Journal of Gerontological Nursing

Feature Article 

Can Older Adults with Dementia Accurately Report Depression Using Brief Forms? Reliability and Validity of the Geriatric Depression Scale

Helen W. Lach, PhD, RN; Yu-Ping Chang, PhD, RN; Dorothy Edwards, PhD

Abstract

The Geriatric Depression Scale (GDS) is a commonly used screening tool, but its use in older adults with cognitive impairment has been controversial. This study compared the short forms of the GDS with clinician diagnosis of depression using standard criteria (Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revision) in people with and without dementia. Sensitivity and specificity were acceptable for all forms of the GDS. These results build evidence for using the short GDS 5- and 15-item versions in populations that include people with mild to moderate dementia, increasing the ease of depression screening so it can be performed more frequently in clinical settings.

Abstract

The Geriatric Depression Scale (GDS) is a commonly used screening tool, but its use in older adults with cognitive impairment has been controversial. This study compared the short forms of the GDS with clinician diagnosis of depression using standard criteria (Diagnostic and Statistical Manual of Mental Disorders, 4th edition, text revision) in people with and without dementia. Sensitivity and specificity were acceptable for all forms of the GDS. These results build evidence for using the short GDS 5- and 15-item versions in populations that include people with mild to moderate dementia, increasing the ease of depression screening so it can be performed more frequently in clinical settings.

Dr. Lach is Associate Professor, Saint Louis University School of Nursing, St. Louis, Missouri; Dr. Chang is Research Assistant Professor, State University of New York at Buffalo, Buffalo, New York; and Dr. Edwards is Associate Professor of Kinesiology, University of Wisconsin, Madison, Madison, Wisconsin.

The authors disclose that they have no significant financial interests in any product or class of products discussed directly or indirectly in this activity, including research support.

Address correspondence to Helen W. Lach, PhD, RN, Associate Professor, Saint Louis University School of Nursing, 3525 Caroline Mall, St. Louis, MO 63104-1099; e-mail: lachh@slu.edu.

Received: July 06, 2009
Accepted: November 05, 2009
Posted Online: May 06, 2010

The need to screen and treat depression in older adults is well documented in the geriatric literature (Blazer, 2003; Lebowitz et al., 1997; Suter, Suter, & Johnston, 2008) and is particularly important in individuals with dementia (Brown, Raue, Halpert, Adams, & Titler, 2009; Steffens et al., 2006). Depression can affect the diagnosis and treatment of other common physical and psychiatric conditions in later life, as well as older adults’ function and quality of life; therefore, depression screening is important in all settings. The Geriatric Depression Scale (GDS) was developed specifically for screening older adults and has been tested extensively against other measures and in various settings (Brink et al., 1982; Yesavage et al., 1983). However, even with effective tools, lack of time in the busy clinical setting is often a barrier to performing depression screening. In response to the need for more rapid screening tools, a 15-item version of the GDS (GDS-15) has also gained wide acceptance (Burke, Roccaforte, & Wengel, 1991; Friedman, Heisel, & Delavan, 2005; Sheikh & Yesavage, 1986), as it is shorter and easier to use in clinical practice. In addition, an even shorter 5-item version of the GDS (GDS-5) has been published that takes less than 3 minutes to administer (Hoyl et al., 1999).

Although the short forms of the GDS and other self-report instruments are widely available, clinicians are concerned about their validity in patients who also have dementia (Gallo & Wittink, 2006; Watson & Pignone, 2003). Screening those with dementia is important because they have a high risk of depression. In response to concerns about the accuracy of self-reports, alternative questionnaires have been developed. One option is the Cornell Scale for Depression in Dementia (CSDD) (Alexopoulos, Abrams, Young, & Shamoian, 1988), which is completed by a clinician on the basis of knowledge of the older adult’s symptoms during the prior week or using input from a knowledgeable informant. The GDS has a version that can be completed by a collateral source (Burke, Rangwani, Roccaforte, Wengel, & Conley, 1997; Nitcher, Burke, Roccaforte, & Wengel, 1993). However, in a busy clinical setting, there may not be time to gather additional data, clinicians may not have observed the patient over time, and patients may not have a family member or friend available for questioning. As a result, patients with dementia may not receive adequate depression screening.

