Keywords

Sex is a key component of long-term relationships ; research consistently links higher levels of sexual frequency (DeLamater et al. 2008; DeLamater and Moorman 2007), sexual satisfaction (DeLamater et al. 2008; Gott and Hinchliff 2003b; Sprecher and Cate 2004), sexual desire (Skultety 2007), and an absence of sexual dysfunction (Laumann et al. 2008) to greater relationship satisfaction and stability (Sprecher and Cate 2004). The causal relationships between these variables are likely bidirectional (Sprecher and Cate 2004). For example, one longitudinal study found that sexual satisfaction positively influences marital quality (Yeh et al. 2006), yet it is just as likely the case that higher levels of relationship satisfaction lead to more satisfying and frequent sexual interactions (Sprecher and Cate 2004). In short, the quality and frequency of sexual experiences are an integral part of long-term relationships.

Most studies on sex in long-term couples, particularly those based on survey research, rely on measures of vaginal intercourse (although sometimes oral and anal sex are measured) (Lodge and Umberson forthcoming; Peplau et al. 2004). This is problematic, however, for a few reasons. First, although some studies suggest that heterosexual couples consider only penetrative, vaginal intercourse to be “real sex” (Lodge and Umberson 2012; Waite and Das 2010), other studies suggest that when vaginal intercourse is either not possible or desired, couples redefine the meaning of sexuality to include other physically intimate experiences (e.g., kissing, holding hands, cuddling) (Gott and Hinchliff 2003a, b; Lodge and Umberson 2012). Second, survey questions that ask about instances of “intercourse” may not adequately measure lesbian sexuality, and as a result estimates of sexual frequency in lesbian couples may be inaccurately low (Peplau et al. 2004). There is therefore a need for future studies to develop more inclusive and expansive measures of sexuality in long-term relationships.

Definitions of long-term relationships also vary, although many studies define long-term relationships as those lasting at least 7 years or more, based on the fact that the median marital duration for heterosexual divorcing couples is 7 years (Elliott and Umberson 2008; Lodge and Umberson 2012). Similar standards have been used to define long-term gay and lesbian cohabiting couples (Umberson et al. 2015b), although given that relationship duration may vary among different couple types (i.e., married versus cohabiting; gay versus lesbian versus heterosexual) (Moore and Stambolis-Ruhstorfer 2013), it is unclear if this is an appropriate benchmark for all types of relationships.

In this chapter, I first summarize the current state of knowledge on sexuality in long-term relationships. Second, I discuss the leading methodological approaches to studying sex in long-term relationships and suggest innovative methods for future research. Third, I discuss limitations of existing research on this topic and suggest directions for future research. The majority of research on sex in long-term relationships is descriptive and based on survey measures of sexual frequency, sexual satisfaction, sexual desire, and sexual dysfunction . Further, most research on this topic focuses on individuals in heterosexual marital relationships and as a result little is known about how individuals in other types of long-term relationships experience sex. As I discuss in this chapter, there is a need for more theoretically-informed studies on sex in long-term relationships, qualitative studies, dyadic research (i.e., research that studies both partners in a relationship), research on non-heterosexual relationships, non-marital relationships, and research on how social class and racial/ethnic diversity shape experiences of sex in long-term relationships.

1 Literature Review

1.1 Why Sex Matters for Relationships

Numerous studies suggest that sex is an integral component of same- and different-sex long-term relationships. Most couples remain sexually active into deep old age (Lindau et al. 2007), and the quantity and quality of sexual activities are linked to several indicators of relationship quality—including relationship satisfaction, feelings of love, commitment, and relationship stability (Sprecher and Cate 2004). Causal direction for these variables has been difficult to determine, although it is likely that these associations are bidirectional—individuals who are happier in their relationships are also more likely to have more frequent and satisfying sexual interactions at the same time that frequent and satisfying sexual interactions reinforce individuals’ positive feelings about their partner and relationship (Lodge and Umberson forthcoming; Sprecher and Cate 2004). However, it is important to note that some long-term relationships are characterized by infrequent or nonexistent sexual activity or low levels of sexual satisfaction, but high levels of relationship satisfaction (and vice versa); more research is needed on these “outlier” couples (Sprecher and Cate 2004).

1.2 Descriptive Studies

Kinsey et al. (1948, 1953) pioneered research on human sexuality. As a result, most contemporary social science research reflects this Kinseyian tradition in that is descriptive and survey-based, focusing most often on sexual frequency, levels of sexual satisfaction and sexual desire, sexual attitudes and recently—alongside what has been termed the “medicalization of sex” (Tiefer 1996)—incidence of sexual dysfunctions .

1.2.1 Sexual Frequency

Sexual frequency varies considerably in relationships, depending on a number of factors including relationship duration, age, union status (i.e., marital versus cohabiting relationship), and whether the relationship is composed of two men, two women, or a man and a woman. Most studies on sexual frequency in long-term relationships have focused on heterosexual marital relationships. Although estimates vary, studies based on nationally representative samples suggest that heterosexual married couples have sex on average between 6 and 7 times a month (Call et al. 1995; Laumann et al. 1994; Michael et al. 1994).

As previously noted, although most surveys ask about frequency of vaginal intercourse, other studies ask about the frequency of other sexual activities (Laumann et al. 1994). For example, data from the National Health and Social Life Survey (NHSLS) found that although most respondents reported having vaginal sex at the last instance of sex (95 %), 30 % reported having oral sex and 1 to 2 % reported having anal sex the last time they had sex (Laumann et al. 1994). Further, most respondents said they had engaged in oral sex at some point in their lifetime and 10 % reported engaging in anal sex at some point in their lifetime (Laumann et al. 1994). Blumstein and Schwartz (1983) found higher estimates of oral sex; their non-representative data suggested that 50 % of gay couples, 39 % of lesbian couples, and 30 % of heterosexual couples usually or always engage in oral sex. Laumann et al. (1994) also found that young adults, White adults, and adults with higher levels of education were more likely to report engaging in oral and anal sex, compared to older adults, Black and Hispanic adults, and individuals with lower levels of education. More recent data from the National Survey of Sexual Health and Behavior (NSSHB) indicate that more than 50 % of men and women ages 18–49 report having oral sex in the past year, while 20 % of men ages 25–49 and women ages 20–39 report having anal sex in the past year (Herbenick et al. 2010).

