The caregiver burden in male romantic partners of women with non-metastatic breast cancer: The protective role of couple satisfaction

Introduction

Breast cancer is the most common cancer among women (Bray et al., 2004). An international epidemiological study showed that its incidence rose by 30%–40% worldwide from the 1970s to the 1990s (Althuis et al., 2005). The growing need for assistance and care for patients, and the impossibility for public health to face this increasing demand, has progressively shifted the burden of patient care from hospitals to informal caregivers (Feldman and Broussard, 2006). The term ‘informal caregiver’ is used in the literature to define a person from the close environment (e.g. partner, adult child or other relatives, friends and neighbours) who does not possess any training in health or psychological care and who provides unpaid care for the patient (Carretero et al., 2009). In most cases, the romantic partner is the principal relative involved in this role (Li et al., 2013).

The partners of women with breast cancer are in charge of providing emotional support and helping the woman to manage her negative emotions such as despair, fear and sadness (Bozo et al., 2009; Lethborg et al., 2003). Caregiving also implies assuming the management of the home and children (if any), while assisting the woman in the daily management of her disease and continuing to carry out work and social commitments (Bakas et al., 2001; Bigatti and Wagner, 2003). The role of informal caregiver thus includes multiple responsibilities.

In the literature, this set of care tasks was referred to as an ‘objective burden’, which is distinguished from the ‘subjective burden’, defined as the feeling of intrusiveness and disruption of the illness and caregiving in caregivers’ lives (Zarit et al., 1980, 1986). Studies showed that the subjective burden (in general measured in terms of the caregiver’s subjective perception of the impact of caregiving on his or her life in different domains) is a strong predictor of caregiver’s quality of life and psychological well-being, over and above the objective burden (in general measured in terms of the number of hours the caregiver spends on caregiving and the number and kinds of tasks he or she is in charge of) (Braun et al., 2007; Haley et al., 2003).

To date, empirical investigations on the informal caregiver of the cancer patient are still relatively rare and mostly focused on advanced, metastatic and palliative stages of cancer (e.g. Braun et al., 2007; Grunfeld et al., 2004). Existing data show that in these situations, caregiving is significantly linked with decreased psychological and physical well-being of the caregiver, with problems such as depression, anxiety, excessive fatigue, sleep disturbance and chronic pain. Between 20% and 30% of caregivers meet the criteria for a psychiatric diagnosis (Baronet, 1999; Li et al., 2013; Pitceathly and Maguire, 2003; Stenberg et al., 2010).

Few studies have so far investigated the impact of non-metastatic breast cancer on romantic partners in terms of caregiver burden (e.g. Lopez et al., 2012; Wagner et al., 2011). Moreover, the evolution over time of the caregiver burden has rarely been studied, although it is known that the care tasks, the burden felt by the caregiver and his well-being may fluctuate following possible changes in the patient’s health (Nijboer et al., 1998; Stenberg et al., 2010). Evolution over time is all the more important to consider when studying non-metastatic breast cancer, where the medical condition of the patient can rapidly change following the treatment phase and where, in most cases, patients enter a rehabilitation phase after a relatively short period. In this context, the caregiver’s subjective burden may also rapidly change in accordance with the medical condition of the patient.

