Going the Distance

Marriage and Health

Across diverse populations, and across multiple domains of health, being in a romantic relationship is associated with better health (Borland, Robinson, Crozier, & Inskip, 2008; Graham, Francis, Inskip, & Harman, 2006; Helbig, Lampert, Klose, & Jacobi, 2006). Being married is associated with the best mental and physical health, compared to being single or in an unmarried relationship (Brown, Bulanda, & Lee, 2005; Stack & Eshleman, 1998). Providing one way to understand partnership and health associations, married individuals have better health behaviors (e.g., substance use, diet) than unmarried ones (Schone & Weinick, 1998; Wilson & Oswald, 2005). Longitudinal evidence indicates marriage is protective of one's health (e.g., Johnson, Backlund, Sorlie, & Loveless, 2000). Therefore, some researchers describe the marriage–health association as marriage protection. Cross-sectional research on this topic continues to frame this phenomenon as the marriage–health association to clarify that results of these studies cannot be causally interpreted.

Relationship Variables: Satisfaction, Maintenance, Stress, and Sex

Not all romantic relationships are equally, positively related to health. Several relationship-specific factors influence the strength of the marriage–health association. These include relationship satisfaction, maintenance strategies, stress, and sex.

First, among heterosexuals, high marital quality—also conceptualized as relationship satisfaction—is associated with lower stress, less depression, higher satisfaction with life, and better health among individuals with chronic disease (Bradbury, Fincham, & Beach, 2000; Holt-Lunstad, Birmingham, & Jones, 2008; Patrick et al., 2013; Umberson, Williams, Powers, Liu, & Needham, 2006; Zhou et al., 2011). Next, relationship maintenance strategies are behaviors that couples use to maintain their relationship. These include frequency, type, and valence of communication and are important because they relate to sustaining a relationship over time (e.g., Canary & Stafford, 1992). Better and more frequent maintenance behaviors (e.g., positivity, networking, and shared tasks) between individuals in a couple relate to less psychological distress for both partners (Badr & Taylor, 2008). Stress—either internal or external to the relationship—is associated with both increased health risks and reduced health benefits for individuals in relationships (Barnes, Brown, Krusemark, Campbell,& Rogge, 2007; Berry & Worthington, 2001; Neff & Karney, 2004; Orth-Gomer et al., 2000). Finally, research indicates direct, positive associations between sex variables (frequency, satisfaction), and the variables sexual-, life-, and relationship satisfaction and more broadly, mental and physical health (Brody & Costa, 2009; Butzer & Campbell, 2008; Yeh, Lorenz, Wickrama, Conger, & Elder, 2006).

Long-Distance Relationships

It is unknown whether individuals in long-distance relationships (LDRs) demonstrate this positive marriage–health association. A recent study compared LDR individuals to those in proximal relationships (PRs) on a psychological distress index, finding no significant differences (Dargie, Blair, Pukall, & Goldfinger, 2015). Otherwise, research on marriage and health has focused on PRs. This is surprising, given that LDRs are increasingly prevalent. Up to 40% of college students (Merolla, 2010), and approximately 1,000,000 other U.S. couples, report being in an LDR or in a dual-career commuter couple (Rhodes, 2002).

Although research on LDRs and health is limited, some studies have compared LDRs and PRs in terms on satisfaction, maintenance, stress, and sex variables. Results indicate that these four variables are important to both PRs and LDRs. Regarding satisfaction, in a nationally representative U.S. study (N = 870), those in LDRs reported higher relationship quality in several domains (e.g., love for partner, fun with partner, and conversational quality), compared to those in PRs (Kelmer, Rhoades, Stanley, & Markman, 2013). Extant research on maintenance behaviors demonstrates no difference in frequency of use between the two types of relationships though different ones are utilized. This is most likely due to individuals in LDRs compensating for lack of face-to-face contact with other modes of communication. Individuals in LDRs and PRs experience similar amounts of stress, although types of stressors vary between the types of relationships. Couples in LDRs are more likely to experience stress related to separation, travel, career, and relationship decision making, while couples in PRs experience the stress of daily hassles, for example, job stress, household, and child-rearing divisions of labor (Pistole, Roberts, & Chapman, 2010). Finally, LDRs have not been studied extensively, as they relate to sexual frequency or satisfaction. Extant research states those in LDRs and PRs report similar levels of sexual communication and satisfaction (Kelmer et al., 2013).

