Modeling the Distress of Spousal Caregivers of People with Dementia

Abstract

Background: The progressive mobilization of spouse caregivers who take care of a person with dementia (PWD) can lead to situations of distress.

Objective: The current study sought to investigate the influence of the characteristics of the caregiving context on spousal caregiver distress.

Methods: 125 spousal caregivers participated in this study. The characteristics of the caregiving context were assessed using questionnaires. We examined a moderated-mediator model (Step 1) in which we hypothesized that PWD and caregiver characteristics and dyadic determinants contribute to spousal caregiver distress. This model was compared based on the age at onset of the disease and the gender of the caregiver (Step 2).

Results: The model revealed that poor self-rated health and a lack of family support accentuated spousal caregiver distress, whereas the feeling of being prepared and level of confidence decreased spousal caregiver distress. Moreover, the quality of couple adjustment affected spousal caregiver distress, and this effect was mediated by the severity of the PWD’s symptoms. Regarding the age at onset of the disease, the path between Couple Adjustment and the Care recipient’s impairments was more important for caregivers of person with early-onset dementia (PEOD). Female caregivers who reported poor self-rated health experienced greater distress.

Conclusions: It would be interesting to create a support program that would incorporate these three areas of intervention regarding the progression of the disease: first, “preparedness modules”; second, “dyadic modules” (especially for caregivers of PEOD); and third, “family modules”. Specific attention should be given to female caregivers who report poor self-rated health.

Keywords

Caregivers dementia early onset Alzheimer disease late onset Alzheimer disease psychological model spouses

INTRODUCTION

Caring for a person with dementia (PWD) athome requires an ever-increasing amount of care. This care is generally provided by the family, usually the children or the spouse [1]. As the disease develops, the PWD requires increasingly moreemotional, psychological, physical, and financial support from the caregiver. Their progressive and permanent mobilization leads to situations of physical and psychological distress [2]. Bandeira et al. [3] assessed burden, depression, anxiety, and Stress as aspects of distress that can be influenced by a variety of factors.

The behavioral and cognitive symptoms of the PWD [4] and changing communication with the PWD [5] contribute to caregiver distress. Moreover, with the evolution of the disease, caregivers are progressively faced with the loss of their freedom [5]. They sacrifice their leisure pursuits, they restrict time with friends and family, and they give up or reduce employment, which causes financial problems [6]. Other factors, such as a lack of social support [6], poor-quality family relationships [7], poor self-rated health [8], a lower sense of confidence in the caregiver role [6], and reduced caregiving self-efficacy regarding the management of the PWD’s symptoms [9], are associated with greater depressive symptoms and burden among caregivers. Conversely, preparedness for caregiving is related to lower levels of caregiver-role strain [10, 11] as well as higher caregiving rewards and better mental health [12].

Studies have shown that caregiving situations are experienced differently depending on the type of relationship with the PWD (e.g., spouse, child, or other) [13, 14]. Distress is more important for spousal caregivers than adult-child caregivers [15, 16], who reported higher levels of physical burden, financial burden, and relationship strain [17]. Spousal caregivers are most likely to live with the PWD, they provide more hours of support, and they experience less respite from the caregiver role [17]. Moreover, a spouse’s diagnosis of dementia leads to specific losses [18]: the spousal caregiver can no longer count on the partner’s help or mutual partnership due to problematic behaviors and cognitive deterioration [19], and changes occur in the partners’ roles, identities, and projects. These losses within many areas of couples’ relationships, including sexuality [20], relationship rituals, communication, closeness, reciprocity, intellectual stimulation, and assistance with decision-making, are associated with spouse caregiver distress [21, 22]. A relationship [23, 24] with unresolved conflicts and low marital intimacy [25, 26] assessed as poor before the disease is associated with the perception of more severe symptoms among caregivers as well as a higher level of burden and depression. However, a satisfying premorbid marital relationship is a protective factor [27, 28].

These studies demonstrate that PWD characteristics (cognitive function disorders and behavioral disturbances), caregiver characteristics (physical and mental health, intrapsychic strains, and social support) and dyadic determinants are directly linked to caregivers’ distress. However, spousal caregiver distress likely involves several intermediate pathways in response to multiple stressors [29]. Van der Lee et al. [30] suggested that complex models are expected to provide a comprehensive view and supply the greatest insight into the importance of determinants of distress. Studies must consider the multidimensional nature of predictors and the ways in which multiple factors involved in caregiver distress interact with each other [31]. Moreover, strengths and vulnerabilities will permit the suggestion of interventions that must target different nuanced aspects to provide optimal benefits for individuals [30, 32].

The review of multivariate models by van der Lee et al. [30] highlighted the importance of including different types of PWD and caregiver determinants in the prediction of distress. This review included PWD determinants such as behavioral and cognitive symptoms and need for support and caregiver determinants such as social support, caregiver physical health, sense of competence, and self-efficacy. In this review, one model that was specifically oriented toward spousal caregivers included both of these categories of determinants: the model of van den Wijngaart et al. [33]. This model showed that PWD’s behavioral problems significantly determined caregiver burden. A limitation of this model was that it did not assess the quality of the couple relationship, which is an important variable associated with distress in previous studies.

These results emphasize the importance of building a model of caregiver distress that is specifically oriented toward the spouse, including PWD and caregiver characteristics as well as dyadic determinants. This study used the conceptual framework of van der Lee et al. [30]. This conceptual framework is based on an extensive review of empirically tested models and categorizes predictors of distress in both categories: PWD and caregiver characteristics. This framework guided the selection of the variables investigated in this study. These variables were completed by dyadic assessment.