The overlap between symptoms of dementia and depression, as well as the belief that patients with dementia may lack the insight needed to answer questions reliably has been of concern to researchers and clinicians alike. These concerns have prompted a number of studies examining performance of the GDS in samples of older adults who have cognitive impairment; these studies have shown conflicting results (Bedard et al., 2003; Burke, Houston, Boust, & Roccaforte, 1989; Burke et al., 1997; Feher, Larrabee, & Crook, 1992; Korner et al., 2006; Muller-Thomsen, Arlt, Mann, Mass, & Ganzer, 2005; O’Riordan et al., 1990; Snow et al., 2005). Several investigators have reported poor predictive ability of the GDS in patients with dementia (Bedard et al., 2003; Burke et al., 1989), whereas others have found that individuals with mild to moderate Alzheimer’s disease are able to reliably report depressive symptoms (Feher et al., 1992; O’Riordan et al., 1990; Snow et al., 2005). Some studies have included only cognitively intact participants (Friedman et al., 2005).

Because the question of whether people with dementia can accurately report depression symptoms persists, this study was conducted in a sample of patients with a wide range of cognitive abilities. The purpose of the study was to compare the different versions of the GDS with the “gold standard”: concurrent diagnosis of depression by a clinician. The outcome was to determine the reliability and validity of the various versions of the GDS and provide recommendations to make depression screening efficient for clinical practice.

Method

The data reported in this study were obtained through a retrospective chart review of all patients attending a university-based interdisciplinary outpatient geriatric assessment program during a 15-month period. Each patient was mailed the GDS along with other standard clinical forms to complete at home and bring to the assessment visit. The assessment included a complete medical evaluation, cognitive testing, and collection of information about functional status and psychosocial and cognitive problems. Separately from the physician, an advanced practice nurse or a social worker interviewed patients to review their responses on the GDS. The physician or a nurse practitioner interviewed the patient and a collateral informant and used the Diagnostic and Statistical Manual of Mental Disorders, fourth edition, text revision (DSM-IV-TR) criteria (American Psychiatric Association, [APA], 2000) to determine a depression diagnosis.

Presence and severity of dementia for each patient were documented using the Clinical Dementia Rating scale (CDR) (Berg et al., 1998; Morris, 1993), which was completed at a team meeting after the patient was evaluated. The CDR is a global rating for staging dementia using scores for impairments in six cognitive domains. Staff were trained in rating patients using the tool, which has shown an interrater reliability of 80% for the global CDR rating (Schafer et al., 2004).

After obtaining Institutional Review Board approval, data from patients with completed forms in the chart were included in this study. A total of 169 patients had complete records for the 15-month study period. Only 9 patients had severe dementia (CDR level = 3) and complete records at the time of the evaluation; they were excluded from the study because of the small sample and concern about their ability to answer the questions. Thus, the final sample was 160 patients and informants.

Measures

The DSM-IV-TR criteria for depression was used as the criterion or gold standard for diagnosis of depression. For a diagnosis of clinical depression, patients must have a depressed mood and/or anhedonia for 2 or more weeks, as well as at least four of eight additional criterion symptoms of depression causing clinically significant distress (APA, 2000).

The 30-item long form GDS (GDS-30) was used in this study. This form includes all of the items needed to create the 5- and 15-item versions of the scale. The CDR was used to identify and stage the severity of cognitive impairment and includes five levels: 0 = normal, 0.5 = very mild dementia/questionable, 1 = mild dementia, 2 = moderate dementia, and 3 = severe dementia. Each patient was rated by the geriatric assessment team as to their dementia severity after a complete evaluation of the patient and interview with the collateral informant.

Data Analysis

Descriptive statistics were computed for each variable. One-way analyses of variance were used to determine differences across dementia severity groups (CDR levels) followed by post-hoc Tukey tests. Internal consistency analyses were computed separately for each scale using Cronbach’s alpha coefficient. The physician’s DSM-IV-TR diagnosis of depression was used to categorize patients into depressed and non-depressed groups. For patients with cognitive impairment, separate logistic regression analyses were computed for each of the three versions of the GDS (30, 15, 5). A log likelihood statistic was used to determine the fit of each model. Receiver operating curve (ROC) analyses were used to establish cut-off points. According to these cut-off points, sensitivity, specificity, and positive/negative predictive value coefficients were determined for each scale to evaluate their effectiveness for screening. SPSS version 13 was used for all analyses.

