| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||||||||||||||||||||
a University of Maryland School of Social Work, Baltimore
Correspondence: Kathryn Betts Adams, PhD, LCSW-C, LCSW-C, 13000 Cleveland Drive, Rockville, MD 20850. E-mail: MDAKBA{at}aol.com.
Decision Editor: Laurence G. Branch, PhD
 |
Abstract |
|---|
 TOP Abstract When Is Geriatric "Depression"...MethodResultsDiscussionReferences |
|---|
Key Words: Depletion • Depression • Disengagement theory • the Geriatric Depression Scale • Socioemotional selectivity
This article links two phenomena: One, developmental changes suggested by the disengagement theory of aging (Cumming and Henry 1961
), the early and influential theory that stated there is a mutual social and affective withdrawal between the older adult and his or her social environment, and by disengagement's newer theoretical "cousins," socioemotional selectivity theory (Frederickson and Carstensen 1990; Carstensen 1992) and gerotranscendence (Tornstam 1989, Tornstam 1997, Tornstam 2000), which also describe the narrowing of the older person's social world and decreasing investment in activities that were important in younger years. The second phenomenon is a postulated subtype of geriatric depression that has been termed depletion, represented by an aggregation of items in measurement studies using well-known depression screening scales (Newmann, Engel, and Jensen 1991; Gallo, Anthony, and Muthen 1994). The current study looks specifically at a group of items on the Geriatric Depression Scale (GDS; Brink, Yesavage, Lum, Heersema, Adey, and Rose 1982) that compose a dimension of Withdrawal/Apathy and Lack of Vigor (WAV; Parmalee, Lawton, and Katz 1989), and how this dimension behaves in relation to other dimensions of the GDS and to health and demographic measures in a convenience sample of community-dwelling elderly adults. The goals of the study include confirmation of a WAV subscale of the GDS and exploration of the behavior of the subscale in light of disengagement-related and depletion theories.
|
When Is Geriatric "Depression" Not Depression? |
|---|
|
TOPAbstractWhen Is Geriatric "Depression"...MethodResultsDiscussionReferences |
|---|
; Lewinsohn, Rohde, Seeley, and Fischer 1991; Fogel and Fretwell 1985). Overdiagnosis of depression in older persons may lead to the prescription of unnecessary or inappropriate medications, labeling individuals with mental health diagnoses unnecessarily, wasting scarce and costly health and mental health resources, and disseminating inaccurate research results. In a controversial article in The Atlantic Monthly titled "Overselling Depression to the Old Folks," the late psychotherapist Stanley Jacobson, then age 70, asserted that mental health professionals have overemphasized the diagnosis of depression in older adults.Health professionals want to reduce the struggle of the old to an illness. They label it "depression," search for biological explanations of the "disease," and call the psychological, spiritual, and social aspects of the conflict mere "risk factors." Unfortunately, by colluding in the denial that the fact of our mortality is significant to our mental health in late life, they contribute more to the prevalence of depression than to its cure. (Jacobson 1995
Among the issues Jacobson referred to in his critique is the challenge of differentiating somatic symptoms that are due to physical illness or aging from those that represent depression. This difficulty means that an item on a depression screening scale may represent a symptom of depression, but it may alternatively be a "symptom" of old age. For example, a study of over 4500 adults concluded that a number of health and activity variables, including cognitive dysfunction, poorer health, reduced independent living skills, and less engagement in pleasant activities, were associated with both old age and depression (Lewinsohn et al. 1991). In addition, sleep disturbances and poor appetite are quite common in older people and have been found to be poor discriminators for clinical depression in this group (Dorfman, Lubben, Mayer-Oakes, et al. 1995).
