tab

Note: L1 = Level 1 person-centered values; L2 = Level 2 person means. Nobservations = 224.

Support Erosion

There was no evidence for the support-erosion hypothesis. Lag-1 rumination was not associated with spousal support (Supplementary Table 2, p > .250, path A), nor did spousal support predict sleep quality (p > .250, path B). Thus, the indirect effect of rumination on sleep through spousal support was also nonsignificant (p > .250, 95% CI −0.122, 0.132). However, the direct effect of lag-1 rumination on sleep quality (path C′) was significant, such that more rumination was associated prospectively with worse sleep quality (estimate = −0.284, SE = 0.093, p = .003). Likewise, the total effect of rumination on sleep quality (i.e., the sum of the indirect and direct effects) was also statistically significant in the same direction (estimate = −0.278, SE = 0.114, p = .015, 95% CI −0.501, −0.054).

As shown in Supplementary Table 3, findings for family/friend support mirrored those of spousal support. Lag-1 rumination did not predict family/friend support (p > .250, path A), nor was family/friend support directly associated with sleep (p > .250, path B). Thus, the indirect effect of lagged rumination through family/friend support on sleep was not significant (p > .250, 95% CI −0.067, 0.115). The direct effect of lagged rumination on sleep quality remained statistically significant, such that greater rumination was associated with poorer sleep at the next time point (estimate = −0.361, SE = 0.095, p = .0002). The total effect of rumination on sleep was also significant (estimate = −0.338, SE = 0.106, p = .001, 95% CI −0.548, −0.133).

Rumination Reduction

Lag-1 spousal support did not predict rumination (Supplementary Table 4, p > .250, path A), but consistent with prior results, greater rumination was associated with poorer sleep (estimate = −0.342, SE = 0.101, p = .002, path B). Spousal support did not have a significant direct effect (p > .250), indirect effect (p > .250, 95% CI −0.213, 0.253), or total effect (p > .250, 95% CI −0.462, 0.508) on sleep.

Likewise, lag-1 family/friend support shared no prospective association with rumination (Supplementary Table 5, p > .250, path A), but greater rumination was associated with poorer sleep (estimate = −0.338, SE = 0.100, p = .002, path B). Family/friend support did not have a significant direct effect (p > .250), indirect effect (p > .250, 95% CI −0.068, 0.216), or total effect (p > .250, 95% CI −0.277, 0.398) on sleep quality.

Discussion

Existing research suggests that perceived support may play three different roles in the relationship between rumination and sleep quality. First, perceived support may buffer negative effects of rumination on sleep quality (i.e., support-buffering hypothesis). Second, perceived support may curtail rumination, which may then promote sleep quality (i.e., rumination-reduction hypothesis). Third, rumination may erode support, which may then undermine sleep (i.e., support-erosion hypothesis). Findings from the current study lend credence only to the support-buffering hypothesis. Negative effects of high-trait rumination on sleep quality were attenuated for older adults who reported high, stable levels of support from their spouses. Perceived family/friend support did not yield the same protective effect.

Support Buffering

In line with the support-buffering hypothesis, social support has been shown to weaken negative effects of rumination on psychological well-being (e.g., Puterman et al., 2010). Findings from the current study suggest that spousal support also weakens the degree to which rumination undermines sleep quality, even after accounting for psychological well-being (i.e., depressive symptoms). Further, our use of longitudinal data afforded the opportunity to parse out the degree to which enduring support versus time-varying support served a protective role. We found that only persistently high (i.e., enduring) support successfully buffered the negative effect of rumination on sleep quality, whereas unique occasions of higher support (i.e., time-varying support) were not sufficient. This finding underscores the protective value of stable—as opposed to fluctuating—perceptions of available support, which may act as a source of comfort that enhances sleep (Pow et al., 2017). Additionally, enduring support may facilitate the “entrainment of circadian rhythms,” thereby helping older adults maintain more consistent sleep habits and schedules (Troxel et al., 2010, p. 460). For example, regularly conversing with a supportive spouse about one’s day before bed may benefit sleep not only by providing comfort, but also by “reinforcing a consistent sleep and wake routine” (p. 460).

