Eva Kahana

Depressive Symptoms and 3-Year Mortality in Older Hospitalized Medical Patients

Hospitalization is associated with a long-term increased risk for death, especially in older persons (1-4). Although the mediators for this increased risk have not been fully elucidated, depression may play an important role. Because depression is common in hospitalized older patients, an association between depression and mortality in this population would be of significant clinical importance (5-13). The hypothesis that depression may be a mediator of death in hospitalized patients is supported by studies demonstrating that depressive symptoms are associated with increased mortality in community-dwelling patients and in highly selected groups of hospitalized older patients, such as those with acute myocardial infarction (14-18). These studies have often been limited by inadequate accounting for the complex interrelations between depressive symptoms and other predictors of death, such as acute physiologic impairment, chronic comorbid illness, functional impairment, and cognitive impairment. Because depressive symptoms are clearly correlated with and may partly be the result of these other factors (5, 19-20), improving our understanding of the relation between depression and death requires use of standardized methods to measure and adjust for confounders. We tested the hypothesis that depressive symptoms are associated with long-term mortality in hospitalized older patients. We demonstrated previously that depressive symptoms are strongly associated with adverse health status outcomes in hospitalized medical patients through 90 days after admission (21). However, our initial study found no association between depressive symptoms and mortality during the first 90 days after admission (21). In this report, we extend mortality follow-up to 3 years by merging our data with a national mortality database. Furthermore, we adjusted for standard measures of physiologic impairment, comorbid illness, and functional impairment at hospital admission to control for the possibility that higher levels of these confounders in patients with more depressive symptoms affect the association between depressive symptoms and death. Methods Patients Patients were drawn from serial, prospective longitudinal studies of functional change in older hospitalized patients on the general medical service of University Hospitals of Cleveland. The inclusion and exclusion criteria for these studies are described elsewhere (3, 21, 22). The first study enrolled 206 patients 75 years of age or older who were admitted between March 1990 and July 1990. The second study, a controlled trial of an intervention to improve functional outcomes, enrolled 651 patients 70 years of age or older who were admitted between November 1990 and March 1992. The first study, which was a pilot study for the second study, enrolled consecutive patients. The second study randomly assigned patients to an intervention designed to improve functional outcomes in older persons or to usual care (22). In each study, patients admitted to the intensive care unit, telemetry service, or oncology service were excluded. Data collection procedures in both studies were almost identical. Other than a slightly higher mean patient age in the first study cohort, the demographic, clinical, and functional characteristics of patients in the first study, the control group of the second study, and the intervention group of the second study were similar. Additional analyses that adjusted for whether patients were in the first study cohort or the control group compared with the intervention group of the second study cohort yielded results that were almost identical to the results reported here. Of 857 older patients enrolled in the two studies, 284 were excluded from the current study because they were too ill or confused to be interviewed about depressive symptoms at the time of admission (n=164), were admitted from nursing homes (n=38), were not available for interview (n=37), declined interview (n=27), or died before being approached (n=18). We excluded patients admitted from nursing homes because interview data were less consistently obtained from these patients. Thus, the analytic sample for this study comprised 573 patients. Assessment of Depressive Symptoms Within 48 hours of admission, patients were interviewed by using the 15-item Geriatric Depression Scale to assess depressive symptoms over the past week (23, 24). The Geriatric Depression Scale is well suited for use in acutely ill older persons because it focuses on symptoms of depression that are less likely to be directly influenced by somatic illness. Examples of items on this scale include feeling bored, dropping activities and interests, feeling helpless, feeling worthless, feeling that life is empty, feeling that others are better off, preferring to stay at home instead of doing new things, and feeling hopeless. We divided patients into those reporting five or fewer symptoms and those