Lewinsohn and colleagues 1991 have also noted that certain symptoms of depression, such as lack of social interest and greater self-involvement, mirror attributes of "normal" older adults according to disengagement theory (Cumming and Henry 1961), the controversial theory that in its most well-known form was presented as a comprehensive theory of normal aging, and, by implication, a model of "successful" aging. Disengagement theory has been out of favor with gerontologists for over 3 decades (Achenbaum and Bengtson 1994; Marshall 1994), although a number of writers have acknowledged that disengagement seems to occur in some older persons (Neugarten 1968; Youmans 1969; Cath 1975; Hochschild 1975; Steinkamp and Kelly 1987). More recently, Johnson and Barer 1992 reported finding clear signs of disengagement in approximately 50% of their sample of community-dwelling adults aged 85 years and older. Two newer theories, socio-emotional selectivity and gerotranscendence, each also assert that losing interest in some activities, not wanting to go out, and the like, may be a part of normal aging, reflecting a natural increased comfort with being alone and having a slower pace of life. Socio-emotional selectivity posits that the older adult becomes gradually less interested in forging new relationships as the focus turns toward conservation of energy for prioritized activities and comfortable, reliable relationships (Carstensen 1992), and has found support in recent research (e.g., Potts 1997). In gerotranscendence, the disengagement that occurs is interpreted as a transcendence of physical frailties and materialist concerns; being alone with time to contemplate is found to be more appealing to the older person (Tornstam 2000).
Another reason assessment of late-life depression is so complex lies in the difference between a clinical diagnostic method of identification of depression, requiring a certain duration of symptoms, level of functional impairment, and combination of symptoms to be classified as depressed, and a measurement approach, for example, use of a depression screening scale based on endorsement of symptoms that are added up to reach a depression score. Because they are relatively inexpensive and nonintrusive means to identify elders who may be suffering from depression, brief screening instruments for geriatric depression have received a great deal of attention in the medical and mental health literature and continue to gain in popularity (Blazer 1994; Brink et al. 1982; Burke, Nitcher, Roccaforte, and Wengel 1992; Koenig, Meador, Cohen, and Blazer 1992). Focusing on this issue, Newmann and associates (Newmann et al. 1991; Newmann, Klein, Jensen, and Essex 1996) found that diagnoses often differ between a clinical diagnostic approach and a measurement approach. The authors speculated that this may explain why the prevalence of major depressive disorder is found to be lower among older persons than for younger adults, whereas scores on standard screening scales of depressive symptomatology are often higher for older persons (Newmann et al. 1996).
Newmann and her colleagues (1996) went on to identify two distinct symptom clusters in the elderly—those that strictly adhere to clinical diagnostic criteria, which they term "depression syndrome," and a unique cluster of symptoms normally associated with depression but without some of the criteria necessary for a depression diagnosis by the usual standards, which they call "depletion syndrome." Using factor analyses to identify loadings of particular symptoms onto the two syndromes, Newmann and colleagues 1996 reported that some symptoms are unique or nearly unique to each syndrome. Whereas both syndromes share the symptoms of low energy, feeling things are an effort, feeling blue, and trouble falling asleep, the depression syndrome includes feelings of guilt, worthlessness, and self-blame and crying easily. Loneliness, lack of interest, and lack of appetite load more strongly on the depletion syndrome. The depletion syndrome increased linearly with age in this study, whereas the depression syndrome did not.
In another study, Gallo and colleagues 1994 found that older adults were less likely than younger adults to endorse items about feeling sad, blue, or, specifically, depressed, in the Diagnostic Interview Schedule (DIS) based on the Diagnostic and Statistical Manual of Mental Disorders, DSM; 3rd ed., rev. (American Psychiatric Association 1987) criteria for major depression, when they nevertheless endorsed other symptoms such as sleep difficulty, lack of hope, and thoughts of death. These authors concluded that this "depression without sadness," or "depletion syndrome" appears to fit a proportion of their sample (Gallo et al. 1994.) It is interesting that the same term was used 20 years earlier in a critique of disengagement theory, in which Cath 1975 discussed the apparent similarity of disengagement to depression in older persons and described a kind of combination of the two by the term, "normal depletion ... something which will occur if one but lives long enough" (p. 212). Fogel and Fretwell 1985 posited the same term, "depletion syndrome," to describe the depression of medically ill elderly persons whose symptoms do not fit neatly into the usual symptom clusters for younger persons. The empirical difference in symptoms noted by all of these researchers raises the question: Are the majority of these older adults who endorse some of the purported depressive symptoms other than sadness actually depressed, or is there a different category into which we should be placing them?