Therefore, relationship satisfaction, maintenance, stress, and sex are relevant in LDRs just as they are in PRs. But, these variables may affect LDRs differentially compared to PRs, with resultant differential effects on individual health. Specifically, the unique nature of LDRs, and the resultant unique relationship of the above variables, may change the marriage–health association. For example, LDRs may be more satisfying due to the fact that being together is more valued when there is so much time apart, or when the couple is together, it feels more like a vacation than when PR couples are together. As a result, we could hypothesize that LDR status could be more strongly associated with positive health, than PRs (Kelmer et al., 2013). Conversely, other research suggests the attenuation of the marriage–health association in LDRs compared to PRs is based on other factors. Merolla (2010) reported several LDR-specific factors could precipitate stress for individuals in the relationships—the distance itself, related difficulties (e.g., managing household and parenting tasks; fewer opportunities for sexual intercourse), and potential increased financial costs.

The Present Study

Despite the theory and empirical evidence highlighted earlier, no study has assessed the associations between being in an LDR and individual health status. This study aims to fill that gap by answering two primary research questions. First, do those in LDRs show the same marriage–health association as those in PRs? To test this, we will compare the psychological and physical health of individuals in PRs and LDRs. Second, does relationship type (PR vs. LDR) predict relationship variables such as satisfaction, maintenance, stress, and sex and health outcomes such as anxiety, depression, health behaviors, fatigue, and pain? This study was approved by the affiliated institution's Institutional Review Board (Project # STU00096201).

Participants

Potential participants completed a screener to determine whether they met study inclusion criteria, which included age (21 years or older), relationship status (married), and relationship arrangement (proximal or long distance; those endorsing a combination of the two arrangements were excluded). Participants were screened into the “proximal” group if they reported that, in a typical month they could see their partner daily, and in a typical week spent 0–2 days more than 50 miles from their partner. Participants were screened into the “long-distance” group if they reported seeing their partner less than daily in a typical month and spending 3 or more full days weekly more than 50 miles apart. Borrowing and adapting criteria from related research, we defined a full day as from the time participants woke up to the time they went to sleep (Dargie et al., 2015).

Procedure

Participants were recruited using multiple online recruitment sources including Amazon's Mechanical Turk (MTurk), Craigslist, and listservs. MTurk is an online labor market used to recruit participants into various research studies. Research indicates the results of studies using MTurk data are not significantly different than results obtained using other online domains (Birnbaum, 2000; Paolacci, Chandler, & Ipeirotis, 2010).

Participants who clicked on our survey link, from any recruitment source, were directed to a Qualtrics survey. Participants first completed the Informed Consent page, and then were directed to the survey, which included programmed skip patterns to guide the participant through the survey. To discourage low-quality responses, we included several reliability check questions, for example, “Regardless of where you live, please select Chicago, IL from the list below,” followed by four city, state options including Chicago, IL. Kittur and colleagues (2008) found a significant increase in the quality of the data obtained when they included verification questions. Survey completion time was approximately 15 min. Participants received US$2.50 for completing our survey—issued either directly through MTurk or via Amazon gift cards.