The present study seeks to remedy these previous limitations by examining the mediating role of PWD and caregiver characteristics as well as dyadic determinants in the predication of spousal caregiver distress. We examined factors such as the severity of the PWD symptoms perceived by the caregiver, the effect of the disease on their lives, their sense of confidence and preparedness, their self-rated health, the quality of their familial and social support, and the adjustment between both partners. These measures of the caregiving context will initially allow us to examine the predisposing, mediating, and variable outcomes within the same investigation. We examined a model (Step 1) in which we hypothesized that PWD and caregiver characteristics as well as dyadic determinants contribute to spousal caregiver distress. This model participates at the understanding of the implications of the role of the spousal caregiver by highlighting the variables that play important roles in the experience of distress. Furthermore, this model emphasizes important targets for psychosocial interventions among the spousal caregivers of PWD [28]. Second, some studies have recommended accounting for caregiver moderator variables such as gender [30] or age at onset of the disease [34] to offer services tailored to the needs of each group [34]. The model was compared based on the age at onset of the disease and the gender of the caregiver (Step 2).In fact, the literature supports the hypothesis of the specificity of experience based on these variables [34 –38]. The international criteria use a cut-off of 65 years old to distinguish early-onset from late-onset dementia [39]. Thus, it would be interesting to compare the experiences of spousal caregivers of person with early-onset dementia (PEOD) and those of people with late-onset dementia (PLOD) as well as the experiences of female and male caregivers. These comparisons will permit us to consider the appropriate support needed in terms of these characteristics (if any) [34] to improve the efficiency of this support [40, 41].

MATERIALS AND METHODS

Participants

This study recruited 201 spouse caregivers; 51 refused to participate due to incompatibilities with their schedules or the difficulty of discussing the disease in front of the PWD when the questionnaires were completed at home. In total, 150 caregivers agreed to participate in the study; however, 25 incomplete questionnaires were removed before analysis (Fig. 1). Non-parametric comparisons (Mann-Whitney U tests) of the sociodemographic and manifest variables (MVs) between the complete and incomplete questionnaires did not reveal significant differences (all ps > 0.05).

Therefore, the final sample consisted of 125 spousal caregivers. In this investigation, as in many previous studies [42], the participants were mostly women (59.2%). Caregivers had an average age of 67.2 years (SD = 6.5 years). Pinquart and Sorensen [17] demonstrated that relative to other caregivers, spousal caregivers were less likely to be employed. In this study, only 16% of the participants were gainfully employed. The participants were highly educated (M = 14, SD = 4.33); relatively wealthy, with an average income of 83.27 euros per day (SD = 34.28 euros); and were already using a relatively high number of services, mainly housekeeping (43.2%). PWD had an average age of 69.25 years (SD = 5.06 years) and exhibited moderate cognitive impairment on average (M = 14.69, SD = 9.12). The disease had evolved for an average of 7.1 years (SD = 4.45 years).

The early-onset dementia group consisted of 52 caregivers, and the late-onset dementia group consisted of 73 caregivers (Table 1). The mean ages of the two groups were significantly different for the caregivers and for the PWD: 58.67 (5.64) for the caregivers of PEOD compared with 75.73 (7.33) for the caregivers of PLOD and 59.08 (4.44) for the PEOD compared with 79.42 (5.69) for the PLOD. Both groups were comparable regarding the Mini-Mental State Examination (MMSE) of the PWD and the time between the first signs and the time of the interview. Moreover, caregivers of PEOD were married for less time than caregivers of PLOD were.

Procedure

Participants meeting the following selection criteria were recruited from the Regional University Hospital Center of Lille in northern France. Caregivers had to be married to a person with Alzheimer’s disease or a related type of dementia (e.g., mixed dementia, frontotemporal dementia, and Lewy body dementia) and have lived with the PWD.

Caregivers were selected from the list of persons followed by the center for 12 months and were called by the project coordinator to explain the study objectives and solicit their participation. Upon agreeing to participate but before completing the questionnaire, the participants received information about the study and signed an informed consent document. Then, the sociodemographic data of the caregivers were gathered. An interviewer administered the questionnaires at the Regional University Hospital Center or at home depending on the participant’s choice. The average completion time of questionnaires was 120 minutes (SD = 37 minutes).

Ethical issues

The study was approved by the national ethics committee, and adhered to the tenets of the Helsinki Declaration of 1975. All participants were given an informational letter and a verbal explanation of the study details. Upon agreeing to participate in the study, participants signed an informed consent document. The caregivers were assured that their information would be kept confidential and that their identity would not be revealed in publishedreports.

Measures

First, the spousal caregivers responded to a set of self-administered questions that included items addressing their sociodemographic status (i.e., sex, age, number of years of marriage, income, employment, and years of education) and services received (e.g., housekeeping, home-delivered meals, personal care, nursing care, in-home respite care, and daycenter).

Several measures were used to assess the characteristics of the caregiving context including the severity of the PWD symptoms perceived by the caregiver, the effect of the disease on their lives, their sense of confidence and preparedness, their self-rated health, the quality of their social support, the adjustment between both partners and spousal caregiver distress.

The severity of the PWD symptoms perceived by the caregiver

The caregivers assessed the functional autonomy of the PWD using two instruments primarily focusing on functional autonomy. The first focused on activities of daily living (ADL) and consisted of a 9-item scale [43] in which each item was rated using a 5-point Likert scale from 1 (entirely able to perform ADL) to 5 (not at all able). The overall score can range between 9 and 45. The internal reliability (Cronbach’s alpha) was 0.82 in the French version [34] and 0.95 in the study sample. The second instrument focused on instrumental ADL (IADL) and was measured by caregivers using a 4-item questionnaire [44] in which each item was rated using a 4-point Likert scale from 1 (entirely able) to 4 (entirely unable). This score can range from 4 to 16. The internal reliability (Cronbach’s alpha) was 0.73 in the French version [34] and 0.78 for our sample. Caregivers evaluated the severity of the cognitive impairment of the PWD using the 8-item Cognitive Status Scale [45]. This evaluation is based on items from the MMSE [46]. This instrument’s score can range from 0 to 32. The internal reliability (Cronbach’s alpha) was 0.89 in the French version [47] and 0.89 in this study.

The effect of the disease on their lives

The Caregiver Reaction Assessment (CRA) [48] specifically assesses the caregivers’ experiences. Five positive and negative dimensions of caregiver reactions were identified: impact on disrupted schedule, financial problems, lack of family support, health problems, and the effect on self-esteem. This questionnaire contains 24 items that employ a 5-point response scale ranging from 1 (not at all confident) to 5 (extremely confident). The internal reliabilities (Cronbach’s alpha) for the five scales were between 0.75 and 0.85 in the French version [49] and 0.71 (impact on disrupted schedule), 0.79 (financial problems), 0.74 (lack of family support), 0.75 (health problems), and 0.88 (the effect on self-esteem) in our study.