In light of the convergence of several theories of aging with a distinct response pattern found on depression screening scales described in the literature, the present study examined a group of "symptoms" from the GDS that may be normal for older adults. The items in question comprise two dimensions, Withdrawal/Apathy and Vigor, identified in a principal components analysis (PCA) of the scale (Parmalee, Lawton, and Katz 1989). Three of these items were also featured in another study using the short form of the GDS (GDS–S; Sheikh and Yesavage 1986). These withdrawal items—low energy, lack of interest in going out, dropping activities and interests—had a relatively lower positive correlation with the other two factors in the shortened scale—Dysphoria and Life Satisfaction—than they did with one another (Mitchell, Mathews, and Yesavage 1993). The authors interpreted this lower correlation between factors to be indicative of clearly distinct dimensions of depression for older adults (Mitchell et al. 1993).
The current study sought first to replicate the existence of a WAV dimension in the GDS in an independent sample of community-dwelling elders. Then, the endorsement rates and relationship of GDS factor scores to one another and to selected demographic and health characteristics were examined. Because the WAV items were hypothesized to reflect developmental change associated with age, their endorsement rates in a relatively well, high-functioning group of older adults were expected to be higher than those of the Dysphoria items. It was also anticipated that the WAV score would have a higher direct correlation with age than would the Dysphoria score from the GDS. We further hypothesized that WAV items contribute disproportionately to the identification of depression in this sample when using the score of 11 or above as a cut-off. A final goal for this study was to explore the WAV items in light of the above theories of aging and depletion. We reasoned that if WAV represents a depletion syndrome, then its score should be associated with health conditions and functional impairment as well as age. If WAV represents socio-emotional selectivity and gerotranscendence, then one would expect less association with dysphoria, anxiety, or stress, as well as positive association with age.
|
Method |
|---|
|
TOPAbstractWhen Is Geriatric "Depression"...MethodResultsDiscussionReferences |
|---|
Measures
The questionnaire for the survey was developed from a selection of scaled measures with the addition of demographic and health questions written specifically for the study. The final mailing included the questionnaire booklet; cover letters from the University of Maryland and from the HMO's Department of Psychiatry; a stamped return envelope; and a "decline" postcard, also postage paid, to be returned if the respondent did not wish to participate in the study. The cover letter offered completers of the questionnaire $5.00 cash on receipt of their materials. Unfortunately, no follow-up mailings could be sent. Some of the scaled measures included in the questionnaire were as follows:
The GDS.
The GDS (Brink et al. 1982), a 30-item self-rating scale, was developed specifically to screen older individuals for depressive symptoms and to correct some of the problems that researchers working with older persons had found with the existing depression scales up to that point (Yesavage, Brink, Rose, et al. 1983). For example, some of the well-known depression screening instruments, notably the Zung Self-Rating Depression Scale (SDS; Zung 1965) and the Minnesota Multiphasic Personality Inventory Depression Scale, tend to overdiagnose depression with elderly respondents because they include many questions about physical symptoms (Blazer and Williams 1980; Brink et al. 1982). Thus, the GDS was written with much less emphasis on somatic complaints. In addition, most of the existing scales, including the Center for Epidemiologic Studies Depression Scale (Radloff 1977), the Beck Depression Inventory (Beck, Ward, Mendelson, Mock, and Erbaugh 1961) and the Zung SDS, use a Likert response system with three or more anchor points in answer to each question. This response format has been found to be confusing to a proportion of aged individuals, who, whether from mild dementia or the effects of depression, appear to misunderstand it (Yesavage et al. 1983), or to perseverate on the worst response, artificially elevating scores (Lyons, Strain, Hammer, Ackerman, and Fulop 1989). The GDS was created using "yes/no" questions to reduce the problem of too many choices and to make the scale as simple as possible for self-rating (Brink et al. 1982; Yesavage et al. 1983). Finally, the GDS was written to maximize its acceptability to older adults in the cohort of the times, who were found to react negatively to questions about sexual interest or suicidal ideation on general depression scales (Yesavage et al. 1983).