Measures

Planned Analyses

Our first research question—how do PRs and LDRs compare on relationship and health variables—was assessed via t-tests comparing the means of these outcome measures across relationship groups. Our second research question—does relationship type (PR vs. LDR) predict responses on relationship and health outcome variables—was assessed using stepwise linear regression models. Using SPSS 21.0 (IBM Corp, 2012), we entered covariates in the first step, and relationship type in the second. These analyses were exploratory, that is, we made no specific predictions regarding the findings for two reasons: No previous study has attempted to answer these questions, and previous, related research has delivered mixed findings. Recognizing that these are exploratory analyses including a high number of statistical tests, we used the conservative p value of .01 as a cutpoint for determining significance.

Health

Primarily, we compared health outcomes between individuals in PRs and LDRs. Overall, and somewhat surprisingly, LDR participants looked “healthier” on more outcomes than PR participants. LDR participants ate healthy with their partner, and exercised both with and without their partner, more frequently than PR participants. And, LDR participants reported better overall health; greater satisfaction with their social role; and less anxiety, depression, and fatigue.

LDR participants reported eating healthier with their partners. Previous research comparing healthy eating in married and unmarried individuals has been mixed (Jeffery & Rick, 2002; Schone & Weinick, 1998; Wilson & Oswald, 2005). One study found that eating habits of one spouse are often interdependent with the other spouse. As such, each spouse's eating behavior potentially could have a positive, or negative, impact on the other (Hommish & Leonard, 2008). Similar in nature, the concept of “partner effects” states that an individual in a marital relationship may influence his or her partner's health behaviors and, therefore, health outcomes (Lewis & Butterfield, 2007). Relatedly, marriage may act as a social control that promotes beneficial health behaviors (Umberson, 1987, 1992).

The notion of marriage as social control may explain another finding—that individuals in LDRs exercise more frequently with their partner than do PR individuals. In this case, the relationship is a social control that promotes positive norms for health behavior. However, when the partners are apart, individuals in LDRs still exercise more frequently. This requires an alternate explanation. A study on marital transitions and health for women found that widowhood and divorce are associated with an increase in physical activity. Underlying this is an increased amount of time one has to devote to oneself (Lee et al., 2005). Without the time constraints that marriage maintenance requires, it is reasonable to expect that LDR individuals have more free time. Alternatively, hormone levels may explain our exercise findings. A study comparing individuals in PRs and LDRs found that men in LDRs have higher testosterone levels than men in PRs. And, both single men and women have higher testosterone levels than those in relationships (van Anders & Watson, 2007). Testosterone is associated with competitive behaviors (Carré & McCormick, 2008), which include exercise and maintaining physical attractiveness (van Anders & Watson, 2007). Our findings suggest that individuals in LDRs are engaging in more competitive (here, exercise) behaviors perhaps as a result of being away from their spouse and thus having an experience that physiologically looks more single.

On the PROMIS health scale, individuals in LDRs reported better overall scores than their PR counterparts. This finding provides a new contribution to the literature, as this is the first known research on long distance marriages and health correlates. There are a number of reasons that this may be the case. Possibly, LDR partners reap the benefits of both the social control of marriage and the freedom of independence and time. In doing so, they receive an optimal balance of positive health benefits related to their partner as well as benefits of structuring their own time and being alone (Long & Averill, 2003). PROMIS subscale scores may provide more insight into the mechanisms of such “protection.” For example, individuals in LDRs displayed lower anxiety and depression scores. One might expect physical proximity to one's spouse to be more beneficial to one's mental health via the advantages of social support (Wilson & Oswald, 2005), and the lack of additional stressors that being apart might include. However, perhaps there is an alternate directionality in that couples who have better mental health are more comfortable in pursuing LDRs as postulated by the marriage selection hypothesis (Stutzer & Frey, 2006). This hypothesis states that individuals who are healthy and/or “evolutionarily fit” are selected into partnerships.