Sense of confidence and preparedness

The degree of caregiver preparedness to provide care was assessed using the 8-item Preparedness for Caregiving Scale [10]. Preparedness was defined as a perceived readiness for multiple domains of the caregiving role, such as providing physical care, providing emotional support, establishing in-home support services, and coping with the stress of caregiving. The responses ranged from 1 (not at all prepared) to 5 (very well prepared). The internal reliability (Cronbach’s alpha) was 0.89 in the French version [50] and 0.88 in this study.

The role that confidence plays was evaluated using the 15-item Self-Efficacy Scale [51]. The caregivers rated their level of confidence regarding coping with caregiving on a 5-point response scale ranging from 1 (not at all confident) to 5 (extremely confident). The internal reliability (Cronbach’s alpha) was 0.89 in the French version [50] and 0.88 for our sample.

Planning for future care needs and knowledge of formal services were evaluated using two questionnaires developed by Sörensen and Pinquart [52] that have response scales ranging from 1 (not at all true for me) to 5 (completely true for me). The Planning for Future Care Needs Scale surveyed the decisions made by caregivers to meet their relative’s future care needs (6 items). The familiarity of caregivers with formal services was assessed using the 7-item Knowledge of Services Scale. The internal reliabilities (Cronbach’s alpha) in the French version for the scales were 0.83 for the Planning for Future Care Needs Scale and 0.88 for the Knowledge of Services Scale [50]. In our sample, the reliability coefficients were 0.82 for the Planning for Future Care Needs Scale and 0.86 for the Knowledge of Services Scale.

Self-efficacy was measured using the Revised Scale for Caregiving Self-Efficacy [53]. This scale is comprised of 3 subscales that specifically measure a caregiver’s ability to obtain respite from family and friends (5 items); to control disturbing thoughts about the caregiving role (5 items); and to respond to a relative’s disruptive behaviors (5 items). The respondents rated their degree of self-efficacy via scores ranging from 0 (absolutely incapable) to 100 (fully capable). Pearson correlation coefficients between the overall scale and the three dimensions ranged from 0.66 to 0.73, and the internal reliability (Cronbach’s alpha) in the French version for the scale was 0.86 [50] and 0.83 (respite care), 0.83 (control disturbing thoughts) and 0.90 (respond to relative’s behaviors) inthis study.

Self-rated health

Based on the SF-36 [54], health was assessed using two subjective questions: “Would you say that your health is...” for the general health [54] and “When you compare yourself with other people your age, would you say your health is...” for health compared with people of the same age. The responses ranged from 1 (very bad) to 5 (very good). The internal reliability in this study was 0.88 (Cronbach’s alpha).

The quality of social support

The Inventory of Socially Supportive Behaviors [55] was used to determine the frequency of the support received by caregivers from their family, friends, and neighbors (informal social support). The 3 subscales address emotional support (11 items), informational support (7 items), and instrumental support (9 items). The items are rated on a 4-point scale ranging from 1 (never/almost never) to 4 (very often/always). The internal reliabilities (Cronbach’s alpha) in the French version were 0.83 (informational support), 0.83 (instrumental support), and 0.84 (emotional support) [56] and 0.77 (informational support), 0.71 (instrumental support) and 0.84 (emotional support) in this study.

The existence of family conflicts concerning care decisions was assessed using the 15-item Family Caregiver Conflict Scale [57]. The caregivers were asked to report the existence of disagreements over general aspects of caregiving using a 7-point Likert scale ranging from 1 (not at all true) to 7 (very true). The internal reliability (Cronbach’s alpha) in the French version was 0.89 [50] and 0.91 for our sample.

Adjustment between both partners

The Dyadic Adjustment Scale (DAS) is the most frequently used evaluation of marital adjustment [58]. In this study, the DAS-16 [59] allowed us to explore two dimensions: the degree of agreement among the couple and the quality of their dyadic interactions. The items were rated on a 6-point scale. The internal reliabilities (Cronbach’s alpha) in the French version for the two scales were 0.89 for the degree of agreement dimension and 0.75 for the quality of their dyadic interactions dimension [59]. The reliability coefficients obtained in our study were 0.92 for the degree of agreement and 0.76 for the quality of interactions.

Spousal caregiver distress

The Hospital Anxiety and Depression Scale (HADS) [60] is a self-assessment for depression and anxiety. This scale is composed of 14 items rated using a 4-point scale from 0 to 3. Seven questions are related to anxiety (total A), and seven are related to depression (total D); thus, two scores are yielded (maximum score for each dimension = 21). To screen for anxious and depressive symptomatology, the following interpretation was proposed for the A and D scores: 7 or less = absence of symptoms; 8–10 = doubtful symptomatology; and 11 or more = certain symptomatology. The internal reliabilities (Cronbach’s alpha) in the French version for the two scales were 0.78 for the depression dimension and 0.81 for the anxiety dimension [61]. The reliability coefficients obtained in our study were 0.75 for the depression dimension and 0.73 for the anxiety dimension in our study.

The 14-item Psychological Distress Index, an adapted version of the Psychiatric Symptom Index [62], measures the level of psychological distress experienced. The respondents rated how often they felt distressed over the past week, from 1 (never/almost never) to 4 (very often/always). The internal reliability (Cronbach’s alpha) in the French version was 0.89 [63] and 0.87 in this study.

Statistical analyses

In the step 1 of this study, the relationships between the characteristics of the caregiving context were investigated among the 125 participants using partial least squares (PLS) and a principal components analysis (PCA) based on a non-linear iterative (NIPALS) algorithm [64, 65] as well as a PLS-path model (PM), which represents a PLS approach to structural equation modeling (SEM) [66, 67].

Similar to covariance-based SEM (CBSEM), PLS-PM is a method for studying complex multivariate relationships among observed (MVs) and latent variables (LVs). However, CBSEM is oriented toward testing a strong theory (confirmatory analysis), whereas PLS-PM is more suited for predictive applications and theory building (exploratory analysis) [68, 69].