Consistent with the classic stress-buffering hypothesis (Cohen & Wills, 1985), findings from the current study suggest that support does not directly predict sleep quality, but rather that it buffers negative effects of stress (i.e., rumination) on sleep quality. This pattern is interesting given that there is an inconsistent link between support and sleep quality in the existing literature. Some studies indicate that general perceptions of support directly promote better sleep (Chung, 2017). However, research focused specifically on partnered older adults indicates that spousal support does not directly affect sleep quality after accounting for other psychological factors, such as depressive symptoms (Chen et al., 2015), or when trying to predict change in sleep quality over time (Lee et al., 2017). Thus, spousal support may act as a stress buffer, rather than a direct predictor of sleep quality. As a perseverative cognitive coping strategy, rumination creates and sustains acute stress responses, thereby contributing to chronic stress (Smyth et al., 2013) and increased stress reactivity (Moberly & Watkins, 2008). In the context of stress, interpersonal behaviors characterized by affiliation (e.g., trust) are associated with both lower presleep cognitive arousal and better sleep quality (Gunn et al., 2017).

Our finding that perceived support from spouses—but not family/friends—attenuates negative effects of high-trait rumination on sleep quality underscores the salient role that marital quality plays in predicting health and well-being later in life. More specifically, perceiving that you can open up to your spouse about worries, and that your spouse understands, may lower arousal and facilitate restful sleep (Gunn et al., 2017). Other research similarly indicates that self-disclosure (Kane et al., 2014) and perceived partner responsiveness (Selcuk et al., 2017) are specific relational processes that both promote sleep quality.

Limitations and Future Research Directions

We did not find support for either the rumination-reduction or the support-erosion hypothesis, perhaps because they required prospective associations to tease apart temporal ordering necessary to test for mediation. Thus, the threshold for detecting support for these hypotheses was likely higher than it was for the support-buffering hypothesis, particularly in light of the strong correlations over time. Perceived support (spousal or family/friend) was also not significantly correlated at the bivariate level with either rumination or sleep, suggesting that, at least in our sample of high-functioning older adults, support may not alter sleep quality in all conditions, nor does it seem to share direct associations with rumination. As previously discussed, other studies of partnered older adults have similarly failed to detect a direct association between spousal support and sleep quality (Chen et al., 2015; Lee et al., 2017).

Given the stability of trait-level rumination and its consistent effect on sleep across all of our models, it was informative and somewhat surprising to see no evidence of support erosion. Indeed, both perceived spousal and family/friend support were highly skewed toward maximum levels. Selection effects may have excluded ruminators whose maladaptive habits eroded their relationships severely enough to not remain partnered; our sample consisted of older adults in predominantly long-term relationships. Future research would benefit from re-examining the support-erosion hypothesis within a sample of newly partnered/married couples in order to assess potential support erosion early in the relationship. In addition, future research on the support-erosion hypothesis may benefit from examining received—rather than perceived—support. Received support may provide a better depiction of actual support exchanges that unfold in the context of stress, rather than individuals’ global perceptions of support that is available to them should they need it. Finally, future research would benefit from including a measure of social network size. Doing so would enable researchers to disentangle the degree to which rumination predicts declines in social support (i.e., support erosion) versus decreases in social network size.

We may have failed to find support for the rumination-reduction hypothesis because our data captured trait-level rumination, which is not as malleable as day-to-day or momentary rumination (i.e., state-level rumination). Even among trait ruminators, there is within-person variability in rumination (Delongis et al., 2010). For example, trait ruminators tend to ruminate more when they experience stressors (Moberly & Watkins, 2008). Future studies should capture momentary fluctuations in rumination to tease apart the degree to which real-time exchanges of support curtail rumination on particularly stressful days, thereby indirectly promoting sleep quality later that night.

Although social ties outside of marriage play a uniquely important role in predicting well-being later in life (Huxhold et al., 2014), we did not find evidence that support from family and friends buffered effects of rumination on sleep quality when men and women were examined together. This may be because effects of support are most salient in the evening, prior to going to sleep, when older adults are perhaps more likely to interact with their partners, as opposed to family and friends with whom they do not cohabitate. It may also be that family/friend support plays a uniquely important role for women given that older women have significantly larger social networks than men do (McLaughlin et al., 2010). Men also tend to rely more exclusively on their spouses for intimacy and support, whereas women also rely on other social ties (Santini et al., 2016; Stronge et al., 2019). In fact, for older women, loneliness that stems from a lack of engagement with friends and relatives is a stronger predictor of depressive symptoms than is simply living alone (Park et al., 2013). Future research should reexamine the current study’s hypotheses within a larger sample of older adults to test gender differences with sufficient power. Future studies would additionally benefit from re-evaluating these hypotheses within a more diverse sample. Existing research demonstrates that there are differences in older adults’ support networks that may attributed both to race and structural factors (e.g., education, income; Peek & O’Neill, 2001). Further, recent research illustrates that there are racial/ethnic differences in older adults’ sleep quality (George et al., 2020).

