Patients presenting with depression (DEP) and cognitive impairment (CI), represent a unique, understudied population that is difficult to diagnose, treat and estimate prognosis. Our pilot data, supported by the literature, suggest that many DEP-CI patients show cognitive decline and often convert to dementia, primarily Alzheimer's disease (AD). In DEP-CI, there is a lack of data on treatment response of mood symptoms to antidepressant treatment and particularly of cognitive deficits to cognitive enhancer treatment. Our initial pilot data in a double-blind study showed that donepezil was superior to placebo in improving memory in antidepressant-treated DEP-CI patients. In a second pilot study, open label es-citalopram plus memantine treatment led to a low rate of conversion to dementia.

In this proposed pilot clinical trial, the investigators will evaluate, treat and follow a broad sample of 80 DEP-CI patients at NYSPI/Columbia University Medical Center (N = 40) and Duke University Medical Center (N = 40). Recruitment will be from clinics and/or advertisements. In the treatment protocol, all 80 DEP-CI patients will receive baseline mood and memory assessments and open antidepressant treatment with citalopram for 8 weeks. At 8 weeks, repeat assessment will occur and patients whose depression has responded to citalopram will be randomized to add-on donepezil or placebo. Non-responders to citalopram will receive open treatment with venlafaxine and will be randomized 8 weeks later (16 weeks of open antidepressant treatment) to add-on donepezil or placebo. Patients will be followed for a total period of 18 months with continuous open antidepressant treatment during the trial.

Donepezil is being studied in order to increase the likelihood of obtaining a signal. If the results are positive, the investigators can begin clarifying the mechanism(s) in subsequent trials. Baseline apolipoprotein E e4 genotype, odor identification deficits, and MRI hippocampal and entorhinal cortex atrophy will be explored as predictors of donepezil response in the 18-month trial. Improving cognition and delaying conversion to a clinical diagnosis of dementia in this high risk group will enhance quality of life, reduce family burden, and markedly diminish overall health care costs.

In the elderly, the most common neuropsychiatric disorders are depression (DEP) and cognitive impairment (CI). Their co-occurrence (DEP-CI) may exceed chance. The lack of research in DEP-CI, the need to determine early prognostic indicators, and to develop optimal treatment strategies, was emphasized by a NIH consensus panel. Since depression in patients with CI increases the risk of conversion to dementia, treatment strategies for DEP-CI have longer-term implications beyond acute antidepressant treatment response. However, in DEP-CI, there is a lack of data on treatment response of mood symptoms to antidepressant treatment and of cognitive deficits to cognitive enhancer treatment, and the long-term prognosis in these patients remains unclear. The investigators have initial pilot data showing that donepezil was superior to placebo in improving memory performance in DEP-CI patients, and pilot data showing that patients with DEP-CI treated with combined escitalopram and memantine have a low conversion rate to dementia, primarily Alzheimer's disease (AD), over one year. This proposal is for the first study to explicitly examine cognitive change, including conversion to dementia, in a randomized, double-blind, placebo-controlled donepezil treatment trial in DEP-CI patients treated openly with antidepressants. The investigators will conduct systematic follow-up to evaluate 18-month outcome. In addition to apolipoprotein E ε4 genotype, the investigators will explore MRI hippocampal and entorhinal cortex atrophy and odor identification deficits as biomarkers that moderate response.

This pilot trial will enroll about 80 DEP-CI patients who present to the departments of Psychiatry, Neurology and Internal Medicine at NYSPI/Columbia University and Duke University medical centers, ensuring broad representation for clinical relevance. In the treatment protocol, all 80 DEP-CI patients will receive open antidepressant treatment with citalopram for 8 weeks. At 8 weeks, citalopram responders will continue to be treated on citalopram, while non-responders will switch to venlafaxine treatment for an additional 8 weeks. At 16 weeks, all subjects will be randomized to add-on donepezil or placebo (N.B. Patients with a prior history of nonresponse to both citalopram and venlafaxine will be enrolled in the protocol and treated with bupropion and subsequently with doctor's choice of antidepressant). Patients will be followed for a total period of up to 18 months in the trial, with antidepressant treatment adjusted as needed based on clinical response and side effects, i.e., open antidepressant treatment is continuous for all subjects throughout the trial and is not the experimental intervention. The investigators chose to study antidepressant plus donepezil compared to placebo based on our pilot data and to increase the likelihood of obtaining a signal.

