For many years, the problem of accurate diagnosis of depression among elderly patients with cognitive impairment has focused on patients with major depression who are misdiagnosed as having dementia—cases of so-called depressive pseudodementia. Yet the more common, and possibly more difficult, predicament for the clinician is making the diagnosis of depression for the patient with established dementia. This difficulty is often exacerbated when a diagnosis is needed for an institutionalized patient with severe dementia, yet reliable information about subjective affective symptoms is lacking due to the patient's cognitive status. This paper discusses ways to improve diagnostic accuracy in treating patients with depression and dementia and reviews treatment options for these patients.
Variability of prevalence rates
The problems of diagnosis are reflected in the wide variation in prevalence rates of depression among patients with dementia reported in the world literature. Reported rates of depressive symptoms for these patients range from 0 to 86 percent, although the mean rate for actual depressive disorders has been calculated as 19 percent (1). Lower rates have been reported in community samples compared with clinical samples (2).
One reason for the variation lies in the interpretation of symptom patterns, partly due to the considerable overlap in the symptoms and behaviors of depression and dementia. Apathy, sleep and appetite changes, poor concentration, psychomotor changes, loss of interests, social withdrawal, self-neglect, irritability, and anxiety can occur in both conditions. To examine the impact that this overlap might have on scores obtained on the 21-item Hamilton Rating Scale for Depression (3), three psychiatrists from our department who are experienced in treating depression in dementia estimated that only three items could be attributed to depression only—depressed mood, guilt, and suicidal ideation.
Disagreement also exists between self-report ratings of depression and observer-rated depression—for example, in cases where persons with cognitive impairment underreport symptoms and caregivers misattribute dementia symptoms to depression (4,5). To complicate matters, there is no certainty that the depression syndrome seen among elderly persons with dementia is the same as the depression experienced by elderly persons who are not demented.
Behavioral disturbances that may be due to depression among elderly persons with dementia include vocally disruptive behavior, aggression, refusal to eat, agitation, and uncooperativeness (6,7). Yet these behaviors may themselves be due to numerous other problems.
Improving diagnostic accuracy
Several reasons to improve diagnostic accuracy in cases of depression in dementia can be given. Research suggests that depression causes excess disability among patients with dementia and that this additional disability can be reversed with successful treatment of the depression (8). Further, depression causes subjective distress in many people with dementia. Finally, caregivers are more likely to be psychologically distressed and have difficulties coping when caring for a person who is depressed or behaviorally disturbed (9). Premature institutionalization may result.
To improve diagnostic accuracy, the clinician first needs to be sensitive to the possibility that depression may develop at any stage of the course of dementia. The risk of depression is increased among patients with vascular and Lewy-body dementias and among those with a history or family history of depression (1). Life events involving loss may precipitate depression among elderly persons with dementia, as they may among those without dementia.
Education of family and professional caregivers about changes in mood and behavior that are associated with depression may assist in identification. In particular, a relatively sudden decline in function over a few weeks requires explanation. Of course, other conditions besides depression, most notably delirium, may be responsible for such changes.
As yet there is no proven gold standard for diagnosing depression in dementia, although several tools are useful. The judicious consideration of all sources of information is required to assist in diagnosis. Both the Neuropsychiatric Inventory (10), which contains a depression subscale, and the Cornell Scale for Depression in Dementia (11) use the best available information from interviews with the patient and the reports of caregivers. The use of self-report scales, such as the Geriatric Depression Scale (12), and patient interviews without other collateral information is problematic when the patient has moderate to severe cognitive impairment. Misdiagnosis can occur in both directions—transient states of dysphoria can be labeled as a depressive syndrome due to an inaccurate longitudinal report, and persistent depressive syndromes may be minimized by patients' poor recall of their effects on themselves.
What are the most reliable clinical indicators of depression among patients with dementia? The answer may depend on the severity of the dementia. Among patients with mild cognitive impairment, depression probably differs little from that seen among elderly patients without dementia. Among some patients, depression may occur in the context of psychological adjustment to the diagnosis of dementia, although insight has not been found to be a reliable marker for depression. Thoughts of death and suicidal ideation are usually part of a depressive syndrome warranting treatment (13).