Conclusion

Findings from this longitudinal study suggest that perceived spousal support is a psychosocial resource that mitigates negative effects of trait rumination on older adults’ sleep quality. Although spousal support did not change the amount of trait-level rumination older adults engaged in, it did appear to render the effects of such rumination less harmful. Taken together, these findings suggest that interventions aimed at mitigating maladaptive outcomes of rumination on sleep quality should consider older adults’ perceptions of their spouse as an understanding confidant with whom they can share their worries.

Supplementary Material

gbaa230_suppl_Supplementary_Material
Click here for additional data file. (35.8KB, docx)

Acknowledgments

If requested, the data utilized in the current study are available from Dr. Lynn M. Martire (e-mail: lmm51@psu.edu) and will be provided on a case-by-case basis. Information about the larger project from which this study was conducted can be found on the Open Science Framework (https://osf.io/gjdev/), which includes detailed information about the study’s measures and timepoints. The specific hypotheses for this study were not preregistered.

Funding

This work was supported by the Center for Healthy Aging at the Pennsylvania State University and the National Institute on Aging at the National Institutes of Health (R03 AG064360 to C. M. Marini; R00 AG056667 and L30 AG06025 to S. J. Wilson; and T32 AG049676).

Conflict of Interest

None declared.

References

  1. Bauer, D. J., Preacher, K. J., & Gil, K. M. (2006). Conceptualizing and testing random indirect effects and moderated mediation in multilevel models: New procedures and recommendations. Psychological Methods, 11(2), 142–163. doi: 10.1037/1082-989X.11.2.142 [DOI] [PubMed] [Google Scholar]
  2. Bauer, D. J., Sterba, S. K., & Hallfors, D. D. (2008). Evaluating group-based interventions when control participants are ungrouped. Multivariate Behavioral Research, 43(2), 210–236. doi: 10.1080/00273170802034810 [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Buschke, H., Kuslansky, G., Katz, M., Stewart, W. F., Sliwinski, M. J., Eckholdt, H. M., & Lipton, R. B. (1999). Screening for dementia with the Memory Impairment Screen. Neurology, 52(2), 231–238. doi: 10.1212/wnl.52.2.231 [DOI] [PubMed] [Google Scholar]
  4. Buysse, D. J., Yu, L., Moul, D. E., Germain, A., Stover, A., Dodds, N. E., Johnston, K. L., Shablesky-Cade, M. A., & Pilkonis, P. A. (2010). Development and validation of patient-reported outcome measures for sleep disturbance and sleep-related impairments. Sleep, 33(6), 781–792. doi: 10.1093/sleep/33.6.781 [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Carney, C. E., Harris, A. L., Moss, T. G., & Edinger, J. D. (2010). Distinguishing rumination from worry in clinical insomnia. Behaviour Research and Therapy, 48(6), 540–546. doi: 10.1016/j.brat.2010.03.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Chen, J. H., Waite, L. J., & Lauderdale, D. S. (2015). Marriage, relationship quality, and sleep among U.S. older adults. Journal of Health and Social Behavior, 56(3), 356–377. doi: 10.1177/0022146515594631 [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Chung, J. (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. doi: 10.1016/j.sleh.2016.10.003 [DOI] [PubMed] [Google Scholar]
  8. Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310–357. [PubMed] [Google Scholar]
  9. DeLongis, A., Holtzman, S., Puterman, E., & Lam, M. (2010). Spousal support and dyadic coping in times of stress. In Sullivan K. T. & Davila J. (Eds.), Support processes in intimate relationships (pp. 153–174). Oxford University Press. [Google Scholar]
  10. Fiori, K. L., Smith, J., & Antonucci, T. C. (2007). Social network types among older adults: A multidimensional approach. The Journals of Gerontology, Series B: : Psychological and Social Sciences, 62, P322–P330. doi: 10.1093/geronb/62.6.P322 [DOI] [PubMed] [Google Scholar]