Numerous studies suggest that sexual frequency declines over time in all types of long-term relationships (Willetts et al. 2004). Research suggests that the most important reason for this decline is habituation—that is decreased interest in sex resulting from the predictability of sex with a particular partner (Call et al. 1995; Peplau et al. 2004). One study on marital duration and sexual frequency, however, found that the most precipitous decline in sexual frequency occurs during the first year of marriage (Call et al. 1995). This phenomenon is typically referred to as the “honeymoon effect,” whereby levels of sexual frequency become more routine and predictable (Call et al. 1995).

In addition to habituation, age is a major reason that sexual frequency declines over time in long-term relationships (DeLamater and Moorman 2007; Kontula and Haavio-Mannila 2009; Lindau et al. 2007). For example, one study, using nationally representative survey data, found that whereas heterosexual married couples had sex on average 6.3 times per month, couples under the age of 24 reported having sex 11.7 times per month and the frequency of sex declined with each subsequent age group to 3 times a month for couples over the age of 65 (Call et al. 1995). However, while age is associated with a decline in vaginal and oral sex, age is not associated with frequency of kissing, hugging, caressing, and sexual touching (AARP 2005). Evidence further suggests that although cohort or generational differences may explain some of the decline in sexual frequency over the course of relationships (because older cohorts tend to engage in less sex in mid and later life than more recent cohorts), age is a more important predictor of sexual frequency (DeLamater and Moorman 2007; Edwards and Booth 1994). Although some of the age-related decline in sexual frequency in long-term relationships stems from physical health problems experienced by one or both partners that limit sexual activity (DeLamater et al. 2008; DeLamater and Moorman 2007), age remains an independent and significant correlate of lower levels of sexual frequency in long-term relationships (Karraker and DeLamater 2013; Karraker et al. 2011).

Several studies suggest that sexual frequency varies by relationship type and union status. For example, gay male couples and heterosexual cohabiting couples have higher rates of sexual frequency than heterosexual married couples, who in turn have higher rates of sexual frequency than lesbian couples (Peplau et al. 2004; Willetts et al. 2004). It is not clear why these differences exist, although sexual frequency may be higher in cohabiting compared to marital relationships because of the less traditional characteristics of cohabiting relationships. It is also plausible that couples who have higher levels of sexual frequency are also couples who are less likely to marry. As discussed later in this chapter, differences in levels of sexual frequency between lesbian, gay, and heterosexual couples may stem from—in part—different gender compositions within these couples. Greater sexual frequency among gay couples, compared to heterosexual married couples, may also be attributable to the fact that gay couples have traditionally been denied access to the institution of marriage (given that cohabiting couples have higher levels of sexual frequency compared to married couples). In other words, it is plausible that cohabiting gay couples have higher rates of sexual frequency compared to married gay couples and because surveys have most often relied on samples of cohabiting gay couples that this may explain some of the difference in sexual frequency between gay couples and heterosexual married couples, although research has not explored this possibility.

Other factors associated with sexual frequency include relationship satisfaction (as previously discussed) (Sprecher and Cate 2004; Willetts et al. 2004) and gender (Willetts et al. 2004). In terms of gender, data from the NHSLS indicate that heterosexual married men report having sex (partnered or unpartnered) 6.9 times a month, compared to 6.5 times a month for heterosexual married women (Laumann et al. 1994; Michael et al. 1994). These differences may be due to gender differences in reporting (whereby men may overestimate sexual frequency and/or women may underestimate sexual frequency) and the fact that men may be more likely than women to have sex outside of the marital relationship (Willetts et al. 2004). Some studies also suggest that living in a rural area, being Catholic, and having a demanding job are associated with lower levels of sexual frequency, although race, ethnicity, social class, and religion generally do not appear to be correlated with levels of sexual frequency (Willetts et al. 2004).

1.2.2 Sexual Satisfaction

As previously noted sexual satisfaction is a key component of relationship satisfaction. Sexual satisfaction is rarely defined in the literature but instead is typically measured with one subjective question: “How satisfied are you with your sex life together?” (Schwartz and Young 2009). Levels of sexual satisfaction are positively related to levels of love, commitment, and relationship quality (Sprecher and Cate 2004). The causal relationship between these variables is not well established, although some longitudinal evidence suggests that changes in sexual satisfaction are linked to changes in relationship satisfaction (Sprecher 2002). Further, longitudinal evidence links lower levels of sexual satisfaction to subsequent relationship dissolution (Sprecher and Cate 2004). Indeed, given evidence that most people in committed relationships are sexually satisfied—Laumann et al. (1994) found that 88 % of heterosexual married respondents report being extremely or very sexually satisfied—it is likely that relationships wherein one or both partners are not sexual satisfied are less likely to last.

Sexual frequency is consistently correlated with levels of sexual satisfaction; individuals who have more frequent sex also report greater levels of sexual satisfaction (Smith et al. 2011; Sprecher and Cate 2004). Again, the causal direction of this relationship is unclear, but likely bidirectional. Individuals who have positive feelings about their sexual encounters are likely to want to have more frequent sex, at the same time that more frequent sex is likely to result in greater frequency of orgasm and in turn, greater sexual satisfaction.