reporting six or more symptoms; these are commonly recommended cutoffs on the 15-item Geriatric Depression Scale (24-26). Measurement of Mortality We determined mortality and date of death during the 3 years after hospitalization by merging our files with the National Death Index, a database of all deaths in the United States generated from state death certificates. Its sensitivity and specificity have been reported to be 98% and 100%, respectively (27). Measurement of Potential Confounders Shortly after admission, we surveyed each patients primary nurse about the patients independence in six activities of daily living (dressing, bathing, grooming, toileting, transferring, and eating) based on the scale of Katz (28). Within 48 hours of admission, we administered to patients the first 21 items of the 30-item Folstein Mini-Mental State Examination (29). To minimize respondent burden, we used only the first 21 items. Scores on the 21-item instrument have previously been shown to correlate highly (r=0.9) with scores on the 30-item instrument and to have construct validity on the basis of their strong association with functional outcomes (30). Data gathered from medical records included the reason for admission, the components of the Acute Physiology and Chronic Health Evaluation (APACHE) II score (31), and the components of the weighted Charlson comorbidity index of illness (32). The APACHE II score is a commonly used measure of physiologic severity, and the Charlson score is often used as a measure of the burden of comorbid illness. Statistical Analysis For our primary set of analyses, we compared patients who had six or more depressive symptoms (depressed patients) with patients who had five or fewer symptoms (nondepressed patients). We used the chi-square test or t-test to compare the characteristics of patients in each category at hospital admission. Survival curves describing mortality in the 3 years after hospitalization in each group were prepared by using the method of Kaplan and Meier. We used Cox regression to determine whether depressive symptoms were independently associated with mortality over 3 years. In the first model, we measured the unadjusted association between depressive symptoms and mortality. In the next four models, we determined the association between depressive symptoms and mortality after controlling for APACHE II scores, Charlson comorbidity index scores, dependence in activities of daily living, or cognitive function. In the sixth model, we controlled for all of these potential confounders as well as age, sex, ethnicity, and whether the patient lived alone. We did two secondary analyses. First, we modeled depression scores as the number of depressive symptoms on admission. We also determined the hazard ratio associated with multiple cut-points on the Geriatric Depression Scale. Results The mean age of the 573 patients was 79.9 years; 67.8% of patients were women and 39.4% were African-American. About half were independent in all activities of daily living at hospital admission. Table 1 lists the 10 most common reasons for hospital admission, classified by using the method of Charlson (33). The mean number of depressive symptoms at hospital admission was 4.5, and 34% of patients reported six or more symptoms. At admission, patients with six or more depressive symptoms had higher comorbidity scores, were more likely to have congestive heart failure or chronic obstructive pulmonary disease, had lower cognitive function scores, and were dependent in more activities of daily living (Table 2). Three years after admission (Figure), the mortality rate was higher among patients with six or more depressive symptoms than among patients with five or fewer depressive symptoms (56% compared with 40%; P<0.001). Of the 376 patients with five or fewer depressive symptoms at admission, 78 (21%) died during the first year of follow-up, 121 (32%) died during the first 2 years, and 151 (40%) died during the 3 years. Of the 197 patients with six or more depressive symptoms on admission, 58 (29%) died in the first year, 88 (45%) died during the first 2 years, and 110 (56%) died during the 3 years. The unadjusted hazard ratio over 3 years of follow-up for patients with six or more depressive symptoms was 1.56 (95% CI, 1.22 to 2.00). Patients with six or more depressive symptoms were also slightly more likely than patients with five or fewer symptoms to be discharged to a nursing home (8.3% compared with 4.9%; P=0.11). Table 1. Most Common Reasons for Hospital Admission (n=573) Table 2. Characteristics of Patients at Hospital Admission Figure. Mortality over 3 years (1095 days) in patients who had six or more depressive symptoms compared with patients who had five or fewer symptoms. Although adjustment for physiologic severity, comorbid illness, dependence in activities of daily living, and cognitive function each reduced the strength of the association between depressive symptoms and mortality, in each case this association remained statistically

