• unlimited access with print and download
    $ 37 00
  • read full document, no print or download, expires after 72 hours
    $ 4 99
More info
Unlimited access including download and printing, plus availability for reading and annotating in your in your Udini library.
  • Access to this article in your Udini library for 72 hours from purchase.
  • The article will not be available for download or print.
  • Upgrade to the full version of this document at a reduced price.
  • Your trial access payment is credited when purchasing the full version.
Buy
Continue searching

Hospital depressive symptoms and ADL disability in older adults: A longitudinal analysis of course and associations

Dissertation
Author: Carrie A. Ciro
Abstract:
Depressive symptoms and disability in activities of daily living (ADL) often increase in older adults during hospitalization and for many persist post-discharge. However, little is known about the psychological and functional response of older adults admitted to an Acute Care for Elders (ACE) unit. Questions remain about the association between depressive symptoms and ADL disability and factors that moderate these associations are unknown. Objectives of this study were to investigate: (1) change in depressive symptoms and ADL function from hospital to 3 month follow-up; (2) the association between depressive symptoms in hospital and ADL function 3 months postdischarge; and (3) moderators of the depression-ADL association. A tri-ethnic (white, black and Hispanic) sample of 403 older adults within an ACE database contributed subjective and objective information related to depressive symptoms, clinical variables and activity/participation measures across two time frames, admission and three months post discharge. A large minority reported high depressive symptoms in hospital and over half reported ADL disability. Across both assessment periods, risk factors for having high depressive symptoms were being unmarried and having any level of ADL disability. Conversely, risk factors for ADL disability were pain and depressive symptoms. At 3 months post discharge, the recovery rate from depression and incident ADL disability was high. Positive change in depression was significantly associated with positive change in ADL status. Increasing severity of hospital depression was associated with increased odds of ADL disability at the 3 month follow-up. Neither gender, marital status, pain nor medical history moderated this relationship. This study indicates that while older adults experience higher depressive symptoms and ADL disability while hospitalized, resolution of symptoms occur for many. This research contributes to the literature by extending our knowledge of the course and associations between depressive symptoms and ADL disability in hospitalized, older adults. Future research which focuses on interventions to minimize depressive symptoms and ADL disability is warranted.

Table of Contents

Table of Contents......................................................................................................viii   List of Tables.............................................................................................................xiv   List of Figures............................................................................................................xvi   Chapter 1: Introduction................................................................................................1   Specific Aims...........................................................................................................1   Significance of Research.........................................................................................2   Outline of Dissertation..............................................................................................2   Section 1: Depressive Symptoms............................................................................3   A. Construct Definitions........................................................................................3   Major Depression..............................................................................................3   Minor Depression..............................................................................................3   Depressive Symptoms......................................................................................4   B. Measurement of Depressive Symptoms in Older Adults..................................4   Center for Epidemiologic Studies-Depression Scale (CESD)...........................4   Clarification of Study Terminology....................................................................5   C. Prevalence of High Depressive Symptoms in Older Adults.............................5   Community-Dwelling Older Adults....................................................................5   Hospitalized Older Adults..................................................................................6   D. Factors Associated with Depressive Symptoms..............................................6   Sociodemographic Characteristics....................................................................7   Clinical Characteristics......................................................................................8   E. Longitudinal Change in Depressive Symptoms in Older Adults.......................9   Community-Based Older Adults........................................................................9   Hospitalized Older Adults..................................................................................9   Factors Associated with Change in Depressive Symptoms............................10   F. Section 1 Summary........................................................................................10   Section 2: Activities of Daily Living (ADL)..............................................................11   A. Construct Definition........................................................................................11   B. Measurement of ADL in Epidemiological Studies...........................................11  