Although sexual frequency predicts sexual satisfaction, and relationship duration and age are negatively related to sexual frequency, numerous studies find that age and relationship duration are not related to sexual satisfaction (McKinlay and Feldman 1994; Sprecher and Cate 2004; Ventegodt 1998). In terms of age, however, research has produced inconsistent results; some studies find that sexual satisfaction increases with age (Gullette 2011; Vares et al. 2007), while still other studies have found that sexual satisfaction declines with age (AARP 2005, 2010). One possibility for these discrepancies may be that cohort and age are confounded in several studies on this topic and that cohort differences—not age—explain declines in sexual satisfaction (Carpenter et al. 2009). Younger or later-born cohorts (i.e., the baby boomer generation) tend to employ a wider range of sexual techniques (e.g., incorporation of oral sex and genital touching) than older or earlier-born cohorts (i.e., the silent generation and the greatest generation) (Edwards and Booth 1994) and later-born cohorts also have higher levels of sexual satisfaction than earlier-born cohorts (Beckman et al. 2008). This may be because greater variation in sexual techniques allows individuals to maintain high levels of sexual satisfaction even as they experience age-related physical changes that interfere with the ability to have (frequent) sex.

Although some studies find no gender differences in levels of satisfaction (e.g., Blumstein and Schwartz 1983), other studies have found such differences. For example, the AARP survey of midlife and older adults (2005) found that partnered men are more likely than partnered women to report that they are dissatisfied or somewhat satisfied with their sexual relationship, whereas partnered women are more likely than partnered men to report that they are extremely satisfied or neutral with respect to their sex life (similar gendered patterns were found among the unpartnered). Further, some evidence suggests that sexual satisfaction is more closely linked to relationship satisfaction for men than it is for women; one longitudinal study found that low levels of sexual satisfaction predict relationship dissolution for men, but not women (Sprecher 2002).

Research is also unclear on whether sexual satisfaction levels differ for individuals in gay, lesbian, and heterosexual relationships. One study found no differences between gay, lesbian, and heterosexual couples (Kurdek 1991), while the American Couples Study (Blumstein and Schwartz 1983) found that gay men were less likely to report that they were sexually satisfied, compared to individuals in heterosexual and lesbian relationships. Further, research indicates that women in lesbian relationships have more frequent orgasms—which is a predictor of sexual satisfaction (Sprecher and Cate 2004)—compared to women in heterosexual relationships (Peplau et al. 2004). However, for all couples—gay, lesbian, and heterosexual—sexual satisfaction is closely linked to relationship satisfaction (Peplau et al. 2004; Schwartz and Young 2009).

Although limited information exists on racial/ethnic differences in sexual satisfaction, results from the AARP (2005) survey of mid and later life adults suggests that Asian Americans have lower levels of sexual satisfaction than Whites, Blacks, and Hispanics (AARP 2005). Further, partnered Blacks and Hispanics are more likely than Whites and Asian Americans to believe that their partner is very satisfied with their sexual relationship and to discuss sexual satisfaction with their partner (AARP 2005). The AARP survey does not offer any explanations for these racial/ethnic differences; thus, future research should examine if these variations exist in other samples as well as explore explanations for racial/ethnic differences in sexual satisfaction in long-term relationships.

Other factors related to higher levels of sexual satisfaction include higher levels of orgasmic frequency, greater sexual communication, higher levels of accepted sexual initiations, low levels of sexual conflict, similarity between partners in terms of sexual behavior preferences, sexual desire, and sexual attitudes (Sprecher and Cate 2004), better physical health (AARP 2005), higher socioeconomic status (Castellanos-Torres et al. 2005), and higher levels of physical activity (AARP 2005).

1.2.3 Sexual Attitudes

Most adults regard sexuality as an important component of relationships. For example, a recent nationally representative survey of adults ages 45 and older found that 60 % believe that sexual activity is critical for relationship quality (AARP 2010). There may be gender differences in sexual attitudes; for example, middle-aged and older men are more likely than middle-aged and older women to report that sex is important for quality of life and relationship satisfaction (AARP 2010). Men’s and women’s views on sexual activity may converge with age, however; one study found that as men age they place less importance on sexual activity (Wiley and Bortz 1996).

Some research suggests that sex becomes less important for some couples over the course of long-term relationships, as both men and women come to view emotional intimacy as more important than sexual intimacy (Lodge and Umberson 2012; Umberson et al. 2015b). This change in the meaning of sex in relationships may be adaptive given that sexual frequency declines over time in long-term relationships and as one or both partners face physical (e.g., menopause, erectile dysfunction) or social changes (e.g., transition to parenthood) that make sex either less desired or feasible. On the other hand, other research suggests that sex does not decrease in significance for some mid- to later life couples—particularly gay couples—and that a reduction in sexual frequency may be experienced as a threat to a masculine identity for both gay and straight men (Lodge and Umberson 2013; Slevin and Mowery 2012). Cohort differences may also matter for attitudes toward sex; research reveals that younger cohorts of older adults have more positive attitudes toward sexuality than older cohorts (Beckman et al. 2008), which may in part reflect a gradual shift away from cultural discourses that define older adults as asexual to discourses that emphasize the importance of remaining sexually active as a marker of healthy and successful aging (Gott 2005; Katz and Marshall 2003).

1.2.4 Sexual Desire

Sexual desire is a complex phenomenon that encompasses biological drives, psychological motivations, and personal and social expectations, beliefs, and values (Kingsberg 2000). Individuals who are partnered report higher levels of sexual desire than individuals who do not have a partner (DeLamater and Sill 2005; Skoog 1996). However, a variety of relationship characteristics appear to be important for levels of sexual desire, perhaps particularly for women—which fits with theory and research that suggests that women’s sexual desire is more fluid and sensitive to relational context (Diamond 2009; Peplau 2001). For example, one study found that relationship duration is negatively related to sexual desire for women, but not men (Kontula and Haavio-Mannila 2009). Relationship quality also matters a great deal for sexual desire; research suggests that women who are able to talk to their partners about how to facilitate their sexual desire report having more sexual desire (Wood et al. 2007). A lack of sexual desire is also associated with low expectations about the future viability of their current relationship for women, but not men (Laumann et al. 2005). Other relationship characteristics, such as conflict and partner discrepancies in desire, may also affect levels of sexual desire (that is, individuals may try to match their level of desire to that of their partner) (Skultety 2007).

