Pet Ownership as a Protective Factor Supporting the Emotional Well-Being of Cancer Patients and Their Family Members. Victoria H. Raveis, Ph.D., Principal Investigator, Frances Mesagno, Ph.D., Co-Principal Investigator, Dan Karus, M.S. Programmer-Analyst, Eileen Gorey, M.P.H., Research Associate. Final Report Year 2, August 15, 1993, Memorial Sloan-Kettering Cancer Center, Department of Social Work, Research Unit, 1275 York Avenue, New York, NY 10021.
Abstract
The goal of this two-year research project is to examine the efficacy of the presence of pets as a protective factor supporting the psychosocial well being of cancer patients and their family members, through both secondary analyses of previously collected data and analyses of data on human-animal interaction to be newly collected over the two-year study period. Specifically, the research aims are:
To examine the association of the presence of pets-dogs, cats and other household animals-with psychosocial well-being (as measured, e.g., by depression, anxiety, self-esteem and, for children, peer-relationships, social competence and school performance) of cancer patients and their family members.
To examine the efficacy of the presence of pets in mediating the impact of stressful life situations (i.e., serious illness and/or death) on the psychological well-being of cancer patients and their family members.
To assess the importance of various aspects (i.e., extent and quality) of human-animal interactions in contributing to the psychosocial well-being of cancer patients and their family members.
To compare the effect of human-animal interactions on the psychosocial well-being of patients and their family members at different stages of the illness course (i.e., diagnosis, active treatment, terminal period); in different types of households (i.e., intact families with young children, families with grown children, single parent households, and adults living alone); and in varying age-groups (i.e., latency-aged children, adolescents, young adults, middle-aged and older adults).
We will accomplish these research aims through conducting secondary analyses of data from two studies we have recently completed which focused on seriously-ill or terminally-ill cancer patients and their families as well as collecting additional information on human-animal interactions as part of a third cancer study which is still ongoing. The first recently completed study was a multi-site, longitudinal investigation of adult cancer patients' needs for illness-related assistance and the impact on the family of providing this assistance. The second recently completed study examined the psychosocial adjustment of children who experienced recent parental death from cancer. The third study, which is still ongoing, is a longitudinal investigation of children experiencing the terminal illness and subsequent death of a parent to cancer. We have now added an extensive pet history questionnaire to the well parent and child interview schedules to enable us to examine the role of companion pets in human adjustment to loss in greater detail.
Secondary analyses conducted during Year 1 focused on addressing Aims 1 and 2. These findings suggest that having a pet in the household may well be associated with better psychological adjustment for caregivers whose spouse is seriously-ill. While the current findings pertain only to spousal caregivers, they suggest that the presence of companion animals, specifically, cats and dogs, may have an beneficial impact in situations of stress and burden, such as experienced when caring for a spouse who is seriously-ill.
Background
Theoretical Framework
The Research Unit of the Department of Social Work at Memorial Sloan-Kettering Cancer Center (MSKCC) has conducted a number of studies designed with the objective of learning more about problems and issues which confront seriously ill cancer patients and their family members or caregivers. In our work, we have become increasingly aware of the broad range of resources that people use to cope with stress in their lives and the importance of less commonly recognized sources of emotional support, such as companion animals, particularly dogs and cats.
In clinical work with patients and their families we have observed that the companionship of pets can facilitate a child's healthy adaptation to a parent's serious illness and death, by providing unconditional love and continuity at a time when children's needs for love, affection and support are enormous. Furthermore, children need to talk about their painful and conflicted feelings regarding parental loss, and often they are able to confide these feelings more readily to their pets than to humans. Taking care of a pet can also enhance children's adjustment by providing environmental consistency and maintaining the structure of their daily routines, so important in the face of the vast disruption that occurs during the parent's terminal illness and in the post-death period (Siegel, Mesagno & Christ, 1990). Children's feelings of competence and control are also enhanced by their fulfilling a necessary family role as the pet's caretaker. Their self-esteem can be enhanced by experiencing a feeling of being needed by the pet. In our interviews, children frequently list their pets as family members and both children and their surviving parents report that their companionship has been enormously comforting.
In our work with other patient populations, we have also observed that household pets are frequently spontaneously mentioned by AIDS patients as important in helping them cope when they feel distressed. These patients have remarked on the unconditional love and affection they receive from their pets and the consistency with which this affection is expressed. Some have acquired a pet since becoming ill because they find that they have more time to themselves. For them, the pet is a source of companionship and a means of overcoming the loneliness they are experiencing. These patients have also remarked that having a pet provides them with a reason to remain active and life-focused (e.g., to take care of the pet, interact with the pet).