Results

The demographic characteristics of the sample are presented in Table 1. Of the 160 individuals included in the study, 16.3% had normal cognition (CDR level = 0) and the remainder had some cognitive impairment (CDR level = 0.5 to 2). The majority of the participants were women (71.9%) and Caucasian (79.4%). The mean age of the sample was 79.24 (SD = 9.80 years), and nearly half had high school or college degrees. Approximately half of all participants lived alone. In addition, 91 (56.9%) were diagnosed with a depression using DSM-IV-TR criteria.

Demographic Profile of the Sample by Clinical Dementia Rating (CDR) Scale Level

Table 1: Demographic Profile of the Sample by Clinical Dementia Rating (CDR) Scale Level

The mean scores and standard deviations of the three scales are shown in Table 2. There were no significant differences in scores across CDR levels. Internal consistency was computed using Cronbach’s alpha coefficient. As expected, the longer version had the highest Cronbach’s alpha coefficient (0.92), and the GDS-15 had an acceptable Cronbach’s alpha coefficient of 0.87. The Cronbach’s alpha coefficient of 0.68 for the GDS-5 was a little lower than the desirable level of 0.70 (Polit & Beck, 2004); however, when looking at only patients who had cognitive impairment (CDR level > 0, n = 134), the Cronbach’s alpha coefficients were slightly better (0.93 for CDR = 0.5, 0.88 for CDR = 1, and 0.70 for CDR = 2).

Participants’ Mean Scores (SD) on Different Versions of the GDS by CDR Level

Table 2: Participants’ Mean Scores (SD) on Different Versions of the GDS by CDR Level

Logistic regression and ROC analyses were computed to determine optimal cut-off points, sensitivity, specificity, positive predictive values (PPV), negative predictive values (NPV), and overall misclassification rates. Patients were classified as depressed or not depressed on the basis of the physician’s DSM-IV-TR diagnosis. The criterion validity indices for the three scales are shown in Table 3. These analyses were computed separately for the GDS-30, GDS-15, and GDS-5.

Validity Indices (%) of Different Versions of the GDS Using Depressiona as the Criterion for Case Definition (N = 135)

Table 3: Validity Indices (%) of Different Versions of the GDS Using Depression as the Criterion for Case Definition (N = 135)

When examining only those patients with cognitive impairment, scores of 11 or higher on the 30-item GDS yielded a sensitivity of 84.2% and a specificity of 72.7%. The data supported the cut-off score of 11, which is commonly recommended for the GDS-30, although a cut-off point of 10 had slightly better sensitivity. The cut-off score of 4 for the GDS-15 was slightly better than the recommended cut-off score of 5, with a sensitivity of 91.2% and specificity of 59.5%, but the usual score was acceptable. The recommended cut-off score of 2 for the GDS-5 showed sensitivity of 93% and specificity of 62.3%. These results support the criterion validity of the GDS as a screening measure of depression in cognitively impaired older adults. As expected, the sensitivity increased as the number of scale items decreased, and there was a modest loss of specificity from 76.6% for the GDS-30 to 62.3% for the GDS-5. All of the results were similar when including both individuals with and without dementia.

We further analyzed the scores of all patients on the three GDS scales to explore possible response differences associated with the gender of the patient (Table 4). For the short forms of the GDS, the sensitivity was higher for men than women, as was the specificity. Similar analyses were computed to examine the sensitivity and specificity of the scales by race. Sensitivity scores ranged from 81.6% to 93.9%, and specificity scores ranged from 53.8% to 66.7% in the Caucasian patients. For African American patients, sensitivity scores ranged from 85% to 90% and specificity scores ranged from 83.3% to 100%.

Performance (%) of Different Versions of the Gds According to Gender and Race

Table 4: Performance (%) of Different Versions of the Gds According to Gender and Race

Discussion

Despite the high prevalence and substantial impact of depression, screening is not consistently performed (Charney et al., 2003; Pignone et al., 2002), and studies have shown that usual care fails to recognize 30% to 50% of depressed patients. When depression goes unrecognized, patients cannot be appropriately treated, and systematic screening has been advocated as a means of improving detection, treatment, and outcomes of depression. Guidelines for practice now emphasize the importance of screening as part of elder care to facilitate treatment (Brown et al., 2009; Kurlowicz & Harvath, 2007; Piven, 2001; Registered Nurses Association of Ontario, 2003). Systematic screening is most likely to be implemented by nurses and other health care providers if the process is easy and can be streamlined to meet the time and resource demands of the practice environment.

