Family Relationships and Troubled Sleep among U.S. Adults

Abstract

Sleep is essential for health and daily functioning, and social relationships may be a key social factor influencing sleep, yet sleep has been understudied in the literature on social relationships and health. This study used data from the National Survey of Midlife Development in the United States to examine associations between troubled sleep and family contact, social support, and strain. Results show that having strained family relationships is associated with more troubled sleep, while supportive family relationships are associated with less troubled sleep. Family strain is more consequential for sleep than support, and sleep troubles are greatest when family relationships are highly strained and provide inadequate emotional support. Family strain is also more harmful to sleep among individuals who are in frequent contact with family members. These findings underscore the importance of focusing on both negative and positive aspects of relationships and highlight the significance of family relationships for sleep.

Keywords

families health behaviors sleep social relationships stress

The importance of sleep for individual health and well-being cannot be overstated. Adequate sleep is essential for daily functioning, optimal physical and mental health, and longevity (Cappuccio et al. 2009; Dinges 1989; Knutson et al. 2006; Mallon, Broman, and Hetta 2002; Phillips and Mannino 2007; Schwartz et al. 1998). Over 50 million Americans suffer from chronic sleep problems (National Institutes of Health, National Center on Sleep Disorders Research 2003), and according to data from the National Sleep Foundation’s Sleep in American Polls, there has been an upward trend in the prevalence of reported sleep problems since 1999, when their first poll was conducted (National Sleep Foundation 2005).

The growing magnitude of sleep problems in the United States has stimulated interest in identifying the causes of poor sleep. Most sleep research has focused on individual biomedical or psychological risk factors for poor sleep (Espie 2002). However, a growing area of research suggests that social factors also shape sleep (Patel 2007), and social relationships in particular may be key to getting a good night’s sleep (Cacioppo, Hawkley, and Berntson 2002). Although sociological theory and research highlight the importance of family relationships for shaping health behaviors (Umberson, Crosnoe, and Reczek 2010), and the nascent sociological literature on sleep suggests that gendered responsibilities in families influence sleep (Burgard 2011; Hislop and Arber 2003; Maume, Sebastian, and Bardo 2009, 2010; Venn et al. 2008), little is known about how family relationships matter for sleep. The few existing sociological studies of sleep in the context of the family have largely focused on how characteristics of spousal relationships and caregiving for children influence sleep. Other family relationships, such as those with parents and siblings, and the importance of the quality of relationships with family members have yet to be studied in relation to sleep.

To address this gap in the literature, the current study focuses on how family contact and relationship quality relate to self-reported sleep problems among U.S. adults. Rather than focus on spousal/partner relationships and relationships with children living in the household, as in previous research, we use a broader definition of family that encompasses relationships with other family members (e.g., parents, siblings, and both young and grown children). Drawing on the theoretical and empirical literature on social relationships and health, we develop expectations for how families influence sleep and examine the role of three key aspects of family relationships: frequency of contact, perceived emotional support, and perceived strain. Identifying how multiple aspects of family relationships influence sleep may help us better understand the role of social relationships in shaping an essential health behavior.

Background

Prior Research on Social Relationships and Sleep

Although seminal work on the health benefits of social relationships included sleep among the key behaviors relevant for health promotion (Berkman and Breslow 1983), sleep has since been relatively understudied in the literature on social relationships and health. There are, however, several recent studies of social factors related to sleep that highlight the importance of family relationship structure for sleep duration and quality.

Prior research indicates that marital status and the presence of children in the home are associated with sleep. Several studies of U.S. adults have found that unmarried men and women are more likely to report insufficient and poor quality sleep (Grandner et al. 2010; Hale 2005; Krueger and Friedman 2009). The significance of parental status for sleep is less clear. Some studies have found that having children in the household is associated with longer and better quality sleep (Lauderdale et al. 2006; Troxel et al. 2009), whereas other studies have shown that parents of young children tend to sleep less and experience more sleep interruptions (Burgard 2011; Krueger and Friedman 2009). These studies suggest that relationships with spouses and household children are important for sleep, but do not address whether the amount of contact with other close family members, such as parents, siblings, or adult children, may also matter for sleep. These other types of family relationships may represent the closest social relationships available to unmarried individuals or those who do not have children living at home.

In addition to lacking a focus on the broader family context, previous studies offer only limited insight into the importance of qualitative aspects of relationships for sleep. Studies of marital quality and sleep have found associations between being happy and satisfied in one’s marriage and reporting less frequent sleep problems (Prigerson, Maciejewski, and Rosenheck 1999; Troxel et al. 2009). In addition, prior research has found that older adults who characterized their significant social relationships positively (e.g., warm, trusting, and satisfying) also reported having better sleep (Friedman et al. 2007). However, a study of married women at midlife found no association between sleep quality and having relationships with others who provide companionship and support (Troxel et al. 2009). The measure of relationship quality in that study pertained to any social relationships and therefore did not necessarily capture family relationship quality. With the exception of these scattered studies, there is a dearth of research specifically examining links between sleep and the quality of family relationships, particularly those with nonspousal family members.

Another shortcoming of existing research on social relationship quality and sleep is the exclusive focus on positive aspects of relationships. This is an oversight because in addition to providing support, social relationships can also be a source of conflict and demands. Both social support and strain are considered key mechanisms through which social relationships influence health behaviors (Umberson et al. 2010). Moreover, theory and evidence indicate that negative aspects of social relationships may be more consequential for well-being than positive aspects (Mavandadi et al. 2007; Newsom et al. 2005). However, we know of no study that has examined the importance of both positive and negative aspects of family relationships for sleep.

How Do Families Influence Sleep?