Aging may also negatively impact sexual desire (DeLamater and Sill 2005; Laumann et al. 2005) in long-term relationships. Some research suggests that the negative relationship between age and sexual desire is stronger for men (DeLamater and Sill 2005) or only holds for men (Laumann et al. 2005). However, women may have lower levels of sexual desire than men: One study of Finnish adults found that at the age of 60 one-half of women reported a somewhat frequent lack of desire, compared to only 15 % of men (Kontula and Haavio-Mannila 2009). Some women also experience a decreased level of desire during the menopausal transition (Basson 2005). However, it remains unclear whether these decreased levels of desire are attributable solely to menopause , as the social context in which women find themselves profoundly shapes how they experience and express desire during menopause (Wood et al. 2007). For example, a loss of reproductive capacities may negatively affect women’s sense of femininity and sexual identity, and thus levels of sexual desire (Kingsberg 2000). Moreover, women who believe that the physical signs of aging make them unattractive may experience a reduced level of sexual desire (Kingsberg 2000). Postmenopausal women’s relationships with their partners also profoundly shape how they experience and express sexual desire (Wood et al. 2007). Beliefs about age and appropriateness of sexual activity may also be important: for women a lack of interest in sex is associated with the belief that aging reduces sexual desire and activity (Laumann et al. 2005). Additional factors that have a negative effect on sexual desire include poor health (Kontula and Haavio-Mannila 2009; Laumann et al. 2005), high blood pressure (DeLamater and Sill 2005), depression (Laumann et al. 2005), and low socioeconomic status and levels of education (DeLamater and Sill 2005). There is also some evidence that among partnered women in the U.S., White and Hispanic women are more likely than Black women to report low levels of desire (West et al. 2008). Despite research on levels of sexual desire, still little is known about the lived experience of sexual desire in long-term relationships or how these lived experiences differ for different social groups.

1.2.5 Sexual Dysfunction

As part of a broader shift of the medicalization of sexuality (Tiefer 1996) (see Cacchioni, Chap. 24, this volume), a great deal of research has focused on sexual dysfunction. Research on sexual dysfunction reflects the Diagnostic and Statistical Manual of Mental Disorders’ classification scheme, whereby sexual problems fall into four categories: (1) sexual desire disorders, (2) sexual arousal disorders, (3) orgasmic disorders, and (4) sexual pain disorders. Sexual dysfunction is important because it may cause depression (Araujo et al. 1998) and marital and relationship conflict (Rust et al. 1988), and is associated with an overall diminished quality of life (Laumann et al. 1999). Moreover, one sexual dysfunction may precipitate another dysfunction; for example, men who experience erectile dysfunction (ED) are more likely to later report low levels of sexual desire (Kingsberg 2000).

Being in a long-term, committed relationship is somewhat protective of sexual dysfunction (Laumann et al. 1999, 2005, 2008). For example, among 57 to 85 year-olds, married men are less likely than widowed or never married men to experience a lack of sexual pleasure and less likely to experience performance anxiety in comparison to separated or divorced men (Laumann et al. 2008). Additionally, among 40 to 80 year-old men, being in an uncommitted relationship is associated with a greater likelihood of erectile difficulties (Laumann et al. 2005). Similarly, among women ages 40 to 80, those who believe or worry that their current relationship is unlikely to last are more likely to report an inability to orgasm (Laumann et al. 2005). This resonates with recent research that suggests that college women are more likely to experience orgasms in committed heterosexual relationships than they are in heterosexual casual relationships (i.e., “hookups”) (Armstrong et al. 2012). Relationship satisfaction is also predictive of sexual functioning (Laumann et al. 2008). Women who are dissatisfied with their relationship are more likely to experience a lack of sexual pleasure and an inability to orgasm, while men who are dissatisfied with their relationship are more likely to experience a lack of sexual interest (Laumann et al. 2008). Further, leaving an unsatisfactory relationship for a new, satisfying relationship may positively impact sexual functioning for both men and women (Kontula and Haavio-Mannila 2009).

1.3 Feminist Studies

1.3.1 Performativity Studies

Recent qualitative research has applied feminist theoretical perspectives to the study of sex in long-term relationships . One such perspective is the “doing gender” perspective; this perspective was originally developed by West and Zimmerman (1987) to refer to the performance of gender—whereby individuals “do” gender in response to culturally constructed notions of masculinity and femininity. In doing so, they reproduce—although sometimes contest—dominant cultural ideologies about gender. Recently, some studies have examined how men and women perform gender in the context of long-term sexual relationships. For example, Lodge and Umberson (2012) found that aging married men and women attempt to perform gender in line with cultural ideals of feminine (i.e., passive, lower levels of sexual desire) and masculine (i.e., active, high levels of sexual desire) sexuality, even as aging presents challenges to these ideals. Specifically, when husbands experience lower levels of desire women often resist initiating sex (even when they desire sex) because it goes against cultural beliefs about feminine sexuality (Lodge and Umberson 2012). Umberson et al. (2015b) also applied a doing gender perspective to the topic of sexuality in long-term relationships to show that women in both heterosexual and same-sex relationships were more apt than heterosexual or gay men to view sex and emotional intimacy as integrally linked.