Literature Review
The contribution of companion animals to human psychological well-being and interpersonal development has received increasing attention in recent years. Levinson's (1969, 1972, 1978) groundbreaking work on the role of pets in human development heightened behavioral scientists' awareness of the varied beneficial consequences of animal-human interaction. While much of the earlier work in this area relied on anecdotal reports and clinical experiences, recent work represents efforts to more systematically explore the role of companion animals.
The work of Katcher, Beck and colleagues (e.g., Beck & Katcher, 1983; Friedmann, et al., 1983; Katcher & Beck, 1986; Katcher et al., 1983; Katcher, Beck & Levine, 1988) has provided evidence that contact and interaction with animal companions increases relaxation and reduces arousal in both children and adults, especially in stressful situations (Katcher, Segal & Beck, 1984). Others (Baun et al., 1984; Grossberg & Alt, 1984) have noted similar findings. In a one-year follow-up study of coronary heart disease patients, patients with a household pet had a better survival rate than those without pets (Friedmann et al., 1980).
Current investigations have primarily focused on the beneficial health aspects of animal companions to the elderly and, to a lesser extent, on the role of pets in childhood development. Kidd and Feldmann (1981) reported that pet owners scored higher than non-owners on measures of responsibility, dependability, nurturance, helpfulness, benevolence and lack of egotism or self-centeredness. Mugford and M'Comisky (1975) reported longer survival among elderly retirees who were given pet birds. In a study of widowers, Akiyama et al. (1987) observed that pet owners had significantly fewer psychosomatic symptoms than those who did not have a pet. Others have not been able to demonstrate statistically significant effects of pet ownership on health outcomes (e.g., Ory & Goldberg, 1983; Lago, Knight & Connell, 1983; Lawton, Moss & Moles, 1984; and Robb & Stegman, 1983); however, their findings do suggest that the influence of animal companionship is complex and may be influenced by other elements, such as personality characteristics or situational factors.
In a longitudinal investigation of health outcomes among the elderly, Lago et al. (1989) posited that the effect of pets on health was indirect, that is, pets improved morale, which more directly influenced the respondents' self-reported health and functioning levels. Garrity et al. (1989) observed that pet ownership in an elderly sample of community respondents was not related to depression when one or more confidants were available, but among a subgroup of recently bereaved elderly with minimal confidant support, pet ownership was associated with less depression. Additionally, for the sample as a whole, strong attachment to a pet was associated with less depression. Siegel (1990) has observed that the presence of pets served as a moderator of the effect of stressful life events on the health of the elderly.
In other empirical work, children have been found to develop strong affective ties with their pets, especially dogs and cats. In a survey of 300 children between the ages of 3 and 13, which explored differences in children's attitudes by age, sex, and type of pet, Kidd and Kidd (1985) found that children defined their pets as playmates and companions. Children without siblings may be particularly motivated to keep pets for play and care (Siegmund & Biermann, 1988). Children regularly engage in mutual activities with pets, and, although predominantly recreational, even preadolescent children assume significant caretaking tasks (Davis, 1987a). There is some evidence that children's pet ownership is associated with their increased self-esteem and sociability (Covert et al., 1985; Siegmund & Biermann, 1988). Levinson (1972) has further argued that through interaction with pets, children can achieve a balance between independence and dependence with his/her family. Covert et al. (1985), however, focusing on the role of adolescent pet ownership on the family system, did not find pet owners and nonowners to differ significantly on the quality of the parent-child relationship. Finally, childhood experience with pets has been found to influence individuals' likelihood to own pets as adults (Serpell, 1981). Poresky et al. (1988) similarly found adults' retrospective reports of pet bonding during childhood to predict contemporary attitudes toward pets and that this relationship was affected by the age at which earliest bonding had occurred.
In summary, although the existing research provides some support for concluding that companion animals make an important contribution to human well-being throughout the life cycle, the results also indicate that this relationship is complex and requires further investigation. Clearly, additional research needs to be done in which various aspects of pet ownership are examined, such as type of pet, and the nature and strength of the pet-human relationship. Moreover, an area which has not been as systematically examined is the role of companion animals in mediating the impact of stressful situations, such as being seriously ill, having a seriously ill family member, or experiencing the death of a family member.
Study
Methodology
We will conduct secondary analyses of data from two studies we have recently completed, which focused on seriously-ill or terminally-ill cancer patients and their families. In addition, during the two-year project period we will collect additional information on human-animal interactions as part of a third cancer study, which is still ongoing to examine this issue in greater detail. All three research studies are described below.