The GDS has continued to gain prominence and acceptance in the gerontology literature. For example, it has been recommended as a standard instrument for research on depression in older adults to facilitate comparison and replication of studies (Koder, Brodaty, and Anstey 1996). The GDS is also included as part of the Geriatric Review Syllabus curriculum for geriatric medicine (Beck 1991). On the other hand, the GDS has received criticism for its simple dichotomous responses (Blazer 1994) and, depending on the sample characteristics, for obtaining lower than desirable sensitivity (Parmalee et al. 1989), specificity (Blazer 1994), or both (Koenig et al. 1992; Kafonek, Ettinger, Roca, Kittner, Taylor, and German 1989). In addition, Norris and Woehr 1998 item analysis of the GDS with a sample of 69 elderly medical patients obtained four item-to-scale correlations in their study that were quite low (two below .3, and two below .4), suggesting these items are not adequately related to the scale. Kafonek, Ettinger, Roca, and colleagues (1989) have noted that some GDS items "lack face validity" for older persons who have recently entered a long-term care facility, such as the item about giving up one's activities and interests. Additionally, despite the fact that the GDS avoids reference to purely somatic complaints, there are several items that appear to be susceptible to somatic conditions (Koenig et al. 1992).
Using the 30 items as a single additive scale with one point each for "depressed" responses, there are two cut-off scores for the GDS to distinguish depressed from nondepressed respondents. Brink and colleagues 1982 recommended the use of 11 points and above to designate depressed as the most conservative cut-off, with an obtained sensitivity of 84% and specificity of 95%, a determination based on responses of a sample of 20 nondepressed elderly persons and 51 who were receiving treatment for depression (Brink et al. 1982). Nevertheless, use of a 14-point cut-off has also been reported (Kafonek et al. 1989; Lyons et al. 1989).
Instrumental Activities of Daily Living (IADL).
Among noninstitutional elderly persons, it is appropriate to measure the degree of impairment in those everyday activities that go beyond self-care to functioning in and out of the home, referred to as IADLs (Lawton and Brody 1969). IADLs can be considered a rough measure of global physical and cognitive functioning. The measure of IADLs used here was taken from the Duke University Older Adults Multidimensional Assessment (OARS; The Duke University Center for the Study of Aging and Human Development 1978) study, a measure that has been used numerous times in large and small studies of older adults. The OARS IADL scale consists of seven items, rated as needing no help (2 points), needing some help (1 point), or being unable to do the item (zero). The summed score, therefore, reflects higher capabilities for higher scores.
Sense of Control.
The degree to which individuals feel they have some control, whether over their everyday life or over their health and health care, has been shown to be an important predictor of health and mental health outcomes. Internal or external locus of control can be formally measured, but scales for doing so are lengthy and complicated. In contrast, a one-item measure of perceived control for this study, taken from Menec, Chipperfield, and Perry 1999, has been found to predict adjustment to illness in prior research, and was related to health and functional disability status in their study. The question reads: "Thinking about life in general, some people generally feel out of control and helpless, while others feel in control and able to cope. How do you generally feel?" with a 0-to-10 rating.
Lubben Social Network Scale.
A brief scale assessing the respondent's living situation, contacts with friends and family, and the number of each that they feel close to, are able to confide in, and receive help from, the Lubben Social Network Scale (LSNS; Lubben 1988) consists of 10 questions scored from 0 to 5. The scale was developed to be used as a screening instrument in health settings, under the widely supported assumption that social support and social networks provide a benefit to elders' mental and physical health and recovery from health problems (Lubben 1988). Scores range from 0 to 50, with higher points indicating more social contact. A score of 20 or lower is considered to place an older person at risk for social isolation associated with extended hospital stays.
|
Results |
|---|
|
TOPAbstractWhen Is Geriatric "Depression"...MethodResultsDiscussionReferences |
|---|
Descriptive information on the sample is presented in Table 1 . Of those who responded, 88.3% were Caucasian (8.9% were African American) with an age range of 65 to 94 and an average age of 73.2 years (SD = 6.2 years). About two thirds of the sample, 212 persons, were from 65 to 74 years old (the "young old") whereas 115 were 75 and over (the "older old"). The majority of respondents were Protestant (55.5%), had a spouse or steady partner (68.8%), and had at least one living child or stepchild (93.6%). Fourteen percent reported that they live in the same household with one of their children, 53.0% reported that at least one child lived within 5 miles, and only 12.2% reported the nearest child lived over an hour away.
|
Health and Functioning.