ix

Katz ADL Scale...............................................................................................11   Bias in Self-Report Measurement...................................................................12   C. Prevalence of ADL disability..........................................................................12   Community-Dwelling Adults............................................................................12   Hospitalized Older Adult.................................................................................14   D. Factors Associated with ADL Status..............................................................17   Sociodemographic Characteristics..................................................................17   Clinical Characteristics....................................................................................19   E. Longitudinal Change in ADL Disability in Older Adults...................................19   Community-Dwelling Older Adults..................................................................20   Hospitalized, Older Adults...............................................................................20   F. Summary of Section II....................................................................................21   Section 3: The Association between ADL and Depressive Symptoms...................22   A. International Classification of Functioning, Disability and Health (ICF)..........22   B. Direct Pathways of Association......................................................................22   Associations in Prevalence and Incidence......................................................23   Association by Severity of Symptoms.............................................................23   Association in Recovery from Disability..........................................................24   Influence of Contextual Factors......................................................................24   C. Indirect Pathways of Association...................................................................25   Evidence of Mediator Pathways......................................................................25   Potential Interaction Effects............................................................................25   D. Summary of Section 3....................................................................................27   Chapter 3: Methods...................................................................................................28   A. Specific Aims.....................................................................................................28   Specific Aim 1.....................................................................................................28   Representative Hypotheses:...........................................................................28   Specific Aim 2.....................................................................................................28   Representative Hypotheses:...........................................................................28   Specific Aim 3.....................................................................................................29   Representative Hypotheses:...........................................................................29  

x

Specific Aim 4.....................................................................................................29   Representative Hypothesis:............................................................................29   B. Conceptual Model..............................................................................................29   C. Design and Sampling........................................................................................31   Setting................................................................................................................31   Recruitment and Screening Procedure...............................................................31   Inclusion Criteria.................................................................................................32   Exclusion Criteria................................................................................................32   Informed Consent...............................................................................................32   Participant Selection...........................................................................................33   Ethical Considerations........................................................................................34   In-Hospital Assessments....................................................................................34   Follow up Interviews...........................................................................................35   D. Measures...........................................................................................................35   E. Data Collection..................................................................................................36   F. Data Analysis.....................................................................................................37   Specific Aim 1 Data Analysis..............................................................................37   Specific Aim 2 Data Analysis..............................................................................38   Specific Aim 3 Data Analysis..............................................................................38   Specific Aim 4 Data Analysis..............................................................................38   Chapter 4: Results.....................................................................................................40   Specific Aim 1........................................................................................................40   A. Overview of Hospital Depressive Symptoms..................................................40   Distribution of CESD Scores...........................................................................40   Prevalence of High Depressive Symptoms.....................................................41   Factors Associated with Depressive Symptoms: Bivariate.............................41   Factors Associated with High Depressive Symptoms: Multivariate.................46   B. Overview of Three-Month Follow-up Depressive Symptoms..........................48   Distribution of CESD Scores...........................................................................48   Prevalence of High Depressive Symptoms.....................................................48   Factors Associated with Depressive Symptoms: Bivariate.............................49  

xi

Factors Associated with High Depressive Symptoms: Multivariate.................53   C. Overview of the Trajectory of Depressive Symptoms....................................55   Change in Depressive Symptoms: Categorical...............................................55   Change in Depressive Symptoms: Continuous CESD score..........................55   Factors Associated with Change in CESD Score: Bivariate............................56   Factors Associated with Change in CESD Score: Multivariate.......................58   Specific Aim 2........................................................................................................63   A. Overview of Hospital ADL Status...................................................................63   Prevalence of ADL Disability...........................................................................63   Prevalence of ADL Disability by ADL Category...............................................63   Factors associated with ADL Status: Bivariate................................................64   Factors Associated with ADL Status: Multivariate...........................................65   B. Overview of Three-Month Follow-up ADL Status...........................................67   Prevalence of ADL Disability...........................................................................67   Prevalence of ADL Disability by ADL Category...............................................67   Factors Associated with ADL Status: Bivariate...............................................68   Factors Associated with ADL Disability: Multivariate.......................................69   C. Overview of the Trajectory of ADL.................................................................71   Change in Prevalence of ADL Disability.........................................................71   Factors Associated with Change in ADL Score: Bivariate...............................73   Factors Associated with Change in ADL Score: Multivariate..........................76   Specific Aim 3........................................................................................................78   A. Association between Depressive Symptoms and ADL Function: Bivariate....78   Hospital Depressive Symptoms and Hospital ADL Status..............................78   Hospital Depressive Symptoms and 3 Month Follow-up ADL Status..............79   B. Association between Depressive Symptoms and ADL Status: Multivariate...80   Background for analysis..................................................................................80   Linear Regression Model using the Continuous CESD Score........................81   Linear Regression Model using Categorical CESD Scores............................83   Comparison of Fit between the Two Models...................................................85   Specific Aim 4........................................................................................................87  