Research Studies for the Secondary Data Analyses
Study 1: Cancer patients and their familial caregivers
The first recently completed study was a longitudinal investigation of adult cancer outpatients and their familial caregivers, conducted in collaboration with Brown University and Hershey Medical Center and supported by a grant from the National Cancer Institute. Dr. Raveis was the Project Director at the Memorial Hospital site. Data were gathered from patients from three geographic locations: New York, Rhode Island and Pennsylvania. The principal objectives of this study were to:
identify the home care needs of patients initiating outpatient chemotherapy and radiation treatment,
determine how these needs change over time, and
determine the kinds of burdens experienced by cancer patients' informal caregivers.
Enrolled patients were asked to identify the principal person who provided or would provide informal (unpaid) assistance to them when necessary. The person they nominated was then approached to become part of the caregiver sample.
Patients and their caregivers were interviewed twice. The initial patient interview was conducted shortly after the patient began a new or initial treatment regimen of chemotherapy or radiation. Their informal caregiver was also interviewed at the same time. Half of the patient sample and their caregivers were randomly selected to be re-interviewed 3 months after the initial interview; the remainder were re-interviewed after 6 months. Structured questionnaires were used to collect information from the patients and their caregivers. The interviews were generally administered over the telephone. The patient interview schedule contained basic socio-demographic data, information on needs for assistance, source and extent of care received, symptom status, healthcare utilization and psychological distress. The caregiver interview schedule included demographic items, measures of burden in various domains (social, financial, physical, time, and employment), substance abuse, availability of social resources, psychological well-being, physical health, and the presence and type of pet in the household. A total of 629 adult cancer outpatients and 483 informal caregivers participated in the interviews.
Early in Year 1 of the study, as part of preparing the data files for the secondary data analyses, we went back to the original questionnaires and coded additional information concerning pet ownership that was obtained during the interview, but not coded into the data files. Specifically, during the interview with the patient's informal familial caregiver, the caregiver was asked if there was a pet in the household. If they reported having a pet, the interviewer then asked what type of pet they had. If more than one type of pet was in the household, i.e., "cat and dog" or "dog, fish and horse", the information was written down in the interview schedule but only the fact that it was a multiple pet household was coded into the data file. The exact composition of the pets in the household was not coded into the computer data files. We felt that it was important to the analyses we planned to carry out to determine not only if pet ownership, per se, was important for psychological well-being, but whether specific types of pets were more strongly associated with reduced distress than other pets, we decided to go back to the original questionnaires and code this missing information. Since the data for this study was obtained from a multi-site project, we contacted the investigators at the other research sites and arranged for them to provide us with the detailed data concerning the exact composition of the pets in a household as well. We then merged this extra information with the data in our existing data files
Study 2: The psychosocial adjustment of bereaved children
The second recently completed study examined the psychosocial adjustment of children who experienced recent parental death from cancer. Dr. Mesagno was the Co-Principal Investigator and Project Director of this study, supported by a grant from the American Cancer Society. It was undertaken to describe the nature, intensity, and determinants of psychological symptomatology in normal bereaved children, aged 6-16 years, during the 16 months following a parent's death and to identify determinants of healthy adaptation to the loss. The study's quasi-longitudinal design, assessing children at various points post-parental death, allows us to document the natural history of childhood bereavement including both the acute and chronic phases of mourning. Data were collected from the children themselves and their surviving parents and teachers, using self-administered instruments and in-depth, face-to-face interviews by experienced clinicians. We assessed children's psychosocial adaptation in a number of domains including their psychiatric symptomatology, competencies, peer relationships and school performance. The study also examined factors, which have been associated with children's psychosocial adjustment, including background characteristics (e.g., child's age, sex, and birth order, parent's sex, socioeconomic status); illness-related factors (e.g., length of parent's illness, number and length of hospitalizations); and characteristics of the family environment (e.g., quality of the parent-child relationship, the surviving parent's depressive symptomatology). The data include the parent's reports of the acquisition or death of a pet since the patient's death and children's reports of what kinds and for how long they have had their current pets. Post-death data are available on 65 families, consisting of an intensive interview with the surviving parent and a randomly-selected index child per family.