Respondents were asked to check off whether or not they currently suffered from a list of chronic health conditions. Most indicated two (31.5%) or one (26.6%) of these conditions. The mean number of conditions was 2.1. Among these, "arthritis that causes pain or limits activities" received the highest endorsement indicated by 45.3% of the sample. High blood pressure affected 44.3% of the sample. Heart disease received 22.9% endorsement and diabetes received 16.5%. Fewer indicated problems with emphysema or other breathing disorder (11.0%), depression (8.3%), anxiety or nerves (7.3%), and cancer (6.4%). Very small numbers, from 1% to 3%, endorsed effects of a stroke, history of alcoholism, Parkinson's disease, or liver disease.
General health-related problems that were listed separately included the experience of pain that is significant (endorsed by 25.1%), early morning wakening (18.1%), difficulty falling asleep (15.6%), failing eyesight that causes difficulty reading or driving (12.5%), and hearing difficulties (10.7%).
The sample rated their health from poor to excellent. The mean rating was 3.15, or a little above the "good" rating of 3.0. Only 25.1% of this community-dwelling sample rated themselves to be in "fair" or "poor" health. Forty-five percent indicated they visited a doctor from three to six times in the past year, whereas 43% said they saw a doctor seven or more times in the year. The remaining 12% said they saw the doctor never, once, or twice during the year.
There was little variance within the sample on the IADL scale. The majority of respondents (64.0%) reported not needing help with any of the activities. Twenty-six percent reported needing help with one, two, or three IADLs. Only 10% required help with four to seven of the activities. Housework was the most highly endorsed IADL with which respondents required help, followed by preparing meals, then shopping, and going places independently.
Social and Mental Health.
Responses to the one-item sense of control question were highly skewed toward the "in control" end of the range from 0 (out of control) to 10 (completely in control). The median rating was 8. Only 10 respondents indicated they felt less in control than the midpoint; 26 (8%) placed their answer at the midpoint, neither in control nor out of control; and the remaining 89% placed their response as more in control than out of control.
LSNS scores in this sample ranged from a low of 9 points to a high of 47. Cronbach's alpha for this sample (N = 306 with complete data) was .6824, and 35 respondents (10.7%) scored in the at-risk range for social isolation.
Full-scale scores on the GDS were available for 272 out of the 327 respondents. Most of the 14.4% of the sample with missing data on the scale had one or two items missing. An average of 7.2 cases had missing data per item. Cases were retained for all areas in which their data were complete. Mean GDS score for the nonmissing sample was 4.92, with a range from 0 to 26 (where 30 is the maximum possible on the scale). Using the recommended cut-off of 0 to 10 for not depressed and 11 or above as suggestive of depression, 35 of the complete cases (12.9%) fell into the depressed category, whereas 87.1% were nondepressed. Using the more stringent cut-off of 0 to 13 versus 14 or above, 21 persons (7.7%) would be considered depressed. Cronbach's alpha was .8674 for the full scale for the 272 respondents in this sample.
A PCA was performed on the GDS with the nonmissing sample to ascertain whether it would be appropriate to use the subscales from Parmalee and colleagues 1989. Two subsets of GDS items from their PCA are of particular interest in this study, the 14-item Dysphoria factor and the Withdrawal/Apathy/[Lack of] Vigor factors with a total of seven items. A PCA with Pearson's r correlation matrices was used, as results have been shown to be nearly identical to those using the ideal polychoric type for dichotomous items (Liang 1984; cited in Parmalee et al. 1989). This PCA on the GDS resulted in nine eigenvalues over 1, but Cattell's scree plot suggested a final solution of six components. Varimax rotation converged in nine iterations.
PCA items and their factor loadings appear in Table 2 . The PCA explains 50.4% of the variance in the items. The first and fifth factors together comprise 13 of the 14 items of the Dysphoria factor in the Parmalee et al. 1989 analysis. (The omitted item is No. 27, "Enjoy getting up in the morning.") Because the first and fifth factors both relate to depressive affect, and they have a sufficiently robust Cronbach's alpha reliability in this sample (r = .8360), they were combined into this 13-item Dysphoria score for further analyses. The second factor in the PCA for this sample consists of six of the seven items that formed the Withdrawal/Apathy and Vigor factors in the Parmalee and colleagues study. The WAV items are: "Have you dropped many of your activities and interests?";"Do you prefer to stay at home, rather than going out and doing new things?"; "Do you find life very exciting?"; "Is it hard for you to get started on new projects?"; "Do you feel full of energy?"; and "Do you prefer to avoid social gatherings?". The missing item, "Mind as clear as it used to be," appears to relate less to disengagement from activities or interests than to trouble with memory or concentration. The six WAV items obtained Cronbach's alpha reliability of .7456, and are used as the WAV variable in further analyses. The third factor here consisting of four items relating to worry is an Anxiety factor (r = .6352). The fourth factor in this PCA is Mental Impairment (r = .6880). Three remaining items comprise an unreliable sixth factor, that might be called Agitation (r = .2723).
