Couple-oriented interventions for chronic illness

Develop an intervention that works for multiple chronic conditions

An intervention that works for multiple patient populations, similar to the Chronic Disease Self-Management Program (CDSMP; Lorig et al., 1999), would expand the reach of CIs. As noted earlier, the chronic diseases that are leading causes of morbidity and mortality share common features. In addition, the high prevalence of physical comorbidities within patients (e.g., cardiovascular disease in persons with diabetes) argues for a cross-disease approach. Cardiovascular disease, cancer, diabetes, and arthritis are the most common and costly chronic health problems and would seem to be good candidates for a cross-disease intervention. Such an intervention could include modules that address issues common to multiple conditions (e.g., improving diet and exercise through dyadic problem-solving and goal-setting activities) and modules that address issues unique to these diseases (e.g., information for the couple regarding effects of cancer treatment on sexual function). Alternatively, a set of modules could be developed to target broad domains that are common across disease populations, with content tailored to the specific issues faced within each domain.

Tailor interventions to the needs of couples

With only one exception, all of the CIs included in our recent review seemed to provide couples with the same materials regardless of their specific needs in different domains. Interventions are likely to be more impactful if tailored to the needs of individual couples. In a tailored approach, the dosage of a specific intervention module depends on characteristics of the couple and could also change for that couple over the duration of the intervention. The potential advantages of tailored interventions include increased participant engagement and increased potency of the intervention. An example of a tailored intervention for couples is the FAMCON (family consultation) intervention for smoking cessation, which has shown cessation and maintenance rates that are comparable to patient-focused interventions (Shoham, Rohrbaugh, Trost, & Muramoto, 2006). In FAMCON, couples receive up to 10 consultation sessions that address couple dynamics which contribute to smoking, such as ironic processes (i.e., spouses’ inadvertent reinforcement of smoking), and that help couples realign their relationship in ways that are not organized around tobacco use. Recommendations are made to the couple based on information collected during the first few meetings.

Explore innovative methods of intervention delivery

Tailored interventions should be delivered in ways that help couples to practice new skills in their daily lives, thereby improving the chance of long-term gains. Ecological momentary intervention (EMI) is an innovative behavioral method that uses mobile technology to deliver interventions to individuals as they go about their daily lives and has been shown to be effective for a variety of health behaviors and psychological or physical symptoms (Heron & Smyth, 2009). EMI involves delivering tailored content in real time (e.g., text messages via a smartphone) based on data that were collected prior to or during the intervention. This method of delivery seems especially promising with regard to physical activity (e.g., King et al., 2008) and other concrete behavioral goals for diet and medication adherence, and may be especially appropriate for couples who cannot travel easily. This application of EMI would be dyadic in that it would be based on within-couple associations observed through either pre-intervention Ecological Momentary Assessment (EMA) or a clinical interview with the couple (e.g., data on spousal communications and overt behaviors that precede changes in patient physical activity). Ideally, dyadic EMI would be preceded by one or more meetings with an interventionist to learn basic concepts and skills and establish a collaborative working relationship.

Identify mechanisms of successful CIs

Studies testing CIs often lack a strong foundation in conceptual models and, relatedly, are rarely designed to examine mechanisms of change. Interventions that successfully modify relationship-related mechanisms (e.g., partner support) and then measure change in patient functioning would enhance our understanding of how close relationships do or do not affect health. Moreover, these interventions could provide critical information regarding the unique and shared pathways leading to clinical outcomes in different diseases (Miller, Chen, & Cole, 2009; Uchino, 2006).

Compare CIs to evidence-based patient-oriented interventions

In our recent review, only half of the studies compared CI to patient-oriented behavioral intervention whereas the remainder had a usual care control group. The comparison of patient- and couple-oriented approaches can help to confirm that greater improvements in patient health achieved by CI are due at least in part to change in dyadic processes and not only due to change in patient psychological well-being or health behaviors (Martire & Schulz, 2007). Thus, it is important for the fields of relationship science and behavioral medicine to compare the efficacy of CIs to that of proven patient interventions for specific conditions or multiple conditions (i.e., the CDSMP; Lorig et al., 1999). An additive treatment design could be used, where a couple’s component is added to a standard patient intervention. Alternatively, researchers could adopt an evidence-based patient intervention, modify it for couples, and compare the relative efficacy of the two approaches (e.g., Martire, Schulz, Keefe, Rudy, & Starz, 2007). Of course, there are challenges in creating two interventions that differ in patient versus dyad focus but are otherwise sufficiently equivalent on dimensions such as length and intensity, but tackling this issue will be worthwhile.