Research Study in Which Additional Data Will Be Collected
Study 3: Facilitating bereaved children's adaptation to the terminal illness and death of a parent
The ongoing study, in which we propose to examine the role of companion pets in human adjustment to loss, is a true longitudinal study of children's adjustment to parental serious illness and death to cancer, and a comprehensive evaluation of the relative efficacy of two preventive interventions designed to facilitate children's healthy adaptation to their loss. Dr. Mesagno is the Co-Principal Investigator and Project Director for this 5-year study supported by a grant from the National Institute of Mental Health. Efficacy of the intervention is assessed in terms of children's ability to maintain competence as demonstrated by their levels of school performance, social competence, self-esteem, psychiatric symptomatology, and behavior problems. Study participation is offered to intact, English-speaking families in which there is at least one child between 7 and 16 years of age and a parent with advanced cancer and an expected survival time of 4-6 months. This study employs a true experimental design with a target sample of 110 families assigned to Treatment Group 1 who are offered participation in an intensive clinical intervention program, and 110 families assigned to Treatment Group 2 who receive enhanced social work services. Baseline (T1) data are collected during the parent's terminal illness, at the time of accrual into the study, prior to the initiation of any intervention. Group differences in children's functioning are also assessed at 7 and 14 months (T2 and T3) following the death of the parent.
Face-to-face research interviews are conducted with the well parents and all eligible children at T1, T2, and T3. The data include self-report, parental-report, and teacher-report measures. Comprehensive assessment of children's psychosocial adaptation is made with standardized measures of their depression, anxiety, behavior problems, competencies, peer relationships, and school performance. In addition, several parent and child report measures have been developed to evaluate family environment and quality of the parent-child relationship, particularly the extent of open parent-child communication, the child's perception of the adequacy of his/her support, and the degree of consistency in his/her environment. We also assess several risk factors expected to be associated with a child's adjustment to parental death (e.g., his/her awareness of the severity of illness, experience of other stressful life events, and number and length of separations from the family). Finally, we assess the psychological adjustment of the well/surviving spouses who must now raise the children as single parents.
During Year 1 we developed an expanded set of pet history questions to be administered to the families participating in the study. In designing the pet history questionnaire we utilized, whenever possible, pre-existing scales and inventories. Specifically, the pet history questionnaire collects information on the history of pet ownership, change in pet ownership over time, strength of the human-animal bond, mutual pet-human activities, and attitudes toward pets.
In developing this questionnaire we incorporated the pet history section from the Pet Attitude Inventory [PAI] (Wilson, Netting & New, 1987) and the pet attachment scale and pet interaction checklist utilized by Albert and Bulcroft (1987). We also included several additional items drawn from measures developed by Poresky et al. (1987), Davis (1987) and Templer et al. (1981), revising the wording in these additional items to make them compatible with the format and metric of the other questionnaire items. We also included extra items that tapped areas of pet relationships which we know from our work and research in bereavement and chronic illness that might be especially important to children experiencing stressful life changes.
In constructing the questionnaire it was necessary to make slight modifications in the pet measures. In some instances we revised the response categories to make them compatible with the format and metric of the Childhood Bereavement Interview Schedule. We also altered the question stems of some items so that they could be answered about any type of pet, not just dogs or cats. Because the pet questions will be asked as part of a face-to-face interview, we opted for obtaining a more comprehensive or precise response to some of the questions rather than giving the response a forced choice alternative. For example, for duration questions, the exact length of time would be recorded rather than coding the answer into a category which spanned a range of time (e.g., 1-5 years, 6-10 years). Also, questions which asked the respondent to give one or more reasons for an attitude or behavior were left open-ended to obtain a more comprehensive answer before being coded into broad categories.
Both an adult version and a child version of the pet questionnaire were constructed. Skip instructions for follow-up interviews were developed so that the pet history questions would be administered to each parent and child only once. Other skip instructions ensure that the appropriate items to be asked of current pet owners (parent/child), former pet owners (parent/child), and non- pet owners (parent/child).
Because of the age range of the population being surveyed (i.e., children aged 7-16 as well as adults), we pre-tested the various versions of the pet questionnaire on different age-groups of children and adults, checking for question flow and sequencing, clarity of wording, comprehension and age-appropriateness before using it with the families in the study. Once we finalized the pet questionnaire items we then incorporated it into the Childhood Bereavement Interview Schedule (Adult and Child versions) and began administering the questionnaire in Year 1 as a family came due for an initial or follow-up interview. Data collection will continue throughout Year 2 of the project. All well/surviving parents and children between 7-16 years old who are accrued into the study, regardless of the type of intervention they receive, are asked to participate in the research interviews. We are estimating that by the completion of the data collection in Year 2 approximately 200 families will have completed individual interviews containing the detailed pet questions (i.e., individual interviews with the well parent and each age-eligible child).
Description of Year 1 Secondary Data Analyses
The secondary analyses that were conducted during Year 1 focused on research aims 1 and 2. Specifically, we examined the association of the presence of a household pet with the psychological well-being of adults who were providing informal caregiving to a family member seriously-ill with cancer. We also examined the efficacy of the presence of a household pet in mediating the stressful consequences of the provision of informal caregiving to cancer patients.