Expanding on West and Zimmerman’s (1987) formulation, Laz (1998) developed the concept of the performance of age, whereby individuals “act their age”—that is behave in line with cultural ideas about what is age appropriate. Lodge and Umberson (2012) applied this perspective to explain why later life—but not midlife—couples deemphasize the importance of sex for their relationships. Studies suggest that maintaining an active sex life is increasingly conceptualized as a part of “successful aging,” (Gott 2005; Katz and Marshall 2003), but that the targets of these discourses are largely midlife individuals. In contrast, there remains considerable cultural ambivalence about later life adults’ sexuality (Frankowski and Clark 2009). Thus, not having sex in later life is more culturally normative than it is for midlife couples, which may be why later life couples deemphasize the importance of sex for their relationships. There is a need for more studies to examine how cultural ideas about age shape the experience of sexuality across the life course as individuals “do” age in their sexual relationships, as well as how cultural ideas around gender, race/ethnicity, and other social statuses intersect with ideas about age to shape the experience of sexuality in unique ways for different groups.

Further expanding on this theoretical tradition, Elliott and Umberson (2008) developed the concept of the “performance of desire,” to refer to a process of “managing feelings around one’s sexual relationship according to how one thinks desire should be both felt and performed” (p. 394). They find that because cultural discourses emphasize the importance of sex for marital relationships, married women often attempt to feel and be more sexual in an attempt to match their level of desire to their husband’s, whereas married men often attempt to feel and be less sexual in an attempt to match their level of desire to their wife’s (2008). The recognition of sexual desire as profoundly shaped by social context and meanings and something that one “performs” is important and future research should explore how cultural meanings around sex and desire shape how men and women in a variety of relational contexts “do” desire.

Theoretical work by Jackson and Scott (2007) has also interrogated the ways in which orgasms are interactionally performed in heterosexual relationships. In particular, they note that because masculine sexuality is based on a performance ethic in which men must demonstrate potency and virility, women must in turn convincingly perform orgasms—that is demonstrate that they are experiencing desire and pleasure (2007). Although research suggests that women “fake” orgasms (Jackson and Scott 2007), empirical studies have not examined the interactional work that goes into doing so in long-term relationships nor how this may change over the course of relationships. Future research should thus examine this question among diverse couples.

1.3.2 Emotion Work and Sex

A few recent studies working from a feminist perspective have examined emotion work around sex in long-term relationships. Emotion work was originally defined by Hochschild (1979) to refer to labor that involves managing one’s emotions to conform to the “feeling rules” of a particular context. A number of studies demonstrate that women do substantially more emotion work than men and that this is particularly true in heterosexual relationships, whereby women undertake emotion work in order to promote relationship quality (Duncombe and Marsden 1993; Erickson 2005; Hochschild 2003). Theoretical work by Jackson and Scott (2007) further suggests that because successful performances of masculine (hetero)sexuality require that men demonstrate an ability to sexually please women, women undertake considerable emotion work to convince their partners that they are experiencing sexual pleasure. Elliott and Umberson (2008) in turn applied these insights to study sex in long-term heterosexual marital relationships and found that because sex tends to be viewed as an integral component of marriage and sex is a frequent source of marital conflict, women undertake emotion work in an attempt to be more desiring of sex—either by initiating sex or by being more receptive to their husband’s sexual advances. They further found that husbands expect their wife to perform such emotion work. Although less common, some husbands undertake emotion work to repress their sexual desires to avoid marital conflict.

Another recent study examined emotion work around sex in long-term lesbian, gay, and heterosexual couples and found that the division of emotion work and type of emotion work around sex varies based on both an individual’s gender as well as the gender of their partner (Umberson et al. 2015b). For example, both heterosexual and lesbian women described emotion work directed toward being more desiring of sex because they view sex as integral to emotional intimacy and relationship quality, but this was less common for women in lesbian relationships because divergent levels of sexual desire were less common in lesbian relationships compared to heterosexual relationships. Further, the division of emotion work directed toward enhancing sexual desire was more equal in lesbian relationships than heterosexual relationships, because both partners in lesbian relationships often reported undertaking such work. In contrast, men were less likely than women to report that they viewed as sex as linked to emotional intimacy, and this was particularly the case for men in gay relationships. Because of the common view in gay relationships that sex and emotional intimacy are unrelated, emotion work in gay relationships often entailed one partner (who desired monogamy) working to accept their partner’s view that sexual nonexclusivity is acceptable as long as it does not involve emotional intimacy. However, all couples reported that over the course of their relationships sexual frequency has declined and in turn they engaged in emotion work to see intimacy as unrelated to sex (Umberson et al. 2015b) .

1.3.3 Gay and Lesbian Couples

There has been relatively little research on sexuality in long-term gay and lesbian couples; in part this may stem from the fact that researchers are wary of reproducing stereotypes about the hypersexuality of gay men or the sexual deviancy of sexual minorities generally (Peplau et al. 2004). However, research has consistently demonstrated that, like heterosexual long-term relationships, same-sex couples experience declines in sexual frequency with relationship duration. However, men in gay relationships consistently report higher levels of sexual frequency than individuals in other couple types, which stems in part from the fact that a significant amount of gay men in coupled relationships supplement their sex lives with outside partners (Schwartz and Young 2009). Age is also negatively correlated with sexual frequency for gay and lesbian couples, although, as in heterosexual relationships, relationship duration exerts a stronger negative impact on sexual frequency. Research also consistently demonstrates that gay couples have higher levels of sexual frequency than heterosexual couples, who in turn have higher levels of sexual frequency than lesbian couples (Peplau et al. 2004).