|
Table 3 displays the 30 GDS items, in descending order of their endorsement rates in this sample—that is, the percentage of respondents who answered the item in the depressed direction. Factor assignments are indicated with the initial after the item number (W = WAV, D = Dysphoria, A = Anxiety and M = Mental Impairment; there are 3 unassigned items). Item-to-total correlations are given in the far right column. The six WAV items had acceptable item-to-scale correlations, ranging from .3655 to .5848. The seven most highly endorsed items on the scale in the depressed direction were the six WAV items and No. 30, the "Mind as clear as it used to be" item that was grouped with these items by Parmalee and colleagues, 1989. The WAV items ranged from 24.8% endorsement (80 respondents) for Item 28, "Do you prefer to avoid social gatherings?" to 58.5% endorsement (186 respondents) on Item 21, "Do you feel full of energy?" One quarter to over one half of the respondents answered in the depressed direction on each of the WAV items. Of the 13 least endorsed items, 11 were from Dysphoria and 2 from the Anxiety subscale.
|
The absolute value and the significance level of the intercorrelations among the four main subscales of the GDS in this sample are shown in Table 4 . The correlation between the Withdrawal/Apathy/Vigor score and Dysphoria score was found to be moderately high, positive, and significant (r = .448, p < .001). The correlation between Anxiety and Dysphoria is higher in value than that between Anxiety and WAV, although both are significant and in a positive direction.
|
|
Another question was how the WAV factor affects "case identification" in the GDS. Without the six WAV items, it was anticipated that fewer respondents would be classified as depressed. Using endorsement of up to one third of the responses for the nondepressed cut-off (0–10), I found that the prorated normal range for the 24-item scale without WAV items was 0 to 8 points out of 24. Removing the WAV items and prorating the scoring criteria did not affect the number of persons originally identified as not depressed by the full-scale score, but the cross-tabulation analysis in Table 6 shows that without the six WAV items, only 22 persons, 8.1% of the sample, would have scores in the new, prorated "depressed" range of 9 to 24 points, compared with the 35 persons, 12.9% of the sample, who scored in the depressed range using the full-scale GDS score. Thus, WAV items were responsible for approximately 37% (13 out of the 35) of those identified as depressed by the full-scale score. Further analysis revealed that 6 of these 13 individuals were from the young-old category and 7 were from the older old category.
|
|
Discussion |
|---|
|
TOPAbstractWhen Is Geriatric "Depression"...MethodResultsDiscussionReferences |
|---|
This study supports earlier findings that the GDS, originally presented and frequently used as a unidimensional measure, encompasses several subscales. The endorsement rates differed markedly between two subscales of the GDS examined here, suggesting that one set of items, those representing Withdrawal, Apathy, and Lack of Vigor (WAV), are more commonly experienced in this sample of community-dwelling elders. Unlike a calculation of hit rates using the scale score and cut-off as a benchmark against which to compare the performance of each item, or item-to-total correlations that use responses from all the scale's other items, simple item endorsement rates are not influenced by the answers to the other items on the scale. Thus, despite a positive correlation between the WAV and Dysphoria subscales, the difference in endorsement rates and mean subscale scores of the two tells us that low energy, having difficulty starting new projects, and so on are experienced more frequently than sadness, crying, hopelessness, and the like. In fact, subscale scores for WAV showed that only 25% of the sample did not have any WAV score, a lower proportion than that for Dysphoria or the other subscales.