Aim 1 (primary aim of the study). To assess change in cognitive status over 18 months in antidepressant-treated DEP-CI patients comparing donepezil to placebo.

Hypothesis 1. Antidepressant-treated DEP-CI patients on donepezil will show a lower rate of conversion to dementia, primarily AD, compared to antidepressant-treated DEP-CI patients on placebo by the end of the 18-month trial.

Hypothesis 2 (secondary). Compared to the placebo group, the donepezil group will show better cognitive outcome by the end of the 18-month trial (SRT total recall: primary measure; modified ADAS-cog: secondary measure).

Hypothesis 3 (secondary). At the end of 24 weeks on add-on donepezil or placebo, the donepezil group will show better cognitive outcome than placebo (SRT total recall: primary measure; modified ADAS-cog: secondary measure).

Aim 2: To evaluate moderators of treatment on cognitive change in the 18-month donepezil-placebo trial, based on the view that patients with incipient AD brain pathology will have superior cognitive outcome on donepezil. These Aim 2 hypotheses are considered exploratory.

Hypothesis 1. Patients with the apolipoprotein E ε4 allele (homozygote or heterozygote), compared to patients without this allele, will have better cognitive outcome on donepezil compared to placebo.

Hypothesis 2. Lower scores on the UPSIT (odor identification test) at baseline will be associated with better cognitive outcome on donepezil compared to placebo.

Hypothesis 3. Smaller MRI hippocampal and entorhinal cortex volumes (atrophy) will be associated with better cognitive outcome on donepezil compared to placebo.

Inclusion Criteria:

• Of either sex, age 55-95 years old with minimum 8 years of education who meet criteria for both depression and cognitive impairment as described below.

• Study Criteria for "depression":

  i. Patients who meet DSM-IV symptom criteria for Major Depression or Dysthymia for a minimum of 6 months (2 year duration DSM-IV TR criterion not required for dysthymic disorder in this study). ii. 24-item HAM-D ≥14.

• Study Criteria for "cognitive impairment":

  i. Subjective memory or other cognitive complaints. ii. Score ≤ 11 on the Logical Memory II (Delayed Paragraph Recall, Paragraph A) test from the Wechsler Memory Scale - Revised OR a score that is ≥ 1.5 standard deviations below the norms on the FC SRT

• Folstein Mini Mental State (MMSE) score ≥ 21 out of 30.

• Clinical Dementia Rating (CDR) of 0.5 on the memory item and global rating of 0.5 indicating questionable dementia

• Willing and capable of giving informed consent

• A family member or close friend who consents to serve as informant during the study; this can be a telephone informant in the case of patients who do not have a live-in informant or close significant other.

Exclusion Criteria:

• Meets Criteria for dementia (DSM-IV) or probable Alzheimer's disease (NINCDS-ADRDA criteria)

• Meets DSM IV TR criteria for:

  1. schizophrenia, schizoaffective disorder, psychotic depression or other psychosis, or bipolar I disorder
  2. alcohol or substance dependence or abuse (current or within past 6 months)

• Active suicidal ideation or suicidal attempt in last 6 months.

• Clinical stroke with residual neurological deficits.

• Use of medications known to have a negative impact on cognition: benzodiazepines in lorazepam equivalents ≥ 2 mg daily, narcotics, or anticholinergics. (N.B. Medications that may be associated with cognitive impairment but are rarely considered the likely etiology, e.g, theophylline, nifedipine, Beta blockers, will not be excluded.)

• An acute, severe or unstable medical condition. For cancer, acutely ill patients (including those with metastases) are excluded, but past history of successfully treated cancer does not result in exclusion.

• Presence of any of the following disorders: a) CNS infection, with CSF evidence of meningitis, encephalitis, or other infectious process; b) Post-traumatic dementia, defined as dementia with a clear temporal relationship to a severe head injury where consciousness was lost; c) Huntington's disease; d) Multiple sclerosis; e) Parkinson's disease; f) Other neurologic disorders with focal signs, e.g., amyotrophic lateral sclerosis; g) Mental retardation.

• Contra-indication to MRI scan: pacemaker, metal implants following surgery, any other contraindication to MRI (e.g., ferromagnetic aneurysm clips, heart valves). For patients with possible claustrophobia, they can do the MRI with adjunct lorazepam 0.5 mg to reduce anxiety. Patients who cannot do the MRI scan will still be eligible for the clinical trial, i.e., MRI is optional.