Relationships with close family members are among our most important social relationships and prior research suggests that they have important implications for sleep. However, we lack an established conceptual framework that details the mechanisms by which families influence sleep. We draw on theory and empirical research on social relationships and health to develop expectations for how family relationships affect sleep. The literature on social relationships highlights three key mechanisms linking family relationships to health: provision of social support, relational demands and conflicts, and companionship (House, Umberson, and Landis 1988; Thoits 2011).

Social support, an aspect of relationship quality that has received considerable attention in research on social relationships and health, is one potential mechanism through which family relationships influence sleep. Supportive family relationships provide emotional care and comfort that can promote health or buffer against the harms of stressful events and chronic problems (Cohen, Gottlieb, and Underwood 2000; Thoits 2011). Previous research has found that supportive family relationships encourage good health habits, such as physical activity and quitting smoking (Murray et al. 1995; Treiber et al. 1991), and family support may also facilitate good sleep habits. In addition, individuals who can rely on family members to help them cope with stressful events and situations may be less likely to experience emotional distress (Kawachi and Berkman 2001; Thoits 1986), thereby limiting the effects of stressors that would otherwise undermine sleep quality (Morin, Rodrigue, and Ivers 2003). We therefore expect that supportive family relationships will be associated with less trouble sleeping.

We also consider whether negative qualities of family relationships are harmful to sleep. Relationships characterized by conflict and demands represent a major source of stress in people’s lives (Burg and Seeman 1994; Rook 1984), and relationship stress may contribute to poor health habits (Umberson, Liu, and Reczek 2008). Using data from sleep diaries and sleep laboratory-based measurements, Hall and colleagues (2000) found that stress-related intrusive thoughts and subjectively assessed stress burden were significant contributors to sleep disturbances. Individuals who are upset about family conflicts or are thinking about how to handle the demands of family members may find it difficult to fall asleep or stay asleep. Thus, our expectation is that strained family relationships will be associated with more troubled sleep.

Interactions with family members likely have both positive and negative content, and family support and strain may jointly influence sleep. Prior research shows that psychological distress is greatest when relationship demands are high and emotional support is low (Durden, Hill, and Angel 2007). Thus, strained and demanding family relationships may be particularly harmful to sleep in the absence of the buffering influence of supportive family relationships. We therefore expect that the association between family strain and troubled sleep will be greatest in the absence of social support and that family support will be most beneficial to sleep when family relationships have minimal strain.

In addition to incorporating a broader range of relationship types and more attention to the positive and negative components of relationship quality, we also consider the amount of contact individuals have with their family members in assessing how these relationships are associated with sleep. There is considerable theory and evidence for the health benefits of social connections and the harms of social isolation (Berkman et al. 2000; House et al. 1988). The companionship that results from social contact may enhance health, while the loneliness that arises from lack of companionship may be detrimental to health and well-being (Thoits 2011). Prior research has shown that feelings of loneliness are associated with worse sleep (Cacioppo, Hawkley, and Crawford 2002). Frequent contact with family members may provide companionship and reduce feelings of loneliness (Shiovitz-Ezra and Leitsch 2010), thereby decreasing the likelihood of experiencing poor sleep. This leads us to expect that more frequent family contact will be associated with less trouble sleeping.

Although we hypothesize that family contact will generally benefit sleep, we also recognize that interactions with family members can consist of both positive and negative exchanges and that the association between family contact and sleep may depend on the extent of support and strain present in family relationships. For example, more frequent contact with supportive family ties may be beneficial to sleep, while more frequent contact with stressful family ties may be detrimental for sleep. Therefore, we further expect that family contact will only benefit sleep for individuals whose family relationships are characterized as being highly supportive, but that family contact will be associated with more trouble sleeping among individuals whose family relationships are highly strained.

Data and Methods

Data

We use the 1995 survey of Midlife in the United States (MIDUS), a nationally representative random-digit-dial sample of noninstitutionalized adults aged 25 to 74, with oversampling of men and older adults. The overall response rate for completion of the phone and self-administered interview is estimated to be about 60.8 percent. We limit our analysis to the 3,034 respondents who completed both the phone interview and the self-administered questionnaire. After excluding 163 respondents missing information on the dependent variable and other covariates, our final analytic sample consisted of 2,871 individuals. Compared to Current Population Survey data (Bureau of Labor Statistics 1994), the study sample slightly underrepresents minorities and those with low income and low education.

Measures

Troubled sleep

Troubled sleep is assessed with the question: “During the past 30 days, how often have you experienced trouble getting to sleep or staying asleep?” The response categories were not at all, once a month, several times a month, once a week, several times a week, and almost every day. Reports of difficulty initiating or maintaining sleep have been used in other large, population-based studies (Grandner et al. 2010) and have been shown to be reliable measures for distinguishing between good and poor sleep quality (Buysse et al. 1989). Prior studies of sleep quality dichotomized responses to compare those who rarely have sleep problems to those who report more frequent experiences of sleep problems (Burgard and Ailshire 2009; Grandner et al. 2010). We further differentiate among respondents who report more frequent sleep problems by coding responses to our troubled sleep measure into three categories: not at all or never (reference), monthly (once a month-several times a month), or weekly-daily (once a week–almost every day).

Family contact and relationship quality

We consider respondent assessments of the frequency of contact with family members who do not live with them as well as assessments of emotional support and strain from nonspousal family members. Family contact is measured with the question: “How often are you in contact with any members of your family—that is, any of your brothers, sisters, parents, or children who do not live with you—including visits, phone calls, letters, or electronic mail messages?”¹ The response categories were: never or hardly ever = 1, less than once a month = 2, about once a month = 3, two or three times a month = 4, about once a week = 5, several times a week = 6, about once a day = 7, and several times a day = 8. We treat family contact as a continuous measure (ranging from 1 to 8), with higher scores indicating more frequent family contact. Although categorical specifications of the family contact measure yielded substantively similar results, we choose to use the continuous measure because it facilitated interpretation of interactions in the analytic models.