Additionally, individuals in LDRs reported less fatigue than PR individuals. This finding corroborates the notion that LDR individuals have more time to allocate to other activities. Maintaining a marriage is time consuming and may contribute to overall fatigue and individuals in LDRs may have fewer time-consuming and fatigue-inducing experiences, by nature of having fewer in-person interactions with their romantic partners. Finally, PROMIS “social role” constitutes an individual's satisfaction with their ability to perform in various settings, such as work and home. Higher reported satisfaction with social role by LDRs may be attributed to individuals in LDRs feeling comfortable with the division of labor and the compartmentalization of their different roles. This is consistent with past empirical findings that dual-career couples report little to no role strain (Anderson & Spruill, 1993).

Conversely, some of our results support past research findings (e.g., Groves & Horm-Wingerd, 1991) that distance between partners attenuates the positive marriage–health association. Individuals in PRs indeed reported “better” health on some indices—sexual frequency, medication adherence, physical functioning, and individual stress—but these are few compared to the number of health outcomes tested in this study.

Relationships

In addition to health outcomes, we also compared relationship variables between groups. PR and LDR participants did not differ significantly on relationship satisfaction ratings, in contrast to previous research showing that LDR individuals report higher relationship quality than their PR counterparts (Kelmer et al., 2013). We also assessed relationship maintenance using two measures and found contrasting results. On the RCCU, which assessed maintenance behaviors during periods of noncopresence, the two groups did not report any significant differences. Conversely, on the RSRMS that measured routine relationship strategies, PR participants trended toward higher scores than LDR participants. This suggests that PR and LDR individuals do not differ in communication continuity when they are apart, but PR individuals may more regularly demonstrate some key maintenance behaviors, for example, openness and positivity, assurances and advice to their partners, and sharing of social networks. Perhaps PR participants score better on the relationship strategies measure because of the increased time spent together and overlapping of social networks inherent in PRs.

We also tested relationship type as a predictor of health and relationship outcomes, using stepwise linear regressions. Overall, relationship type predicted 1 of 4 relationship variables and 7 of 15 health variables. Being in a PR predicted higher sexual frequency, and “healthier” scores on subscales assessing social role and fatigue. Being in a LDR predicted higher perceived stress and higher relationship stress. Simultaneously, however, it also predicted better diet and exercise habits. This is an interesting dichotomy. LDR individuals report more stress but not related unhealthy behaviors. Perhaps this is because the specific stressors experienced in LDRs, such as extra expenses for travel, do not have significant negative implications for health behaviors (Mietzner & Lin, 2005). Overall, studies that compare LDRs and PRs acknowledge that LDR-specific stress has been underresearched (Pistole et al., 2010).

Interpretations and Future Research Directions

The findings from the RRSS, which measured stress within the relationship, showed that PR participants reported lower scores than LDR participants. In conjunction with our previous findings that PR individuals report lower levels of overall stress than LDR individuals, we can conclude that being in an LDR is correlated with higher stress levels, both within and outside the relationship. Research on the challenges of being in an LDR points to increased pressure that the individuals feel when reunited to make the most of their short time together (Sahlstein, 2004), which may account for our finding. Alternatively, because time spent together is so precious, individuals often devote entire weekends to partner activities, which may result in additional stress over undone professional work and may provoke a feeling of being overwhelmed (Sahlstein, 2004). When apart, individuals in LDRs may be unable to moderate their partner's stress. For example, the stress/social support model proposes marital happiness may moderate the effects of stressors on an individual (Robles & Kielcolt-Glasser, 2003). However, those in LDRs may not be able to spend as much time with their partner face-to-face due to distance. Ditzen et al. (2007) found that women who received positive physical interaction with their partner before stress showed significantly lower levels of cortisol and lower heart responses than those who received no social interaction or only verbal support. Those in LDR may not have as much physical contact with their partners as their PR counterparts, which may lead to higher levels of stress individually as well as in the relationship. Future studies can pinpoint the mechanisms underlying stress relief in long-distance couples by clarifying the role of technology, for example, FaceTime, Skype, Google Chat, and so on, that is commonly used in LDRs (Hertlein, 2012; Tong & Walther, 2011). For individuals in LDRs, technology use, especially to establish “face-to-face” contact, may serve as a protective factor against stress—tantamount to individuals in PR spending time together. Alternatively, technological “face-to-face” contact may not provide the same stress relief, and more broadly health benefits, as actual in-person contact.