For many years treatment for patients with cancer involved long and frequent hospitalizations. More recently, however, care has increasingly been administered on an outpatient basis (Mor, Guadagnoli and Wool, 1987). This has resulted in family members performing a growing variety of complex time-consuming and stressful tasks to meet the outpatients needs for assistance (Siegel, Raveis, Mor and Hoots, 1991). While the practical and emotional benefits that accrue to the recipients of familial support and care are well recognized (Shanas, 1979), there is substantial evidence that the burdens of providing informal practical and emotional support are far-reaching and diverse, negatively impacting on numerous aspects of caregiver's lives (see Siegel, Raveis, Mor and Hoots, 1991, for review).
The data for these analyses was drawn from the recently completed collaborative study conducted by MSKCC, Brown University and Pennsylvania State University College of Medicine. The research, a longitudinal investigation of the caregiving needs of cancer patients and their family members (see Study 1).
In the analyses presented below, we examine the contribution of the presence of a pet (specifically a cat or a dog) in the household to caregiver psychological well-being in association with caregiver burden and other aspects of the caregiving situation. The sample is restricted to the 243 spousal dyads for which all the data were available. Among married patients, the spouse is almost always the primary provider of emotional support and personal care (Shanas, 1979; Stone et al., 1987). Spousal caregivers also appear to be particularly at risk for caregiver burden; typically providing the most extensive and comprehensive care, they maintain the role longer, tolerate greater levels of disability than other caregiver groups, experience more severe lifestyle adjustments and exhibit lower levels of well-being than other caregiver groups (Cantor, 1983; Soldo and Myllyluoma, 1983; George and Gwyther, 1986).
A comparison of households containing a dog or cat with those that do not
Before a multivariate analysis of the effect of the presence of a pet in the household on caregiver depression, it was important to look for ways in which pet owners and non-pet owners differed. Forty percent of the households included in the analysis contained at least one pet. Thirty-seven percent of all households contained either a dog or a cat (with or without another pet). Of the eight households which had a pet, but did not contain either a dog or a cat, three contained only a bird(s); three contained only fish and/or turtles; and two contained a pet(s) of some other unspecified species. For all subsequent analyses, the measure of pet ownership used was whether or not a cat and/or a dog was present in the household. Given the wide variety of species included in respondent households it is important to provide at least some control for the type of pet(s) included in the household. This restricted definition of pet ownership provides an important control in that any effect associated with the measure can be said to apply only households known to contain at least one or more mammals (all households containing rabbits or horses also contain a cat or dog). The analysis of whether different pets are associated with different effects on adjustment will be conducted during Year 2. Cats and dogs are the only two species of pets residing in a sufficient number of respondent households to analyze separately and are therefore the only two species for which these analyses will be run.
Although caregivers who were also pet owners of cats or dogs were more likely to be males-46% of the caregivers who were also pet owners were male compared to 66% of the non-owners, this difference was not statistically significant. Pet owners were significantly more likely to be white-100% of the pet owners were white compared to 93% of non-pet owners (p. =.021). Pet owners, as a group, were significantly younger than non-owners (mean age 56.0 vs. 61.0, p<.001). They were also more likely to be employed (49% vs. 39%), although this difference was not statistically significant. Both pet owners and non-owners were fairly well-educated (40% and 38%, respectively, had attended college, with 20% and 27% having graduated college). The distribution of household incomes was similar for both groups even though the median household income was higher for those households containing a cat or dog than those that did not ($32,239 and $27,500, respectively). There was no significant difference between pet owners and non-pet owners in terms of any measure of the patient's health status, type of cancer, or stage of illness at the time of the interview.
Measurement of variables used in the regression analyses
A variety of measures, based on a review of the literature on informal caregiving, have been shown to be associated with caregiver psychological distress (see Siegel, Raveis, Mor and Hoots, 1991 for a review) and were selected for inclusion in the regression analysis in addition to the presence of a cat or dog in the household. These groups of variables include demands and characteristics of the caregiving situation, caregiving burdens, and various sociodemographic characteristics.
Demands and characteristics of the caregiving situation
The patient's need for illness-related assistance with each of the following activities was identified: personal care (bathing, dressing, or getting around); instrumental activities (cooking and shopping, light and heavy housekeeping); transportation (to doctor or for other purposes), home health care, child care, and administrative tasks. The total number of domains of activities for which they required assistance because of their illness was then summed to form a count of the total number of patient needs. Patients were also asked the total number of days during the prior two weeks that they spent entirely or most in bed or had to curtail their activities because of their illness or illness-related treatment. The responses were summed (with a cap of 14 days) and trichotomized as 0 days, 1 to 13 days, and 14 days, representing a measure of reduced patient activity days. Familial caregivers were asked to assess the extent to which the patient's illness made it difficult to establish a daily routine or plan activities. Caregivers also reported how confident they felt that they could personally provide the patient with additional care should the need arise.