Social support is a four-item scale assessing perceived emotional support from family members (α = .82). Respondents were asked how much family members (not including spouses or partners): (1) really care about you? (2) understand the way you feel about things? (3) can be relied on for help if you have a serious problem? and (4) can be opened up to if you need to talk about your worries? The four response categories ranged from a lot to not at all; responses were reverse-coded so that higher scores indicated more support. Strain is a four-item scale assessing negative interactions with family members (α = .80). Respondents were asked how much family members (not including spouses or partners): (1) make too many demands on you? (2) criticize you? (3) let you down when you are counting on them? and (4) get on your nerves? The four response categories ranged from often to never; responses were reverse-coded so that higher scores indicated more strain. The support and strain scales are described in greater detail by Walen and Lachman (2000).

In addition to the independent effects of support and strain on troubled sleep, we also consider the combination of support and strain. We created a categorical measure capturing combinations of family support and strain. First, we created two dichotomous indicators representing either high support or high strain that provide a contrast between those reporting median or above median scores on the support or strain scales and those who report lower than median scores for support or strain. We then combined the low/high dichotomous indicators into a four-category measure of high support/low strain (reference), high support/high strain, low support/low strain, and low support/high strain.

Control variables

Multivariate analyses include adjustments for age, gender, race-ethnicity, marital status, parental status, education, household income, and employment status. Age is measured in years and treated continuously. Gender is coded 0 for males and 1 for females. We use dichotomous variables for race that distinguish between whites (reference), blacks and/or African Americans, Hispanics, and “other” race-ethnic groups.

Marital status is a categorical variable that distinguishes among those who were married (reference), divorced or separated, widowed, or never married. Parental status is assessed with respect to both whether one has children and the ages of the children and is coded into three categories: no children (reference), any children under 18 years of age, and only children ages 18 or older.

Educational attainment is coded categorically as less than high school, high school, and some college or more (reference). Annual household income is measured in 1995 dollars and treated continuously in $10,000 increments. We also include an indicator of whether the respondent was unemployed or not in the labor force and therefore not working for pay at the time of the survey.

We also adjust for self-rated physical health and depression, which are significant health risk factors for poor sleep (Vgontzas and Kales 1999) that may also influence perceptions of family relationships. Self-rated physical health is treated as a continuous measure and is assessed with the question: “In general, would you say your physical health is poor = 1, fair = 2, good = 3, very good = 4, or excellent = 5?” Given robust associations between sleep disturbances and major depressive episodes (Ohayon 2002), we also include a measure of depression, adapted from the Composite International Diagnostic Interview (CIDI), that is a count of zero to six symptoms, experienced for two weeks or longer in the year prior to the interview, that typically accompany depression and are used in clinical settings to diagnose major depression; symptoms include problems with lack of interest, appetite, energy, concentration, feelings of self-worth, and suicidal thoughts (Kessler, Mickelson, and Williams 1999).²

Analytic Strategy

The influence of family relationship contact and quality on sleep troubles is assessed through multinomial logistic regression analysis, which estimates the log-odds of being in a more frequent sleep trouble category (monthly, weekly-daily) compared with the reference category (no sleep trouble). This allows us to identify family relationship characteristics that are associated with more frequent troubled sleep, compared to the ideal situation of not experiencing sleep troubles. We first examine the independent associations of family contact, strain, and support with troubled sleep, as well as the joint contribution of support and strain to sleep troubles. We then examine interactions between frequency of family contact and support, strain, and support/strain combinations. Multivariate models adjust for sociodemographic and health characteristics. All analyses are weighted to account for differential probability of selection and nonresponse and were conducted using Stata/SE 11.

Results

Descriptive statistics

Table 1 shows descriptive statistics for all analysis variables. Slightly less than half of respondents (48 percent) reported never experiencing trouble sleeping, while 29 percent had problems sleeping on a monthly basis and 23 percent had problems sleeping on a weekly or daily basis. On average, respondents reported being in contact with family members several times a week (mean = 5.9), perceived high levels of emotional support from their family relationships (mean = 3.4), and rarely perceived their family relationships to be strained and demanding (mean = 2.1). About one-third of respondents reported having a combination of high support/low strain relationships and another one-third reported having a combination of low support/high strain relationships. The remaining respondents reported low support/low strain (21 percent) or high support/high strain (15 percent) combinations.

Table 1.

Descriptive Statistics, Midlife in the United States (MIDUS) 1995 (N = 2,871)

	Mean	SD
Troubled sleep frequency
Never	.48
Monthly	.29
Weekly-daily	.23
Family relationship characteristics
Family contact	5.9	1.6
Relationship quality
Support	3.4	.6
Strain	2.1	.6
High support/low strain	.32
Low support/low strain	.21
High support/high strain	.15
Low support/high strain	.31
Control variables
Age, years	45.3	13.5
Female	.56
Race-ethnicity
White	.81
Black	.10
Hispanic	.06
Other	.03
Marital status
Married	.69
Divorced/separated	.16
Widowed	.05
Never married	.11
Parental status
No children	.17
Children under 18 years	.45
Children 18 years or older	.38
Education
Less than high school	.13
High school or equivalent	.39
Some college or more	.49
Household income, dollars	39,536	36,622
Not working for pay	.28
Self-rated physical health	3.4	1.0
Depression	.6	1.6

Notes: Means and standard deviations are presented for continuous variables; proportions are shown for categorical measures. Numbers are weighted.

The average age of respondents was 45, women constituted just over half the sample, and a majority of respondents were white, married, had children, and had at least a high school level of education. The average household income was about $39,500 and nearly 28 percent of respondents were not working for pay at the time of the interview. Respondents reported their physical health as good to very good (mean = 3.4) and reported experiencing less than one depressive symptom on average.