  11. George, K. M., Peterson, R. L., Gilsanz, P., Mungas, D. M., Glymour, M. M., Mayeda, E. R., & Whitmer, R. A. (2020). Racial/ethnic differences in sleep quality among older adults: Kaiser Healthy Aging and Diverse Life Experiences (KHANDLE) study. Ethnicity & Disease, 30(3), 469–478. doi: 10.18865/ed.30.3.469 [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Guastella, A. J., & Moulds, M. L. (2007). The impact of rumination on sleep quality following a stressful life event. Personality and Individual Differences, 42, 1151–1162. doi: 10.1016/j.paid.2006.04.028 [DOI] [Google Scholar]
  13. Gunn, H. E., Critchfield, K. L., Mackaronis, J. E., Rau, H. K., Cribbet, M. R., Troxel, W. M., & Williams, P. G. (2017). Affiliative interpersonal behaviors during stress are associated with sleep quality and presleep arousal in young, healthy adults. Sleep Health, 3(2), 98–101. doi: 10.1016/j.sleh.2016.12.004 [DOI] [PubMed] [Google Scholar]
  14. Harvey, A. G. (2002). A cognitive model of insomnia. Behaviour Research and Therapy, 40(8), 869–893. doi: 10.1016/s0005-7967(01)00061-4 [DOI] [PubMed] [Google Scholar]
  15. Holliday, S. B., & Troxel, W. (2017). Relationship quality: Implications for sleep quality and sleep disorders. In McHale S. M., King V., & Buxton O. M. (Eds.), Family contexts of sleep and health across the life course (pp. 53–84). Springer. [Google Scholar]
  16. Huxhold, O., Fiori, K. L., Webster, N. J., & Antonucci, T. C. (2020). The strength of weaker ties: An underexplored resource for maintaining emotional well-being in later life. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 75(7), 1433–1442. doi: 10.1093/geronb/gbaa019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Huxhold, O., Miche, M., & Schüz, B. (2014). Benefits of having friends in older ages: Differential effects of informal social activities on well-being in middle-aged and older adults. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 69(3), 366–375. doi: 10.1093/geronb/gbt029 [DOI] [PubMed] [Google Scholar]
  18. Ibarra-Rovillard, M. S., & Kuiper, N. A. (2011). Social support and social negativity findings in depression: Perceived responsiveness to basic psychological needs. Clinical Psychology Review, 31(3), 342–352. doi: 10.1016/j.cpr.2011.01.005 [DOI] [PubMed] [Google Scholar]
  19. Johnson, D. P., & Whisman, M. A. (2013). Gender differences in rumination: A meta-analysis. Personality and Individual Differences, 55(4), 367–374. doi: 10.1016/j.paid.2013.03.019 [DOI] [PMC free article] [PubMed] [Google Scholar]
  20. 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. doi: 10.1037/hea0000077 [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. King, D. B., & DeLongis, A. (2014). When couples disconnect: Rumination and withdrawal as maladaptive responses to everyday stress. Journal of Family Psychology, 28(4), 460–469. doi: 10.1037/a0037160 [DOI] [PubMed] [Google Scholar]
  22. Kirkegaard Thomsen, D., Mehlsen, M. Y., Christensen, S., & Zachariae, R. (2003). Rumination—relationship with negative mood and sleep quality. Personality and Individual Differences, 34, 1293–1301. doi: 10.1016/S0191-8869(02)00120-4 [DOI] [Google Scholar]
  23. Ledermann, T., & Kenny, D. A. (2017). Analyzing dyadic data with multilevel modeling versus structural equation modeling: A tale of two methods. Journal of Family Psychology, 31(4), 442–452. doi: 10.1037/fam0000290 [DOI] [PubMed] [Google Scholar]
  24. 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(5), 821–831. doi: 10.1007/s10865-017-9850-2 [DOI] [PubMed] [Google Scholar]
  25. Lepore, S. J., Ragan, J. D., & Jones, S. (2000). Talking facilitates cognitive-emotional processes of adaptation to an acute stressor. Journal of Personality and Social Psychology, 78(3), 499–508. doi: 10.1037//0022-3514.78.3.499 [DOI] [PubMed] [Google Scholar]