xii

A. Linear Regression Models..............................................................................87   Chapter 5: Discussion................................................................................................88   A. Purpos...............................................................................................................88   B. Specific Aim 1 Discussion..................................................................................88   Prevalence of High Depressive Symptoms: Hospital..........................................88   Factors Associated with High Depressive Symptoms.........................................89   Trajectory of Depressive Symptoms...................................................................89   Factors Associated with Positive Change in Depressive Symptoms..................90   C. Specific Aim 2 Discussion.................................................................................90   Prevalence of ADL Disability: Hospital..............................................................90   Factors Associated with ADL Status...................................................................91   Trajectory of ADL Disability................................................................................92   Factors Associated with Change in ADL............................................................92   D. Specific Aim 3 Discussion.................................................................................93   Association between Depressive Symptoms and ADL Status in Hospital..........93   Association between Hospital Depressive Symptoms and ADL Post-Discharge 93   Association between Hospital Depressive Symptoms and Subcategories of ADL Post-Discharge...................................................................................................94   E. Specific Aim 4 Discussion..................................................................................95   Moderator Analysis............................................................................................95   Gender............................................................................................................95   Marital Status..................................................................................................95   Pain.................................................................................................................96   Medical Conditions..........................................................................................96   F. Summary...........................................................................................................97   G. Study Strengths.................................................................................................97   H. Study Limitations...............................................................................................97   I. Future Directions for Research...........................................................................98   Appendix A..............................................................................................................100   Appendix B..............................................................................................................101   Appendix C..............................................................................................................103  

xiii

References..............................................................................................................104   Biosketch.................................................................................................................118  

xiv

List of Tables

Table 1. Data Type, Source and Operational Definition............................................................35

Table 2. Data Collection Time Points for Study Variables.........................................................37

Table 3. Sociodemographic characteristics of sample by depressive symptoms during hospitalization.............................................................................................................................43

Table 4. Clinical characteristics of sample by depressive symptoms during hospitalization.....45

Table 5. Modified Poisson regression models assessing sociodemographic and clinical characteristics associated with risk of having high depressive symptoms in hospital.................47

Table 6. Sociodemographic characteristics of sample by high and low depressive symptoms at the 3 month follow-up .................................................................................................................50

Table 7. Clinical characteristics of sample by low and high depressive symptoms at the 3 month follow-up...........................................................................................................................52

Table 8. Linear regression model using a negative binomial distribution to assess sociodemographic and clinical variables associated with follow-up CESD score.......................54

Table 9. Change in depressive symptoms at the 3 month follow-up.........................................55

Table 10. Sociodemographic and clinical characteristics associated with change in CESD.....57

Table 11. Linear regression model assessing sociodemographic and clinical variables that predict change in CESD .............................................................................................................60

Table 12. Logistic regression models assessing sociodemographic and clinical characteristics associated with having a positive change in CESD score (v. negative change).........................62

Table 13. Sociodemographic and clinical variables associated with hospital ADL status.........65

Table 14. Logistic regression models assessing sociodemographic and clinical characteristics associated with having at least one ADL limitation in hospital (vs. no limitations)......................66

Table 15. Sociodemographic and clinical characteristics associated with follow-up ADL.........69

Table 16. Logistic regression models assessing sociodemographic and clinical characteristics associated with the risk of having at least one ADL limitation at the 3 month follow-up.............70

Table 17. Course of ADL status from hospital to 3 month follow-up.........................................71 

xv

Table 18. Sociodemographic and clinical characteristics associated with change in ADL........75

Table 19. Linear regression model assessing sociodemographic and clinical variables that predict change in ADL.................................................................................................................77