Caregiver burden
Based on the literature, caregiver burden was conceptualized as falling into five principal domains: employment, financial, physical, social and time. Using items contained in the caregiver interview, separate measures were constructed of objective (i.e., empirically verifiable or observable consequences of caregiving) and subjective (i.e., perceived or felt consequences of caregiving) burden for each domain. A separate summary measure of global objective and global subjective burden was constructed for each domain (see Siegel, Raveis, Mor and Hoots, 1991), consistent with other caregiver investigations (Raveis, Seigel and Sudit, 1990).
Presence of a cat or dog in the household
The measure of pet ownership used in the regression analyses was whether or not a cat and/or a dog was present in the household as discussed above.
Description of the sample and distribution of variables used in the regression analyses
In 62% of these dyads the husband was the patient, while in the remainder the wife was the patient. The majority of patients and their spouses were 45 and older, with a sizable proportion 65 and older, 37% and 35%, respectively. Nearly all respondents (patients and caregivers) were white (96% and 95% respectively). About one quarter of the patients and 43% of the spousal caregivers were currently employed. Median household income was $29,149. As stated earlier, in 37% of the households a cat or dog was present. All patients reported needing assistance with tasks in at least one domain of activity, with most needing help with three or more domains. While none of the patients reported needing assistance with all six domains of activities this is due to the fact (given the age distribution) that only 7% of the households contained a child under 12 years of age, rather than indicating a lack of need. Thirty-nine percent of the patients had had to curtail their activities everyday during the past two weeks while 29% reported not having to curtail their activities at all. Over three-fifths (62%) of the caregivers reported their caregiving activities had disrupted their daily routines at least "a little." Most were "quite" or "very" confident they would be able to provide more care if the patient needed it in the future, but one-fourth were only "somewhat" or "not at all" confident they would be able to do so. A sizable proportion (30%) of the spousal caregiver sample scored at or above the cutoff for clinical caseness for depression on the CES-D scale. The mean CES-D for the entire spousal caregiver sample was 11.4 (S.D.-9.8).
The predictors of depression included in these analyses fall into four major domains: sociodemographic characteristics, measures of the ill spouse's level of need and current health status, measures of caregiver burden (global objective and subjective, and the spousal caregiver's assessment of ability to provide additional assistance if needed), and whether there is a cat and/or dog in the household.
Since the analysis is restricted to spousal dyads, caregiver and patient demographics are highly correlated. As a result only the spousal caregiver demographic characteristics [i.e., sex, age, employment status, and health status-spousal caregiver has a health condition(s) which make it difficult to provide care] are being included in multivariate analyses. Spousal caregiver's depression score is significantly correlated (p<.05) with all but two of the independent measures-spousal caregiver's employment status and the presence of a cat or dog in the household. The spousal caregiver's depression score is most strongly correlated with all measures of caregiver burden (all p<.001), especially the global measure of subjective burden). Spousal caregiver depression is also significantly correlated with patient's illness-related need for assistance and reduced activity days.
As one would expect, based on the prior research on informal caregiving and caregiver burden, higher levels of spousal caregiver burden and patient need for assistance are associated with higher levels of caregiver depression. In addition, consistent with the epidemiological research on gender an age differences in depression, the caregiving husbands reported lower levels of depression and younger spousal caregivers had higher levels of depression. Those spousal caregivers who reported having a limiting health condition also tended to have higher depression scores. Although the zero-order correlation between spousal caregiver depression and the presence of a cat or dog in the household was not significant, there was a trend towards lower levels of depression among those caregivers with a cat or dog in their household (r=-.08, p=.l0).
Correlations between the independent variables are also as expected with measures in each domain (for the most part) significantly correlated to each other. The measures of patient need are significantly and positively correlated with higher levels of burden. There are only scattered correlations between the measures of caregiver characteristics and the burden measures. The presence of a cat and/or dog in the household is significantly correlated with caregiver sex (pets are more likely in households where the husband is the caregiver) and caregiver age (pets are more likely in younger households.