This study was designed to determine the sensitivity and specificity of the short versions of the GDS in order to guide depression screening in clinical practice. Earlier researchers have reported conflicting findings regarding the ability of individuals with dementia to accurately report depressive symptoms. Congruent with the findings of Snow et al.’s (2005) study, the current study results provide convincing evidence that individuals with mild to moderate dementia can provide valid information about their mood using standardized self-report questionnaires. The clinic population for this study provided a range of ages and educational levels, and the representation of women and African American older adults approximates the metropolitan region where the clinic is located.

The results show that all three versions of the GDS perform well at screening for major depression across groups. As a result, the short forms of the GDS could be used to screen patients, saving valuable clinical time. In addition, in this study, the forms were completed at home and brought to the appointment, a procedure that could be even more efficient. Alternatively, forms could be completed by the individual in an office waiting room or in their hospital room.

This study had several strengths. First, the GDS scores were compared with concurrent clinician diagnosis of depression using DSM-IV-TR criteria. This design is superior to prior studies that compared the GDS with other depression screening tools (Bedard et al., 2003; Feher et al., 1992; Muller-Thomsen et al., 2005; Weeks, McGann, Michaels, & Penninx, 2003). Although prior studies showed that the CSDD performed more accurately than the GDS in screening depression among individuals with dementia (Korner et al., 2006; Muller-Thomsen et al., 2005), the CSDD requirement of a reliable informant or clinical observation over a period of time before completion can be a major limitation in practice. This study indicated that the GDS is valid and reliable for self-report in older adults with mild to moderate dementia and has satisfactory sensitivity, specificity, and internal consistency.

Another strength of the study was the use of the CDR rating scale. This provided a better staging of the severity of dementia in the sample than scales used in prior studies (Feher et al., 1992; Korner et al., 2006; Muller-Thomsen et al., 2005; Snow et al., 2005). The final strength was that the sample included a larger percentage of minority older adults than prior studies. While the percentage of African Americans was only 20%, this was still higher than other studies and reflects the population of the metropolitan area where the study was conducted. Different forms of the GDS functioned well among African American participants suggesting the GDS versions can be used with this population.

Limitations

There were several limitations to this study. First, because the data were collected as part of a clinical program, the clinician making the diagnosis of depression was not necessarily blinded to the GDS scores. The clinic procedure called for the physician or nurse practitioner to interview the patient and rate the DSM-IV-TR criteria independent of the GDS scores. A different health professional reviewed the GDS forms with the patient in a separate interview. Sometimes, the professionals discuss the case and possibly the GDS scores during the clinic visit; as a result, the scores may have influenced the diagnosis of depression. However, the physicians did not routinely examine the GDS scores or use them during their assessment and rating of the DSM-IV-TR criteria. Second, the different forms of the GDS were not administered separately; the short form information was extracted from the 30-item version, so results might have differed if participants had completed all three forms.

Another limitation is that the GDS forms were completed at home by the patient, so a nonstandardized administration format was used. It is unknown how much the collateral informants assisted the patient in completing the forms or whether they looked at the patient’s answers in completing their forms. However, a related study looked at a collateral informant version of the GDS (Lach & Chang, 2005) and found consistently higher scores from collateral sources, suggesting that family members were not completing the forms for the patients. Those who did not complete the forms and those with severe dementia were excluded from this study. Therefore, the instrument may not work in these groups. However, these results suggest that if people are able to answer the GDS questions and do not have severe dementia, then they are able to provide clinically useful information about the presence of depression. Future studies should provide standardized testing conditions and use all three GDS versions to confirm these findings.

Conclusion

Effective screening for depression can be performed in the clinical setting using the short forms of the GDS in less than 3 minutes. Use of short forms may increase the use of screening in busy clinical practices. This study verified the accuracy of self-reported depression in patients with mild to moderate dementia so that nurses can quickly and confidently administer depression screening and identify older adults who may have depression.