Associations between family relationship characteristics and troubled sleep frequency

Results from multinomial logistic regression models for the influence of family contact and relationship quality on troubled sleep frequency are shown in Table 2. Relative risk ratios are presented, which indicate the probability of experiencing troubled sleep either monthly (first column for each model) or weekly-daily (second column for each model) divided by the probability of never experiencing troubled sleep. Ratios between zero and one indicate less risk of experiencing troubled sleep, and ratios greater than one indicate more risk of experiencing troubled sleep.

Table 2.

Multinomial Logistic Regression Estimates of the Risk of Monthly and Weekly Sleep Trouble Relative to No Sleep Trouble (N = 2,871)

	Model 1		Model 2		Model 3		Model 4		Model 5
	Monthly	Weekly-Daily	Monthly	Weekly-Daily	Monthly	Weekly-Daily	Monthly	Weekly-Daily	Monthly	Weekly-Daily
Family relationship characteristics
Contact	1.03	1.09*	1.06	1.13**	1.03	1.08*	1.05	1.10*	1.05	1.10*
	(.03)	(.04)	(.04)	(.04)	(.03)	(.04)	(.04)	(.04)	(.04)	(.04)
Support			.82*	.76**			.86	.89
			(.07)	(.07)			(.08)	(.09)
Strain					1.18	1.49***	1.11	1.42**
					(.10)	(.15)	(.10)	(.16)
Support/strain (reference = high support/low strain)
Low support/low strain									1.36*	1.32†
									(.19)	(.22)
High support/high strain									1.16	1.34†
									(.18)	(.24)
Low support/high strain									1.34*	1.74***
									(.18)	(.26)
Constant	1.15	1.21	1.79	2.21	.79	.46	1.28	.66	.78	.70
	(.52)	(.58)	(.90)	(1.17)	(.39)	(.25)	(.74)	(.43)	(.38)	(.36)
Log likelihood	−2,867		−2,861		−2,859				–2,862

Notes: Estimates of relative risk are presented. Standard errors are in parenthesis. Models are adjusted for age, gender, race-ethnicity, marital status, parental status, education, income, working status, self-rated physical health, and depression.

†

p < .10. *p < .05. **p < .01. ***p < .001.

Model 1 shows the association between family contact and troubled sleep, adjusted for all control variables. Contrary to our expectations, increasing frequency of family contact is associated with greater risk of experiencing weekly-daily sleep troubles, though it is not associated with risk of moderately frequent (i.e., monthly) troubled sleep. Models 2 through 5 further add relationship quality indicators. The association between family contact frequency and troubled sleep persists even after accounting for levels of relationship support and strain.

Results from model 2 show that having more supportive family relationships is associated with lower risk of troubled sleep. By contrast, model 3 shows that family strain is associated with greater risk of troubled sleep, particularly very frequent troubled sleep. In model 4 we include both support and strain to determine how they relate to troubled sleep independently of each other. When support and strain are considered together, only strain continues to be associated with troubled sleep, suggesting that the greater risk of troubled sleep associated with family strain persists even among individuals reporting similar levels of family support. Finally, model 5 suggests that family support has a buffering effect on the association between family strain and troubled sleep. Compared to relationships characterized as being high in support and low in strain, all other combinations of support and strain are associated with a greater risk of having troubled sleep. Relationships low in support but high in strain carry the greatest risk of very frequent (i.e., daily) sleep troubles.

Associations between control variables and troubled sleep were consistent across models and with findings reported in previous studies (coefficients are not shown but full results are available upon request). Inclusion of covariates did not change the associations between family relationship characteristics and troubled sleep frequency, although the magnitude of the coefficients for family support and strain were slightly reduced with the adjustment for self-rated physical health and depression.

Interactions between family contact frequency and relationship quality

In Table 3 we test how the association between relationship quality and troubled sleep varies according to family contact frequency by including interactions between contact and support, strain, and support/strain combinations. All models are adjusted for the control variables but we only present coefficients for the main effects and interaction terms (full results available upon request). As model 1 shows, the interaction between family contact and family support is not significant. However, model 2 shows a significant interaction between contact and strain, and this interaction is particularly strong with respect to weekly-daily frequency of troubled sleep. We plot this interaction in Figure 1.³ The interaction suggests that for individuals who never experience family strain, the probability of having troubled sleep declines as family contact becomes more frequent. However, as the frequency of contact increases among those who have some or frequent strain in their family relationships, the risk of troubled sleep increases. For example, among individuals who report often experiencing family strain, the probability of having very frequent troubled sleep increases from 24 percent when family contact occurs only a few times a month to 46 percent when family contact occurs several times a day. Consequently, the lowest and highest probability of frequent troubled sleep occurs among individuals who are in contact with family members multiple times a day, suggesting the importance of considering both the positive and negative aspects of family exchange.

Table 3.

Multinomial Logistic Regression Estimates of Interactions between Family Contact and Relationship Quality (N = 2,871)

	Model 1		Model 2		Model 3
	Monthly	Weekly-Daily	Monthly	Weekly-Daily	Monthly	Weekly-Daily
Family relationships characteristics
Contact	1.32†	1.34*	.86	.79†	.92	.93
	(.20)	(.19)	(.09)	(.10)	(.06)	(.07)
Support	1.17	1.01
	(.30)	(.25)
*Contact	.93	.95
	(.04)	(.04)
Strain			.73	.65
			(.21)	(.21)
*Contact			1.09†	1.15**
			(.05)	(.06)
Support/strain (reference = high support/low strain)
Low support/low strain					.31*	.57
					(.17)	(.35)
*Contact					1.29**	1.14
					(.12)	(.12)
High support/high strain					.87	.13*
					(.62)	(.11)
*Contact					1.05	1.43**
					(.12)	(.18)
Low support/High strain					.52	.41
					(.28)	(.25)
*Contact					1.17†	1.27*
					(.10)	(.13)

†

p < .10. *p < .05. **p < .01. ***p < .001.