More precisely, unlike CBSEM, the PLS-PM does not use maximum likelihood estimations of the parameters based on a variance-covariance matrix; rather, it is a full-information procedure that iteratively performs a series of linear regressions. The PLS-PM uses ordinary least squares, enabling models to be tested with fewer distributional assumptions and smaller samples than SEM. Moreover, the PLS-PM is less sensitive to normality problems and more suitable for avoiding indeterminacy problems [70]. PLS-PM loadings (w) of manifest variables (MVs) on each latent variables (LVs) are similar to PCA regression loadings, and its path coefficients (β) are similar to the standardized beta coefficients in a classic regression analysis. The assessment of model quality is related to (a) the adjustment of the MVs to their respective LVs by examining the unidimensionality of each LV via Cronbach’s alpha (α), Dillon-Goldstein’s rho (DG-rho) and cross-loadings (i.e., the loadings of an MV with the other LVs); and (b) the proportion of explained variance between the LVs by examining the R² determination coefficient. Moreover, the goodness-of-fit (GoF) index was used as a global criterion to evaluate the general performance of the model. This index is usually interpreted as the predictive power of the model; however, no threshold exists enabling us to determine its significance, and no guidance has been provided regarding what number should be considered an acceptable GoF value[67, 71].

In the step 2, we computed and compared models based on the onset of dementia (early versus late) and the gender of the spousal caregiver (female versus male). We used a resampling approach with a permutation procedure (10,000 replicates) that does not rely on normality or similar sample size assumptions (unlike the bootstrapped t-test).

For analysis, we used the packages plsdepot (version 0.1.17) and plspm (version 0.4.7) [72] available in R (version 3.2.1) [73]. The results are shown as path coefficients with their bootstrapped (10,000 replicate) 95% confidence intervals (CIs).

Regarding sample size, the requirements of PLS-PM are generally much smaller than those of CBSEM [74, 75]. The references on this topic [69] recommend specifying the effect size for each regression analysis in the model (due to the core algorithm of PLS-PM that iteratively performs a series of linear regressions) and looking up Green’s approximation [76] of Cohen’s power tables [77]. For a model with a maximum of 4 manifest variables related to a latent variable and a maximum of 4 structural paths (relationships between latent variables), we found that we would initially need a minimum sample of 84 cases with the aim of obtaining a power of 0.80 and an alpha level of 0.05.

RESULTS

Step 1: Moderated-mediator model of spousal caregiver distress

Outer model

The outer model describes the relationships between each LV and its respective MV. Each LV was created to represent a unique construct emerging from several measures. Initially, we used PLS-PCA outputs to visually determine similar groups of measurements. When a measurement was dismissed (from a group of variables), we attempted to integrate it into another group of variables until complete withdrawal. The second stage checked the quality of the outer model: all MVs that did not contribute to the unidimensionality criterion of the LV (α <0.64; DG-rho <0.84) or showed cross-loadings (an MV that was more correlated with an independent LV) were discarded. The final outer model included 16 MVs loaded onto 6 LVs that were labeled by 3 experts in the field. These variables are described in Table 2along with the unidimensionality and loading information.

Table 2 shows that all of the loadings of the outer model (see also Fig. 2) were acceptable (w > 0.73).

Inner model

The inner model (Fig. 2) describes the relationships between the LVs and was constructed to explain the spousal caregiver distress LV. The remaining five LVs were combined to maximize the GoF index. The final moderated-mediator model had a GoF index of 0.51 and explained 56.5% of the variance of the spousal caregiver distress LV. The path coefficients are reproduced in Fig. 2, and Table 3 displays the original and bootstrapped means, standard deviations and 95% CIs.

This step led to a 4-branch structure. One branch contained an LV mediator. In fact, the moderated-mediator model suggests that the LVs poor self-rated health, family divergence, preparedness and confidence affected the spousal caregiver distress LV. Moreover, the effect of couple adjustment on spousal distress was mediated by the care recipient impairments LV.

Step 2: Moderated-mediator model comparisons

Early- versus late-onset dementia

The path coefficient comparison of the relationship between the couple adjustment and care recipient impairment LVs reached significance (p = 0.03), with a stronger effect for the early-onset dementia model than the late-onset dementia model (–0.57 versus –0.22, respectively).

Wife versus husband caregivers

The path coefficient comparison of the relationship between poor self-rated health and spousal caregiver distress was significant (p = 0.02), with a stronger effect for the female model than the male model (0.47 versus 0.08, respectively).

DISCUSSION

This study proposed a moderated-mediator model of spousal caregiver distress by studying the variables thought to be either independently or linearly related specifically to spousal caregivers. The variables that influence spousal caregiver distress are of major interest to orient and improve care because spousal caregivers are a vulnerable population [33].

Even if van den Wijngaart et al. [33] had used PWD and caregiver characteristics in their analysis, dyadic determinant did not assess in their model. However, some studies [4, 21] showed that these variables predict the psychological distress of the caregiver. The moderated-mediator model developed in this study provides a comprehensive view of spousal caregiver distress by accounting for the limitations of previous studies and examining the role of many categories of variables associated with PWD and caregiver characteristics as well as dyadic determinants.

Thus, the current study highlights three variables that have direct relationships with spousal caregiver distress. These results were consistent with the findings of van der Lee et al. [30], who reported that social functioning, social support, caregiver physical health, and self-efficacy were significant predictors of caregiver distress. One of the salient findings of the study was a mediated relationship. Below, we present the three variables that directly influence spousal caregiver distress: poor self-rated health, family divergence, and preparedness and confidence. Then, we discuss the relationship of the quality of couple adjustment with spousal caregiver distress, which is mediated by the caregivers’ perceptions of the PWD’s symptoms.

The first variable with a positive direct relationship effect is Poor self-rated health. Studies have shown that spousal caregivers are more depressed, express higher levels of negative affect, have more symptoms of psychological distress, and report higher health problems than non-spousal caregivers [8, 78]. The current study suggests a relationship not only between poor self-rated health and depressive symptoms but also between poor self-rated health and other variables that contribute to spousal caregiver distress such as the effect on health, the organization of activities and psychological distress. Furthermore, the literature shows that this distress is greater for female caregivers [29 , 36–38] because of their difficulties with delegation and their sense of personal responsibility [36]. This study highlights an indicator (Poorer self-rated health) to identify female caregivers who may present with greater distress. This indicator can be a major asset to the support.