Lesbian couples’ low levels of sexual frequency, often referred to as “lesbian bed death” may stem from several factors. First, research suggests that because of gender socialization women are less attuned to their sexual desires as well as less likely to initiate sex than men, and this effect may be amplified in a relationship with two women (Peplau and Fingerhut 2007; Peplau et al. 2004). Another possibility is that women simply have lower levels of sexual desire than men, the effect of which is again amplified in a relationship involving two women (Peplau and Fingerhut 2007; Peplau et al. 2004). A third possible explanation is that many surveys fail to accurately capture the realities of lesbian sex; survey questions often ask respondents about instances of “intercourse,” thus underestimating the frequency of sexual activity in lesbian relationships (Peplau and Fingerhut 2007; Peplau et al. 2004). Thus, it remains unclear if “lesbian death bed” actually exists or if lower levels of sexual frequency among lesbian couples in survey research reflect the questions that researchers ask and how they ask them. Regardless, recent research suggests that some lesbian couples perceive “lesbian death bed” to be a real phenomenon and actively seek to avoid it in their relationship by attempting to feel or “do” desire and engage in sexual intimacy (Umberson et al. 2015b) .

Another theme from previous research is that individuals in gay couples are less likely than individuals in heterosexual and lesbian couples to believe that sexual exclusivity is important as well as less likely to be sexually exclusive (Peplau and Fingerhut 2007; Peplau et al. 2004). However, nonmonogomous gay couples often establish sexual contracts that set rules around extradyadic sex, including rules about safe sex and emotional attachment to other partners (see Sheff and Tesene, Chap. 13, this volume). Importantly sexual exclusivity is not related to levels of relationship satisfaction, commitment, closeness, or relationship satisfaction for gay couples (Peplau et al. 2004).

2 Methodologies

The dominant methodological strategy for studying sexuality in long-term relationships has been survey methods , which typically rely on individual outcomes (rather than dyadic outcomes). In the past two decades, several nationally representative surveys (e.g., NHSLS, The National Survey of Sexual Health and Behavior [NSSHB], AARP Sexuality at Midlife and Beyond surveys, National Survey of Families and Households [NSFH], and the National Social Life, Health, and Aging Project [NSHAP]) have included measures on sexual frequency, sexual satisfaction, sexual desire, sexual attitudes, among other variables. Thus, we now have an impressive foundation of cross-sectional knowledge in terms of these outcomes. To advance knowledge in the study of sex in close relationships, more innovative methods are needed, however. In this section, I discuss some of these, including the need for more dyadic research, longitudinal surveys (including daily diary methods), qualitative studies, and the need for nationally representative data on gay and lesbian relationships.

2.1 Dyadic Data

An important avenue for future research on sex in long-term relationships is to study dyads (i.e., both partners in a relationship), rather than individuals in relationships (DeLamater and Hyde 2004; Perlman and Campbell 2004). Dyadic data allow researchers to compare partners’ perspectives and behaviors, identifying points of overlap and difference (Umberson et al. 2015a). For example, researchers might compare partners’ levels of sexual satisfaction to examine how partner similarity and/or discrepancy predict sexual frequency or relationship dissolution. Dyadic data further allow for validity checks—that is, by comparing partners’ reports (e.g., reports of sexual frequency or change in sexual frequency over time) (Umberson et al. 2015a). Partner discrepancies can also reveal valuable information about relationships. For example, Mitchell et al. (2012) collected dyadic data to study concordance around sexual agreements or contracts around extradyadic sex among gay couples and found that couples who were more congruent about having and adhering to sexual agreements had higher levels of relationship satisfaction. Another important characteristic of dyadic data is that it can yield relationship-level data; for example researchers might ask whether the division of labor (a relationship-level variable) is related to sexual frequency (another relationship-level variable). It is important that researchers conduct both quantitative and qualitative dyadic research.

2.2 Longitudinal Methods

Research on sexuality has historically been plagued by funding issues (Perlman and Campbell 2004). As a result, few nationally representative longitudinal studies have included measures on sexuality. One recent exception to this is NSHAP, which includes numerous questions about sex. However, this survey is limited to adults ages 57 to 85. As I discuss later, however, longitudinal research is particularly important for research that incorporates a life course perspective , given that only longitudinal research can fully address questions around relationship and sexual turning points and histories (see Carpenter, Chap. 5, this volume). Collecting longitudinal qualitative data is also important. Diamond (2009) collected longitudinal data to examine women’s transitions between same- and different-sex unions as well as transitions in women’s sexual identities. Similarly, researchers could collect qualitative longitudinal data to examine changes in the meanings and importance of sex over time within relationships as well as how meanings around sex change during and after relationship and other life course (e.g., becoming a parent) transitions.

One particularly fruitful avenue for longitudinal research is daily diary studies, in which respondents fill out surveys for several consecutive days (usually over a period of several weeks). Daily diary studies are increasingly common in family research generally, but have less frequently been applied to the study of sex. Such surveys could, however, yield important information about how sexual frequency, satisfaction, desire and other variables related to sex fluctuate daily, as well as how they may fluctuate in response to other relationship variables (which likely also fluctuate daily), such as levels of relationship conflict, perceived emotional support, and the division of labor. Dyadic daily diary studies, whereby both partners (independently) fill out daily questionnaires that ask questions about their sexual relationships are a particularly useful avenue for future research in this regard. For example, Ridley et al. collected daily diary data to reveal that daily fluctuations in positive and negative feelings towards one’s partner were associated with fluctuations in sexual thoughts and behaviors, although in different ways for gay, lesbian, and heterosexual couples (Ridley et al. 2008). A more recent daily diary study with heterosexual couples found that daily fluctuations in sexual desire and partner discrepancy in levels of desire were associated with quality of sexual experience (Mark 2014). Dyadic daily diary studies thus provide an opportunity to address a range of questions concerning sex in coupled relationships.