Implications for Marriage and Family Practice

Our findings have several clinical implications. Working with LDR couples presents a challenge in that, based on our findings, we now know that on average they experience a significantly higher level of relationship-related stress than those in PRs. Clinical work should focus on stress management and relief in these relationships through techniques that promote and develop coping mechanisms, for example, coping-oriented psychoeducation (Bodenmann, 1997) and Couples Coping Enhancement Treatment (Bodenmann & Shantinath, 2004). Providing psychoeducation on effective conflict resolution tactics, either preemptively or as conflict arises, could be apropos for LDR couples. Related, current conflict resolution tactics, such as the Prevention and Relationship Enhancement Program (Renick, Blumberg, & Markman, 1992), could be adapted for the LDR couples. Such adaptations could include conflict management, communication, and supportiveness when apart; and, setting boundaries and guidelines for time together, for example, not expecting time together, although rare, to be “perfect.”

Clinicians also can help LDR couples navigate distance with technology. For example, setting reminders to check-in with each other can help partners feel connected as can exploring personal preferences on what types of technology to use for connection when apart (e.g., text, phone call or video calls). Further, clinicians may help reconcile the dual identities that LDR individuals face by strategizing with clients on how to incorporate the partner into the individual's daily life, as well as how to maintain and build intimacy, even when partners cannot be together physically. Helping the couple identify, acknowledge, and resolve these constraints may help ameliorate some stress inherent in LDR.

LDR may affect not just two members of the couple but other family members, namely, children (Johnson et al., 2000). Family therapy for LDR couples can be used as, or as a supplement to, structured family meetings—times for the long-distance parent to check-in on their children's academic standing, interpersonal issues, behaviors, etc and to connect emotionally and communicate effectively with their children. In this way, the ritual of family therapy could counterbalance the long-distance parent not consistently being in the home. Moreover, family therapy sessions can be used to collaboratively generate other rituals for the LDR family to engage in, for example, nightly phone calls between the long-distance parent and the family members. Additionally, children may want to use family therapy session time to process their own emotional reactions to the LDR arrangement. Family therapists can model for parents effective responses to these disclosures; and, can teach stress reduction and other coping techniques to the child (Rhodes, 2002).

Limitations

Our findings should be interpreted in the context of study limitations. First, the sample was relatively homogeneous—composed of mostly White heterosexuals. This is consistent with recent published studies of LDR individuals and couples (e.g., Dargie et al., 2015; Kelmer et al., 2013). Our, and other studies’, relatively homogenous sample may represent high prevalence of this demographic group among those in LDRs. Alternatively, our homogenous sample may represent the more global phenomenon of disproportionate participation of, for example, Whites over non-Whites in online research (Du Bois, Johnson, & Mustanski, 2012). Fortunately our sample was large, therefore providing us the power to detect significant differences between not only LDR and PR, but also, for follow-up analyses, subgroups of these, for example, LDR males versus LDR females. Second, our study is cross-sectional, thus prohibiting any causal interpretations of the data. Therefore, we did not test “marriage protection” purely, given that we cannot state with confidence that being in PR or LDR causes, for example, relationship satisfaction, or less fatigue. Certainly, longitudinal or experimental designs are preferred, in part because they can help determine causality. However, the purpose of this study was to provide preliminary information about how LDR status was associated with relationship and health indices; and then, for future studies to refine our design and help to establish causal relationships between the variables we found to be significantly related here. Finally, with regard to analyses, we did not prohibit both spouses from completing the survey. Therefore, some data dependency may exist in this data set (e.g., Farr & Patterson, 2013). However, we did prohibit from completing the survey any repeat IP addresses; therefore, if partners used the same computer to complete the survey, the second partner's data were not included in the data set.