The second variable with a direct positive relationship was Family divergence. The moderated-mediator model developed in this study suggested that the differences in the decisions related to care and a lack of family support in everyday life especially contributed to spousal caregiver distress. The children of the caregivers of PLOD or the siblings of PEOD can also express their disagreements about decision making by the caregiver concerning PWD. Therefore, to avoid conflicts or questions about one’s caregiving behaviors, spousal caregivers prefer not to ask for help at the risk of spending more time to care for the PWD and increase their levels of depression and burden [79]. The literature shows that the quality of family relationships [7] is an important predictor of the burden that caregiving spouses feel. Deimling et al. [80] and Heru and Ryan [7] identified the presence of family conflicts as a part of the depression with which caregivers are confronted.

The third variable with a direct negative relationship is Preparedness and confidence. One interesting aspect of this study was the relationship between these variables: The more a spousal caregiver feels competent, confident, and well prepared for his or her new role, the less distress he or she experiences. Spousal caregivers provide a level of care at home that is often equivalent to that of professional nurses; however, they often feel unprepared and lack the knowledge and skills to provide care [81]. Campbell et al. [29] highlighted the importance of the variables related to decreases in self-concept (including senses of competence and efficacy), and found that caregivers who report a high level of preparedness with regard to caregiving experience lower levels of caregiver strain [10, 11]. The current study confirmed that major interest exists in guiding spousal caregivers to better prepare for their new roles and that greater confidence is needed to prepare for the multiple domains of caregiving, such as providing physical care and emotional support, establishing in-home support services, and coping with the stress of the caregiving situation.

Concerning the mediated relationship, our model found that the Care recipient’s impairments mediated the relationship between Couple adjustment and Spousal caregiver distress. Previous studies have shown the importance of the quality of the couple relationship with regard to the subjective assessment of symptoms and sense of burden [23 , 82]. The advantage of the current moderated-mediator model is that it suggests how these three variables fit together: Higher quality marital relationships are associated with the less severe perception of PWD’s symptoms and less caregiver distress. In addition, the existing relationship between the quality of adjustment of the couple and the caregiver’s perception of the severity of the PWD’s symptoms varies with age at onset of the disease and was more important for caregivers of PEOD than those of PLOD. This result can be explained by the fact that the DAS-16 focuses on the frequency and amount of positive interactions that become impaired or non-existent as the disease progresses, sometimes plunging the home into silence. However, studies of couples that include a PEOD [21 , 84] have shown that communication is an important dimension in the experience of the caregiving situation. Thus, the current moderated-mediator model suggests that the preservation of communication in couples with a PEOD is important, and this result is a major asset regarding the orientation of the support for caregivers of PEOD.

Importantly, numerous variables were not selected in the analyses used to create the model. In fact, variables such as Respite from family and friends, Response to a relative’s disruptive behaviors, Control disturbing thoughts, Anxiety, Financial problems, Planning for future care needs, Knowledge of formal services, Emotional support, Instrumental support, and Informational support did not show strong relationships with other LVs in the model. These exclusions raise the possibility of a gap between the experiences of caregivers and the knowledge of medical and psychosocial professionals as well as highlight the need for an informal support network and respite solutions, future planning, and information concerning formal support to alleviate the distress related to the caregiving situation.

Limitations

The aforementioned variables that did not show strong relationships with other LVs did not contribute to our moderated-mediator model of spousal caregiver distress; it would be interesting to assess their relationships with more positive variables such as quality of life or the sense of well-being related to the caregiving situation. In fact, recent studies have highlighted the gratifications related to the caregiving situation, the positive aspects of the role of the caregiver [32 , 85–87], and the variables that affect individuals’ responsiveness to interventions [32]. Cohen et al. [87] observed that the ability of caregivers to identify positive aspects within a relationship is a protective factor against the negative effects of caregiving. It would be interesting to assess quality of life using questionnaires [88] or scales derived from positive psychology [89 –91] because these measures would allow us to consider the positive and negative experiences of caregivers.

Finally, this study created a moderated-mediator model of spousal caregiver distress that contributes to better comprehension of caregiver distress, confirming age at onset of the disease and gender as moderators, and highlighting PWD’s symptoms as a mediator. Because the sample size of this study is modest compared with the number of variables, one recommendation would be to increase the sample size to increase the scope and generalizability of the results. This increase may also facilitate the use of covariance-based structural equation modeling (CBSEM) as a confirmatory analysis that provides several measures of fit that are widely acknowledged by the scientific community.

CONCLUSIONS

This study has examined a moderated-mediator model to evaluate variables that contribute to spousal caregiver distress. This model was used for comparisons based on the age at disease onset and caregiver gender. The study demonstrated that preparedness, family support, self-rated health, and the quality of the couple relationship contribute to spousal caregiver distress. Moreover, dyadic determinants were more important for caregivers of PEOD.

These findings offer implications for clinical practice. Spousal caregivers should be provided with adequate supportive care, training, and anticipatory guidance regarding the trajectory of dementia [92]. Many effective interventions are already available for caregivers that address many of these findings. For example, Ducharme et al. [50] tested the efficacy of a psychoeducational individual program. This seven-session program allowed caregivers to feel better prepared to deal with situations related to caregiving and to acquire skills to help caregivers adapt to their new role. Whitlach et al. [93] provided a protocol, the Early Diagnosis Dyadic Intervention (EDDI) program. Each member of the dyad expresses his or her preferences and concerns about the care situation. The objective of the nine-session program is to work together to develop a mutual plan for coping with the disease in the long term. Concerning family conflicts, studies encourage the use of family-centered interventions in dementia [94].

This study is of major interest because it accounts for specific elements regarding the provision of support among spousal caregivers. It would be interesting to create a support program that would incorporate these three areas of intervention regarding the progression of the disease. The first modules, available as early as possible after diagnosis, would be oriented toward preparation for their new role and the couple relationship (variables that have the greatest weight on distress in our model). The “preparedness modules” would facilitate discussion of caregivers’ perceptions of the disease and the caregiving situation; the development of communication skills with the PWD; the development of coping strategies for dealing with difficulties; the identification of formal services; and plans for the future. The “dyadic modules” would increase understanding of the care preferences and expectations of each dyad member and would facilitate exploration of the emotional significance and relationship issues caused by the disease for both partners. These dyadic modules would be particularly important for caregivers of PEOD. Reflection on family relationships that includes family dynamics and interpersonal conflict should be considered in a third time (variable that has the least weight on distress in our model), especially because the literature showed that family conflicts occurred most often in mild to moderate dementia [94]. These “family modules” would permit us to assess the potential support of the family and the limitations of their interventions and would provide information about the disease for a better understanding of each member of the family. Specific attention should be given to female caregivers who report poor self-rated health.