Altruism, helping, and volunteering: pathways to well-being in late life.

Objectives: We examined the influence of prosocial orientations including altruism, volunteering, and informal helping on positive and negative well-being outcomes among retirement community dwelling elders. Method: We utilize data from 2 waves, 3 years apart, of a panel study of successful aging (N = 585). Psychosocial well-being outcomes measured include life satisfaction, positive affect, negative affect, and depressive symptomatology. Results: Ordinal logistic regression results indicate that altruistic attitudes, volunteering, and informal helping behaviors make unique contributions to the maintenance of life satisfaction, positive affect and other well being outcomes considered in this research. Predictors explain variance primarily in the positive indicators of psychological well-being, but are not significantly associated with the negative outcomes. Female gender and functional limitations are also associated with diminished psychological well-being. Discussion: Our findings underscore the value of altruistic attitudes as important additional predictors, along with prosocial behaviors in fostering life satisfaction and positive affect in old age.

Long-term impact of preventive proactivity on quality of life of the old-old.

Objective This research explored the long-term benefits of engaging in proactive health promotion efforts among old-old residents of Sunbelt retirement communities to empirically test components of the Preventive and Corrective Proactivity (PCP) Model of Successful Aging. Specifically, we examined the contributions of exercise, tobacco use, moderate alcohol use, and annual medical checkups to multidimensional quality of life indicators of physical health, psychological well-being, and mortality. Method Data were obtained from a longitudinal study of adaptation to aging. Annual in-home interviews were conducted with 1000 older adults over a 9-year period. Whether health promotion behaviors at baseline predicted quality of life outcomes 8 years later was examined, controlling for the baseline outcome, sociodemographic variables, and, as an additional test, baseline health conditions. Results Exercise was predictive of fewer IADL limitations and greater longevity, positive affect, and meaning in life 8 years later. Avoiding tobacco was predictive of longevity. Before controlling for health conditions, exercise predicted decreased risk of basic activities of daily living limitations and having more goals; moderate alcohol use predicted longevity; annual health checkup predicted more IADL limitations; and having once smoked predicted having more IADL limitations and negative affect. Conclusions Among the old-old, exercise had long-term and multifaceted benefits over an 8-year period. Tobacco avoidance also contributed to long-term positive outcomes. These results lend support to the long-term preventive value of health-promoting proactivity spontaneously engaged in by old-old persons proposed in the framework of the PCP model.

The Personal and Social Context of Planning for End-of-Life Care

Objectives: To examine the potential facilitators of or deterrents to end‐of‐life planning for community‐dwelling older adults, including personal (health‐related and sociodemographic) and social (physician and family) influences.

Job Commitment and Turnover among Women Working in Facilities Serving Older Persons

This study presents a model linking personal and job-related factors to job satisfaction, job commitment, and turnover. Responses from the staff of six nursing homes and 12 community facilities serving older adults were included. Using a modified version of the causal model of turnover developed by Price and Mueller, three sets of predictors were tested to explain the causes for turnover: personal characteristics, job characteristics, and attitudes. The best predictor of turnover was the employees intention to leave, followed by the length of employment (shorter), and age (younger). Intention to leave was, in turn, predicted by age (younger), length of employment (shorter), job dissatisfaction, and the type of agency for which the employee worked (community).Dissatisfaction seems to be a major factor that results in a desire to leave the job and may lead to either turnover or continued dissatisfaction with the job. Implications for enhancing employee morale and reducing job turnover are discussed.

Family caregiving across the lifespan

Introduction - Eva Kahana, David E Biegel, and May L Wykle PART ONE: PARADIGMS FOR CAREGIVING Developmental Challenges and Family Caregiving - Eva Kahana et al Bridging Concepts and Research Someone to Watch Over Me - Joan Aldous Family Responsibilities and Their Realization Across Family Lives Altruism Through the Life Course - Elizabeth Midlarsky Optimal Use of Formal and Informal Systems Over the Life Course - Eugene Litwak, Dorothy Jones Jessop, and Heather J Moulton PART TWO: ILLNESS AND LIFE STAGE: CHALLENGES FOR CAREGIVING Caregiving and Children - Joan M Patterson and Barbara J Leonard Caregivers of Persons Living With AIDS - Patricia Flatley Brennan and Shirley M Moore Predictors of Caregiver Burden Among Support Group Members of Persons With Chronic Mental Illness - David E Biegel, Li-yu Song and Venkatesan Chakravarthy Psychoeducational Programs - Catherine F Kane From Blaming to Caring The Home Care of a Patient With Cancer - Charles W Given and Barbara A Given The Midlife Crisis Caregiving Issues After a Heart Attack - Rosalie F Young and Eva Kahana Perspectives on Elderly Patients and Their Families PART THREE: THE INTERFACE BETWEEN FORMAL CARE PROVIDERS AND CAREGIVING FAMILIES Care at Home - Lucy Rose Fischer and Nancy N Eustis Family Caregivers and Home Care Workers The Caregiver as the Hidden Patient - Jack H Medalie Challenges for Medical Practice Caregiving Issues in Families of Children With Chronic Medical Conditions - Anne E Kazak and Dimitri A Christakis Relationships Between the Frail Elderlys Informal and Formal Helpers - Linda S Noelker and David M Bass Conclusion - Eva Kahana, David E Biegel, and May L Wykle