Figure 1.

Probability of Weekly-Daily Sleep Trouble by Family Relationship Strain and Contact Frequency

We also find significant interactions between family contact and support/strain combinations. As shown in model 3, the interactions are strongest for weekly-daily troubled sleep, with the exception of the interaction between contact and low support/low strain, which is stronger for monthly troubled sleep. Figure 2 illustrates the interactions for weekly-daily troubled sleep.⁴ Among individuals whose family relationships are high in support but low in strain, the probability of having troubled sleep declines from about 20 percent to 17.5 percent with increasing family contact. Individuals with family relationships low in both support and strain have a similar probability of troubled sleep regardless of the frequency with which they are in contact with their family. Among individuals whose family relationships are high in strain, the probability of troubled sleep increases with more frequent family contact.

Figure 2.

Probability of Weekly-Daily Sleep Trouble by Family Contact Frequency and Combinations of Family Relationship Support and Strain

Additional analyses

We include a number of control variables in our analyses to reduce the likelihood that unmeasured confounders influence the results presented previously, but we were unable to include two key potential confounders in our analyses: preexisting insomnia and the use of sleep aids. Although MIDUS respondents were not asked about preexisting insomnia specifically, they were asked if they had experienced or been treated for chronic sleeping problems in the 12 months prior to their interview. In addition, the data do not include measures of sleep aid use but do include a measure of whether respondents used sedatives, including sleeping pills, on their own in the prior 12 months. Reliable measures of insomnia and sleep aid use were not available in the data, but we replicated all analyses with the additional inclusion of these indicators of prior experience of chronic sleep problems and sedative use. The coefficients for family relationship characteristics were somewhat attenuated with the inclusion of these additional covariates, but the results were otherwise unchanged.

One additional concern was whether the associations we observed were consistent across age and by gender. Thus we also examined interactions between all family relationship characteristics and both age and gender. We did not find any significant interactions in fully adjusted models, but in models unadjusted for health covariates there were significant interactions between frequency of family contact and age (rr = 1.1, p = 0.02) and family strain and female gender (rr = 1.4, p = 0.07) for weekly/daily sleep troubles.

Discussion and Conclusion

Efforts to identify the social determinants of poor sleep have not fully considered the role of family relationships, and there has been a notable absence of attention to sleep in much of the literature on social relationships and health. This is an oversight because sleep is a major daily activity that everyone engages in and one that could be highly responsive to the content of social exchanges. Exchanges with family in particular may have considerable importance for sleep quality. In this study we bring together theoretical and empirical research on social relationships with evidence from the growing sleep literature to highlight the importance of considering sleep in the context of the family and to gain insights into how family relationships influence a key health behavior.

The findings from this study show that family relationship characteristics influence sleep in ways largely consistent with prior theory and research on social relationships and health. Family support was associated with less troubled sleep, and family strain was associated with more troubled sleep. Importantly, family strain was a stronger predictor of experiencing troubled sleep than was family support. However, we also found that family strain is most harmful to sleep when support is low, suggesting that the stress of family demands and conflict may be somewhat offset by supportive exchanges with family members. This finding is consistent with prior research that has found that psychological distress is greatest among people whose relationships are both overly demanding and lack emotional support (Durden et al. 2007).

In the absence of emotional support from family members, strained and demanding family relationships may contribute to sleep problems in several ways. Demanding family relationships may be particularly distressing when individuals feel they cannot rely on family members for assistance in dealing with their own serious problems. Furthermore, the combination of dealing with demands from family members in addition to one’s own problems may compound feelings of worry and anxiety that can disrupt sleep (Hall et al. 2000). Family influences on sleep are complex because family exchanges include both positive and negative content. This study underscores the importance of considering the joint distribution of both positive and negative aspects of relationship quality.

Our expectation that increased family contact would benefit sleep was not supported by the results, which showed that more frequent family contact was associated with greater risk of troubled sleep. Although social relationships are generally assumed to promote health and well-being, not all social contact benefits health and some forms of social contact may undermine health and health-related habits (Burg and Seeman 1994; Umberson et al. 2010). Our data demonstrate this point nicely, showing that the risk of troubled sleep associated with family strain increased as family contact became more frequent. Although we thought family contact would provide companionship and buffer against feelings of loneliness, frequent contact with family members can also be accompanied by increased demands and irritations that may interfere with sleep.

Our findings also suggest that family contact can be beneficial to sleep when relationships are not strained. This is more consistent with the prevailing perspective that having social ties and frequent social contact is beneficial to health and well-being. Thus, this study demonstrates how the role of social contact in promoting health largely depends on the content of the social exchange and provides support for the perspective that there are limits to the health benefits of social ties.

The study conclusions should be considered in light of several limitations. First, our data are cross-sectional and thus we are not able to assess the temporal ordering between sleep quality and family relationship characteristics. Prior research using longitudinal data suggests that relationship demands and support influence subsequent psychological well-being (Durden et al. 2007), and at least one study has shown that relationship quality affects subsequent sleep quality (Prigerson et al. 1999). However, the association between social relationships and sleep quality is likely reciprocal (Troxel et al. 2007). Stressful relationships may interfere with sleep, and as a result of inadequate sleep relationships may become more strained.

In addition, we were unable to disentangle the role of health status in the association between family relationships and sleep. We wanted our results to be robust to potential confounders and therefore included adjustments for physical and mental health characteristics that are known risk factors for poor sleep and that might influence frequency of family exchanges and the assessment of the content of social exchanges. However, it is possible that our health controls are actually on the causal pathway linking family relationships to sleep. It is also plausible that sleep mediates the association between social relationships and health (Cacioppo, Hawkley, and Crawford 2002). Longitudinal studies are needed to assess the temporal ordering of social exchanges, sleep quality, and health.