  26. Madrid-Valero, J. J., Martínez-Selva, J. M., Ribeiro do Couto, B., Sánchez-Romera, J. F., & Ordoñana, J. R. (2017). Age and gender effects on the prevalence of poor sleep quality in the adult population. Gaceta Sanitaria, 31(1), 18–22. doi: 10.1016/j.gaceta.2016.05.013 [DOI] [PubMed] [Google Scholar]
  27. Martire, L. M. & Scheier, M. F. (2000). Physical comorbidity index. School of Medicine, University of Pittsburgh. Unpublished manuscript. [Google Scholar]
  28. McLaughlin, K. A., & Nolen-Hoeksema, S. (2011). Rumination as a transdiagnostic factor in depression and anxiety. Behaviour Research and Therapy, 49, 186–193. doi: 10.1016/j.brat.2010.12.006 [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. McLaughlin, D., Vagenas, D., Pachana, N. A., Begum, N., & Dobson, A. (2010). Gender differences in social network size and satisfaction in adults in their 70s. Journal of Health Psychology, 15(5), 671–679. doi: 10.1177/1359105310368177 [DOI] [PubMed] [Google Scholar]
  30. Moberly, N. J., & Watkins, E. R. (2008). Ruminative self-focus, negative life events, and negative affect. Behaviour Research and Therapy, 46(9), 1034–1039. doi: 10.1016/j.brat.2008.06.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Morin, C. M., Rodrigue, S., & Ivers, H. (2003). Role of stress, arousal, and coping skills in primary insomnia. Psychosomatic Medicine, 65(2), 259–267. doi: 10.1097/01.psy.0000030391.09558.a3 [DOI] [PubMed] [Google Scholar]
  32. Neff, L., & Karney, B. (2005). Gender differences in social support: A question of skill or responsiveness? Journal of Personality and Social Psychology, 88, 79–90. doi: 10.1037/0022-3514.88.1.79 [DOI] [PubMed] [Google Scholar]
  33. Neikrug, A. B., & Ancoli-Israel, S. (2010). Sleep disorders in the older adult—a mini-review. Gerontology, 56(2), 181–189. doi: 10.1159/000236900 [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Nolen-Hoeksema, S. (1991). Responses to depression and their effects on the duration of depressive episodes. Journal of Abnormal Psychology, 100(4), 569–582. doi: 10.1037//0021-843x.100.4.569 [DOI] [PubMed] [Google Scholar]
  35. Nolen-Hoeksema, S., & Davis, C. G. (1999). “Thanks for sharing that”: Ruminators and their social support networks. Journal of Personality and Social Psychology, 77, 801–814. doi: 10.1037/0022-3514.77.4.801 [DOI] [PubMed] [Google Scholar]
  36. Nolen-Hoeksema, S., Wisco, B. E., & Lyubomirsky, S. (2008). Rethinking rumination. Perspectives on Psychological Science: A Journal of the Association for Psychological Science, 3(5), 400–424. doi: 10.1111/j.1745-6924.2008.00088.x [DOI] [PubMed] [Google Scholar]
  37. Park, N. S., Jang, Y., Lee, B. S., Haley, W. E., & Chiriboga, D. A. (2013). The mediating role of loneliness in the relation between social engagement and depressive symptoms among older Korean Americans: Do men and women differ? The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 68(2), 193–201. doi: 10.1093/geronb/gbs062 [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Peek, M. K., & O’Neill, G. S. (2001). Networks in later life: An examination of race differences in social support networks. International Journal of Aging & Human Development, 52(3), 207–229. doi: 10.2190/F1Q1-JV7D-VN77-L6WX [DOI] [PubMed] [Google Scholar]
  39. Pennebaker, J. W., & O’Heeron, R. C. (1984). Confiding in others and illness rate among spouses of suicide and accidental-death victims. Journal of Abnormal Psychology, 93(4), 473–476. doi: 10.1037//0021-843x.93.4.473 [DOI] [PubMed] [Google Scholar]
  40. Pilkonis, P. A., Choi, S. W., Reise, S. P., Stover, A. M., Riley, W. T., & Cella, D. (2011). Item banks for measuring emotional distress from the Patient-Reported Outcomes Measurement Information System (PROMIS®): Depression, anxiety, and anger. Assessment, 18(3), 263–283. doi: 10.1177/1073191111411667 [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Pow, J., King, D. B., Stephenson, E., & DeLongis, A. (2017). Does social support buffer the effects of occupational stress on sleep quality among paramedics? A daily diary study. Journal of Occupational Health Psychology, 22(1), 71–85. doi: 10.1037/a0040107 [DOI] [PubMed] [Google Scholar]