References

  • Alexopoulos, G.S., Abrams, R.C., Young, R.C. & Shamoian, C.A. (1988). Cornell Scale for Depression in Dementia. Biological Psychiatry, 23, 271–284. doi:10.1016/0006-3223(88)90038-8 [CrossRef]
  • American Psychiatric Association. (2000). Diagnostic and statistical manual of mental disorders (4th ed., text rev.). doi:10.1176/appi.books.9780890423349
  • Bedard, M., Molloy, D.W., Squire, L., Minthrorn-Biggs, M.B., Dubois, S & Lever, J.A. et al. (2003). Validity of self-reports in dementia research: The Geriatric Depression Scale. Clinical Gerontologist, 26, 155–163. doi:10.1300/J018v26n03_13 [CrossRef]
  • Berg, L., McKeel, D.W. Jr.. , Miller, J.P., Storandt, M., Rubin, E.H. & Morris, J.C. et al. (1998). Clinicopathologic studies in cognitively healthy aging and Alzheimer’s disease: Relation of histologic markers to dementia severity, age, sex, and APOE genotype. Archives of Neurology, 55, 326–335. doi:10.1001/archneur.55.3.326 [CrossRef]
  • Blazer, D.G. (2003). Depression in later life: Review and commentary. Journals of Gerontology. Series A, Biological Sciences and Medical Sciences, 58, 249–265.
  • Brink, T.L., Yesavage, T.L., Lum, O., Heersema, P., Adey, M.B. & Rose, T.L. (1982). Screening tests for geriatric depression. Clinical Gerontologist, 1, 37–43. doi:10.1300/J018v01n01_06 [CrossRef]
  • Brown, E.L., Raue, P., Halpert, K.D., Adams, S. & Titler, M.G. (2009). Detection of depression in older adults with dementia. Journal of Gerontological Nursing, 35(2), 11–15. doi:10.3928/00989134-20090201-08 [CrossRef]
  • Burke, W.J., Houston, M.J., Boust, S.J. & Roccaforte, W.H. (1989). Use of the Geriatric Depression Scale in dementia of the Alzheimer type. Journal of the American Geriatrics Society, 37, 856–860.
  • Burke, W.J., Rangwani, S., Roccaforte, W.H., Wengel, S.P. & Conley, D.M. (1997). The reliability and validity of the collateral source version of the Geriatric Depression Rating Scale administered by telephone. International Journal of Geriatric Psychiatry, 12, 288–294. doi:10.1002/(SICI)1099-1166(199703)12:3<288::AID-GPS474>3.0.CO;2-1 [CrossRef]
  • Burke, W.J., Roccaforte, W.H. & Wengel, S.P. (1991). The short Geriatric Depression Scale: A comparison with the 30-item scale. Journal of Geriatric Psychiatry and Neurology, 4, 173–178. doi:10.1177/089198879100400310 [CrossRef]
  • Charney, D.S., Reynolds, C.F. 3rd. , Lewis, L., Lebowitz, B.D., Sunderland, T. & Alexopoulos, G.S. et al. (2003). Depression and Bipolar Support Alliance consensus statement on the unmet needs in diagnosis and treatment of mood disorders in later life. Archives of General Psychiatry, 60, 664–672. doi:10.1001/archpsyc.60.7.664 [CrossRef]
  • Feher, E.P., Larrabee, G.J. & Crook, T.H. (1992). Factors attenuating the validity of the Geriatric Depression Scale in a dementia population. Journal of the American Geriatrics Society, 40, 906–909.
  • Friedman, B., Heisel, M.J. & Delavan, R.L. (2005). Psychometric properties of the 15 item Geriatric Depression Scale in functionally impaired, cognitively intact, community dwelling elderly primary care patients. Journal of the American Geriatrics Society, 53, 1570–1576. doi:10.1111/j.1532-5415.2005.53461.x [CrossRef]
  • Gallo, J. & Wittink, M.N. (2006). Depression assessment. In Gallo, J., Bogner, H.R., Fulmer, T. & Pavenza, G.J. (Eds.), Handbook of geriatric assessment (4th ed., pp. 153–174). Boston: Jones and Bartlett.
  • Hoyl, M.T., Alessi, C.A., Harker, J.O., Josephson, K.R., Pietruszka, F.M. & Koelfgen, M. et al. (1999). Development and testing of a five-item version of the Geriatric Depression Scale. Journal of the American Geriatrics Society, 47, 873–878.
  • Korner, A., Lauritzen, L., Abelskov, K., Gulmann, N., Marie Brodersen, A. & Wedervang Jensen, T. et al. (2006). The Geriatric Depression Scale and the Cornell Scale for Depression in Dementia. A validity study. Nordic Journal of Psychiatry, 60, 360–364. doi:10.1080/08039480600937066 [CrossRef]
  • Kurlowicz, L.H. & Harvath, T.A. (2007). Depression. In Capezuti, E., Zwicker, D., Mezey, M., Fulmer, T., Gray-Miceli, D. & Kluger, M. (Eds.), Evidence-based geriatric nursing protocols for best practice (3rd ed., pp. 57–82). New York: Springer.