Third, our measure of sleep quality is based on self-reports of sleep trouble and may be subject to reporting error. However, research suggests that self-reports of diminished sleep quality are valid indicators of clinical sleep problems that may not be captured using objective sleep assessments (Regestein et al. 2004). Moreover, we use a measure of sleep quality that has been shown to differentiate between “good” and “poor” sleepers in population-based studies (Buysse et al. 1989; Levine et al. 2003). This measure does not identify individuals with clinically significant sleep disorders, such as insomnia or sleep apnea, but rather is a general measure of sleep quality that can be used to identify those experiencing troubled sleep in the larger U.S. population of both normal and disordered sleepers. However, our results were robust to the inclusion of an indicator of chronic sleeping problems, suggesting the effects of family relationships are not limited to those with sleep disorders.

A fourth limitation is that we lacked potentially important information on sleep-related behaviors. In 1999 an estimated one-quarter of Americans used prescription or over-the-counter medication to help them sleep (National Sleep Foundation 1999). Individuals experiencing family stress may be more likely to use sleep aids and as a result may report less troubled sleep than they would in the absence of these medications. Our data lacked optimal measures for exploring this possibility, but our findings were very similar in sensitivity analyses that included an approximate indicator of sleep medication use.

However, use of medication for sleep is more common among women than men. According to a 1999 Sleep in America Poll, 29 percent of women reported taking medication to help them sleep compared to only 19 percent of men (National Sleep Foundation 1999). The same report found that women were more likely to say that stress affected their sleep, suggesting that family-based stress may be more consequential for women’s sleep. Yet we found only weak evidence of a gender difference in the association between family strain and troubled sleep, though the interaction suggested family strain was more consequential for women’s sleep. However, because we were unable to fully account for sleep medication use in the analysis of gender interactions, we may have underestimated the association between family strain and troubled sleep among women.

Another limitation is that the measures of family characteristics do not specifically identify particular family members. Whereas most of the previous literature on social relationships and health and nearly all of the previous research on sleep has focused on marital and parental relationships, we examined nonspousal family relationships, such as those with parents, siblings, and grown children. Our study findings are thus applicable to people who are not married and who both do and do not have children. However, our measures of family relationship characteristics are not specific to particular family members and we thus cannot identify which family relationships are most beneficial or harmful to sleep. For instance, demands from children may be more detrimental to sleep than a general sense of irritation with siblings. In addition, we are not able to determine if assessments of support and strain are based on interactions with one family member or multiple family members. However, it may be the case that it does not matter who the source of conflict or support is, but simply that one’s family relationships are perceived as being supportive or strained.

This study makes important contributions to research on social relationships and health and the burgeoning literature on the social determinants of sleep. Whereas many studies of the health effects of social ties, and nearly all research on families and sleep, have focused on the role of marriage and parenting, our study assessed the influence of family relationships more generally and demonstrated the importance of these understudied relationships. In addition, by analyzing the independent and interactive effects of support and strain, rather than focusing on their separate effects, our analysis provides a more balanced view of how negative and positive aspects of relationships combine to influence sleep quality. Finally, by highlighting the importance of family relationships for sleep, our study provides further evidence that social relationships may be a key factor contributing to sleep problems among U.S. adults.

Footnotes

Acknowledgements

We thank the editor and reviewers for their helpful comments. A public-use version of Midlife in the United States (MIDUS) is available from the Inter-university Consortium for Political and Social Research at the University of Michigan. An earlier version of this article was presented at the 2009 meeting of the American Sociological Association.

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by grants P30AG17265 and T32AG0037 from the National Institute on Aging to the University of Southern California.

Notes

Bios

Jennifer A. Ailshire is a National Institute on Aging postdoctoral fellow in the Multidisciplinary Training Program in Gerontology at the University of Southern California and the USC/UCLA Center for Biodemography and Population Health. Her research focuses on neighborhood context, social relationships, and health and well-being over the life course. She recently received funding from the National Institute on Aging to study the psychosocial and physiological pathways linking neighborhoods and social relationships to health in older adults.

Sarah A. Burgard is associate professor of sociology and epidemiology and research associate professor at the Population Studies Center at the University of Michigan. Her research focuses on the ways that stratification by race-ethnicity, gender, and socioeconomic position influence people’s opportunities in life and how this influences their health. Burgard uses large-scale social surveys and other data to examine a variety of health outcomes that chart the boundaries of the healthy life course. She is currently engaged in several data collection projects aimed at understanding links between the Great Recession and population health.

References

Berkman

Lisa F.

Breslow

Lester

. 1983. Health and Ways of Living: The Alameda County Study. New York: Oxford University Press.

Berkman

Lisa F.

Glass

Thomas

Brissette

Ian

Seeman

Teresa E.

2000. “From Social Integration to Health: Durkheim in the New Millennium.” Social Science & Medicine 51:843–57.

Bureau of Labor Statistics. 1994. Current Population Survey. Washington, DC: Bureau of the Census.

Burg

Matthew M.

Seeman

Teresa E.

1994. “Families and Health: The Negative Side of Social Ties.” Annals of Behavioral Medicine 16:109–15.

Burgard

Sarah A.

2011. “The Needs of Others: Gender and Sleep Interruptions for Caregiving.” Social Forces 89:1189–215.

Burgard

Sarah A.

Ailshire

Jennifer A.

2009. “Putting Work to Bed: Stressful Experiences on the Job and Sleep Quality.” Journal of Health and Social Behavior 50:476–92.

Buysse

Daniel J.

Reynolds

Charles F.

Monk

Timothy H.

Berman

Susan R.

Kupfer

David J.

1989. “The Pittsburgh Sleep Quality Index: A New Instrument for Psychiatric Practice and Research.” Psychiatry Research 28:193–213.