Theoretical and Research Perspectives on Grandparenthood

It has been argued that the status of the aged is highest in those societies where they continue to perform useful and valued functions (Cowgill, in press). With some 70 percent of older people in the United States having living grandchildren, grandparenthood represents an area of potential usefulness and social value for a large segment of the elderly. What then is the significance and meaning of grandparenthood for the old and not-so-old? How do grandparents influence their grandchildren-what is the nature of their interactions with one another? How does the role and its meaning vary for diverse segments of the elderly population? The focus of scientific interest in intergenerational relationships is usually on the broader social scene. Yet the prototypes of these relationships originate within the family. The critical nature of parental influences on child development is well known. In a recent work on the family (Anthony, 1970), the important influences of children on their parents have also received considerable attention. But what is the influence of the childs experience or lack of experience with his grandparents on shaping his personality? How does it affect his later views of older people and hence the status of the elderly in society? Conversely, does the grandchild have any important effects on his grandparent? Is the relationship between grandchildren and grandparents a real one or simply a formality or ritual as some have suggested? Studies of grandparenthood are thus far largely pilot investigations, based on relatively small samples and yielding data which may or may not be generalizable to larger populations. Yet when one considers all the evidence so far available, one gets a glimpse of the complexities, as well as the regularities now known, relating to the mutual influences between the grandparent and grandchild generations. We will attempt to summarize some of the issues and problems which emerge, based on a review of the literature, and based on several exploratory studies that we have conducted with diverse groups of grandparents and their grandchildren.

Proactive aging: A longitudinal study of stress, resources, agency, and well-being in late life

Objectives: Using the Proactivity Model of Successful Aging, we examined how internal and external resources contribute to the maintenance of psychological well-being and social activities among older adults who experience normative stressors of aging. Outcome variables in this study are collectively referred to as quality of life (QOL). We also examined the mediating role of proactive adaptations between internal and external resources and QOL indicators. Method: Based on five annual interviews of a sample of 1000 community-dwelling older adults in Florida (effective N = 561), we tested the lagged effects of stressors on two indicators of QOL, four years later. In the full longitudinal model, using structural equations, we estimated the direct effects of internal and external resources on QOL, along with indirect effects through proactive adaptations. Results: Stressors negatively influenced QOL four years later. Internal and external resources led to better QOL four years later, both directly and indirectly through proactive adaptations of marshaling support and planning for the future. Conclusion: These findings lend support to the Proactivity Model of Successful Aging by documenting the value of proactive adaptations (i.e., exercise, planning ahead, and marshaling support) as proximate influences on QOL outcomes (i.e., depressive symptomatology and social activities). Findings suggest that older adults can maintain successful aging even in the face of health-related and social stressors by invoking accumulated resources to deal actively with the challenges of aging.

Psychological well-being among Holocaust survivors and immigrants in Israel

This research assessed predictors of psychological well-being among 180 survivors of the Holocaust and among a comparison group of 160 immigrants of similar sociocultural background living in Israel. Four variables: better health, higher instrumental coping, lower emotional coping, and lesser social concern, were found to be significant predictors of psychological well-being in both groups. Among survivors, these four variables, as well as being married, fewer life crises, communication with co-workers, and not being resigned to fate, accounted for 52% of explained variance in psychological well-being. Among immigrants who served as a comparison group, these four variables, along with a relaxed personality style and good communication with ones spouse, accounted for 36% or explained variance in psychological well-being. Theses factors underscore the importance of current social and psychological adaptation for psychological well-being among survivors and among older persons of similar backgrounds who did not endure the extreme trauma of the Holocaust.

The Effects of Stress, Vulnerability, and Appraisals on the Psychological Well-Being of the Elderly

A conceptual model was developed and tested that examined the relationships between respondent characteristics, stressors, psychological well-being measures, and intervening life domain appraisals. The model was tested using data from elderly members of a Detroit area HMO. An innovative focus of the study was comprehensive consideration of the array of stressors impinging on the elderly. Stressors examined included recent negative life events, cumulative life crises, living with an ill family member, and social isolation. Domain appraisals in the model concerned satisfaction with activities, relationships, health, and income. Support was found for the hypothesis that stressors affect well-being indirectly through domain satisfactions.