Among patients with moderate to severe dementia, symptoms that have a predominantly affective quality—guilt, hopelessness, helplessness, worthlessness, sadness, unhappiness, and mood-congruent delusions—should be given greater weighting in the assessment. However, at the level of moderate to severe dementia, behavioral disturbances, including agitation, vocal disruption, and refusal to eat, are often the presenting problem, even though accompanying depressive symptoms are usually present. When the clinician is in doubt, a clinical trial of antidepressant therapy may be warranted.
Given the uncertainties of diagnosis, a higher proportion of treatment failures than would otherwise be expected is likely to occur. Caregivers should be suitably informed of the prognosis, including the degree of improvement that is likely to occur with treatment response.
The decision about which therapeutic approach to take should generally follow the same principles as those used in treating elderly patients without dementia. For treatment of mild to moderate depression associated with adjustment to the diagnosis of dementia for a patient with good insight, supportive psychotherapy alone or in combination with antidepressants should be considered (14).
However, few psychosocial treatments for depression among patients with dementia have been studied. One exception is a recent controlled trial of behavioral treatments using pleasant events and problem solving for patients living in the community and their caregivers (15). The intervention was associated with a significant improvement in depressive symptoms among both the patients and the caregivers, and the improvement was maintained for six months. The study results imply that treatment for the caregiver may be as important as treatment for the patient.
Patients who meet criteria for major depression, particularly with melancholia, and those with atypical depression, should be considered for antidepressant therapy. The choice of antidepressant should largely be based on tolerability.
Limited evidence from controlled trials and case reports suggests that SSRIs such as citalopram are effective in reducing depressive symptoms and improving associated behavioral disturbances, including agitation, irritability, and restlessness among patients with dementia (16,17). Sertraline has been reported to improve symptoms of food refusal and affective symptoms among patients with severe dementia for whom use of traditional diagnostic criteria for depression is not possible (7). Tricyclics may be effective, but they have been reported to have adverse cognitive effects (18).
Among patients with severe depression, particularly with psychosis, electroconvulsive therapy (ECT) should be considered. The outcome is similar to that found in the treatment of elderly persons without dementia, although patients with dementia may experience increased post-ECT confusion (19).
Other pharmacological agents could be used to treat depressive symptoms. Anticonvulsants such as carbamazepine, valproate, and gabapentine have long been used in treatment of bipolar disorder and have been found to reduce agitated behaviors among patients with dementia (20), although their usefulness in the treatment of depressive disorders is uncertain. Some reports have suggested that antidementia drugs may improve neuropsychiatric symptoms that include anxiety and apathy among patients with Alzheimer's disease (21). Again, it is unclear at this stage if such improvement is simply a general effect that ameliorates subsyndromal depressive symptoms or whether the drugs have a role in the treatment of comorbid depressive disorders. Psychostimulants have been recommended, particularly for patients for whom apathy is the main problem, but controlled trials are lacking. It has been suggested that treatment response is usually noted within a few days.
Whatever approach is used to treat the depression, treatment resistance should be managed in the same fashion as it is among elderly patients without dementia. Response to antidepressants may be slower, so an adequate trial duration at a therapeutic dosage of at least six to eight weeks is required. The cases of patients who fail to respond should be reviewed to reaffirm the diagnosis before embarking on a second therapeutic trial. The combination of antidepressants with psychosocial therapies involving the caregiver may be the most effective approach, although controlled studies are lacking.
The diagnosis of depression among patients with dementia is a challenge to most clinicians due to the variability of symptom patterns, differences between self-report and observer ratings, and the frequency of atypical cases. Depression should be treated, because it causes patients excess disability and suffering and is associated with distress among caregivers.
Psychosocial treatments, often involving the caregiver, are indicated for patients with mild to moderate depression, while the choice of antidepressants for treatment of more severe depressions should be based on how the patient tolerates the drug. Treatment resistance is common, and serial treatments involving combinations of psychosocial and physical treatments may be required.
Dr. Draper is senior staff specialist in psychogeriatrics at Prince of Wales Hospital in Randwick, New South Wales 2031, Australia. He is also senior lecturer in the Schools of Psychiatry and Community Medicine at the University of New South Wales in Sydney (e-mail, firstname.lastname@example.org). Marion Z. Goldstein, M.D., is editor of this column.