Table 20. Associations between categories of 3 month follow-up ADL and hospital depressive symptoms ...................................................................................................................................80

Table 21. Linear regression models assessing the association between follow-up ADL status and hospital depression controlling for sociodemographic and clinical variables.......................82

Table 22. Regression models with linear contrast statement assessing the association between follow-up ADL status and categorical hospital CESD scores controlling for sociodemographic and clinical variables...................................................................................................................84

Table 23. Odds ratio and confidence intervals estimating the linear association between follow- up ADL and hospital CESD scores.............................................................................................85

Table 24. Linear regression model results assessing interaction effects with hospital depressive symptoms on the depression-ADL association ..........................................................................87 

xvi

List of Figures

Figure 1. Prevalence of high depressive symptoms in community samples of older adults by country or region of the United States..........................................................................................6

Figure 2. Prevalence of chronic disability by number of ADL limitations, adults aged 65+ (1982- 2005)...........................................................................................................................................13

Figure 3. Comparison of percent reporting ADL disability by age group using NHANES data (1988-2004)................................................................................................................................13

Figure 4. Percent reporting incident ADL disability during hospitalization.................................14

Figure 5. Modified Cascade to Dependency.............................................................................15

Figure 6. Prevalence of disability in five age groups of community-dwelling adults aged 60+ years in Hong Kong (1996-2004)................................................................................................17

Figure 7. Twelve month outcomes in ADL and mortality by ADL discharge status...................20

Figure 8. ICF model illustrating direct and indirect pathways of association between depression and ADL and the potential moderating effect of contextual factors............................................23

Figure 9. Conceptual model (ICF) postulating how depressive symptoms may be associated with disability and the potential influence of contextual factors...................................................30

Figure 10. Flow Diagram of Patient Recruitment for the ACE Unit Study.................................32

Figure 11. Flow chart of ACE unit study participants................................................................34

Figure 12. Frequency distribution of hospital CESD scores .....................................................40 

Figure 13. Prevalence of high depressive symptoms in hospital using two different CESD cut- off points.....................................................................................................................................41 

xvii

Figure 14. Frequency distribution of 3 month follow-up CESD scores......................................48 

Figure 15. Prevalence of depression at 3 month follow-up using two different CESD cut-off points..........................................................................................................................................49 

Figure 16. Frequency distribution of the CESD change score..................................................56 

Figure 17. Scatter plot illustrating the relationship between change in hospital CESD score and change in ADL status .................................................................................................................58 

Figure 18. Percent reporting ADL limitations in hospital ..........................................................63 

Figure 19. Prevalence of ADL disability by ADL category.........................................................64 

Figure 20. Percent reporting ADL limitations at 3 month follow-up...........................................67 

Figure 21. Prevalence of ADL disability by ADL category.........................................................68 

Figure 22. Prevalence of ADL disability across 3 time points...................................................72 

Figure 23. Percent reporting ADL disability by ADL category at three time points ...................73 

Figure 24. Hospital ADL status by hospital CESD scores.........................................................78 

Figure 25. Follow-up ADL status by hospital CESD scores......................................................79 

1

Chapter 1: Introduction SPECIFIC AIMS Depressive symptoms and disability in activities of daily living (ADL) often increase in older adults during hospitalization and for many persist post-discharge. However, little is known about the psychological and functional response of older adults admitted to an Acute Care for Elders (ACE) unit. Questions remain about the association between depressive symptoms and ADL disability and factors that moderate these associations are unknown. Objectives of this study were to investigate: 1) change in depressive symptoms and ADL function from hospital to 3 month follow-up; 2) the association between depressive symptoms in hospital and ADL function 3 months post-discharge; and 3) moderators of the depression-ADL association in a tri-ethnic sample of older adults. Specific aims of this study were to:

1. Determine the trajectory of depressive symptoms from hospitalization to 3 months post discharge. Prevalence estimates in hospital and post-discharge, as well as change in depressive symptoms, will be explored by relevant sociodemographic and clinical characteristics such as age, gender, ethnicity, pain and ADL function. 2. Determine the trajectory of ADL function from hospitalization to 3 months post discharge. Prevalence estimates in hospital and post-discharge, as well as change in ADL function, will be explored by relevant sociodemographic and clinical characteristics such as age, gender, ethnicity, pain and depressive symptoms. 3. Examine the direct associations between hospital depressive symptoms and ADL function 3 months post-discharge, controlling for relevant sociodemographic and clinical variables such as age, gender and pain. 4. Examine the interaction between hospital depressive symptoms and select personal and health characteristics on ADL status 3 months post-discharge. Personal and health characteristics will include measures such as gender, marital status and pain.

