Depression and Dementia

Another factor that complicates the diagnosis of depression in demented persons is that often a demented patient does not communicate his or her depressive symptoms in a way that is amenable to diagnosis. It has been hypothesized that demented patients may not be aware of their depres­sive symptoms (Chemerinski et al. 2001), and in many cases, caregiver accounts are necessary in order to accurately assess whether or not a patient is depressed. Demented patients often report themselves to be less depressed than do their caregivers and clinicians. For instance, a study of 75 outpatients with diagnosed AD compared concordance of ratings of depression between the patient, caregiver, and clinician. They found that patients rated themselves as less depressed than did their caregivers and clinicians and that results did not vary by severity of dementia (Teri and Wagner 1991). Similar results were found in a smaller study of 31 AD patients and their caregivers; caregivers reported more depressive symp­toms than did the patients themselves and concordance between ratings did not vary according to the depression status of the caregiver (Moye, Robiner, and Mackenzie 1993) . Another study examined whether there were specific symptoms of depression that differed between caregiver and patient reports (Mackenzie, Robiner, and Knopman 1989). They found that patient ratings identified only 14% of the sample as depressed while caregiver ratings identified one-half of the sample as depressed. Discor­dance was noted for specific symptoms including patients’ loss of interest or pleasure, irritability, fatigue, and feelings of worthlessness.

Another complexity to diagnosis of depression in demented patients is that many depression screening instruments were not developed spe­cifically for older populations and the validation of these scales among demented older persons has been sparse. The Geriatric Depression Scale (GDS) (Sheikh and Yesavage 1986), Hamilton Rating Scale for Depression (HRSD) (Hedlund 1979), the Cornell Scale for Depression (Alexopoulos et al. 1988), and the DSM-IV criteria (American Psychiatric Association

1994)   have been most widely compared in the published literature. Differ­ences in the rates of depression among demented persons based on these various different screening tools have been compared in a number of stud­ies. A large study of 288 outpatients with dementia found the prevalence of depression to be 8.0% according to the GDS, 7.4% using the HRSD, and 6.3% according to DSM-IV criteria (Brodaty and Luscombe 1996) . Rates of depression also differed by dementia subtype among the scales. Depression was more likely to be diagnosed in vascular dementia than in Alzheimer’s disease using the HRSD and the GDS, whereas rates accord­ing to the DSM-IV criteria did not differ by dementia subtype. Of note is the low prevalence of depression in this study compared to the previously described literature in this post.

Provisional criteria for the diagnosis of depression of Alzheimer’s disease (NIMH-dAD) were developed as part of a workshop sponsored by the National Institute of Mental Health (Olin, Katz, et al. 2002; Olin, Schneider, et al. 2002). The NIHM-dAD criteria are similar to the DSM-IV criteria for major depression, but incorporate modifications to address specific characteristics of depression in AD. For instance, the NIMH- dAD require three or more symptoms of depression rather than the five required for major depression; these criteria include irritability and social isolation or withdrawal as candidate symptoms of depression, and include ‘decreased positive affect or pleasure’ instead of loss of interest or pleasure, and require that the symptoms occur during the prior two-week period and represent change from previous function, but do not require that their symptoms occur every day. Validation studies of the NIMH- dAD are underway.

A recent study compared the NIMH-dAD to the Cornell Scale for Depression in Dementia, the GDS and the DSM-IV criteria among 101 patients with AD (Teng et al. 2008) . They found that the frequency of depression was significantly higher using the NIMH-dAD (44%) than that obtained using the DSM-IV criteria for major or minor depression (36%) or using the established cut-points on the GDS (33%) and the Cornell Scale for Depression in Dementia (30%). The authors suggest that, compared to the DSM-IV criteria and the other scales, the NIMH-dAD criteria are less
stringent with respect to the requirements for frequency and duration of symptoms, resulting in higher prevalence estimates.

One recent study examined the validity of the Cornell Scale for Depres­sion and the GDS in 145 patients over the age of 65 who were either depressed only, demented only, demented and depressed, or control sub­jects (Korner et al. 2006). They found that while the scales were equally valid for assessing depression in an elderly population, the Cornell Depression Scale retained its sensitivity and specificity (93% and 97%, respectively using a cut-point of >6) in demented subjects. In contrast, the validity of the GDS diminished in the demented population. It is possible that the symptoms assessed using the GDS are more ambiguous in dementia or may overlap with symptoms experienced in dementia, altering the valid­ity of this instrument among demented persons.

Overall, these findings suggest that diagnosis of depression in patients with dementia is challenging and may require consideration of results from numerous depression screening tools as well as caregiver reports. Use of depression scales that focus on symptoms not shared by depression and dementia may also enhance reliability and validity of assessment. Validity of screening tools in demented populations should be consid­ered before choosing a screening instrument. Incorporating reports from patients, caregivers, and clinicians may provide the most complete and accurate picture of the patient’s emotional state.