In order to understand the caregiver burden, we referred to the stress process model (Pearlin et al., 1990; Yates et al., 1999), which has been widely used to identify risk and protective factors for caregivers’ well-being. According to this theoretical model, the objective burden (i.e. illness-related factors) and the background and context characteristics of the caregiver (i.e. gender, age, ethnicity, socio-economic status and biopsychosocial history) are at the origin of the subjective burden. Caregivers have personal and social resources that act as protective factors and consequently moderate the effect of the stressors on possible negative outcomes. Individual coping and social support (i.e. specific instrumental and emotional support that the caregiver may receive from his close environment in the context of the illness) are the most studied resources to explain modulations in the caregiver burden in a variety of illnesses such as cancer, dementia or spinal cord injury. Studies have shown that efficient coping and support are both related to a lower subjective burden and better psychological adjustment (Chiou et al., 2009; Papastavrou et al., 2009; Rodakowski et al., 2012). Surprisingly, the general quality of the relationship between the caregiver and the patient has rarely been taken into account. In non-cancer diseases, a few studies have nevertheless shown that the caregiver’s satisfaction with the relationship was significantly linked to the subjective burden: the higher the quality of the relationship, the lower the perception of burden (Quinn et al., 2009; Snyder, 2000; Steadman et al., 2007; Williamson and Shaffer, 2001). A recent study on caregivers of cancer patients entering home hospice care found that the relationship quality significantly predicted the spousal caregiver burden (Reblin et al., 2015). These findings strongly support taking into consideration the quality of the relationship in a model of stress and coping in caregiving for non-metastatic cancer patients.

The first aim of this study was to assess the subjective burden of partners of women with non-metastatic breast cancer and its evolution over time. We thus considered two different phases: the active treatment phase (3 months post-surgery; T1 in the research design) and the rehabilitation phase (12 months post-surgery; T2 in the research design). We hypothesized that the subjective burden would decrease when patients enter the rehabilitation phase (i.e. end of active treatment). The second aim of this study was to examine the unique contribution of couple satisfaction to the subjective burden of romantic partners of women with non-metastatic breast cancer. We hypothesized that higher couple satisfaction would be associated with a lower subjective burden at both assessment points. In accordance with the stress process model, we tested this hypothesis using the socio-demographic data of the partner as the background and context factors and using the medical and psychological conditions of the patient as the primary stressor (i.e. objective burden). We also considered individual coping, social support and couple satisfaction as resources that may moderate the effect of caregiving stress and protect the partner from the subjective burden.

Methods

Results

Couple satisfaction as predictor of subjective burden

Preliminary checks

We assessed first which background and context factor (socio-demographic data of the partner) and primary stressors (medical and psychological conditions of the patient) were associated with subjective burden (see section ‘Participants’ for details about socio-demographic and medical data). The level of subjective burden at T1 was significantly correlated (Pearson’s correlation) with the romantic partner’s age (r = −.42; p = .003), whether the patient underwent ALND (r = .37; p = .012) and the patient’s psychological distress (r = .49; p < .001). These three variables were thus selected and introduced into the regression models as predictors of subjective burden at T1. At T2, no socio-demographic or medical variable was significantly correlated with the subjective burden. Accordingly, none of these variables were tested in the regression model.

Bivariate links between study variables

Results of Pearson’s correlations showed that, according to our second hypothesis, the partner’s couple satisfaction was strongly associated with the level of subjective burden at T1 (r = −.76; p < .001) and T2 (r = −.68; p < .001): the higher the couple satisfaction, the lower the subjective burden reported by the partner. At T2, but not at T1, satisfaction with perceived social support was also related to subjective burden: the higher the satisfaction with social support, the lower the subjective burden (r = −.39; p = .006). No significant association was found between subjective burden and individual coping at either T1 or T2.

Predictors of subjective burden

Two hierarchical multiple regression analyses were performed to examine the role of couple satisfaction in determining the subjective burden at T1 and T2 (see Table 1). Results at T1 showed that age, which was entered at step 1, significantly explained 18% of the variance of subjective burden: the younger the partner, the higher the subjective burden. The medical and psychological condition of the patient (i.e. ALND and psychological distress) introduced at step 2 accounted for a significant additional 26% of the variance of subjective burden: patients having undergone ALND (β = .36; p < .01) and patients having a higher level of psychological distress (β = .40; p  < .01) both significantly induce a higher level of subjective burden. Couple satisfaction, individual coping and social support, introduced at step 3, added a significant additional 29% of the variance, but only the effect of couple satisfaction was significant (β = −.61; p  < .001): the higher the couple satisfaction, the lower the subjective burden. The variables that stayed significant at step 3 were ALND, the patient’s psychological distress and the partner’s couple satisfaction. This final model explained 72% of the variance of subjective burden.