2.3 Qualitative Methods

To date, most studies on sex in long-term relationships have been based on survey data. Thus, there is a need for more qualitative studies on this topic, which is also a particularly important endeavor for theoretical development. Qualitative data are particularly well-suited to revealing meanings (e.g., social, cultural, and individual understandings and perceptions) around sex in long-term relationships. This is important because researchers typically assume meanings in designing survey questions, but the particular questions asked may or may not accurately reflect respondents’ lived meanings around sex. Further, the fact that so much survey research has focused on sexual frequency may reflect a male perspective, to the extent that women may be more concerned with the quality of sexual interactions, as opposed to the frequency of those interactions (Schwartz 2004). Thus, findings from qualitative research can be used to inform future surveys, by revealing important new insights into the experience of sex in long-term relationships. Recent qualitative studies on sex in long-term relationships have revealed, for example, important insights into emotion work around sex in lesbian, gay, and heterosexual relationships (Elliott and Umberson 2008; Umberson et al. 2015b), the performance of sexual desire in marital relationships (Elliott and Umberson 2008), how the meaning of sex changes over time in long-term relationships (Lodge and Umberson 2012; Umberson et al. 2015b), and how individuals construct meanings around the link between sex and emotional intimacy in lesbian, gay, and heterosexual relationships (Umberson et al. 2015b). These findings (which are described above) reveal important theoretical insights that can be used to inform future surveys.

2.4 Nationally Representative Data on Gay and Lesbian Couples

Obtaining nationally representative data on gay and lesbian individuals and couples remains a challenge. This is particularly the case in terms of research on sexuality, given that funding agencies have often been reluctant to fund sexuality research. For example, in order to obtain funding for the NHSLS Laumann et al. had to abandon their plans to include adequate subsamples of gay and lesbian respondents (Perlman and Campbell 2004). Thus, most nationally representative surveys that include data on sexuality do not include sufficient numbers of gay and lesbian individuals or couples. As a result, most of what we know about sex in gay and lesbian couples is based on convenience samples (Peplau et al. 2004). Obtaining nationally representative samples of gay and lesbian couples, however, is important for a fuller understanding of the diversity of sex in long-term relationships. Such samples can also reveal important insights into how gender matters not just at the individual-level but at the relationship-level, given that researchers could compare sexual relationships composed of two women, two men, and one woman and one man .

3 Future Directions

3.1 The Need for Theoretically-Informed Studies

Four broad theoretical perspectives have informed the study of sex in intimate relationships : evolutionary psychology, attachment theory, social exchange theory, and symbolic interactionism (Perlman and Campbell 2004). Of these theories, the only one that it is distinctly sociological is symbolic interactionism and many of the recent theoretically-informed studies on sex in long-term relationships reviewed earlier (e.g., studies that apply a performance perspective) certainly reflect this theoretical tradition. However, many—if not most—studies on this topic are simply descriptive. Thus, one of the most pressing concerns for future research on sex in long-term relationships is the need for theoretical development and theoretically-informed studies. In this section, I suggest that incorporating theoretical insights from the following theories can reveal new insights into the study of sex in long term relationships: (1) a “gendered sexuality over the life course” perspective, (2) a gender-as-relational perspective, (3) critical feminist gerontology, and (4) queer theory . Merging some or all of these different theoretical perspectives into a particular study can reveal new insights into this topic and also responds to calls for the use of more integrative theorizing in the study of sex in relationships (DeLamater and Hyde 2004).

3.1.1 Gendered Sexuality Over the Life Course Perspective

Drawing on gender, scripting, and life course theories, Carpenter (2010, Chap. 5) argues that research on sexuality should incorporate a “gendered sexuality over the life course” theoretical perspective. Specifically, from this perspective: “sexual beliefs and behaviors result from individuals’ lifelong accumulation of advantageous and disadvantageous experiences, and their adoption and rejection of sexual scripts, within specific socio-historical contexts” (Carpenter 2010, p. 157; Montemurro 2014). Further, these gender-specific experiences and scripts give rise to gendered trajectories of sexuality, which are experienced differently at the intersection of race/ethnicity, social class, and sexual identity. Future research could apply this perspective to the topic of sex in long-term relationships by using longitudinal or retrospective data to examine how earlier life course experiences of gendered sexuality matter for later life experiences of gendered sexuality, and how those experiences are inflected by race/ethnicity, social class, and sexual identity. For example, research suggests that in heterosexual relationships men’s sexual pleasure is typically privileged (Armstrong et al. 2012), yet women whose early life course relationships do not conform to this gendered pattern may develop a greater sense of sexual agency and greater expectations of sexual pleasure, which in turn likely shape their experiences of partnered sex throughout the life course—albeit in different ways based on women’s various social locations and access to privilege.

A life course perspective further points to the importance of relationship histories (Cooney and Dunne 2001) and how they intersect with sexual histories. Earlier relationship experiences—both in terms of the current relationship and previous relationships—are likely important for understanding how sex is presently experienced within long-term relationships. For example, although the link between relationship satisfaction and sexual satisfaction is well established, it is unclear if or how levels of relationship satisfaction early in the course of relationships may be related to levels of sexual satisfaction later on or vice versa. Further, from a gendered sexuality over the life course perspective, relational experiences likely occur in particularly gendered ways. Thus, future research on sex in long-term relationships should apply these insights. Doing so would undoubtedly reveal new insights into gendered experiences of sex in long-term relationships as well as how those may differ across social groups .

3.1.2 Gender as Relational

A gender-as-relational theoretical framework builds on insights from a doing gender perspective (West and Zimmerman 1987) to argue that gender is performed in relation to others and individuals do gender differently based on social context (Springer et al. 2012). Recent research applying this perspective to understand gendered experiences of intimate relationships , for example, suggests that men will perform masculinity differently based on whether they are in a relationship with a man or woman. Umberson et al. (2015b) find that meanings and experiences of sex in long-term relationships reflect an individual’s gender in relation to the gender of their partner in that women partnered with women reinforce a view of emotional intimacy and sex as integrally connected, while women partnered with men challenge their partner to adopt this perspective at the same time that their partner challenges this view. In contrast, men partnered with men reinforce one another’s view of intimacy and sex as separate. Researchers could adopt this theoretical lens to investigate a variety of gendered sexual phenomenon within same-sex and different-sex long-term relationships .