Cacioppo

John T.

Hawkley

Louise C.

Berntson

Gary G.

Ernst

John M.

Gibbs

Amber C.

Stickgold

Robert

Hobson

J. Allen

. 2002. “Do Lonely Days Invade the Nights? Potential Social Modulation of Sleep Efficiency.” Psychological Science 13:384–87.

Cacioppo

John T.

Hawkley

Louise C.

Crawford

L. Elizabeth

Ernst

John M.

Burleson

Mary H.

Kowalewski

Ray B.

Malarkey

William B.

Van Cauter

Eve

Berntson

Gary G.

2002. “Loneliness and Health: Potential Mechanisms.” Psychosomatic Medicine 64:407–17.

10.

Cappuccio

Francesco P.

D’Elia

Lanfranco

Strazzullo

Pasquale

Miller

Michelle A.

2009. “Quantity and Quality of Sleep and Incidence of Type 2 Diabetes: A Systematic Review and Meta-Analysis.” Diabetes Care 33:414–20.

11.

Cohen

Sheldon

Gottlieb

Benjamin H.

Underwood

Lynn G.

2000. “Social Relationships and Health.” Pp. 3–28 in Social Support Measurement and Intervention: A Guide for Health and Social Scientists, edited by Cohen

Underwood

Gottlieb

B. H.

, and Fetzer Institute. Oxford: Oxford University Press.

12.

Dinges

David F.

1989. “The Nature of Sleepiness: Causes, Contexts, and Consequences.” Pp. 147–80 in Perspectives in Behavioral Medicine: Eating, Sleeping, and Sex, edited by Thorpy

M. J.

Billiard

Hillsdale, NJ: Lawrence Erlbaum.

13.

Durden

Emily D.

Hill

Terrence D.

Angel

Ronald J.

2007. “Social Demands, Social Supports, and Psychological Distress among Low-Income Women.” Journal of Social and Personal Relationships 24:343–61.

14.

Espie

Colin A.

2002. “Insomnia: Conceptual Issues in the Development, Persistence, and Treatment of Sleep Disorders in Adults.” Annual Review of Psychology 53:215–43.

15.

Friedman

Elliot M.

Love

Gayle D.

Rosenkranz

Melissa A.

Urry

Heather L.

Davidson

Richard J.

Singer

Burton H.

Ryff

Carol D.

2007. “Socioeconomic Status Predicts Objective and Subjective Sleep Quality in Aging Women.” Psychosomatic Medicine 69:682–91.

16.

Grandner

Michael A.

Patel

Nirav P.

Gehrman

Philip R.

Xie

Dawei

Sha

Daohang

Weaver

Terri

Gooneratne

Nalaka

. 2010. “Who Gets the Best Sleep? Ethnic and Socioeconomic Factors Related to Sleep Complaints.” Sleep Medicine 11:470–78.

17.

Hale

Lauren

. 2005. “Who Has Time to Sleep?” Journal of Public Health 27:205–11.

18.

Hall

Martica

Buysse

Daniel J.

Nowell

Peter D.

Nofzinger

Eric A.

Houck

Patricia

Reynolds

Charles F.

Kupfer

David J.

2000. “Symptoms of Stress and Depression as Correlates of Sleep in Primary Insomnia.” Psychosomatic Medicine 62:227–30.

19.

Hislop

Jenny

Arber

Sara

. 2003. “Sleepers Wake! The Gendered Nature of Sleep Disruption among Mid-Life Women.” Sociology 37:695–711.

20.

House

James S.

Umberson

Debra

Landis

Karl R.

1988. “Structures and Processes of Social Support.” Annual Review of Sociology 14:293–318.

21.

Kawachi

Ichiro

Berkman

Lisa F.

2001. “Social Ties and Mental Health.” Journal of Urban Health: Bulletin of the New York Academy of Medicine 78:458–67.

22.

Kessler

Ronald C.

Mickelson

Kristin D.

Williams

David R.

1999. “The Prevalence, Distribution, and Mental Health Correlates of Perceived Discrimination in the United States.” Journal of Health and Social Behavior 40:208–230.

23.

Knutson

Kristen L.

Ryden

Armand M.

Mander

Bryce A.

Van Cauter

Eve

. 2006. “Role of Sleep Duration and Quality in the Risk and Severity of Type 2 Diabetes Mellitus.” Archives of Internal Medicine 166:1768–74.

24.

Krueger

Patrick M.

Friedman

Elliot M.

2009. “Sleep Duration in the United States: A Cross-Sectional Population-Based Study.” American Journal of Epidemiology 169:1052–63.

25.

Lauderdale

Diane S.

Knutson

Kristen L.

Yan

Lijing L.

Rathouz

Paul J.

Hulley

Stephen B.

Sidney

Steve

Liu

Kiang

. 2006. “Objectively Measured Sleep Characteristics among Early-Middle-Aged Adults: The CARDIA Study.” American Journal of Epidemiology 164:5–16.

26.

Levine

Douglas W.

Kripke

Daniel F.

Kaplan

Robert M.

Lewis

Megan A.

Naughton

Michelle J.

Bowen

Deborah J.

Shumaker

Sally A.

2003. “Reliability and Validity of the Women’s Health Initiative Insomnia Rating Scale.” Psychological Assessment 15:137–48.

27.

Mallon

Broman

J. E.

Hetta

2002. “Sleep Complaints Predict Coronary Artery Disease Mortality in Males: A 12-Year Follow-Up Study of a Middle-Aged Swedish Population.” Journal of Internal Medicine 251:207–16.

28.

Maume

David J.

Sebastian

Rachel A.

Bardo

Anthony R.

2009. “Gender Differences in Sleep Disruption among Retail Food Workers.” American Sociological Review 74:989-1007.

29.

Maume

David J.