To test the relative strength of the univariate relationships between predictors and to determine which relationships remained significant once control for the other measures was made, a hierarchiecal linear regression model was tested. Even though spousal caregivers' employment status was not significantly correlated with spousal caregivers' depression score, it was retained because it was strongly correlated to other independent measures and because it was conceptually and theoretically important to retain. After the model was estimated, including all variables but the presence of a cat or dog in the household, it was reestimated with this variable included to see if its inclusion:
added to the explanatory power of the model, and
impacted on spousal caregivers' depression level once the patient's level of need, caregiver characteristics, and the caregiver's level of burden and ability to do more, were controlled for.
Results and discussion of the regression analyses
Model 1 showed that global subjective burden which also had the strongest zero-order correlation with caregiver depression is also its strongest predictor. The fact that no other measure of burden or patient need retains a significant relationship with caregiver depression once global subjective burden is controlled for suggests the effect of these measures on caregiver depression may be indirect - i.e. the patient's level of need, the caregiver's ability to provide help, and the amount of burden the caregiver is already doing impact the patient's level of depression through global subjective burden rather than directly. The main predictor of caregiver depression is not so much the status of the patient or the patient's ability to perform various activities, but caregivers' perception of the situation. These findings suggest the use of LISREL or path analysis to determine which aspects of subjective burden are most closely associated with depression, and the nature of indirect effects patient need and other measures of burden have on caregiver depression. Ironically, the only other significant predictor of caregiver depression in Model 1, after all other independent measures have been controlled for is caregiver employment status, which did not have a significant zero-order correlation with caregiver depression.
Model 2 reestimated the equation predicting spousal caregiver depression with the presence of a cat or dog in the household included. As was the case with caregiver employment status, even though the presence of a cat or dog in the household had a weak zero-order relationship with caregiver depression its relationship is significant once the other independent measures have been controlled for (beta = - .111; p=. 041). The inclusion of the presence of a cat or dog in the household to Model 1 did little to change the betas between any of the other independent measures, but did significantly increase R2 (F change=4.206;p=. 041). As was the case in Model 1, global subjective burden (beta = - .454; p=. 000) and caregiver employment status (beta =- .148; p =. O21) were the only two of the original ten independent variables to have a significant direct effect on caregiver depression. These findings suggest that all other things being equal, the predicted level of depression for a cat and/or dog owner is 2.239 points lower than that of a similar person who does not own a pet.
These findings suggest that having a pet in the household may well be associated with better psychological adjustment for caregivers whose spouse is seriously ill. While the current findings pertain only to spousal caregivers, they suggest that the presence of companion animals, specifically, cats and dogs, may have a beneficial impact in situations of stress and burden, such as experienced when caring for a spouse who is seriously ill.
References
Akiyama, H., Holtzman, J .M. & Britz, W. E. (1986). Pet ownership and health status during bereavement. Omega, 17:187-93.
Albert, A., & Bulcroft, K. (1987). Reviews and research reports: Pets and urban life. Anthrozoos, 1(1): 9-25.
Baun, M.M., Bergstrom, N.F., Lanston, & Thomas, L. (1984). Physiological effects of petting dogs: Influences of attachment. In R.K. Anderson, B.L. Hart and L.A. Hart (Eds.), The pet connection: Center to Study Human-Animal Relationships and Environments, (pp. 162-70). Minneapolis: University of Minnesota.
Beck, A.M., & Katcher, A.H. (1983). Between pets and people: The importance of animal companionship. New York: G.P. Putnam's Sons.
Cantor, M., (1983), Strain among caregivers: A study of experience in the United States. The Gerontologist, 23: 597-604.
Covert, A.M, Whiren, A.P., Keith, J. & Nelson, C. (1985). Pets, early adolescents, and families. Marriage and Family Review, 8(3-4): 95-lO8.
Davis, J.H. (1987a). Pet care during preadolescence: Developmental considerations. Child: Care: Health and Development,13: 269-276.
Davis, J.H. (l987b). Preadolescent self-concept development and pet ownership. Anthrozoos, 1(2): 90-94.
Friedmann, E., Katcher, A.H., Lynch, J.J., & Thomas, S.A. (1980). Animal companions and one-year survival of patients after discharging from a coronary care unit. Public Health Reports, 95: 307-12.
Friedmann, E., Katcher, A.H., Thomas, S .A., Lynch, J .J., & Messent, P.R. (1983). Social Interaction and blood pressure: Influence of animal companions. Journal of Nervous and Mental Diseases, 171: 46l-65.
Garrity, T.F., Stallones, L., Marx, M.B., & Johnson, T.P. (1989). Pet ownership and attachment as supportive factors in the health of the elderly. Anthrozoos, 3(1): 35-44.
George, L.K. and Gwyther, L.P. (1986). Caregiver well-being: A multidimentional examination of family caregivers of demented adults. Gerontologist, 21: 464-70.