Footnotes

ACKNOWLEDGMENTS

The authors wish to thank the two anonymous reviewers for their contributions in improving this article.

The authors disclose receipt of the following financial support for this research: the Fond de Recherche du Québec-Santé (Québec), the Canadian Institutes of Health Research (Canada), the Agence Nationale de Recherche (France) under Grant ANR-10-MALZ-0102-01; This study was supported by the Labex (laboratory of excellence) Distalz (Development of Innovative Strategies for a Transdisciplinary approach to ALZheimer’s disease), the MESHS (Maison Européenne des Sciences de l’Homme et de la Société, Lille, France); and the MEL (Métropole Européenne de Lille).

Authors’ disclosures available online ().

References

Stommel

, Collins

, Given

(1994) The costs of family contributions to the care of persons with dementia. Gerontologist 34, 199–205.

Conde-Sala

, Garre-Olmo

, Turró-Garriga

, Vilalta-Franch

, López-Pousa

(2010) Differential features of burden between spouse and adult-child caregivers of patients with Alzheimer’s disease: An exploratory comparative design. Int J Nurs Stud 47, 1262–1273.

Bandeira

, Pawlowski

, Gonçalves

, Hilgert

, Bozzetti

, Hugo

(2007) Psychological distress in Brazilian caregivers of relatives with dementia. Aging Ment Health 11, 14–19.

Black

, Almeida

(2004) A systematic review of the association between the behavioral and psychological symptoms of dementia and burden of care. Int Psychogeriatr 16, 295–315.

Farran

, Keane-Hagerty

, Salloway

, Kupferer

, Wilken

(1991) Finding meaning: An alternative paradigm for Alzheimer’s disease family caregivers. Gerontologist 31, 483–489.

Brodaty

, Donkin

(2009) Family caregivers of people with dementia. Dialogues Clin Neurosci 11, 217–228.

Heru

, Ryan

, Iqbal

(2004) Family functioning in the caregivers of patients with dementia. Int J Geriatr Psychiatry 19, 533–537.

Harwood

, Barker

, Ownby

, Duara

(2000) Caregiver self-rated health in Alzheimer’s disease. Clin Gerontol 21, 19–33.

Gilliam

, Steffen

(2006) The relationship between caregiving self-efficacy and depressive symptoms in dementia family caregivers. Aging Ment Health 10, 79–86.

10.

Archbold

, Stewart

, Greenlick

, Harvath

(1990) Mutuality and preparedness as predictors of caregiver role strain. Res Nurs Health 13, 375–384.

11.

Wang

Y-N

, Shyu

, Tsai

W-C

, Yang

P-S

, Yao

(2013) Exploring conflict between caregiving and work for caregivers of elders with dementia: A cross-sectional, correlational study. J Adv Nurs 69, 1051–1062.

12.

Shyu

, Yang

C-T

, Huang

C-C

, Kuo

H-C

, Chen

S-T

, Hsu

W-C

(2010) Influences of mutuality, preparedness, and balance on caregivers of patients with dementia. J Nurs Res 18, 155–163.

13.

Covinsky

, Newcomer

, Fox

, Wood

, Sands

, Dane

, Yaffe

(2003) Patient and caregiver characteristics associated with depression in caregivers of patients with dementia. J Gen Intern Med 18, 1006–1014.

14.

Etters

, Goodall

, Harrison

(2008) Caregiver burden among dementia patient caregivers: A review of the literature. J Am Acad Nurse Pract 20, 423–428.

15.

Pinquart

, Sörensen

(2003) Differences between caregivers and noncaregivers in psychological health and physical health: A meta-analysis. Psychol Aging 18, 250–267.

16.

Mohamed

, Rosenheck

, Lyketsos

, Schneider

(2010) Caregiver burden in Alzheimer disease: Cross-sectional and longitudinal patient correlates. Am J Geriatr Psychiatry 18, 917–927.

17.

Pinquart

, Sörensen

(2011) Spouses, adult children, and children-in-law as caregivers of older adults: A meta-analytic comparison. Psychol Aging 26, 1–14.

18.

Robinson

, Clare

, Evans

(2005) Making sense of dementia and adjusting to loss: Psychological reactions to a diagnosis of dementia in couples. Aging Ment Health 9, 337–347.

19.

Perren

, Schmid

, Wettstein

(2006) Caregivers’ adaptation to change: The impact of increasing impairment of persons suffering from dementia on their caregivers’ subjective well-being. Aging Ment Health 10, 539–548.

20.

Eloniemi-Sulkava

, Notkola

I-L

, Hämäläinen

, Rahkonen

, Viramo

, Hentinen

, Kivelä

S-L

, Sulkava

(2002) Spouse caregivers’ perceptions of influence of dementia on marriage. Int Psychogeriatr 14, 47–58.

21.

Wawrziczny

, Antoine

, Ducharme

, Kergoat

M-J

, Pasquier

(2016) Couples’ experiences with early-onset dementia: An interpretative phenomenological analysis of dyadic dynamics. Dementia 15, 1082–1099.

22.

Davies

, Newkirk

, Pitts

, Coughlin

, Sridhar

, Zeiss

(2010) The impact of dementia and mild memory impairment (MMI) on intimacy and sexuality in spousal relationships. Int Psychogeriatr 22, 618–628.

23.

Mahoney

, Regan

, Katona

, Livingston

(2005) Anxiety and depression in family caregivers of people with Alzheimer disease: The LASER-AD study. Am J Geriatr Psychiatry 13, 795–801.

24.

Gold

, Cohen

, Shulman

, Zucchero

, Andres

, Etezadi

(1995) Caregiving and dementia: Predicting negative and positive outcomes for caregivers. Int J Aging Hum Dev 41, 183–201.

25.

Horowitz

, Shindelman

(1983) Reciprocity and affection. J Gerontol Soc Work 5, 5–20.

26.