OPEN IN VIEWER

Table 1.

Hierarchical multiple regression analysis predicting romantic partner’s subjective burden at T1 and T2 (n = 47).

	T1			T2
	Step 1	Step 2	Step 3	Step 1
Age	−.42**	−.30*	−.16
ALND		.36**	.30**
Patient’s distress		.40**	.19*
RAS			−.61***	−.67***
INCOPE-2R			−.04	−.02
SSQ-6			.03	−.01
R 2	.18	.44	.72	.46
ΔR2	.18**	.26***	.29***	.46***

ALND: axillary lymph node dissection; RAS: Relationship Assessment Scale; INCOPE-2R: Individual Coping Questionnaire (revised version); SSQ-6: Social Support Questionnaire (6-item short-form).

*

p < .05; **p < .01; ***p < .001.

At T2, couple satisfaction, individual coping and social support were introduced at step 1, but only couple satisfaction was a significant predictor of subjective burden (β = −.67; p < .001): the higher the couple satisfaction, the lower the subjective burden. The model explained 46% of the variance of subjective burden.

For all regressions, residual analyses revealed no outliers, homoscedasticity and normality of residuals. Predictors were not multi-collinear.

Discussion

In this study, we assessed the subjective caregiver burden in male partners of women with non-metastatic breast cancer. The aims of the study were, on one hand, to assess the subjective burden and its change over time and, on the other, to examine the contribution of couple satisfaction in moderating the impact of caregiving stress on the partner’s subjective burden.

The majority of romantic partners reported a subjective burden 3 months (i.e. active treatment phase, T1) and 1 year (i.e. rehabilitation phase, T2) after surgery; in accordance with our hypothesis, subjective burden was significantly lower at T2. Results from multivariate analyses offer some details to better understand this decrease over time. During the active treatment phase, the patient’s medical and psychological conditions were significant predictors of subjective burden. Indeed, we found that ALND and the patient’s psychological distress explained approximately 26% of the variance of subjective burden at T1. The negative impact of ALND may be explained in terms of an increase in the patient’s physical impairment. This type of surgery may indeed induce several side effects such as pain, swelling, weakness and trouble moving the arm, but also more severe problems such as lymphoedema (Swenson et al., 2002). Patients are thus often debilitated, preventing the use of their arm, and became more dependent on the practical help of their partner. This result was in line with other studies showing that increased physical impairment in a patient is linked to greater caregiver burden (Pinquart and Sörensen, 2003). The link between the patient’s distress and the partner’s subjective burden was not surprising. The literature has indeed shown that among the many tasks that the informal caregiver must assume, providing emotional support proved to be the most time-consuming and difficult (Bakas et al., 2001; Carey et al., 1991). In contrast, at T2, no significant link was found between subjective burden and the patient’s medical and psychological condition in bivariate and multivariate analyses. This could be explained by the fact that at that time, the patients have finished chemotherapy and radiotherapy. The end of these treatments and the gradually physical recovery from surgery marked the beginning of the rehabilitation phase, which is also accompanied by a lower level of psychological distress in patients. This improvement in patient condition was probably at the origin of the decrease in subjective burden reported by the partner between T1 and T2.