3.1.3 Critical Feminist Gerontology

For a fuller understanding of gendered sexuality over the life course and in particular gendered experiences of sexuality in mid and later life, I further suggest that researchers integrate key theoretical insights from critical feminist gerontology. Critical feminist gerontology integrates a feminist perspective (which maintains that gender is a key axis of inequality), with critical gerontology, which emphasizes that ageism—as a cultural and social structural system—is a key axis of inequality (Calasanti 2005). For example, research on sex in long-term relationships applying this perspective has focused on the role of cultural devaluations of the aging body, and in particular the devaluation of the appearance of the aging female body and the devaluation of the functionality of the aging male body, in shaping gendered sexual experiences in long-term relationships (Lodge and Umberson 2012). This perspective is not only useful for understanding gendered experiences of sex in later-life relationships, but gendered sexual experiences in early adulthood and midlife relationships as well, because constructions of age and ageism operate at all points of the life course (albeit in different forms). While previous research has included age and gender as variables in research on sex in long-term relationships , future research can go further to examine how age and gender intersect as cultural and structural systems to shape sexual experiences across the life course. Critical feminist gerontology can also be applied to queer couples: for example, research has applied this perspective to understand why some midlife men in relationships with other men find perceived declines in the attractiveness of their bodies—that is both self-perceptions that their bodies are declining in attractiveness as well as perceptions that others perceive their bodies as declining in attractiveness—as having a negative impact on their sex lives (Lodge and Umberson 2013).

3.1.4 Queer Theory

Queer theory disrupts the heteronormative assumptions (i.e., assumptions based on conventional understandings of gender and heterosexuality) upon which much contemporary research and theory is based upon. Queer theory can therefore be merged with any of the above theories to reveal new insights into the study of sex in long-term relationships that disrupt heteronormative assumptions about gender, sex, and relationships. For example, Brown (2009) suggests that merging queer and life course theories is a particularly useful endeavor because queer theory disrupts the heteronormative assumptions upon which life course research is often based on (e.g., assumptions of marriage and parenthood), at the same time that life course theory provides a framework for examining life experiences as shaped by social structures and relational contexts with others.

In terms of sex in long-term relationships , queer theory could be merged with a gendered sexuality over the life course theoretical framework to examine how gendered experiences of sexuality differ for gender queer individuals (i.e., individuals who do not endorse or conform to conventional masculine and feminine identities and presentations) over the life course. These perspectives could also be merged to examine how transitioning from different-sex to same-sex (and vice versa) relationships or from different gendered identities (e.g., from a man to a woman) over the life course shape relational sexual experiences (see Devor and Dominic, Chap. 11, this volume). Similarly, researchers could merge a gender-as-relational framework with queer theory to ask questions about how a woman partnered with a woman might experience sex differently than a woman partnered with a man or a man partnered with a man in order to potentially queer our understandings of gender, sex, and relationships.

3.2 The Need for More Research on Diversity in Long-Term Relationships

In addition to the need for more research on sex in LGBT relationships, little research has focused on sex in non-marital relationships. Research is needed to understand if and how sex may be experienced differently in non-traditional long-term relationships —including nonmonogamous and living-apart-together relationships (i.e., those where partners maintain separate residences, but often spend the night at one another’s homes). The application of queer and life course theories could be particularly useful for understanding non-traditional long-term relationships.

Additionally, more research is needed on experiences of sex in long-term relationships across different racial-ethnic (DeLamater and Hyde 2004), cultural (Perlman and Campbell 2004) and social class groups, as well as how these experiences may differ based on the racial-spatial and social class organization of specific communities (Laumann et al. 2004). Although previous research suggests that there are differences in sexual experiences for different social groups (Laumann et al. 1999), little is known about why these differences exist (DeLamater and Hyde 2004). Although some descriptive studies include racial/ethnic and/or socioeconomic status as variables in their analytical models, the tasks of systematically examining racial/ethnic and socioeconomic status differences, the effect sizes of such differences, and if meanings and lived experiences of sex in long-term relationships differ across groups have not been adequately performed. This likely connects to the dearth of qualitative research: we don’t know these things because researchers have failed to ask certain questions and most of what we know on the topic of sex in long-term relationships is based on middle-class white samples. These omissions have occurred not only at the individual-level, but at the dyadic level as well. For example, an interesting avenue for future research could be to examine if and how racial/ethnic or socioeconomic difference within a couple shape sexual experiences. In addition to the paltry attention paid to race, ethnicity, and social class, research on this topic has been dominated by a focus on the United States context, virtually ignoring cross-cultural experiences of sex in long-term relationships (Perlman and Campbell 2004). An important exception to the dearth of cross-cultural research is the Global Study of Sexual Attitudes and Behaviors (GSSAB), a survey of 27,500 men and women from 29 different countries (Laumann et al. 2005). However, in addition to survey research that compares different national and cultural groups, we need research that can speak to the links between sexual and relational experiences and specific cultural norms and values (Peplau et al. 2004). These omissions likely stem from the fact that questions of difference in the study of sexuality have overwhelmingly focused on gender (DeLamater and Hyde 2004); thus, questions around race, ethnicity, and culture represent a key area for future research on this topic .

Significant strides have been made in the study of sex in long-term relationships over the past 20 years. Researchers now have access to high quality survey data and we now an impressive foundation of knowledge on topics such as sexual frequency, sexual satisfaction, and the association between these two variables to relationship satisfaction. However, to keep the field moving along at this impressive pace, researchers must pioneer new methodological strategies, move beyond descriptive studies to apply cutting-edge theoretical perspectives to the study of sex in long-term relationships , and shed greater light on the full diversity of sexual experiences in long-term relationships.