Morris

, Morris

, Britton

(1988) The relationship between marital intimacy, perceived strain and depression in spouse caregivers of dementia sufferers. Br J Med Psychol 61, 231–236.

27.

Steadman

, Tremont

, Davis

(2007) Premorbid relationship satisfaction and caregiver burden in dementia caregivers. J Geriatr Psychiatry Neurol 20, 115–119.

28.

Harris

, Adams

, Zubatsky

, White

(2011) A caregiver perspective of how Alzheimer’s disease and related disorders affect couple intimacy. Aging Ment Health 15, 950–960.

29.

Campbell

, Wright

, Oyebode

, Job

, Crome

, Bentham

, Jones

, Lendon

(2008) Determinants of burden in those who care for someone with dementia. Int J Geriatr Psychiatry 23, 1078–1085.

30.

van der Lee

, Bakker

, Duivenvoorden

, Dröes

R-M

(2014) Multivariate models of subjective caregiver burden in dementia: A systematic review. Ageing Res Rev 15, 76–93.

31.

Kim

, Chang

, Rose

, Kim

(2012) Predictors of caregiver burden in caregivers of individuals with dementia. J Adv Nurs 68, 846–855.

32.

Hilgeman

, Allen

, DeCoster

, Burgio

(2007) Positive aspects of caregiving as a moderator of treatment outcome over 12 months. Psychol Aging 22, 361–371.

33.

Van Den Wijngaart

, Vernooij-Dassen

, Felling

(2007) The influence of stressors, appraisal and personal conditions on the burden of spousal caregivers of persons with dementia. Aging Ment Health 11, 626–636.

34.

Ducharme

, Lachance

, Kergoat

M-J

, Coulombe

, Antoine

, Pasquier

(2016) A comparative descriptive study of characteristics of early- and late-onset dementia family caregivers. Am J Alzheimers Dis Other Demen 31, 48–56.

35.

van Vliet

, de Vugt

, Bakker

, Koopmans

, Verhey

(2010) Impact of early onset dementia on caregivers: A review. Int J Geriatr Psychiatry 25, 1091–1100.

36.

Brown

, Chen

, Mitchell

, Province

(2007) Help-seeking by older husbands caring for wives with dementia. J Adv Nurs 59, 352–360.

37.

Kaiser

, Panegyres

(2006) The psychosocial impact of young onset dementia on spouses. Am J Alzheimers Dis Other Demen 21, 398–402.

38.

Takano

, Arai

(2005) Gender difference and caregivers’ burden in early-onset Alzheimer’s disease. Psychogeriatrics 5, 73–77.

39.

Antoine

, Pasquier

(2013) Emotional and psychological implications of early AD diagnosis. Med Clin North Am 97, 459–475.

40.

Zarit

, Femia

(2008) A future for family care and dementia intervention research? Challenges and strategies. Aging Ment Health 12, 5–13.

41.

Nolan

, Grant

, Keady

(1996) Understanding Family Care: A Multidimensional Model of Caring and Coping. Open University Press, London.

42.

Pinquart

, Sörensen

(2006) Gender differences in caregiver stressors, social resources, and health: An updated meta-analysis. J Gerontol B Psychol Sci Soc Sci 61, P33–P45.

43.

Gendron

, Levesque

(1993) Evaluating the functional autonomy of persons with Alzheimer’s disease: A tool for observing four activities of daily living. Am J Alzheimers Dis Other Demen 8, 24–35.

44.

Lawton

, Brody

(1969) Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist 9, 179–186.

45.

Pearlin

, Mullan

, Semple

, Skaff

(1990) Caregiving and the stress process: An overview of concepts and their measures. Gerontologist 30, 583–594.

46.

Folstein

, Folstein

, McHugh

(1975) “Mini-mental state”. A practical method for grading the cognitive state of patients for the clinician. J Psychiatr Res 12, 189–198.

47.

Hébert

, Bravo

, Girouard

(1992) Validation de l’adaptation Française du modified mini-mental state (3MS). Rev Geriatr 17, 443–450.

48.

Given

, Given

, Stommel

, Collins

, King

, Franklin

(1992) The Caregiver Reaction Assessment (CRA) for caregivers to persons with chronic physical and mental impairments. Res Nurs Health 15, 271–283.

49.

Antoine

, Quandalle

, Christophe

(2010) Vivre avec un proche malade: Évaluation des dimensions positive et négative de l’expérience des aidants naturels. Psychol Rev Psychiatr 168, 273–282.

50.

Ducharme

, Lévesque

, Lachance

, Kergoat

M-J

, Legault

, Beaudet

, Zarit

(2011) “Learning to become a family caregiver” efficacy of an intervention program for caregivers following diagnosis of dementia in a relative. Gerontologist 51, 484–494.

51.

Kuhn

, Fulton

(2004) Efficacy of an educational program for relatives of persons in the early stages of Alzheimer’s disease. J Gerontol Soc Work 42, 109–130.

52.

Sörensen

, Pinquart

(2001) Developing a measure of older adults’ preparation for future care needs. Int J Aging Hum Dev 53, 137–165.

53.

Steffen

, McKibbin

, Zeiss

, Gallagher-Thompson

, Bandura

(2002) The revised scale for caregiving self-efficacy: Reliability and validity studies. J Gerontol B Psychol Sci Soc Sci 57, P74–P86.

54.

Ware

, Sherbourne

(1992) The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care 30, 473–483.

55.

Krause

, Markides

(1990) Measuring social support among older adults. Int J Aging Hum Dev 30, 37–53.

56.

Lauzon

, Voyer

(1998) Traduction en Français et validation de la fiabilité de l’ISSB-M. Rapport de recherche inédit. Faculté des sciences infirmières, Université de Montréal.

57.

Clark

, Shields

, Aycock

, Wolf

(2003) Preliminary reliability and validity of a family caregiver conflict scale for stroke. Prog Cardiovasc Nurs 18, 77–82,92.

58.

Spanier

(1976) Measuring dyadic adjustment: New scales for assessing the quality of marriage and similar dyads. J Marriage Fam 38, 15–28.

59.

Antoine

, Christophe

, Nandrino

J-L

(2008) [Dyadic Adjustment Scale: Clinical interest of a revision and validation of an abbreviated form].. Encéphale 34, 38–46.

60.