Develop a cross-disease outcomes battery

A brief battery of patient and spouse outcome measures for use across illness populations would promote greater synthesis of the CI literature. Outcome domains that are important across populations include psychological well-being, health behaviors, relationship functioning, physiological functioning, and patient illness symptoms. Specific measures tapping these domains should be identified for inclusion in an outcomes battery. A cross-disease assessment approach has been adopted with success in the Patient-Reported Outcomes Measurement Information System (PROMIS; Cella et al., 2007) and the NIH Toolbox (Gershon et al., 2010). These projects have yielded brief measures of psychological, physical, and social health functioning for use in clinical trial research, and are a useful starting point for developing a cross-disease outcomes battery for CIs.

Proximal effects of close relationships on patient health

The quality of a patient’s marriage clearly affects long-term outcomes such as recurrent health events, hospitalization, and survival (e.g., Kimmel et al., 2000; Orth-Gomer et al., 2000; Rohrbaugh, Shoham, and Coyne, 2006). But a key question remains largely unanswered: what are the proximal effects of close relationships that lead to these more distal health outcomes? Furthermore, would CIs have larger effects if targeted at these associations? Negative aspects of marital functioning are related to changes in depressive symptoms, cardiovascular reactivity (increased heart rate and blood pressure), and immune response (Kiecolt-Glaser & Newton, 2001). However, other potential pathways linking relationship functioning to health have received little empirical attention. I will provide examples of two domains in which future research would be especially valuable.

First, we currently have little understanding of the spouse’s role in patient health behaviors essential for the management of chronic illness. It is unclear what exactly spouses say and do to affect patients’ daily physical activity, dietary adherence, smoking or alcohol use, medication adherence, and sleep practices. Evidence that obesity and physical activity are related not only to the etiology of disease but also length of post-diagnosis survival puts this issue at the forefront (e.g., Ibrahim & Al-Homaidh, 2010). A second relatively unexplored pathway linking close relationships and health is parasympathetic activation. Recent research shows that marital interactions affect heart rate variability (HRV), which is believed to reflect an individual’s efforts to regulate emotion. Over time, negative interactions may reduce resting HRV, an indicator of self-regulatory capacity, and increase the risk of cardiovascular morbidity and mortality (Butler, Wilhelm, & Gross, 2006; Smith et al., 2011).

Questions about associations between relationship functioning and proximal indicators of health are perhaps best answered with repeated measures designs in which both partners are assessed daily or throughout the day (e.g., EMA). Intensive repeated measurement allows for addressing questions regarding within-couple variability (e.g., on days when a spouse is more critical, is the patient less physically active?), and extends previous research on couples and chronic illness that has taken a between-couple approach to analysis. Intensive repeated measurement studies could also reveal the effects of patient symptoms on partner mood, support behaviors, and physiological responses (Monin et al., 2010).

Moderators of relationship-health linkages

Clinical experience tells us that there is much variability between couples in terms of the strength of within-couple associations such as those described above. In some couples the patient’s treatment decisions and daily medication adherence are strongly affected by the partner’s attitudes and own self-care, whereas in other couples the patient’s decisions and adherence are driven more by their own characteristics or other interpersonal influences. As another example, in some couples the patient’s symptoms have little negative impact on the partner whereas in other couples the partner’s emotional well-being is essentially a barometer for how the patient is feeling. Research aimed at identifying factors that explain these between-couple differences could be used to tailor CIs. More specifically, it would be useful to identify moderators that tell us who may benefit most from a CI, as well as moderators at the within-couple level that tell us the appropriate timing and amount of a CI.

There are many potential between-couple moderators of the relationship-health association (e.g., stage or severity of patient illness, illness in the spouse), but perhaps the most obvious are overall relationship quality and gender-linked traits. The general quality of the relationship may serve as an interpretive backdrop that alters patients’ appraisals of spousal behaviors and therefore the impact of those behaviors on health (Kiecolt-Glaser & Newton, 2001). By extension, individuals with low marital satisfaction may experience greater benefits from a CI than those with high marital satisfaction. Gender-linked traits may also be an important moderator. Studies of cardiovascular disease indicate that, compared to male patients, female patients show stronger linkages between marital stress and outcomes such as length of hospital stay after coronary artery bypass surgery and survival of congestive heart failure (Coyne et al., 2001; Kulik & Mahler, 2006). However, for some conditions (e.g., prostate cancer), sex is confounded with patient/partner status. In addition, for heterosexual partners, being a female patient is confounded with having a male partner. Therefore, it may be more useful to examine gender-linked traits, such as unmitigated agency or communion (e.g., Helgeson & Lepore, 2004) as moderators of relationship-health linkages. This would allow us to examine this moderator with all diseases and to include both opposite-sex and same-sex couples.