2

Data are from a sample of 306 white, black and Hispanic older adults admitted to the Acute Care for Elders (ACE) unit at the University of Texas Medical Branch (UTMB) from 2005-2007 (ACE unit defined: Chapter 2: Section II.C.2.c). Data were collected face-to-face within 24 hours of admission and by telephone interview 3 months post discharge. The UTMB ACE unit specializes in helping older adults recover from acute medical events; it presents unique opportunities for research related to the hospitalization of older adults.

SIGNIFICANCE OF RESEARCH An ACE unit is uniquely structured to promote positive experiences for hospitalized older adults, yet little research is available that defines the outcomes for patients served. Through a broad and longitudinal examination of depressive symptoms and ADL status, we hope to better understand potential psychological and functional benefits of an ACE unit admission. Second, while research describing an association between depressive symptoms and ADL status is available, the review provided in this work brings organization to existing studies within the framework of the International Classification of Function. 1 (Chapter 2: Section 3). Finally, our exploration of novel longitudinal questions and interaction effects will contribute depth to our understanding of the depression-ADL association for researchers and clinicians working with hospitalized older adults.

OUTLINE OF THE DISSERTATION Chapter 1 provides a brief overview of the study, including the specific aims, study population, design and the contributions of this study to understanding issues with hospitalized older adults. Chapter 2 provides background literature on depressive symptoms, ADL and the depression- ADL association. Chapter 3 describes the study design, conceptual model, measures and data analyses. Chapter 4 contains the results for each specific aim via text, tables and figures. Finally, Chapter 5 interprets the study results within the context of current literature, as well as describes study strengths and limitations. References and the author vitae conclude this work.

3

Chapter 2: Background

This chapter provides rationale for our study methodology and context for interpreting the study results. A review of depression and depressive symptoms is provided first by defining constructs, risk factors and course of illness for older adults. Next, an overview of ADL disability is presented, including estimates of disability prevalence, risk factors and course of change in ADL status for older adults admitted for hospitalization. Finally, a summary of the association between depressive symptoms and ADL disability is provided, with background supporting potential moderators of this association. SECTION 1: DEPRESSIVE SYMPTOMS Section 1 provides a broad background for depressive symptoms in older adults. Additional review information is provided in Appendices A-C. A. Construct Definitions A.1. Major Depression Major depression is a diagnosis provided by a psychologist or psychiatrist using criteria listed in the Diagnostic and Statistical Manual, now in its 4 th edition. (DSM-IV). 2

To meet criteria for major depression, a patient must have 1-2 core symptoms (depressed mood and lack of interest) along with 4 or more of the following symptoms for at least 2 weeks: 1) feelings of worthlessness or inappropriate guilt, 2) diminished ability to concentrate or make decisions, 3) fatigue, 4) psychomotor agitation or retardation, 5) insomnia or hypersomnia, 6) significant decrease or increase in weight or appetite and 7) recurrent thoughts of death or suicide. The prevalence of major depression is estimated to range from 1-4% of the general population. 3 In older adults, major depression has been associated with negative health care outcomes such as poorer recovery from illness, increased utilization of health services and mortality. 3, 4,5,6,7-10

A.2. Minor Depression Minor depression, also called subsyndromal or subthreshold depression, is also diagnosed using the DSM-IV when one or more of the core symptoms and 1-3 of the

4

additional symptoms listed above are present. Using DSM-IV criteria, minor depression is estimated to be as high as 9% in the general population. 11