Behaviors and feelings similar to those in the WAV items are described in the disengagement theory of aging, socio-emotional selectivity, and gerotranscendence theories and in work by Newmann 1989, Newmann and colleagues 1991, Newmann and colleagues 1996 and Gallo, Anthony and Muthen 1994, whose "depletion syndrome" differs somewhat from clinical depression in symptom configuration, a constellation of behaviors not identical with the GDS WAV, but with some overlap. In the present study, the number of WAV items endorsed has a significant association with increasing age, as does the depletion syndrome in Newmann and colleagues 1996, whereas the Dysphoria subscale score is unrelated to age. Slight differences in bivariate correlations in which the correlations with health and functioning variables are stronger with WAV than with Dysphoria are also suggestive. WAV's intercorrelation with the GDS Anxiety subscale is also somewhat lower than that of Dysphoria, whereas its relation to the Mental Impairment subscale is somewhat higher. These findings taken together may be cautiously interpreted as suggesting that endorsement of WAV items could be tapping a disengagement-like or depletion condition, something that may naturally occur in elders, particularly those over 75 or in poor health, rather than a mental disorder such as depression. We may speculate that this condition represents the psychological and social adaptations to the physical and perhaps cognitive decrements that eventually come with advancing age.
The six WAV items, one fifth of those on the scale, have contributed disproportionately to the total GDS depression score, pushing 37% of those who were classified as depressed by the total scale over the 10-point cut-off. Without the WAV items, 8.1% of the sample would be categorized as depressed, instead of the 12.9% who were classified as depressed with the full scale. On the GDS, each item counts uniformly as one point toward the depressed category, whereas by another diagnostic method such as the DSM–IV, certain key symptoms are "required" for a depression diagnosis and others merely tend to covary with the required symptoms. The WAV items appear to fall in the latter category of symptoms—they may accompany depression, but they also may be due to illness, extreme old age, or a depletion–disengagement syndrome due to changes and losses of old age. Unfortunately, equal weighting of the items on the GDS and similar screening scales may give clinicians and researchers who use the scale the impression that each item describes an equally important symptom, perpetuating the idea that older persons' development and depression profiles differ little from those of younger adults.
As Stanley Jacobson 1995 wrote in his article on overselling depression, it is wise to evaluate an older person's mood and affect carefully before labeling or medicating him or her for depression. "Depression without sadness," on second look, may be socio-emotional selectivity or gerotranscendence—effects of "normal" aging and ways to cope with those effects. If practitioners and researchers choose to use depression screening scales such as the GDS, their validity may be improved by examining reliable subscale scores and concurrently considering the client's age and functional limitations. Although more work is needed, this article proposes a WAV subscale that appears to tap a unique dimension of the experience of older adults, thus adding to the information provided by a total GDS score.
It will be appropriate to continue to explore this relationship of depletion, depression, and developmental changes of old age and further refine their definitions and measurement. Theories of aging, both new and old, should guide us in this effort.
Practice Concepts
The Forum
Book Reviews
|
Acknowledgments |
|---|
Received for publication April 5, 2001. Accepted for publication July 31, 2001.
|
References |
|---|
|
TOPAbstractWhen Is Geriatric "Depression"...MethodResultsDiscussionReferences |
|---|
This article has been cited by other articles:
 |
|
 |
|
  P. Martin, A. Bishop, L. Poon, and M. A. Johnson Influence of personality and health behaviors on fatigue in late and very late life. J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2006; 61(3): P161 - P166. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. W. Braam, I. Bramsen, T. G. van Tilburg, H. M. van der Ploeg, and D. J. H. Deeg Cosmic transcendence and framework of meaning in life: patterns among older adults in the Netherlands. J. Gerontol. B. Psychol. Sci. Soc. Sci., May 1, 2006; 61(3): S121 - S128. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. B. Adams, H. C. Matto, and S. Sanders Confirmatory Factor Analysis of the Geriatric Depression Scale Gerontologist, December 1, 2004; 44(6): 818 - 826. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 K. B. Adams and S. Sanders Alzheimer's Caregiver Differences in Experience of Loss, Grief Reactions and Depressive Symptoms Across Stage of Disease: A Mixed-Method Analysis Dementia, June 1, 2004; 3(2): 195 - 210. [Abstract] [PDF]
|
|
|
|
 |
|
 D. G. Blazer Depression in Late Life: Review and Commentary J. Gerontol. A Biol. Sci. Med. Sci., March 1, 2003; 58(3): M249 - 265. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