Sebastian

Rachel A.

Bardo

Anthony R.

2010. “Gender, Work-Family Responsibilities, and Sleep.” Gender & Society 24:746–68.

30.

Mavandadi

Shahrzad

Sorkin

Dara H.

Rook

Karen S.

Newsom

Jason T.

2007. “Pain, Positive and Negative Social Exchanges, and Depressive Symptomatology in Later Life.” Journal of Aging and Health 19:813–30.

31.

Morin

Charles M.

Rodrigue

Sylvie

Ivers

Hans

. 2003. “Role of Stress, Arousal, and Coping Skills in Primary Insomnia.” Psychosomatic Medicine 65:259–67.

32.

Murray

Robert P.

Johnston

Janet J.

Dolce

Jeffrey J.

Lee

Wondra Wong

O’Hara

Peggy

. 1995. “Social Support for Smoking Cessation and Abstinence: The Lung Health Study.” Addictive Behaviors 20:159–70.

33.

National Institutes of Health, National Center on Sleep Disorders Research. 2003. “2003 National Sleep Disorders Research Plan” (NIH publication 03-52). Bethesda, MD: U.S. Department of Health and Human Services.

34.

National Sleep Foundation. 1999. Sleep in America: 1999. Washington, DC: Author.

35.

National Sleep Foundation. 2005. 2005 Sleep in America Poll. Washington, DC: Author.

36.

Newsom

Jason T.

Rook

Karen S.

Nishishiba

Masami

Sorkin

Dara H.

Mahan

Tyrae L.

2005. “Understanding the Relative Importance of Positive and Negative Social Exchanges: Examining Specific Domains and Appraisals.” The Journals of Gerontology. Series B, Psychological Sciences and Social Sciences 60:P304–12.

37.

Ohayon

Maurice M.

2002. “Epidemiology of Insomnia: What We Know and What We Still Need to Learn.” Sleep Medicine Reviews 6:97–111.

38.

Patel

Sanjay R.

2007. “Social and Demographic Factors Related to Sleep Duration.” Sleep 30:1077–78.

39.

Phillips

Barbara

Mannino

David M.

2007. “Do Insomnia Complaints Cause Hypertension or Cardiovascular Disease?” Journal of Clinical Sleep Medicine: JCSM: Official Publication of the American Academy of Sleep Medicine 3:489–94.

40.

Prigerson

Holly G.

Maciejewski

Paul K.

Rosenheck

Robert A.

1999. “The Effects of Marital Dissolution and Marital Quality on Health and Health Service Use among Women.” Medical Care 37:858–73.

41.

Regestein

Quentin R.

Friebely

Joan

Shifren

Jan L.

Scharf

Martin B.

Wiita

Brinda

Carver

Judith

Schiff

Isaac

. 2004. “Self-Reported Sleep in Postmenopausal Women.” Menopause 11:198–207.

42.

Rook . 1984. “The Negative Side of Social Interaction: Impact on Psychological Well-Being.” Journal of Personality and Social Psychology 46:1097–108.

43.

Schwartz

Skai W.

Cornoni-Huntley

Joan

Cole

Stephen R.

Hays

Judith C.

Blazer

Dan G.

Schocken

Douglas D.

1998. “Are Sleep Complaints an Independent Risk Factor for Myocardial Infarction?” Annals of Epidemiology 8:384–92.

44.

Shiovitz-Ezra

Sharon

Leitsch

Sara A.

2010. “The Role of Social Relationships in Predicting Loneliness: The National Social Life, Health, and Aging Project.” Social Work Research 34:157–67.

45.

Thoits

Peggy A.

1986. “Social Support as Coping Assistance.” Journal of Consulting and Clinical Psychology 54:416–423.

46.

Thoits

Peggy A.

2011. “Mechanisms Linking Social Ties and Support to Physical and Mental Health.” Journal of Health and Social Behavior 52:145–61.

47.

Treiber

Frank A.

Baranowski

Tom

Braden

David S.

Strong

William B.

Levy

Maurice

Knox

Willie

. 1991. “Social Support for Exercise: Relationship to Physical Activity in Young Adults.” Preventive Medicine 20:737–50.

48.

Troxel

Wendy M.

Buysse

Daniel J.

Hall

Martica

Matthews

Karen A.

2009. “Marital Happiness and Sleep Disturbances in a Multi-Ethnic Sample of Middle-Aged Women.” Behavioral Sleep Medicine 7:2–19.

49.

Troxel

Wendy M.

Robles

Theodore F.

Hall

Martica

Buysse

Daniel J.

2007. “Marital Quality and the Marital Bed: Examining the Covariation between Relationship Quality and Sleep.” Sleep Medicine Reviews 11:389–404.

50.

Umberson

Debra

Crosnoe

Robert

Reczek

Corinne

. 2010. “Social Relationships and Health Behavior across the Life Course.” Annual Review of Sociology 36:139–57.

51.

Umberson

Debra

Liu

Hui

Reczek

Corinne

. 2008. “Stress and Health Behavior over the Life Course.” Pp. 19–44 in Advances in Life Course Research: Stress Processes across the Life Course, edited by Turner

H. A.

Schieman

New York: Elsevier.

52.

Venn

Susan

Arber

Sara

Meadows

Robert

Hislop

Jenny

. 2008. “The Fourth Shift: Exploring the Gendered Nature of Sleep Disruption among Couples with Children.” The British Journal of Sociology 59:79–97.

53.

Vgontzas

Alexandros N.

Kales

Anthony

. 1999 “Sleep and Its Disorders.” Annual Review of Medicine 50:387–400.

54.

Walen

Heather R.

Lachman

Margie E.

2000. “Social Support and Strain from Partner, Family, and Friends: Costs and Benefits for Men and Women in Adulthood.” Journal of Social and Personal Relationships 17:5–30.