Concerning the effect of couple satisfaction on subjective burden, results of multivariate analyses confirm our second hypothesis. Couple satisfaction, after the medical and psychological conditions of the patient were controlled for, explained 29% and 46% of the variance of subjective burden at 3 and 12 months post-surgery, respectively. This means that the negative impact of caregiving largely depends on the quality of the relationship between the caregiver and the care recipient, over and above the patient’s condition and regardless of the phase of treatment. This result was in line with the few studies focusing on the caregiver–patient relationship that showed that the quality of this relationship was crucial in determining subjective burden. When the quality of the relationship was good, caregivers experienced less subjective burden even though an objective burden was present (Snyder, 2000; Steadman et al., 2007). Starting with these results, we can conclude that subjective burden seems to have an important relational component. Characteristics of satisfied couples may explain this effect. First, the partners of highly satisfied couples normally reported an important sense of ‘we-ness’, namely a sense of unity (Gottman and Levenson, 1999; Reid et al., 2006). In the context of a serious illness, where the lives of both partners are likely to be affected, this translates into the perception of the illness as a shared stressor or a ‘we-disease’ (Kayser and Scott, 2008), which involves both partners and not only the patient. From this perspective, we can imagine that caregiving is not perceived by the healthy partner as an external imposition, but as the obvious consequence of the situation and the only thing to do to face the illness together. In addition, according to the equity theory (Walster et al., 1978), when a couple faces a serious illness such as breast cancer, there is a change in the balance of give-and-take between partners (Cutrona, 1996). The healthy partner must take on new caregiving roles and the exchange may become more unidirectional (Kuijer et al., 2004). In distressed couples, partners may feel inequitably treated and this feeling of inequity is generally associated with increased distress. In contrast, in satisfied couples, the imbalance is not experienced as inequitable, but only as a momentary transition (Buunk and VanYperen, 1991; Kuijer et al., 2001).

Contrary to what was assumed in the stress process model, no background and context characteristics of the partner predicted subjective burden. The effect of age, the only variable that was found to be significantly correlated with subjective burden at T1, lost its significance when couple satisfaction was entered into the regression model. In addition, no significant effect was found for individual coping and social support.

This study has several limitations. The modest sample size limits generalizability. The participation rate of romantic partners was not high but is nonetheless satisfactory, considering that participants must complete questionnaires during the acute illness phase, as well as the great difficulty in recruiting couples in psycho-oncology studies in which both the patient and the partner need to agree to participate (Fredman et al., 2009). A second limitation was that we did not have a specific measure of care demands (e.g. number of hours the partner spent on caregiving, number and kinds of tasks that the partner performed). Finally, several studies have shown that caregiving may also have a positive impact on the caregiver. Indeed, caregiving may be experienced as a challenge and may induce feelings of pride, a sense of self-worth and rewards. It may also be accompanied by increased love, affection and commitment between the caregiver and the care recipient (Carey et al., 1991; Lopez et al., 2005; Marks et al., 2002). It would thus be incorrect to consider caregiving only as a stressor linked to distress, fatigue and burden (Nijboer et al., 1998). This aspect would be interesting to integrate in further studies because it could have a protective effect on both the couple relationship and the partner’s psychological well-being.

Notwithstanding these limitations, this study has shown that couple satisfaction is the factor that most explained the subjective burden in romantic partners of non-metastatic breast cancer patients during the first year after surgery. A study of the subjective burden in the context of a loving relationship thus seems to need to take into account the quality of the relationship because it can determine the perception of caregiving. From a clinical perspective, interventions aimed at helping romantic partners to better adjust to the illness and to improve their psychological well-being need to take into account the quality of the relationship. Reducing the partner’s distress by directly acting on the objective burden, for example, by reducing caregiving tasks and providing practical support, would certainly help. In the case of an illness that enters remission such as non-metastatic breast cancer, however, the portion of subjective burden linked to the concrete (objective) caregiving burden probably would have declined naturally over time following the improvement of the medical condition of the patient, while leaving unaltered any portion of burden linked to an unsatisfactory couple relationship. It would thus be more interesting to evaluate the relationship between the caregiver and the patient to identify possible difficulties and, if any exist, to assist caregivers and care recipients in restoring, improving and cultivating their relationship. This could be done, for example, by improving communication and reciprocity and restoring the perception of equity in a give-and-take balance, while resolving and lessening couple conflicts. Acting on a dysfunctional and unsatisfactory relationship would probably have a more effective and durable effect on the subjective burden of the caregiver and consequently on his well-being.