  42. Puterman, E., DeLongis, A., & Pomaki, G. (2010). Protecting us from ourselves: Social support as a buffer of trait and state rumination. Journal of Social and Clinical Psychology, 29, 797–820. doi: 10.1521/jscp.2010.29.7.797 [DOI] [Google Scholar]
  43. Robinette, J. W., & Charles, S. T. (2016). Age, rumination, and emotional recovery from a psychosocial stressor. The Journals of Gerontology, Series B: Psychological Sciences and Social Sciences, 71(2), 265–274. doi: 10.1093/geronb/gbu097 [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Santini, Z. I., Fiori, K. L., Feeney, J., Tyrovolas, S., Haro, J. M., & Koyanagi, A. (2016). Social relationships, loneliness, and mental health among older men and women in Ireland: A prospective community-based study. Journal of Affective Disorders, 204, 59–69. doi: 10.1016/j.jad.2016.06.032 [DOI] [PubMed] [Google Scholar]
  45. Schuster, T. L., Kessler, R. C., & Aseltine, R. H.Jr. (1990). Supportive interactions, negative interactions, and depressed mood. American Journal of Community Psychology, 18(3), 423–438. doi: 10.1007/BF00938116 [DOI] [PubMed] [Google Scholar]
  46. Selcuk, E., Stanton, S. C., Slatcher, R. B., & Ong, A. D. (2017). Perceived partner responsiveness predicts better sleep quality through lower anxiety. Social Psychological and Personality Science, 8, 83–92. doi: 10.1177/1948550616662128 [DOI] [Google Scholar]
  47. Smith, M. T., & Quartana, P. J. (2010). The riddle of the sphinx: Sleep, pain, and depression. Sleep Medicine, 11(8), 745–746. doi: 10.1016/j.sleep.2010.05.004 [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Smyth, J., Zawadzki, M., & Gerin, W. (2013). Stress and disease: A structural and functional analysis. Social and Personality Psychology Compass, 7, 217–227. doi: 10.1111/spc3.12020 [DOI] [Google Scholar]
  49. Stronge, S., Overall, N. C., & Sibley, C. G. (2019). Gender differences in the associations between relationship status, social support, and wellbeing. Journal of Family Psychology, 33(7), 819–829. doi: 10.1037/fam0000540 [DOI] [PubMed] [Google Scholar]
  50. Trapnell, P. D., & Campbell, J. D. (1999). Private self-consciousness and the five-factor model of personality: Distinguishing rumination from reflection. Journal of Personality and Social Psychology, 76(2), 284–304. doi: 10.1037//0022-3514.76.2.284 [DOI] [PubMed] [Google Scholar]
  51. Troxel, W. M., Buysse, D. J., Monk, T. H., Begley, A., & Hall, M. (2010). Does social support differentially affect sleep in older adults with versus without insomnia? Journal of Psychosomatic Research, 69(5), 459–466. doi: 10.1016/j.jpsychores.2010.04.003 [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Yu, L., Buysse, D. J., Germain, A., Moul, D. E., Stover, A., Dodds, N. E., Johnston, K. L., & Pilkonis, P. A. (2012). Development of short forms from the PROMISTM sleep disturbance and sleep-related impairment item banks. Behavioral Sleep Medicine, 10, 6–24. doi: 10.1080/15402002.2012.636266 [DOI] [PMC free article] [PubMed] [Google Scholar]
  53. Zawadzki, M. J., Graham, J. E., & Gerin, W. (2013). Rumination and anxiety mediate the effect of loneliness on depressed mood and sleep quality in college students. Health Psychology, 32(2), 212–222. doi: 10.1037/a0029007 [DOI] [PubMed] [Google Scholar]
  54. Zawadzki, M. J., Sliwinski, M. J., & Smyth, J. M. (2018). Perseverative cognitions and stress exposure: Comparing relationships with psychological health across a diverse adult sample. Annals of Behavioral Medicine, 52(12), 1060–1072. doi: 10.1093/abm/kay009 [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

gbaa230_suppl_Supplementary_Material
Click here for additional data file. (35.8KB, docx)

Articles from The Journals of Gerontology Series B: Psychological Sciences and Social Sciences are provided here courtesy of Oxford University Press

RESOURCES