A.3. Depressive Symptoms Depression is also quantified by the number and frequency of depressive symptoms reported during examination. Using established cut-off points, the person is categorized as having “high” or “low” depressive symptoms. Both minor depression and high depressive symptoms have been termed “clinically-significant” or “clinically-relevant” depression, as both are associated with increased risk for diminished functional outcomes and major depression episodes. 12-15

B. Measurement of Depressive Symptoms in Older Adults A common method for assessing depressive symptoms in epidemiologic studies is through the use of screening tools or depression rating scales. These assessments are performed by licensed practitioners or lay interviewers and are based on examiner report or self-report of depressive symptoms. Appendix A is a table of the most common screening tools with information on purpose, cut-off points for case definition and established sensitivity and specificity. In brief, the Hamilton Rating Scale uses DSM-IV criteria to rate depressive symptoms; cut-off points are provided for clinically significant and severe depression. 16 Sensitivity and specificity have also been studied by neurological categories. 17 The Geriatric Depression Scale (GDS) is widely used in older adult depression research due to its ease of administration and self-report format. 18

The GDS has high sensitivity and specificity when used in hospitalized older adults. The Beck Depression Inventory (BDI) measures the severity of cognitive, affective and somatic symptoms of depression through patient self-report. 19 A broad range of cut-off scores have been published for this measure and specificity and sensitivity information has been established for use with the Psychological Subscale. The Center for Epidemiological Studies-Depression scale, the measure used in our study, is reviewed in greater detail in the next section. B.1. Center for Epidemiologic Studies-Depression Scale (CESD) The CESD is a measure of the severity of depressive symptoms in clinical and community epidemiological samples. 20 This 20 item self-report measures four

5

constructs related to depression: 1) positive affect, 2) negative affect, 3) interpersonal symptoms and 4) somatic symptoms. Numerous published studies have used a cut-off point of ≥ 16 to classify a patient with clinically significant or clinically-relevant depressive symptoms. Minor depression has been operationally defined as a CESD score of 11-15. 3

Beekman et al. have suggested a higher cut-off score of ≥ 20 in hospitalized older adult populations. 21 One rationalization for using the higher cut-off score is the belief that somatic symptoms measured by the CESD may be confounded by other medical conditions with similar symptoms. For example, the CESD measures difficulty with sleep, a side effect of numerous medical conditions. Other researchers dispute the validity of minimizing the impact of somatic complaints, suggesting that body symptoms are just as reliable in classifying depression as cognitive/interpersonal symptoms. 22

Regardless, both cut-off points demonstrate acceptable sensitivity and specificity for hospitalized older adults. Taken together, a number of tools are used to assess depression in hospitalized older adults. Case definition cut-off points are available for classifying high depressive symptoms; sensitivity and specificity scores are available by population sampled, medical condition or diagnostic cut-off score. B.2. Clarification of Study Terminology Patients in this study with a CESD score of ≥ 16 are defined as having high depressive symptoms; low depressive symptoms are quantified by a CESD score of <16. Terms such as clinically-relevant or clinically-significant depression are used interchangeably with high depressive symptoms.

C. Prevalence of High Depressive Symptoms in Older Adults C.1. Community-Dwelling Older Adults High depressive symptoms are estimated to range from 8- 16%, 3 but have been as high as > 37% in samples of older adults that live in community settings. 23 Figure 1 illustrates the prevalence of high depressive symptoms (CESD ≥ 16) in community- based samples of older adults by country or region of the United States. 23-27 Older

6

adults with high depressive symptoms are less likely to participate in research, a thwart to accurate measurement. 28

Figure 1. Prevalence of high depressive symptoms in community samples of older adults by country or region. Source: Data from Herrman et.al 2002.

C.2. Hospitalized Older Adults In a review of the literature over the last 15 years, we found 12 studies that assessed the prevalence of high depressive symptoms in hospitalized older adults (Appendix B). Clinically-significant depression ranged from 3-51%. 9, 29-39 Only one study occurred on an ACE unit with a specific population of people with cancer. 29 The wide range in prevalence estimates is likely due to the use of different assessments or different cut-off points used with similar assessments.