  • Lach, H.W. & Chang, Y.P. (2005). Validity of the Geriatric Depression Scale for dementia patients [Abstract]. The Gerontologist, 45(Special Issue II), 246.
  • Lebowitz, B.D., Pearson, L.S., Schneider, L.S., Reynolds, C.F. 3rd. , Alexopoulos, G.S. & Bruce, M.L. et al. (1997). Diagnosis and treatment of depression in late life. Consensus statement update. Journal of the American Medical Association, 278, 1186–1190. doi:10.1001/jama.278.14.1186 [CrossRef]
  • Morris, J.C. (1993). The Clinical Dementia Rating (CDR): Current version and scoring rules. Neurology, 43, 2412–2414.
  • Muller-Thomsen, T., Arlt, S., Mann, U., Mass, R. & Ganzer, S. (2005). Detecting depression in Alzheimer’s disease: Evaluation of four different scales. Archives of Clinical Neuropsychology, 20, 271–276. doi:10.1016/j.acn.2004.03.010 [CrossRef]
  • Nitcher, R.L., Burke, W.J., Roccaforte, W.H. & Wengel, S.P. (1993). A collateral source version of the Geriatric Depression Rating Scale. American Journal of Geriatric Psychiatry, 12, 143–152. doi:10.1097/00019442-199300120-00007 [CrossRef]
  • O’Riordan, T.G., Hayes, J.P., O’Neill, D., Shelley, R., Walsh, J.B. & Coakley, D. (1990). The effect of mild to moderate dementia on the Geriatric Depression Scale: A preliminary report. Age and Ageing, 19, 57–61. doi:10.1093/ageing/19.1.57 [CrossRef]
  • Pignone, M.P., Gaynes, B.N., Rushton, J.L., Burchell, C.M., Orleans, C.T. & Mulrow, C.D. et al. (2002). Screening for depression in adults: A summary of the evidence for the U.S. Preventive Services Task Force. Annals of Internal Medicine, 136, 765–776.
  • Piven, M.L. (2001). Detection of depression in the cognitively intact older adult. Journal of Gerontological Nursing, 27(6), 8–14.
  • Polit, D.F. & Beck, C.T. (2004). Nursing research: Principles and methods. Philadelphia: Lippincott Williams & Wilkins.
  • Registered Nurses Association of Ontario. (2003). Screening for delirium, dementia, and depression in older adults. Retrieved from the National Guideline Clearing-house website: http://www.guideline.gov/summary/summary.aspx?doc_id=5313&nbr=003636#s3
  • Schafer, K.A., Tractenberg, R.E., Sano, M., Mackell, J.A., Thomas, R.G. & Gamst, A. et al. (2004). Reliability of monitoring the Clinical Dementia Rating in multicenter clinical trails. Alzheimer’s Disease and Associated Disorders, 18, 219–222.
  • Sheikh, J.I. & Yesavage, J.A. (1986). Geriatric Depression Scale (GDS): Recent evidence and development of a shorter version. In Brink, T.L. (Ed.), Clinical gerontology: A guide to assessment and intervention (pp. 165–173). New York: Haworth Press.
  • Snow, A.L., Kunik, M.E., Molinari, V.A., Orengo, C.A., Doody, R. & Graham, D.P. et al. (2005). Accuracy of self-reported depression in persons with dementia. Journal of the American Geriatrics Society, 53, 389–396. doi:10.1111/j.1532-5415.2005.53154.x [CrossRef]
  • Steffens, D.C., Otey, E., Alexopoulos, G.S., Butters, M.A., Cuthbert, B. & Ganguli, M. et al. (2006). Perspectives on depression, mild cognitive impairment, and cognitive decline. Archives of General Psychiatry, 63, 130–138. doi:10.1001/archpsyc.63.2.130 [CrossRef]
  • Suter, P., Suter, W.N. & Johnston, D. (2008). Depression revealed: The need for screening, treatment, and monitoring. Home Healthcare Nurse, 26, 543–350. doi:10.1097/01.NHH.0000338514.85323.35 [CrossRef]
  • Watson, L.C. & Pignone, M.P. (2003). Screening accuracy for late-life depression in primary care: A systematic review. Journal of Family Practice, 52, 956–964.
  • Weeks, S.K., McGann, P.E., Michaels, T.K. & Penninx, B.W. (2003). Comparing various short-form Geriatric Depression Scales leads to the GDS-5/15. Journal of Nursing Scholarship, 35, 133–137. doi:10.1111/j.1547-5069.2003.00133.x [CrossRef]
  • Yesavage, J.A., Brink, T.L., Rose, T.L., Lum, O., Huang, V. & Adey, M. et al. (1983). Development and validation of a Geriatric Depression Screening Scale: A preliminary report. Journal of Psychiatric Research, 17, 37–49. doi:10.1016/0022-3956(82)90033-4 [CrossRef]