Grossberg, J., & Alt, E., Jr. (1984). Interactions with pet dogs: Effects on human blood pressure. Abstract in People, Animals, Environment, 2(2): 20.
Katcher, A.H., & Beck, A.M. (1986). Dialogue with animals. Translations and Studies College of Physicians, 8: 105-12.
Katcher, A., Beck, A .M., & Levine, D. (1988). Reviews and research report: Evaluation of a pet program in prison - the pal project at Lorton. Anthrozoos, 2(3): 175-180.
Katcher, A.H., Friedman, E., Beck, A.M., & Lynch, J.J. (1983). Looking, talking and blood pressure: The Physiological consequences of interaction with the living environment. In A.H. Katcher & A.M. Beck (Eds.), New perspectives on our lives with companion animals, (pp. 351-59). Philadelphia: University of Pennsylvania Press.
Katcher, A.H., Segal, H., & Beck, A.M. (1984). Comparison of contemplation and hypnosis for the reduction of anxiety and discomfort during dental surgery. American Journal of Clinical Hypnosis, 27: 14-21.
Kidd, A.H., & Feldman, B.M. (1981). Pet ownership and self-perceptions of older people. Psychological Reports, 48: 867-875.
Lago, D., Delaney, M., Miller, M., & Grill, C. (1989). Companion animal's attitudes toward pets, and health outcomes among the elderly: a long-term follow-up. Anthrozoos, 3(1): 25-34.
Lago, D.J., Knight, B.L., & Connell, C.M. (1983). Rural elderly relationships with companion animals. Feasibility study of a pet placement program. In A. Katcher, & A. Beck (Eds.), New perspectives on your life with companion animals, (pp. 328-40). Philadelphia: University of Pennsylvania Press.
Lawton, M. P., Moss, H., & Moles E. (1984). Pet ownership: A research note. The Gerontologist 208-11.
Mor, V., Guadagnoli, E., Wool, M. (1987). An examination of the concrete service needs of advanced cancer patients. Journal of Psychosocial Oncology. 5: 1-17.
Mugford, R.A., & M'Comisky, J.G. (1975). Some recent work on the psychotherapeutic value of caged birds with old people. In R.S. Anderson (Eds.), Pet Animals and Society. London: Bailliere Tindall.
Ory, M.G., Goldberg, E.L. (1983). Pet possession and well-being in elderly women. Research on Aging, 5(3): 389-409.
Poresky, R.H., Hendrix, C., Mosier, J.E. & Samuelson, M.L. (1988). Young children's companion animal bonding and adults' pet attitudes: A retrospective study. Psychological Reports, 62: 419-425.
Raveis, V.H., Siegel, K., Sudit, H. (1990) Psychological impact of caregiving on the caregiver: Critical review of research methodologies. In: D.E. Biegel and A. Blum (Eds.), Aging and Caregiving. Newbury Park, CA: Sage.
Robb, S.S., Stegman, C.E. (1983). Companion animals and elderly people: A challenge for evaluators of social support. Gerontologist, 23(3): 277-282.
Shanas, E. (1979). The family as a support system in old age. The Gerontologist, 19: 169-174.
Serpell, J.A. (1981). Childhood pets and their influence on adults' attitudes. Psychological Reports, 49:651-654.
Siegel, K., Mesagno, F.P., & Christ, C. (1990). A prevention program for bereaved children. American Journal of Orthopsychiatry, 60(2): 168-175.
Siegel, K., Raveis, V.H., Mor, V., and Hoots, P. (1991). The relationship of spousal caregiver burden to patient disease and treatment-related conditions. Annals of Oncology. 2:511-516.
Siegel, J.M. (1990). Stressful life events and use of physician services among elderly: The moderating role of pet ownership. Journal of Personality and Social Psychology, 58(6): 1081-1086.
Siegmund, R. & Biermann, K. (1988). Common leisure activities of pets and children. Anthrozoos, 2(1): 53-57.
Soldo, B.J. & Myllyluoma, J. (1983). Caregivers who live with dependent elderly. Gerontologist, 23, pp. 605-611.
Stone, R., Cafferata, G.L., & Sangl, J. (1987). Caregivers of the frail elderly: A National Profile. The Gerontological Society of America, 27, (5), 616-626.
Templer, D., Salter, C., Dickey, S., Baldwin, R., & Veleber, D. (1981). The construction of a pet attitude scale. The Psychological Record, 31: 343-348.
Wilson, C.C., Netting, F.E., & New, J .C. (1987). Reviews and research reports: The pet attitude inventory. Anthrozoos, 1(2): 76-84.