D. Factors Associated with Depressive Symptoms To justify the covariates selected in the statistical models of this study, a brief review of select literature on factors associated with depressive symptoms was conducted. Initially, within this review, studies were separated by community-dwelling and hospitalized samples. However, after review, the correlates of depressive symptoms are remarkably similar, regardless of whether the sample was community or hospital-based; so, while differences cannot be emphasized, when helpful, sample differences will be mentioned.

7

D.1. Sociodemographic Characteristics D.1.a. Gender Gender is associated with depressive symptoms in studies with community and hospital-based samples. Females are 1-3x more likely to be depressed than men. 3, 11, 40- 42 Disability, higher in females due to chronic conditions and longer life span, is thought to contribute to this increased risk. 3

D.1.b. Age Age has been studied extensively as a risk factor for high depressive symptoms due to the belief that depressive symptoms increase with age. In hospital studies, when patients are analyzed by age, younger patients were more likely to be depressed than older ones, when factors such as gender and disability are controlled for. 3,34, 37, 43-45

D.1.c. Ethnicity Study results vary when analyzing the association between ethnicity and depressive symptoms. Blazer et al. 46 found black race to be protective against high depressive symptoms (compared to whites), but other researchers find no significant difference in depressive symptoms between blacks and whites. 3, 25, 34, 40, 44 Huisaini reported that any symptom differences between blacks and whites were due to lower levels of social support, general stress and or more medical issues. 47 In studies comparing community- dwelling Hispanics to whites, Hispanics were more likely than whites to have high depressive symptoms, especially those who were less acculturated. 43, 48-50

D.1.d. Marital Status Marital status is associated with depressive symptoms. Specifically, being unmarried increases the risk of high depressive symptoms in community and hospital-based samples. 11, 40, 43, 44, 51-53 Living alone, loneliness and low social support also increase the risk of depression in community-dwelling and hospitalized older adults. 34-36, 42-45, 54 In a study of adults ≥ 75 years in Finland, those who reported being lonely "often or always" were at 9x the risk for being considered depressed compared to those reporting lower levels of loneliness. 26

8

D.1.e. Education Education is also associated with depressive symptoms. In particular, higher education has been associated with decreased risk of depression in community and hospital samples, but this association may vary in its effect by age. 40, 41, 43, 46, 48, 52, 53, 55, 56 In a longitudinal sample of > 33,000 adults in Norway, the protective effect of higher education decreased with increasing age up to 64 years; then, from ages 65-74 years, the protective effect of higher education increased. 57 On the other hand, higher education has been associated with persistent depression three months post- hospitalization in adults. 58

Considering the sociodemographic characteristics discussed, being female, under 65 years, unmarried, of Hispanic race or having a low educational status are risk factors associated with clinically-significant depression. Therefore, stratification of these variables will be used in our predictor models to delineate risks or protective factors.

D.2. Clinical Characteristics D.2.a. Medical Conditions The presence or history of certain diagnostic categories such as digestive, neurologic and cardiac disorders has been associated with increased risk of high depressive symptoms (OR: 1.79, 1.65 and 1.82 respectively). 51 History of stroke, diabetes, arthritis, immune disorders, chronic obstructive pulmonary disease or any chronic illness is also associated with increased risk of depressive symptoms in community-dwelling older adults. 26, 44, 54, 59, 60 Additionally, history of previous myocardial infarction is associated with high depressive symptoms in hospitalized elders. 34

D.2.b. Pain The prevalence of chronic pain has been reported to be as high as 80% in hospitalized or institutionalized older adults and as high as 50% in community-dwelling older adults. 61, 62 The presence of pain is associated with high depressive symptoms in hospitalized 60 and non-hospitalized samples. 63 Risk of clinically-significant depression is

9

also more likely in community-dwelling adults 85 years and older who use analgesics (OR: 2.7; 95% CI: 1.7-4.4) 54

D.2.c. Body Mass Index (BMI) Being either overweight or underweight has been studied as both a cause and outcome of depression. 64, 65 In one study of hospitalized older adults in Japan, BMI was not significantly associated with high depressive symptoms, 32 but other studies of both hospitalized and non-hospitalized samples have suggested a u-shaped relationship, in which those who were underweight or overweight were at greater risk for high depressive symptoms than those at normal weight. 65