Demographic Profile of the Sample by Clinical Dementia Rating (CDR) Scale Level

CDR Level
Variable 0n= 26 (16.3%) 0.5n= 57 (35.6%) 1n= 48 (30%) 2n= 29 (18.1%) TotalN= 160 (100%)
Mean age (SD) 79.58 (8.13) 78.93 (13) 80.08 (7.67) 78.17 (6.75) 79.24 (9.80)
Gender, n (%)
  Men 5 (19.2) 16 (28.1) 12 (25) 12 (41.4) 45 (28.1)
  Women 21 (80.8) 41 (71.9) 36 (75) 17 (58.6) 115 (71.9)
Race, n (%)
  Caucasian 17 (65.4) 48 (84.2) 38 (79.2) 24 (82.8) 127 (79.4)
  African American 8 (30.8) 9 (15.8) 10 (20.8) 5 (17.2) 32 (20)
  Missing data 1 (3.8) 1 (0.6)
Educational level, n (%)
  Less than high school 7 (26.9) 22 (38.6) 23 (47.9) 11 (37.9) 64 (39.4)
  High school 7 (26.9) 19 (33.3) 18 (37.5) 6 (20.7) 50 (31.3)
  More than high school 11 (42.3) 16 (28.1) 4 (8.3) 10 (34.5) 41 (25.6)
  Missing data 1 (3.8) 3 (6.3) 2 (6.9) 6 (3.7)
Collateral informant, n (%)
  Spouse 7 (26.9) 14 (24.6) 11 (22.9) 9 (31) 41 (25.6)
  Child 12 (46.2) 37 (64.9) 27 (56.3) 18 (62.1) 94 (58.8)
  Other 4 (15.4) 6 (10.5) 9 (18.8) 2 (6.9) 21 (13.1)
  Missing data 3 (11.5) 1 (2.1) 4 (2.5)
Living situation, n (%)
  Alone 12 (46.2) 33 (57.9) 24 (50) 8 (27.6) 77 (48.1)
  Spouse 8 (30.8) 16 (28.1) 15 (31.3) 13 (44.8) 52 (32.5)
  Child 2 (7.7) 7 (12.3) 6 (12.5) 5(17.2) 20 (12.5)
  Other 3 (11.5) 1 (1.8) 3 (6.3) 2 (6.9) 9 (5.6)
  Missing data 1 (3.8) 1 (3.4) 2 (1.3)
Meets DSM-IV-TR depression criteria, n (%)
  Yes 13 (50) 34 (59.6) 24 (50) 20 (69) 91 (56.9)
  No 13 (50) 23 (40.4) 24 (50) 9 (31) 69 (43.1)