Zigmond

, Snaith

(1983) The hospital anxiety and depression scale. Acta Psychiatr Scand 67, 361–370.

61.

Bocéréan

, Dupret

(2014) A validation study of the Hospital Anxiety and Depression Scale (HADS) in a large sample of French employees. BMC Psychiatry 14, 354.

62.

Ilfeld

(1976) Further validation of a psychiatric symptom index in a normal population. Psychol Rep 39, 1215–1228.

63.

Préville

, Boyer

, Potvin

, Perreault

, Légaré

(1992) La détresse psychologique: Détermination de la fiabilité et de la validité de la mesure utilisée dans l’enquête Santé Québec. Québec PQ: Rapport de recherche soumis á Santé Québec.

64.

Tenenhaus

(1998) La régression PLS: Théorie et Pratique. Editions Technip, Paris.

65.

Tenenhaus

(2007) Statistique. Méthodes pour décrire, expliquer et Prévoir. Dunod, Paris.

66.

Vinzi

, Trinchera

, Amato

(2010) PLS path modeling: From foundations to recent developments and open issues for model assessment and improvement. In Handbook of Partial Least Squares, Vinzi

, Chin

, Henseler

, Wang

, eds. Springer Heidelberg, Berlin, pp. 47–82.

67.

Sanchez

(2013) PLS Path Modeling with R. Trowchez Editions, Berkeley.

68.

Gefen

, Straub

, Boudreau

M-C

(2000) Structural equation modeling and regression: Guidelines for research practice. Commun Assoc Inf Syst 7, 1–78.

69.

Roldán

, Sánchez-Franco

(2012) Variance-based structural equation modeling: Guidelines for using partial least squares in information systems research. In Research Methodologies, Innovations and Philosophies in Software Systems Engineering and Information Systems, Mora

, Gelman

, Steenkamp

, Raisinghani

, eds. IGI Global, pp 193–221.

70.

Fornell

, Bookstein

(1982) Two structural equation models: LISREL and PLS applied to consumer exit-voice theory. J Mark Res 19, 440.

71.

Henseler

, Sarstedt

(2013) Goodness-of-fit indices for partial least squares path modeling. Comput Stat 28, 565–580.

72.

Sanchez

, Trinchera

, Russolillo

(2013) PLSPM: Tools for partial least squares Path modeling (PLS-PM). R package version 0.4 1.

73.

R Core Team (2013) R: A Language and Environment for Statistical Computing, R Foundation for Statistical Computing, Austria, Vienna.

74.

Chin

, Newsted

(1999) Structural equation modeling analysis with small samples using partial least square. In Statistical Strategies for Small Samples Research, Hoyle

, ed. age Publishing House, Thousand Oaks, CA, pp. 307–341.

75.

Reinartz

, Haenlein

, Henseler

(2009) An empirical comparison of the efficacy of covariance-based and variance-based SEM. Int J Res Mark 26, 332–344.

76.

Green

(1991) How many subjects does it take to do a regression analysis. Multivar Behav Res 26, 499–510.

77.

Cohen

(1988) Statistical Power Analysis for the Behavioral Sciences. Routledge, –Hillsdale, NJ.

78.

Pruchno

, Potashnik

(1989) Caregiving spouses. Physical and mental health in perspective. J Am Geriatr Soc 37, 697–705.

79.

Cooper

, Katona

, Orrell

, Livingston

(2008) Coping strategies, anxiety and depression in caregivers of people with Alzheimer’s disease. Int J Geriatr Psychiatry 23, 929–936.

80.

Deimling

, Smerglia

, Schaefer

(2001) The impact of family environment and decision-making satisfaction on caregiver depression: A path analytic model. J Aging Health 13, 47–71.

81.

Schumacher

, Beck

, Marren

(2006) Family caregivers: Caring for older adults, working with their families. Am J Nurs 106, 40–49.

82.

Kramer

(1993) Marital history and the prior relationship as predictors of positive and negative outcomes among wife caregivers. Fam Relat 42, 367–375.

83.

Kimura

, Maffioletti

, Santos

, Baptista

, Dourado

(2015) Psychosocial impact of early onset dementia among caregivers. Trends Psychiatry Psychother 37, 213–219.

84.

O’Shaughnessy

, Lee

, Lintern

(2010) Changes in the couple relationship in dementia care: Spouse carers’ experiences. Dementia 9, 237–258.

85.

Roff

, Burgio

, Gitlin

, Nichols

, Chaplin

, Hardin

(2004) Positive aspects of Alzheimer’s caregiving: The role of race. J Gerontol B Psychol Sci Soc Sci 59, P185–P190.

86.

Boerner

, Schulz

, Horowitz

(2004) Positive aspects of caregiving and adaptation to bereavement. Psychol Aging 19, 668–675.

87.

Cohen

, Colantonio

, Vernich

(2002) Positive aspects of caregiving: Rounding out the caregiver experience. Int J Geriatr Psychiatry 17, 184–188.

88.

Baumann

, Erpelding

M-L

, Régat

, Collin

J-F

, Briançon

(2010) The WHOQOL-BREF questionnaire: French adult population norms for the physical health, psychological health and social relationship dimensions. Dépidémiologie Santé Publique 58, 33–39.

89.

Diener

, Emmons

, Larsen

, Griffin

(1985) The satisfaction with life scale. J Pers Assess 49, 71–75.

90.

Scheier

, Carver

, Bridges

(1994) Distinguishing optimism from neuroticism (and trait anxiety, self-mastery, and self-esteem): A reevaluation of the life orientation test. J Pers Soc Psychol 67, 1063–1078.

91.

Peterson

, Park

, Seligman

MEP

(2005) Orientations to happiness and life satisfaction: The full life versus the empty life. J Happiness Stud 6, 25–41.

92.

Rose

, Lopez

(2012) Transitions in dementia care: Theoretical support for nursing roles. Online J Issues Nurs 17, 4.

93.

Whitlatch

, Judge

, Zarit

, Femia

(2006) Dyadic intervention for family caregivers and care receivers in early-stage dementia. Gerontologist 46, 688–694.

94.

Peisah

, Brodaty

, Quadrio

(2006) Family conflict in dementia: Prodigal sons and black sheep. Int J Geriatr Psychiatry 21, 485–492.