Taken as a whole, the presence of specific medical conditions and pain, as well as BMI above or below normal, are potential risk factors for high depressive symptoms and thus will be analyzed in our predictor models. E. Longitudinal Change in Depressive Symptoms in Older Adults E.1. Community-Based Older Adults Recovery of depressive symptoms in community samples is reported by percent of recovery and percent with ongoing depressive symptoms. Recovery occurs for 12-73%, intermittent reoccurrence for 44-84%, while 2-49% have chronic symptoms. 66

26, 67-69

Most improvement occurs within 6 months and people with minor depression generally have more positive outcomes; they were also at greater risk for other psychiatric diagnoses. 68, 69 Overall, most people experience a chronic fluctuation of depressive symptoms rather than a full recovery. 67-69

E.2. Hospitalized Older Adults In hospitalized older adults, change is assessed by prevalence differences at admission and follow-up. Studies in post stroke populations show significant increases in depressive symptoms across follow-up periods after in-patient rehabilitation. 70, 71 By contrast, high depressive symptoms decreased significantly over time in patients admitted to non-rehabilitation units. Initial prevalence rates of 28-34% drop to 17% by 2- 4 weeks post discharge. 30, 34

10

Taken together, results of studies using prevalence estimates of depression indicate high percentages of recovery. These studies tend to provide dichotomous results (depressed versus not depressed) captured at one time point during follow-up. In studies assessing change in depression by percentage of remission, results indicate a fluctuating or chronic course of depression captured by measuring depression at ≥ 2 time points at follow-up. Appendix C provides a table which outlines studies assessing change in depressive symptoms in hospitalized older adults over the last 15 years. E.3. Factors Associated with Change in Depressive Symptoms Factors that contribute to recovery from high depressive symptoms are important to understand. Some physicians argue that hospitalization is the primary predictor of depressive symptoms in hospitalized, older adults and that, for the majority, symptoms resolve after discharge home. 72 According to this hypothesis, other factors, such as sociodemographic characteristics, ADL status or depression intervention, are less relevant for recovery. In this review, no other studies could be found that assessed factors associated with change in high depressive symptoms. However, Koenig et al. identified factors associated with improved time to remission of major and minor depression, finding that less severe depression, less severe medical illness, less intense or no previous use of anti-depressants, black race and higher social support all contributed significantly to shorter remission time. 73

Full document contains 136 pages
Abstract: Depressive symptoms and disability in activities of daily living (ADL) often increase in older adults during hospitalization and for many persist post-discharge. However, little is known about the psychological and functional response of older adults admitted to an Acute Care for Elders (ACE) unit. Questions remain about the association between depressive symptoms and ADL disability and factors that moderate these associations are unknown. Objectives of this study were to investigate: (1) change in depressive symptoms and ADL function from hospital to 3 month follow-up; (2) the association between depressive symptoms in hospital and ADL function 3 months postdischarge; and (3) moderators of the depression-ADL association. A tri-ethnic (white, black and Hispanic) sample of 403 older adults within an ACE database contributed subjective and objective information related to depressive symptoms, clinical variables and activity/participation measures across two time frames, admission and three months post discharge. A large minority reported high depressive symptoms in hospital and over half reported ADL disability. Across both assessment periods, risk factors for having high depressive symptoms were being unmarried and having any level of ADL disability. Conversely, risk factors for ADL disability were pain and depressive symptoms. At 3 months post discharge, the recovery rate from depression and incident ADL disability was high. Positive change in depression was significantly associated with positive change in ADL status. Increasing severity of hospital depression was associated with increased odds of ADL disability at the 3 month follow-up. Neither gender, marital status, pain nor medical history moderated this relationship. This study indicates that while older adults experience higher depressive symptoms and ADL disability while hospitalized, resolution of symptoms occur for many. This research contributes to the literature by extending our knowledge of the course and associations between depressive symptoms and ADL disability in hospitalized, older adults. Future research which focuses on interventions to minimize depressive symptoms and ADL disability is warranted.