Participants’ Mean Scores (SD) on Different Versions of the GDS by CDR Level

CDR Level
Version 0 (n= 26) 0.5 (n= 57) 1 (n= 48) 2 (n= 29) Total (N= 160)
GDS-30 11.73 (6.88) 12.61 (8.33) 11.58 (8.11) 12.28 (8.29) 12.10 (7.97)
GDS-15 5.50 (3.77) 5.95 (4.30) 5.63 (4.13) 5.38 (4.44) 5.58 (4.16)
GDS-5 2.12 (1.53) 2.26 (1.71) 2.19 (1.62) 2.00 (1.60) 2.17 (1.62)

Validity Indices (%) of Different Versions of the GDS Using Depressiona as the Criterion for Case Definition (N = 135)

Version Cut-Off Point Sensitivity Specificity PPV NPV OMR Classified Correctly (%) Area Under ROC Curve p Value 95% CI
GDS-30 9 91.2 64.9 65.8 90.9 23.9 76.1 0.894 0.000 0.839 to 0.948
10 84.2 72.7 69.6 86.2 22.4 77.6
11b 82.5 76.6 72.3 85.5 20.1 79.9
GDS-15 3 96.5 42.9 55.6 94.3 34.3 65.7 0.893 0.000 0.838 to 0.948
4 91.2 59.5 62.7 90.2 26.9 73.1
5b 84.2 68.8 66.7 85.5 24.6 75.4
GDS-5 1 98.2 33.8 52.3 96.3 38.8 61.2 0.897 0.000 0.842 to 0.951
2b 93 62.3 64.6 92.3 24.6 75.4
3 75.4 85.7 79.6 82.5 18.7 81.3

Performance (%) of Different Versions of the Gds According to Gender and Race

Gender Race
Version Validity Index Men (n= 45) Women (n= 115) Caucasian (n= 127) African American (n= 32)a
GDS-30 Sensitivity 83.3 84.6 81.6 90
Specificity 77.8 68.3 66.7 100
PPV 71.4 68.8 60 100
NPV 87.5 84.3 85.2 85.7
OMR 20 24.3 27.5 6.2
GDS-15 Sensitivity 94.4 86.5 87.8 90
Specificity 66.7 54 53.8 83.3
PPV 65.4 60.8 54.4 90
NPV 94.7 82.9 87.5 83.3
OMR 22.2 31.3 33 12.5
GDS-5 Sensitivity 94.4 88.5 93.9 85
Specificity 66.7 58.7 56.4 91.7
PPV 65.4 63.9 57.5 94.4
NPV 94.7 86 93.6 78.6
OMR 22.2 27.8 29.1 12.5

Keypoints

Lach, H.W., Chang, Y.-P. & Edwards, D. (2010). Can Older Adults with Dementia Accurately Report Depression Using Brief Forms? Reliability and Validity of the Geriatric Depression Scale. Journal of Gerontological Nursing, 36(5), 30–37.

  1. Depression screening is important in all settings, as depression can affect the diagnosis and treatment of other physical and psychiatric conditions, as well as older adults’ functioning and quality of life.

  2. The Geriatric Depression Scale (GDS) is a commonly used screening tool, but its use in older adults with cognitive impairment has been controversial.

  3. The results of this study show that all three versions of the GDS (30, 15, and 5 items) perform well at screening for major depression across cognitive levels.

  4. The GDS is valid and reliable for self-report in older adults with mild to moderate dementia and has satisfactory sensitivity, specificity, and internal consistency.

Authors

Dr. Lach is Associate Professor, Saint Louis University School of Nursing, St. Louis, Missouri; Dr. Chang is Research Assistant Professor, State University of New York at Buffalo, Buffalo, New York; and Dr. Edwards is Associate Professor of Kinesiology, University of Wisconsin, Madison, Madison, Wisconsin.

The authors disclose that they have no significant financial interests in any product or class of products discussed directly or indirectly in this activity, including research support.

Address correspondence to Helen W. Lach, PhD, RN, Associate Professor, Saint Louis University School of Nursing, 3525 Caroline Mall, St. Louis, MO 63104-1099; e-mail: .lachh@slu.edu

10.3928/00989134-20100303-01

Sign up to receive

Journal E-contents